INLAND LAKES 
 
 OF 
 
 MICHIGAN 
 
 I. D. SCOTT 
 
Class. 
 
Digitized by the Internet Archive 
 in 2011 with funding from 
 The Library of Congress 
 
 http://www.archive.org/details/inlandlakesofmicOOscot 
 
.MIOHKJAN GEOLOGICAL AND IJIOLOOIOAL SURVEY 
 
 ^ 
 
 IMiblicatiou 30 
 Goolojfical Series 26 
 
 INLAND LAKES OF MICHIGAN 
 
 By 
 I. D. SCOTT 
 
 PUBLISHED AS A PART OF THE ANNUAL REPORT OF THE BOARD OF 
 GEOLOGICAL SURVEY FOR 1020 
 
 LANSING, MICHIGAN 
 AVYNKOOP IIALLENBECK CRAWFORD CO., STATE PRINTERS 
 
 1921 
 

 LIBRARY OF CONOi <f^'S§ 
 
 JAN101922^ 
 
 OOCUIMt'N'rS D ViJIOW 
 
BOARD OF GEOLOGICAL AND BIOLOGICAL SURVEY, 1920. 
 
 ex officio : 
 
 The Governor of the State^ 
 
 HON. ALBERT E. SLEEPER. 
 
 The Superintendent of Public Instruction^ 
 HON. THOMAS E. JOHNSON. 
 
 The President of the State Board of Education^ 
 HON. FRED JEFFERS. 
 
 director^ 
 R. A. SMITH. 
 
 scientific advisors. 
 Geologists.— Dr. L. L. Hubbard, Houghton; Prof. W. H. Hobbs, 
 Ann Arbor; Prof. W. H. Sherzer, Ypsilanti; Prof. E. C. Case, 
 Ann Arbor. 
 
 Botanists. — Prof. E. A. Bessey, East Lansing, Prof. F. C. Newcomb, 
 Ann Arbor. 
 
 Zoologists. — Prof. W. B. Barrows, East Lansing; Prof. J. Reighard, 
 Ann Arbor; Dr. Bryant Walker, Detroit. 
 
LETTER OF TRANSMITTAL. 
 
 To the Honorahle, the Board of Geological and Biological Survey 
 of the State of Michigan: 
 
 Governor Albert E. Sleeper 
 
 Hon. Fred Jeffers 
 
 Hon. Thomas E. Johnson 
 
 Gentlemen : I have the honor to submit herewith a monographic 
 report on the Inland Lakes of Michigan, by Dr. I. D. Scott with the 
 recommendation that it be published and bound as Publication 30, 
 Geological Series 25. 
 
 This monograph is the result of several years of field study and 
 its object is to describe and explain the large lakes whose history is 
 a part of the history of the Great Lakes, as well as those smaller 
 lakes having an economic or aesthetic value. The book should be 
 of value to students and teachers of physiography, to tourists and 
 those desiring to attract tourists to the State, to State, city and 
 town officials seeking park sites and municipal water supplies, as 
 well as to the fisherman and those who seek recreation. 
 
 The field work upon which the greater part of the report is based 
 was made during the summer months of 1913 and 1914. Progress 
 on the writing of the report was interrupted by the War and further 
 field studies and a reconnaisance of a greater number of the smaller 
 lakes were made in the spring and summer of 1920. Earlier pub- 
 lication was desired but compensation for the delay is afforded 
 in the more comprehensive report of a greater number of small but 
 locally important lakes. 
 
 Very respectfully yours, 
 
 R. A. SMITH, 
 
 Director. 
 
 Lansing, Michigan, Dec. 11, 1920. 
 
PREFACE. 
 
 The number of inland lakes in the State of Michigan is not defin- 
 itely known but has been placed by some at greater than five thous- 
 and. They range in area from thirty-one square miles down to 
 small, unnamed ponds and, in the Southern Peninsula alone, more 
 than seventy have an area of one square mile or more. This number 
 is considerably increased when the lakes of the Northern Peninsula 
 are added. It has been estimated that lakes constitute about one- 
 fiftieth of the total area of the State, a percentage so large that one 
 may, with justice, entertain doubt as to their value to the common- 
 wealth. 
 
 This doubt becomes almost a conviction if one considers the well- 
 known fact that most lakes cover land of very high fertility. A 
 classic example of the value of such land is the bed of former Lake 
 Agassiz upon which is grown a large part of the enormous wheat 
 crops of the Dakotas and Minnesota. Assuming that all of the in- 
 land lakes of the State could be drained, more than twelve hundred 
 square miles of land of exceptional value would be opened to culti- 
 vation. In addition, it is probable that many nearby swampy areas 
 would likewise be made available for use, and sanitary conditions 
 be greatly improved by the extinction of the breeding places for 
 disease-spreading insects. Also lake deposits, such as marl and 
 peat, are frequently of considerable value and their exploitation, 
 which is usually destructive to the lake, may be a legitimate enter- 
 prise. But our initial assumption that all of these lakes can be 
 drained is impossible and, inasmuch as the data at hand is not 
 sufficient for a fair estimate of the areas that can be reclaimed in 
 this way, no attempt is made to state definitely their value. It is 
 obvious, however, that it would be enormous. 
 
 Yet, on the other hand, lakes in themselves are a very useful re- 
 source and function in such varied Avays that, although a statement 
 as to their monetary value is impossible, there are many who 
 consider their jiresence, within limits, a valuable asset. Among the 
 functions performed by lakes may be mentioned their service as 
 natural reservoirs. They accommodate the waters of spring freshets 
 and melting snows with comparatively small rise in water level and 
 thus lessen the flooded condition of streams and hinder the strip- 
 ping of the land. Also, by throwing dams across the outlets the 
 
viii PREFACE 
 
 outflow of the lakes may be controlled for a number of purposes, 
 for example, — power, irrigation, logging operations, city water sup- 
 ply, etc. The consideration of lakes as a source of food supply, as 
 highways of commerce and as a tempering effect on climate applies 
 more particularly to the larger lakes and inland seas and is, there- 
 fore, mentioned only in passing. 
 
 But the most important function of lakes is, however, not com- 
 mercial but lies rather in their unique advantages for the recreation 
 of man. Here one may rest 
 
 "Escaped awhile, 
 
 From cares that wear the life away, 
 To eat the lotus of the Nile, 
 And drink the poppies of Cathay, 
 To fling the loads of custom down. 
 Like driftweed, on the sand slopes brown, 
 And in the sea waves drown the restless pack 
 Of duties, claims and needs that barked upon their track." 
 
 WJiittier. 
 
 The pure air, cool temperatures and simple conditions of life 
 stimulate renewed physical and mental vigor. Yet, lakes would 
 fail in their service as recreational centers were opportunities for 
 expression of the revived faculties lacking. This, however, appears 
 contrary to fact, as shown by the ever-increasing numbers which 
 migrate to them each summer. 
 
 The mere mention of the familiar water-sports should be sufficient 
 to emphasize the appeal of lakes to our physical natures. But the 
 appeal is deeper. Lakes are attractive not alone for their beauty 
 but to a large extent because they portray so faithfully our own 
 emotions and intensify .the condition of our physical environment. 
 During periods of calm, winter's solitude is accentuated by the ice- 
 bound expanse and, in summer, tranquility is reflected from the un- 
 broken surface. At times its leaden waters appear sullen, fortelling 
 impending storms, at others boisterous and jubilant, and again, 
 whipped to a state of fury. 
 
 Nor, is the intellectual side wanting. Of the various phases of 
 the study of nature none is more easily observed and readily in- 
 terpreted than Earth Science from the physiographic viewpoint 
 and that part devoted to the study of lakes is one of the most in- 
 teresting. From this viewpoint undrained areas are considered 
 as one of the early phases in the wearing away of a land surface 
 by streams. As the streams deepen their valleys and stretch out 
 tributaries, all parts of the basin become completely drained and 
 lakes are, therefore, considered as transient features of the land- 
 
PREFACE ix 
 
 scape. From a pliysiograpliic standpoint one may study the entire 
 life history of such bodies of water. In this work tlie principle 
 events to be deciphered are the origin of the basin, its develop- 
 ment by the various agencies active upon it and finallj^ its ex- 
 tinction or death. In addition, the study has a much wider appli- 
 cation, for lakes are but oceans in miniature, except for tides, and 
 present similar problems on a more convenient scale. 
 
 From a practical standpoint the physiographic study of a lake 
 gives a more intimate knowledge of and a closer acquaintance with 
 the conditions not only of the shores but the surrounding coiinriy. 
 This knowledge and familiarity cannot fail to be of service to the 
 resorter both in the selection of the lake and the site on it. To 
 illustrate, the larger lakes, although they may often be treacherous 
 in times of storm, have advantages over smaller ones. The summer 
 temperatures are apt to be lower and the very factors which make 
 the lake dangerous, inasmuch as they w^ork on a large scale, are 
 beneficial in various ways. Thus, better and cleaner beaches are 
 built and the submerged terrace is broader and drops into deep 
 water from depths usually greater than a man's height, lessening 
 the danger of accidents due to walking off the ''drop off" or "channel 
 bank". The situation of the lake is important and proximity to 
 other large bodies of water is favorable. The ideal location is to 
 the east of a large lake because the wands, prevailingly from the 
 west, are cooled in their passage over the large expanse of water 
 which has a lower temperature than the air in summer. 
 
 It w^ould seem axiomatic that the shores and surrounding country 
 should be well drained, if the lake is to be useful for summer 
 homes, in order to secure healthful living conditions and to insure 
 a minimum of pests. However, the writer has seen far too many 
 resorts planned on a magnificent scale which exist only on plats 
 executed for the use of distant real estate dealers and has helped in 
 locating some of the properties only to find them situated on an 
 insignificant lake in the midst of a swamp. Physiographic study 
 would eliminate this. An ideal site, according to the writer, is to 
 be found on lakes which have stood for a considerable time at an 
 appreciably higher level — of which Michigan has many — for under 
 these conditions a sandy terrace is now exposed high and dry above 
 the level, surmounted by a cliff of varying height from the base of 
 which springs of cool, pure water often flow. 
 
 As stated above, lakes have served a useful purpose in the storing 
 of w^ater for various projects which, in most cases, necessitates the 
 building of a dam, thereby interfering with the natural level of 
 the lake in question. This may involve a raising or lowering of the 
 
X PREFACE 
 
 level, or both at different times of the year, and results in serious 
 inconvenience and often damage to property along the shores. A 
 lowering of the level means stranded docks and boat houses; a 
 strip of the bottom exposed that often becomes foul from swampy 
 conditions and decaying vegetation unless the lowering is perma- 
 nent. A raising of the level is more serious and results in flooded 
 shores and an increased activity of the waves. The latter is very 
 noticeable on many lakes of the State and various means are em- 
 ployed to stop shore destruction by wave action. These, however, 
 afford only temporary relief and are a source of expense and con- 
 stant attention. In this case the physiographic principles seem to 
 be ignored. 
 
 From the educational standpoint the study is also of importance. 
 Physical Geography, in whole or in part, is quite generally taught 
 in the schools of the State and it is truly educational in scope. 
 Furthermore, the process of reasoning is complete. It puts new 
 meaning in familiar things and only moderate teaching ability is 
 demanded to arouse a lively interest on the part of the pupils. But 
 it is not primarily a text book subject. Illustrative material is a 
 necessity. Pictures may partially supply the need but by far the 
 best illustrations are those obtained by direct observation. Excep- 
 tional indeed are the localities that do not furnish abundant ac- 
 cessible material for field study. Our lakes illustrate one phase of 
 the subject of physiography and, on account of their number and 
 distribution, should be a most valuable asset to the teachers of the 
 State. Even the smallest pond is of some value in this respect and 
 it is urged that advantage be taken of the opportunities. 
 
 It is hoped that from the brief statements concerning the points 
 of view from which lakes may be considered it will be clear that 
 both are well supported. As a matter of fact there are many lakes 
 in our State that might well be exploited commercially but there 
 are others which appear to be of greater value in their natural con- 
 dition. Each lake, then, becomes a problem in itself and a physio- 
 graphic study of the lake seems a prerequisite to its solution. The 
 technicalities of such a study need not be overwhelming. No 
 branch of earth science is more interesting that the study of lakes, 
 and no special equipment other than an active brain and a reason- 
 ably vigorous physique is necessary. It provides both physical and 
 mental recreation of the best type and is profitable as well as in- 
 teresting. Familiar features take on new meaning and the changes 
 taking place are a source of continued interest. The writer is con- 
 vinced that the report of the studies of Michigan's inland lakes, 
 undertaken during the summers of 1913, 1914 and 1920, will be of 
 
PREFACE xi 
 
 greatest service if the needs of the increasing nnmber of summer 
 visitors and of those engaged in educational work are kept in mind. 
 Therefore, the attempt has been made to present the essentials of 
 the subject in as untechnical a way as possible in the introductory 
 chapters. Following the introductory work are detailed descrip- 
 tions and discussions of the physiography of some of the more 
 important lakes. 
 
 It is obvious that all of the lakes of the State could not be in- 
 cluded in this study and therefore a selection was made based on 
 the importance, accessibility, distribution and promise of scientific 
 results. Mistakes, both of omission and commission, appear in "this 
 selection as the work progressed, the principal difficulty being in 
 the matter of distribution. In order to improve this, it was de- 
 cided to include a large number of lakes in a reconnaisance study 
 during the summer of 1920 and the results of this work are given 
 in the final chapters. In these brief reports an attempt has been 
 made to classify the lake basins and to state the type of the adjust- 
 ments that have taken place on the shores. Also some information 
 as to the accessibility of the lake, localities where the adjustments 
 may be easily recognized, and the desirability of the lake as a sum- 
 mer resort maj^ be included. 
 
 It is recognized that there will be some disappointment in the 
 selection of the lakes described in detail but this need not be serious 
 if one of the objects of this report is attained, namely, to present 
 the underlying principles in such a way that they may be applied 
 by those who may study these pages. Often the difficulty is in get- 
 ting a start and it is felt that the final chapters may be of service 
 in this respect. 
 
 The first essential in undertaking a study of this kind is to have 
 a reasonably accurate map of the lake and its surroundings. Pre- 
 ferably this should show relief features; and the best to be ob- 
 tained are the topographic maps made by the United States Geolog- 
 ical Survey in co-operation with the State. These maps are about 
 thirteen by seventeen inches in size and are made on a scale of ap- 
 proximately one inch to one mile for most districts, thus including 
 and area of nearly 220 square miles. They are sold by the Director 
 of the United States Geological Survey, and bj" the Michigan Geo- 
 logical Survej^ at a nominal cost of ten cents and by all means 
 should be procured, if they are available. Unfortunately much of 
 the State is as yet unmapped, but encouraging progress lias been 
 made recently and we look forward to a more rapid production of 
 these most useful maps as the demand increases. 
 
xii PREFACE 
 
 Other maps that are useful are those issued by the United States 
 Lake Survey, Detroit, Mich. They are very accurate, both as to 
 shores and depth of water, but only a limited amount of the sur- 
 roundings is included in the map. These maps are made for naviga- 
 tion purposes and represent navigable waters directly connected 
 with the Great Lakes with one exception, the map of the Inland 
 Route including Crooked, Burt and Mullet Lakes. 
 
 In most cases the only maps available are the United States 
 Land Survey plats which give only the outline of the lake, and 
 this is not accurate. Those used in this report were corrected in 
 a rough way and, although far from satisfactory, are sufficiently 
 reliable for the purpose. Most county maps and atlases are com- 
 piled from these plats and may be relied upon to the same extent. 
 
 The three main problems to be studied are, as indicated above, 
 the origin of the basin, its subsequent development, and its extinc- 
 tion. The first of these involves a knowledge of the topographic 
 features of the region and necessitates an examination of the sur- 
 rounding country. It is often the most difficult to decide, and 
 valuable information will be found in a publication of the Michigan 
 Geological and Biological Survey by Frank Leverett: Publication 
 25, Geological Series 21, Surface Geology of Michigan. "■• The in- 
 serted maps are especially valuable and should be mounted on cloth 
 to save wear and tear. 
 
 The development of the basin and causes working towards ex- 
 tinction are best discovered by making a detailed study of the 
 shores and the off-shore lake bottom. The use of a boat is necessary 
 for the latter and may be serviceable for a traverse of the shores if 
 the lake is large. In general, however, a traverse on foot does away 
 with the inconvenience of landing and gives more satisfactory re- 
 sults for the beginner. The sounding of the shallow water requires 
 some apparatus. The writer found an exhausted dry cell a con- 
 venient weight but does not recommend any weight under twenty 
 pounds for deep water. Accurate soundings involve both depths 
 and locations. The process is tedious, and expensive instruments 
 are necessary, therefore this is not recommended. For our pur- 
 poses the depths of the water over the terrace and the width of the 
 terrace are desirable. The width is the more difficult to obtain 
 but an estimation will answer the purpose unless a detailed and 
 serious study is to be undertaken. 
 
 If this report is successful, the physiographic study of lakes will 
 
 ♦Publication 25 is a revision of two earlier publications, viz. : Pub. 7, Surface 
 Geology of the Northern Peninsula and Pub. 9, Surface Geology of the Southern 
 Peninsula of Michigan. Both these publications are now out of print. 
 
PREFACE xiii 
 
 be the result. Workers may benefit thems.elves and others as well 
 if their results are known, and the writer will be glad to receive 
 suggestions, criticisms, and new developments concerning any lakes 
 of the State whether included in this report or not. The use of 
 the camera is strongly recommended also. 
 
 The illustrations in the report are from drawings and photo- 
 gaphs by the author unless otherwise accredited. Acknowledgments 
 are due to the many individuals who by information and services 
 made much of the field work possible. Mr. Frank Leverett, United 
 States Geologist, has aided the writer in glacial problems both by 
 personal communication and by placing at his disposal valuable 
 data, at the time unpublished. The advice and assistuce of Mr. R. 
 C. Allen, former Director of the Michigan Geological Survey and 
 Mr. R. A. Smith, present Director, have been of especial service in 
 the prosecution of the field work and the preparation of this report. 
 
TABLE OF CONTENTS 
 
TABLE OF CONTENTS 
 
 Letter of Transmittal, R. A. Smith v 
 
 Preface vii 
 
 List of Illustrations xii 
 
 CHAPTER I 
 
 The Origin and Classification of Lake Basins 3 
 
 The Atmosphere, 3; Running Water, 4; Ground Water, 6; Glaciers, 6; Physiography, 
 of the State of Michigan, 13; Northern Peninsula, 13; Highland Province, 15; 
 Lowland Province, 16: Southern Peninsula, 18; Origin of Lake Basins, 22; Rivers, 
 Ox-bows, 23; Groundwater, Sinks, 25; Diastrophism, Faulting, 26; Warping, 27; 
 Glacial Lakes, Scour, 28; Deposition, Morainal, 28; Inter-niorainal, 30; Morainal 
 dam, 31; Fosse, 31; Pits, 32; Waves and Currents, Lagoons, 31; References, 35. 
 
 CHAPTER II 
 
 The Development of Lake Shores and Extinction of Lakes 37 
 
 Characteristics of Waves and Currents, 38; Waves, 38; Currents, 43; The Work of 
 Waves and Currents, 46; The Work of Waves, 47; The Work of Currents, 53; The 
 General Effects of Waves and Currents, 55; The Work of Ice on Lake Shores, 56; 
 Ice Ramparts, 56; Ice-jam, 59; The Extinction of Lakes, 61; The Cycle of Shore 
 Development, 66; References, 67. 
 
 CHAPTER III 
 
 Lakes of the Cheboygan River Basin 69 
 
 Crooked Lake, 72; Burt Lake, 78; Mullett Lake, 87; Black Lake, 99; Douglass 
 Lake, 106. 
 
 CHAPTER IV 
 
 Lakes of the Grand Traverse Region 119 
 
 Walloon Lake, 120; Pine Lake, 134; Torchlight Lake, 143; Elk Lake, 159; Crystal- 
 Lake, 167. 
 
 CHAPTER V 
 
 Interior Lalces of the Southern Peninsula. Western Interlobate Area 177 
 
 Lakes of St. Joseph County, 177; Klinger Lake, 178; Corey Lake, 182; Clear Lake, 
 187; Long Lake, 190; Gun Lake, 195; Lakes Cadillac and Mitchell, 203; Lake 
 Mitchell, 205. 
 
 CHAPTER VI 
 
 Interior Lakes of the Southern Peninsula, Continued. Lakes of the Western Inter-lobate 
 
 Area and of Alpena County 209 
 
 Houghton Lake, 209; Higgins Lake, 215; Portage Lake, Crawford County, 224; 
 Otsego Lake, 227; Hubbard Lake, 235; Long Lake, Alpena County, 242. 
 
 CHAPTER VII 
 
 Interior Lakes of the Southern PeTiinsula, Continued. Lakes of the Eastern Inter- 
 lobate Area 263 
 
 Orchard Lake, 253; Cass Lake, 258; Long Lake, Genesee County, 261. 
 
 CHAPTER VIII 
 
 Lakes of the Northern Peninsula 269 
 
 Brevort Lake, 269; The Manistique lakes, 275; Manistique Lake, 277; Whitefish 
 Lake, 279; Indian Lake, 281; Huron Mountain Lakes, 286; Pine Lake, 289; Conway 
 Lake, 290; Rush Lake, 291; Ives and Mountain Lakes, 293; Lake Michigamme, 296; 
 Portage and Torch Lakes, 303; Lake Gogebic, 310; Chicagon Lake, 318. 
 
Kviii TABLE OP CONTENTS 
 
 CHAPTER IX Page 
 
 Lakes of the Kalamazoo Morainic and Outwash Systems 321 
 
 Sugarloaf Lake, 324; Cavanaugh Lake, 324; Crooked Lake, 325 ; Wampler's Lake, 325; 
 Vineyard Lake, 325; Clark Lake, 326; Devils and Round Lakes, 327; Bawbeese, 
 Coldwater and Marble Lakes, 328; Coldwater Lake, 329; Marble Lake, 329; Goguac 
 Lake, 330; Gull Lake, 330; Crooked Lake, Barry County, 332; Wall Lake, 332; 
 Austin Group (Austin, Long, West, and Gourdneck), 333; Long Lake, 334; Diamond 
 Lake, 334. 
 
 CHAPTER X 
 
 Lakes of the Valpariaso-Charlotte, Morainic and Outwash Systems 335 
 
 Walled Lake, 335; Whitmore Lake, 336; Huron River Group (Portage, Base Line, 
 Strawberry, Zuke5^ etc.), 337; Duck Lake, 337; Chippewa Lake, 338; Hess Lake, 
 339; Reed Lake, 339; Miner Lake, 340; Paw Paw Lake, 340; Cora Lake, 341; Sister 
 Lakes, 342; Indian Lake, 343. 
 
 CHAPTER XI 
 
 Lakes of the Port Huron Morainic and Outwash Systems 345 
 
 Crystal Lake, Montcalm County, 345; Coldwater Lake, Isabella County; 346; Mis- 
 saukee Lake, 347; Manistee Lake, 348; Big Star Lake, 349; Fremont Lake, 350; Big 
 Blue Lake, 351; Twin Lakes, 352; Wolf Lake, 352. 
 
 CHAPTER XII 
 
 Border Lakes and Lakes of Diverse Origin Outside the Port Huron Morainic System. 355 
 Border Lakes, 355; Bass Lake, 357; Portage Lake, Manistee County, 358; Platte 
 Lakes, 359; Glen Lake, 360; Grand Lake, Presque Isle County, 361; Van Etten 
 Lake, 363; Lakes of Diverse Origin, 364; Long Lake, Iosco County, 364; Sage Lake, 
 365; Carp Lake, 366; Grand Traverse Lakes, 367; Green and Duck Lakes, 367; 
 Long Lake, Grand Traverse County, 368; Silver Lake, 369; Bear Lake, 370. 
 
LIST OF ILLUSTRATIONS 
 
 PLATES 
 
 Page 
 Kalamazoo County Frontispiece 
 
 View on Gull Lake, 
 
 Plate I. Sink Hole, Sunken Lake, Alpena County 
 
 Plate II, A. Waves, Burt Lake 
 
 Plate II, B. Whitecaps and breakers, Higgins Lake 
 
 Plate III, A. Ice-push terrace, Athabasca 
 
 Plate III, B. Lake nearing extinction by vegetation Second Sister Lake, near Ann 
 
 Arbor 
 
 Plate IV, A. Sand Dunes, Little Traverse Bay, Round Lake in foreground 
 
 Plate IV, B. Marl Bed, Sturgeon River near Burt Lake 
 
 Plate V, A. Pittsburgh Landing, Burt Lake 
 
 Plate V, B. Boulder-paved Bank, Burt Lake 
 
 Plate VI, A. Nigger Creek, Mullett Lake 
 
 Plate VI, B. Stony Point, Mullett Lake 
 
 Plate VII, A. Hook, Douglass Lake 
 
 Plate VII, B. Raised Beaches, Pine Lake 
 
 Plate VIII, A. Stratified Edge of Built-Terrace, Torchlight Lake 
 
 Plate VIII, B. Raised Boulder Strand, Torchlight Lake 
 
 Plate IX, A. Algonquin Bar, Torchlight Lake 
 
 Plate IX, B. Algonquin Bar, Elk Lake 
 
 Plate X, A. Ice-jam, Crystal Lake 
 
 Plate X, B. " Drop-off," Corey Lake 
 
 Plate XI. Bar, Long Lake, St. .loseph County 
 
 Plate XII, A. White Sands, Lake Mitchell 
 
 Plate XII, B. Ice-jam, Lake Mitchell 
 
 Plate XIII. Stream diverted by current.s. Lake Mitchell 
 
 Plate XIV, A. " Beaver Dam," Otsego Lake 
 
 Plate XIV, B. Ice-formed V-bar, Long Lake, Alpena County 
 
 Plate XV, A. Ice-formed Spit, Long Lake, Alpena County 
 
 Plate XV, B. Ice Rampart, Apple Island, Orchard Lake 
 
 Plate XVI, A. Spit, Cass Lake 
 
 Plate XVI, B. Dissection of ice rampart, Cass Lake 
 
 Plate XVII, A. Ice rampart, Indian Lake 
 
 Plate XVII, B. Vecetation — Lagoon and bar near Indian Lake, Schoolcraft County. 
 
 Plate XVIII. E.^tinction by vegetation, Arm of PiiiC Lake, Huron Mountains. . . . 
 
 Plate XIX, A. Waves breaking over submerged bar, Taire Michiganime 
 
 Plate XIX, B. Storm beach, east ead of Lake Michigamme 
 
 Plate XX, A. Looking North on Lake Gogebic 
 
 P late XX, B. Mountain Lake, Huron Mountains 
 
 26 
 42 
 
 42 
 60 
 
 60 
 71 
 71 
 80 
 80 
 91 
 91 
 114 
 114 
 146 
 146 
 155 
 155 
 170 
 170 
 191 
 206 
 206 
 207 
 2.32 
 232 
 249 
 249 
 258 
 2.58 
 283 
 283 
 289 
 300 
 300 
 313 
 313 
 
 rig. 
 
 1 
 
 Fig. 
 
 2 
 
 Fig. 
 
 3. 
 
 F'ig. 
 
 4. 
 
 Fig. 
 
 o 
 
 Fig. 
 
 6 
 
 Fig. 
 
 7 
 
 Fig. 
 
 8 
 
 Fig. 
 
 9 
 
 Pig. 
 
 10 
 
 Fig. 
 
 11 
 
 Fig. 
 
 12 
 
 Fig. 
 
 13 
 
 Fig. 
 
 14 
 
 FIGURES 
 
 A stage of the Wi.sconsin glacier showing lobate character of the ice front . . 
 
 Relative position of glacial deposits 
 
 Map showing the morainic systems of Michigan 
 
 Map of Lake Algonquin .showing relation to present Great Lakes 
 
 Map of Lake Nipissing showing its relation to the present Great Lakes. . . 
 
 Map showing physiographic provinces of the Northern Peninsula 
 
 Map showing physiographic provinces of the Southern Peninsula 
 
 Map and diagrams showing characteristics of ox-bow lakes 
 
 Map and diagram to show manner of formation of lakes in sink-holes. . . . 
 
 Diagram and map to show characteristics of rift valley lakes 
 
 Diagram showing characteristics of lagoons of the drowned stream type. . 
 Diagrams and maps showing characteristics of morainal (marginal) lakes. 
 
 Lake Antoine. A lake of the morainal dam type 
 
 Sketch and diagram to illustrate mamier of formation of fosse lakes 
 
 7 
 10 
 10 
 12 
 13 
 14 
 19 
 24 
 25 
 26 
 27 
 29 
 30 
 32 
 
XX LIST OF ILLUSTRATIONS 
 
 Paee 
 
 Pig. 15. Map and diagram to illustrate ciiaracteristics of pit lakes 33 
 
 Fig. 16. Diagram illustrating movement of water particles in wa-i.es of oscillation 38 
 
 Fig. 17. Diagram illustrating the ideal development of waves 39 
 
 Fig. 18. Diagram illustrating rapid decrease with depth of motion of water particles in a 
 
 wave 40 
 
 Fig. 19. Diagram to illustrate the formation of white caps 41 
 
 Fig. 20. Diagram of the path of a particle of water affected by a number of successive 
 
 waves .... 41 
 
 Fig. 21. Diagram showing paths of a water particle in waves of translation 43 
 
 Fig. 22. Diagram showing upward and forward, and downward and forward movements 
 
 of a wave of translation 43 
 
 Fig. 23. Diagram to show actual motion of water pajticles striking shore at an oblique 
 
 angle 44 
 
 Fig. 24. Diagram to show currents .set up in a circular lake ., 45 
 
 Fig. 25. Diagram showing conditions under which a shore current is formed 46 
 
 Fig. 26. Farwells Point, an undercut cliff on Lake Mendota, Wisconsin ■. . . 47 
 
 Fig. 27. Profile of a cut and built terrace on a steep rocky shore 49 
 
 Fig. 2k. Profile of a cut and built terrace on a shore formed of loose materials 49 
 
 Fig. 29. Section of a barrier with characteristic steep landward and gentle seaward slope 50 
 
 Fig. 30. Sea-cave's in process of formation along joints 51 
 
 Fig. 31. A sea arch and small caves on the shore of one of the Apostle Islands, Lake 
 
 Superior 52 
 
 Fig. 32. Stack on the shore of Lake Superior 53 
 
 Fig. 33. Diagram illustrating the filling of a lake by vegetation 63 
 
 Fig. 34. Outline map showing lakes of the Cheboygan" River Basin 70 
 
 Fig. 35. Map of Crooked Lake showing configuration of bottom 72 
 
 Fig. 36. Spit oflfset near distal end. Northwest end of Oden Island, Crooked Lake. ... 75 
 
 Fig. 37. Map showing outlines and configuration of the bottom of Burt and Mullet Lakes 78 
 
 Fig. 38. Hook at Greenman Point, Burt Lake 83 
 
 Fig. 39. Diagrammatic profile of the exposed terraces of Burt Lake 83 
 
 Fig. 40. Recession of a flooded shore line due to removal of vegetation 94 
 
 Fig. 41. Outline map of Black Lake, Chebo^'gan and Presque Isle counties. 100 
 
 Fig. 42. Outline map of Douglass Lake 107 
 
 Fig. 43. Sand cusp, Douglass Lake 112 
 
 Fig. 44. Sand cusp, Douglass Lake 112 
 
 Fig. 45. Conventional diagram illustrating the increase in curvature of waves within an 
 
 embayment 113 
 
 Fig. 46. Diagrammatic plan of bars and lagoon at Sedge Point, Douglass Lake 114 
 
 Fig. 47. Diagrammatic plan of bars and lagoons at Diogenes Point, Douglass Lake. . . . 115 
 
 Fig. 48. Outline map of Walloon Lake. 121 
 
 Fig. 49. Spit at the west side of the entrance to the North Arm Walloon Lake 124 
 
 Fig. 50. Spit at east side of the entrance to the North Arm, Walloon Lake 124 
 
 Fig. 51. Ice rampart, Mud Hole, Walloon Lake 129 
 
 Fig. 52. Map showing outline and configuration of the basin of Pine Lake, Charlevoix 
 
 County 134 
 
 Fig. 53. Diagram showing exposed terraces along the shores of Pine Lake 138 
 
 Fig. 54. Map of Torchlight Lake and vicinity 144 
 
 Fig. 55. Diagram showing terraces about the shores of Torchlight Lake 147 
 
 Fig. 56. Conventional sketch of point designated at B on ma,p. Torchlight Lake 150 
 
 Fig. 57. Conventional sketch of point near Y. M. C. A. camp, Torchhght Lake 152 
 
 Fig. 58. Outhne map of Elk Lake ■ • 160 
 
 Fig. 59. Tenement house in process of burial by a moving dune, Elk Rapids. 165 
 
 Fig. 60. Outline map of Crystal Lake, Benzie County 168 
 
 Fig. 61. Small ice rampart of sand. Crystal Lake 170 
 
 Fig. 62. Diagram showing the attachment of the bars at the west end of Crystal Lake 
 
 to the north shore cliffs 173 
 
 Fig. 63. Topographic map of Khnger Lake and surroundings 178 
 
 Fig. 64. Topographic map of Corey, Clear and Long Lakes and surroundings 182 
 
 Fig. 65. Sketch of small island tied to the mainland by two bars or tombolos 185 
 
 Fig. 66. Outline map of Gun I ake, Barry County 194 
 
 Pig. 67. Outline map of Lakes Mitchell and Cadillac, Wexford County 202 
 
 Fig. 68. Map showing the distribution of the glacial formations in the vicinity of 
 
 Houghton and Higgins Lakes 210 
 
 Fig. 69. Outline map of Houghton Lake, Roscommon County .■ 212 
 
 Fig. 70. Outline map of Higgins Lake, Roscommon Couny 216 
 
LIST OF ILLUSTRATIONS xxl 
 
 Page 
 
 Fig. 71. Outlijie map of Portage Lake, Crawford County > 225 
 
 Fig. 72. Map of the glacial formations in the vicinity of Otsego Lake 228 
 
 Fig. 73. Outline map of Otsego Lake 229 
 
 Fig. 74. Map of the glacial formations in the vicinity of Hubbard Lake 235 
 
 Fig. 75. Outline map of Hubbard Lake, Alcona County 237 
 
 Fig. 76. Outline map of Long Lake, Alpena County 243 
 
 Fig. 77. Boulder pushed on-shore by an ice -jam, Long Lake, Alpena Coimty 248 
 
 .Fig. 78. Boulder pushed on-shore by an ice-jam 248 
 
 Fig. 79. Topographic map of Cass and Orchard Lakes and vicinity, Oakland Couhty . . 254 
 
 Fig. 80. Outline map of Long Lakeland neighboring lakes, Genesee County 262 
 
 Fig. 81. Outline map of Brevort Lake, Mackinac County 270 
 
 Fig. 82. OutUne map of the Manistique Lakes, Luce and Mackinac counties 276 
 
 P'ig. 83. Outline map of Indian Lake near Manistique, Schoolcraft County 282 
 
 Fig. 84. Topographic map of the Huron Mountains 288 
 
 Fig. 85. Outline map of Lake Michigamme, Marquette County 297 
 
 Fig. 86. Geological section across the Marquette Iron Range, .showing relation of Lake 
 
 Michigamme to the formation 297 
 
 Fig. 87. Sketch showing plan of the spits on the north shore of Lake Michigamme .... 300 
 
 Fig. 88. Cuspate foreland on the north shore of Lake Michigamme 301 
 
 Fig. 89. Outline map of Portage and Torch Lakes and vicinity, Houghton County .... 304 
 
 Fig. 90. Outline map of Lake Gogebic, Ontonagon and Gogebic counties 311 
 
 Fig. 91. Map of western part of the Northern Peninsula showing the general distribution 
 
 of the Copper and Gogebic Ranges and also the basin of the Ontonagon River 312 
 Fig. 92. Map to show the change in drainage which has taken place within the present 
 
 basin of the Ontonagon River 315 
 
 Fig. 93. Topographic map of Chicagon Lake and vicinity, Iron County 319 
 
THE INLAND LAKES OF MICHIGAN 
 
 I. D. SCOTT 
 
CHAPTER I 
 THE ORIGIN AND CLASSIFICATION OF LAKE BASINS 
 
 Lakes are numerous and the types of basins are many. If the 
 basins are classified according to their manner of formation, com- 
 plications may arise on account of several factors entering into 
 the formation of a single basin. As the study is extended all of the 
 agencies which are shaping the land surfaces of the earth are in- 
 volved and it is, therefore, necessary to become familiar with the 
 broader phases of the work and results of these agents. 
 
 To the beginner, one of the striking facts derived from the study 
 of earth science is that the surface of the earth is slowly but con- 
 stantly changing. Uplift and subsidence of the land are funda- 
 mental conceptions and no longer is the expression "terra firma" 
 strictly applicable. Elevation has lifted the continents higher 
 above the seas, while depression has served to deepen the ocean 
 basins. On the first land waves, currents, and the atmosphere be- 
 gan their work, and with further elevation other agents — running 
 water, ground water, winds, glaciers — became active. In general, 
 the work of these agents is to wear down the land and transport 
 the material elsewhere, eventually to the oceans. The continents 
 occupy only one-fourth of the surface of the earth and are low in. 
 average elevation compared to the depth of the oceans, therefore, 
 if elevation were ineflflcient or not active, they would soon be worn 
 down nearly to sea level. But the continents have stood for ages 
 far beyond the scope of human experience and, with the land as- 
 sured, our interest centers on the agents which are fashioning its 
 surface. The agents which are of most importance for our purposes 
 are the atmosphere, running water, wind, ground water, and 
 glaciers. 
 
 The Atmosphere. One class of work done by the atmosphere has 
 been given the descriptive term weathering. Under weathering is 
 included the action of such agencies as frost, temperature changes, 
 plants, animals, abrasion by the wind, and the chemical action of 
 the gases of the atmosphere, all of which tend to break up the solid 
 rock into smaller and smaller fragments. The comminution con- 
 tinues until the particles are small enough to be removed by the 
 
4 INLAND LAKES OF MICHIGAN 
 
 various ageuts of transijortatiou and, hence, may be considered a 
 process preparatory to transportation. 
 
 The lateral movement of the air, or wind, serves as a transporting 
 agent in addition to its action in the process of weathering. 
 In regions where the earthy material is loosely consolidated 
 and whose surface is unprotected by vegetation, e. g., deserts, sand 
 plains, and the shores of bodies of water, the wind is especially 
 active. The finest particles are picked up and often carried great 
 distances while the coarser sand grains are rolled along the surface, 
 collecting here and there in hills Avhich are called dunes. The ma- 
 terial of the dunes is clean sand, irregularly stratified, and the 
 slopes are gentle on the windward side but steep on the lee. Where- 
 ever the sand is widely distributed, as on the sand plains in our 
 State, the dunes tend to assume a crescentic form, but along the 
 shores of lakes the supply is local and the dunes are heaped in 
 wild confusion, with little regularity except that the slopes are 
 characteristic. The latter are well illustrated along the western 
 cost of the Southern Peninsula from Michigan City to the Straits, 
 and the crescentic type may be seen relieving the monotony of the 
 swampy plains of the eastern portion of the Northern Peninsula. 
 
 Running Water. Running water is one of the most important 
 agents at work on the surface of the land. Wherever rainfall is 
 sufficient the water collects in channels and flows onward, joining 
 other streams, until it reaches a trunk stream which carries it to 
 the sea. On account of their flow streams are able to pick up and 
 to transport the solid material supplied them by weathering. The 
 more swiftly flowing streams are able to carry larger particles 
 and greater amounts of all sizes. It is usually the case that active 
 streams are not supplied with enough disintegrated material to 
 tax their energy to the limit and some of this unexpended energy 
 is used by the suspended particles in filing, or abrading, the beds 
 of the streams. In addition, the solvent action of the water removes 
 material and the two processes working together deepen the stream 
 beds. Early in the formation of valleys the process of weathering 
 attacks the sides and reduces the slopes until tributaries develop 
 along them, repeating the process. Also both the main streams 
 and the tributaries tend to work headward and increase their length, 
 pushing their tentacles farther and farther into the land and tap- 
 ping the un drained areas. Eventually the headward extension is 
 halted by encountering streams flowing in the opposite direction, 
 forming divides and limiting the size of the basin. In this manner 
 sireams expand into great river systems which occupy definite 
 
ORIGIN AND CLASSIFICATION OF LAKE BASINS 5 
 
 basins, and the basins are dissected and lowered bj the constant 
 removal of material. 
 
 Probably the most important factor in determining the velocity of 
 a stream is the slope of its bed, and it is obvions that the slope, 
 and conseqnently the velocity, nuist gradnally decrease as the down- 
 ward cntting proceeds, since the month of the stream is fixed at 
 sea level. Eventnally tlie transporting power is taxed to its limit 
 and the stream can no longer cnt downward becanse all of its energy 
 is nsed in transporting snspended material and in friction. This 
 condition is reached first near its month and develops upstream, al- 
 though there may be local exceptions due to more easily eroded 
 rocks. 
 
 After the limit of downward cutting has been reached any further 
 reduction in velocity is accompanied by a deposition of some of 
 the load. The largest particles are dropped first and, if the decrease 
 in velocity continues, layers of increasingly fine material are added, 
 forming a deposit composed of layers whose constituent particles 
 are assorted in size and graded from coarse at the bottom to fine 
 at the top. The ideal condition is where a stream enters a body of 
 standing water, in which case the velocity begins to decrease at the 
 mouth and becomes zero at some point out in the lake. But the 
 velocity of streams varies at dift'erent times of the year, being great- 
 est at the spring floods, and enables the stream to transport coarser 
 material at this time. Thus, instead of a single layer becoming 
 finer in size of particles otT shore, there is formed a verticle series 
 of strata showing the assortment and gradation mentioned above. 
 This assortment and regular stratification are characteristic of de- 
 posits by running water. 
 
 It must not be assumed that the degrading Avork of a stream is 
 finished when the downward cutting ceases, for, at about this time, 
 the stream begins to sAving laterally, or meander, and develops a 
 flat on both sides which is flooded during high water and is called 
 a flood plain, or better, a valley flat. Also the valley sides are being 
 flattened by rain wash and other agencies until finally, after long 
 periods of time, the areas between the water courses have slopes 
 so flat that the material is not removed. At this time there is the 
 broad valley flat adjacent to the stream and on either side low 
 gently rolling plains stretch outward with almost imperceptible 
 slope toward the sea. Sncli a region is called a jJeneplaui (almost 
 a plain) and represents the cessation of erosion by running water. 
 
 Complete peneplanation is an ideal condition never realized as 
 far as we know because of interruptions of the process by uplift 
 and by the varying resistence of the rocks, some of Avhich stand in 
 
6 INLAND LAKES OF MICHIGAN 
 
 relief above the peneplain and are called nionadnocks. In case of 
 uplift of the land the power of the streams is revived and they 
 renew their attack on the land. It is interesting to note in this 
 connection that the peneplains that have been recognized up to the 
 present time have all been elevated above their normal position, but 
 erosion has not as yet obliterated their features. 
 
 Ground Water. In some regions underground water is a power- 
 ful eroding force, although generally not so effective as surface 
 streams. It is always present in the rocks and its source is rain. 
 Much of the rainfall sinks into the earth and percolates through 
 the interstices and fissures in the rocks until its downward passage 
 is interrupted, when it flows or seeps laterally, finally reaching the 
 surface again. It is interesting and important to note that water 
 is the greatest solvent known and its action is greatly increased 
 when it contains other substances in solution. Thus, limestone is 
 quite readily soluble in water containing carbon dioxide, one of 
 the atmospheric gases, in solution, and in this way funnels are 
 formed in the surface of the earth through which the water passes 
 underground. The water sooner or later assumes a lateral flow, 
 which is usually localized along the fissures and the beds of the 
 limestone, and dissolves definite channels for itself which are called 
 caves or caverns. As the process continues the interlacing chan- 
 nels enlarge and the roofs become weaker until finally they fall, 
 blocking the cave with rubble. 
 
 The surface effects are at first a number of depressions, known as 
 sink holes, which increase in number and extent, forming extremely 
 rough ravines with occasional remnants of the roof standing as 
 natural bridges. The sinks are often clogged with fine material 
 and become lakes. 
 
 Glaciers. Under this heading we wish to include only the work 
 of the great ice masses which spread over the land and replace 
 the variegated landscape with a cold, white, monotonous solitude, — 
 an absolute desert. These continental glaciers advance and retreat 
 over thousands of square miles of the land, grinding and plucking 
 the solid rock, incorporating and carrying forward the disintegrated 
 material, and depositing it near their borders as they melt. The 
 movement is outward from centers and is to some extent independ- 
 ent of the slope of the land. However, large depressions like those 
 occupied by the Great Lakes serve as channels along which the ice 
 movement is accelerated, forming great projections, or lobes, in the 
 ice front. See Fig. 1. Such an ice mass covered northeastern North 
 America in recent geological times and advanced and retreated over 
 
ORIGIN AND CLASSIFICATION OF LAKE BASINS 
 
 the area of the State of Michigan at least five times. The form 
 and distribntion of the material deposited by the ice during its last 
 retreat have determined the present land surface to a very great ex- 
 tent and, since the basins of the inland lakes almost without excep- 
 tion occur in glacial formations, some further consideration of the 
 work of o-laciers seems necessarv. 
 
 
 
 
 Fig. 1. 
 
 of the Wiscousiu glacier showing lobate character of the ice 
 front, (after Taylor and Leverett) 
 
 It is essential to remember that the ice moves forward constantly. 
 The forward movement continues until temperatures are encount- 
 ered which are warm enough to melt the ice effectively and this 
 determines the position of the margin of the ice. Thus, at the border 
 there are two factors active, the forward movement of the ice tend- 
 ing to advance the ice edge, and the melting which has the opposite 
 effect. Whenever the forward movement exceeds the melting, the 
 ice front advances and a continuance of this process causes an ex- 
 tension of the glacier. On the other hand, excessive melting causes 
 a retreat of the ice front and when both factors are equal it remains 
 stationary. The movement of the ice particles and the shifting of 
 the ice front should not be confused. 
 
 Glaciers of this type profoundly affect the land which they over- 
 ride, in places wearing away the rock and in others depositing great 
 
8 INLAND LAKES OF MICHIGAN 
 
 quantities of material which is so characteristic in constitution and 
 form as to be readily recognized. Ice itself has little or no power to 
 wear away the rock over which it passes but, by sinking into the 
 fissures which are universally present in rocks, it grasps the sep- 
 arated blocks and plucks them away in its forward movement. 
 Such blocks of rock when firmly frozen in the base of the glacier 
 become powerful abrading tools which grind away the solid rock 
 leaving smoothed, polished, striated, and grooved surfaces. 
 
 The various glacial deposits to which the general term drift has 
 been given, although differing greatly in form and material, have 
 one predominating characteristic, that of heterogeneity. By this is 
 meant that the material is composed of many different kinds of 
 rock. It is the direct result of the immense size of the glacier which 
 traverses great distances, encountering many different rock forma- 
 tions all of which contributed to its load. Some of the deposits are 
 laid down by the ice alone and these, although varying in form and 
 relief, are readily recognized by the character and disposition of 
 the material. In addition to its heterogeneous constitution, this 
 material, known as houlder clay or till, is of all sizes from the 
 finest "flour" to immense boulders, with no indication of assort- 
 ment or stratification as described for stream deposits. However, 
 stratified and assorted glacial deposits are common and these 
 indicate that the glacial material was worked over by running 
 water. Such deposits offer no diflQaulty of explanation when it is 
 realized that the melting of the ice furnished a great volume of 
 water which flowed away from the ice or was ponded in front of it. 
 
 The characteristics and manner of formation of the glacial de- 
 posits may be best understood by imagining the existence of a 
 glacier. Whenever the ice front remains stationary for a period of 
 time, the constantly forward moving ice with its load of earthy 
 material may be likened to a belt conveyor except that, instead of 
 returning empty, it melts. The earthy material, unevenly distrib- 
 uted in the ice, is carried forward and deposited in hummocks at 
 the margin. The resulting land form, known as a moraine, is a long, 
 curved ridge of till whose surface is composed of irregularly dis- 
 tributed knobs and basins. Its width is relatively narrow but its 
 length may be hundreds of miles. 
 
 At the same time the waters from the melting ice flow forward 
 carrying great quantities of material which is deposited either 
 among the moranic knobs in rounded hills of irregularly stratified 
 sand and gravel, called kames, or just in front of the moraine. In 
 the latter case the streams are often heavily clogged with drift and 
 
ORIGIN AND CLASSIFICATION OF LAKE BASINS 9 
 
 tend to braid rather than keep to definite channels. Under these 
 conditions broad plains are formed which slope gently away from 
 the ice and are composed of assorted and often stratified material. 
 They are linown as ontwash plains. 
 
 If the ice advances, the forms discussed above will be overridden 
 by the ice and obliterated or covered, but if the margin of the ice 
 retreats these forms will remain and, in addition, others which were 
 covered by the ice are revealed. Of these the ground moraine, or 
 till plain, is the most common. As the names signify, it is composed 
 of boulder clay and has some of the characteristics of the moraines. 
 Its surface has a knob and basin topography but the slopes are much 
 more gentle and the relief lower. The expression swell and sag is 
 commonlj^ used in describing these features. 
 
 Another topographic form bared by the ice is the drumlin. These 
 elliptical hills, composed of compact boulder clay, have a smooth, 
 rounded surface and, when vieAved from the side, resemble very 
 closely a plano-convex lens which is resting on the flat side. Their 
 length varies but is usually a mile or less and the relation of the 
 dimensions to each other will be clear from the statement that the 
 height may be measured in feet, the breadth in yards, and the 
 length In rods. An interesting relationship is that of their longer 
 axes which are apparently parallel for local areas but show a radi- 
 al distribution over larger tracts, indicating an alignment along 
 the direction of ice movement. The theories advanced for the 
 manner of their formation are diverse and need not be considered 
 here. 
 
 The last of the forms uncovered by the ice to be considered is 
 the esker, a low serpentine ridge rising above the till plain. It is 
 composed of imperfectly stratified sand and gravel and is usually a 
 few feet high, j^ards in width, but may extend for miles. It is 
 thought to have been formed by deposition by streams running in 
 definite channels underneath the ice. 
 
 The forms discussed above may all, with the exception of drum- 
 lins, be referred to a definite position of the ice front. When the 
 ice border is stationary a moraine is piled up, the strength de- 
 pending on the length of the halt and the amount of material in 
 the ice. At the same time the forms deposited by the water from 
 melting — ontwash plains, kames, eskers — are developed locally in 
 their respective positions in front, near-by, and back of the ice 
 margin, and may or may not be present in a given locality. Under- 
 neath is the till plain on which drumlins may be formed. The rela- 
 tive position of these forms is shown in Fig. 2. 
 
10 
 
 INLAND LAKES OF MICHIGAN 
 M 
 
 Fig. 2. Relative positions of glacial deposits, (after Tarr and Martin). 
 
 As stated above, most of the surface of the State is composed of 
 glacial deposits left by the great ice sheet during its final retreat 
 in recent geological time, and some idea of the nature of this re- 
 treat may be gained by a consideration of the distribution of the 
 deposits. The accompanying outline map, Fig. 3, shows the posi- 
 
 Mg. 3. Map of the principal moranic systems of Michigan. On this map are 
 shown the southern limits of the Wisconsin stage of glaciation ; 2, the Kalamazoo- 
 Mississinawa moranic system ; S, the Valpariso^Charlotte moranic system ; 4, Port" 
 Huron moranic system. 
 
ORIGIN AND CLASSIFICATION OF LAKE BASINS 11 
 
 tions of some of the stronger nioraiiies ol" the Michigan aud Huron- 
 Erie lobes, numbered in the order of their formation. Two things 
 are plainly evident, the lobation of the ice and the duplication of 
 the moraines in roughly parallel sequence. The lobation became 
 more pronounced as the ice retreated and the narrow interlobate 
 areas, areas which lie in the angle made by the junction of two 
 lobes, were regions of excessive accumulation due to the presence 
 of two ice margins in close proximity. 
 
 The duplication of the moraines, a few of which are shown on 
 the map, indicates a gigantic and thoroughly contested struggle be- 
 tween the forward movement and the warmer temperatures which 
 caused the melting of the ice, with the latter victorious. Thus, the 
 ice advanced overwhelming everything in its path until checked 
 by melting, when it entrenched by building a moraine. This posi- 
 tion was held until the margin was forced back to another stand 
 where the process of entrenching was repeated. Again and again 
 this occurred with occasional minor advances which served only to 
 prolong the struggle, and the ice retreated haltingly before the 
 onslaught of the weather. 
 
 Another effect of the recession of the ice was the ponding of 
 great bodies of water between the ice front and the divides. It is 
 readily seen that, once the divide had been uncovered, a flat trough- 
 like depression stood in front of the ice edge whose margins were 
 the divide on one side and the ice on the other. The filling of such 
 depressions with water gave rise to a series of lakes adjacent to 
 or filling the present basins of the Great Lakes. As the ice receded, 
 larger and larger depressions were uncovered and lower outlets 
 were found, forming a succession of lakes each of which, with some 
 exceptions, was larger but stood at a lower level than its predeces- 
 sor. The history of these lakes is complicated and has been fully 
 described in the publications of the Michigan Geological Survey 
 and elsewhere. Yet the history of many of the inland lakes is 
 closely connected with the two stages preceding the present Great 
 Lakes, and a brief description of these is added. 
 
 The earlier of these lakes is known as Lake Algonquin and in- 
 cluded all of the Great Lakes except Ontario, which was covered 
 by the waters of a lake called Iroquois. The relation of its out- 
 line to those of the present lakes is shown in Fig. 4. Its shores now 
 stand at elevations varying from 59G feet above sea level along the 
 southern borders of Lake Huron to 720 feet at the Garden Penin- 
 sula, Big Bay de Noc, Lake Michigan, and 940 feet in the vicinity 
 of Marquette, that is, above and at varying distances back from the 
 
12 
 
 INLAND LAKES OF MIOHIGA>^ 
 
 shores of the present lakes. The map shows that a relatively nar- 
 row strip along the Superior shore of the western part of the North- 
 ern Peninsula was covered by these waters and that only a small 
 portion of the eastern part was uncovered at this time, the land 
 areas being islands. In the Southern Peninsula a considerable 
 area in the northern part was covered by Ldke Algonquin and the 
 shore was very irregular consisting of many bays, promontories, and 
 islands. Farther south the areas covered were narrow strips of 
 land adjacent to the present shores of Lakes Huron and Michigan. 
 The stage immediately preceding the present is known as Lake 
 Nipissing which occupied the same basins as Lake Algonquin but 
 stood at a lower level and was somewhat smaller, in fact was but 
 little larger than the present lakes. Its outlines are shown in Fig 5. 
 
 Fig. 4. 
 
 Map of Lake Algonquin sliowing relation to present Great Lakes, (after 
 Taylor and Leverett). 
 
 On the borders of the Northern Peninsula the Nipissing beaches 
 stand at elevations from 10 to 60 feet above Lake Superior and are 
 usually found a short distance inland. In the Southern Peninsula 
 the beaches drop in level and the areas covered by this lake be- 
 come smaller to the south. As in the case of Algonquin time, the 
 northwestern part of this peninsula was an archipelago, and many 
 of the inland lakes in this region have the shore lines and terraces 
 of Lake Mpissing standing above their present shores. 
 
 It is hoped that the brief statement of the work of these agents 
 given above will enable the reader to comprehend more easily the 
 
ORIGIN AND CLASSIFICATION OF LAKE BASINS 
 
 n 
 
 forces at work on the land, and also aid in nnderstandin<>; the tech- 
 nical terms that are necessary in a report of this natnre. 
 
 niYSIOGRAPHY OF THE STATE OF MICHIGAN 
 
 The State of Michigan is divided natnrally into two distinct 
 parts which have been named the Southern and Northern Peninsu- 
 las on account of their positions with reference to Lakes Michigan, 
 
 Fig. 5. Map of Lake Nipissing, showing relation to present Great Lakes, (after 
 Taylor and Leverett) 
 
 Superior, Huron, and Erie. Thus, the Northern Peninsula is 
 bounded to a large extent by Lake Superior, St. Mary's River, and 
 Lake Michigan, and the eastern part is truly a peninsula. See 
 Fig. 6. Construing the term peninsula rather looselj^, the Menom- 
 inee and Montreal rivers may be included with the lakes, making 
 the land connection less than 70 miles in width for the entire pen- 
 insula. 
 
 The Southern Peninsula, surrounded on all sides except the 
 southern by the waters of Lakes Michigan, Huron, Erie and their 
 connections, is a broad peninsula which bears a resemblance to a 
 great hand with the thumb just east of Saginaw Bay. See Fig. 7. 
 
 Northern Peninsula. The Northern Peninsula, Fig. 6, is a 
 rather narrow strip of land about 330 miles in length and has an 
 average width that is estimated at less than 50 miles. Its outline 
 
14 
 
 INLAND LAKES OF MICHIGAN 
 
 Fig. 6. Map showing physiographic provinces of the Northern Peninsula. 
 
ORICIN AND CLASSIFICATION OF LAKE BASINS 15 
 
 is iiTegiilar, having two proiiiinent projections — Keweenaw penin- 
 sula on tlie north and the Menominee district on the south — and 
 numerous smaller points and bays as well. 
 
 The altitude ranges from 580 feet above tide at the shores of 
 Lake Michigan to more than 2,000 feet at the Porcupine Mountains 
 in the northwestern part of the Peninsula. On the basis of eleva- 
 tion and underlying rocks, the area may be divided into two definite 
 provinces : One, the Highlands, which is underlain by rocks largely 
 of Pre-Paleozoic* age, lies west of a north-south line passing 
 through Marquette, and the other, the Lowlands, extends eastward 
 from this line and is underlain by Paleozoic* rocks. Glacial drift 
 has been deposited over almost all of the Peninsula, the exceptions 
 being more numerous in the Highland province. 
 
 Highland Province. This province extends from the meridian of 
 Marquette westward and southward beyond the boundaries of the 
 State, and stands at an average elevation of 1600-1800 feet above 
 sea level, or 1000-1200 feet above Lake Superior. The region is a 
 table-land which rises rapidly from the Lowlands on the east and 
 north and its surface is covered with a variable thickness of glacial 
 drift through which a relatively small number of hard rock knobs 
 projects. The relief is moderate with differences in elevation which 
 probably do not exceed 500 feet locally and are 100-300 feet on the 
 average. The greatest elevations are rock hills which reach a max- 
 imum height of 2023 feet in the Porcupine Mountains. 
 
 The Highlands are a part of a great uplifted peneplain which 
 was formed in ancient times. The underlying rocks are mainly of 
 Pre-Paleozoic age and consist of crystalline masses and banded 
 rocks the distribution of which greatly influenced the action of 
 streams, causing, thereby, characteristic topographic forms. The 
 erosion was profound and interrupted by several uplifts but, 
 throughout the vast interval of time during which peneplanation 
 was accomplished, the crystalline masses resisted erosion and stood 
 in relief above the peneplain as monadnocks. The banded rocks 
 were tipped on edge and presented alternately weak and resistant 
 layers to the action of the streams. The hard layers resisted ero- 
 sion and stood as mouadnock ridges whose longer axes are roughly 
 parallel to Lake Superior, while the softer layers were bevelled by 
 the surface of the peneplain. The ridges were not continuous but 
 
 ♦Cenozoic includes Present The geological time scale, the main divisions of 
 
 and Glacial times which are given, is tabulated to give the effect 
 Mesozoic of a great column of superposed rocks, the oldest 
 
 Paleozoic at the bottom and the youngest aboTe in the order 
 
 Proterozoic 1 p T>-\Tfyr,rn\n ^^ formation. The position in the scale gives the 
 
 Archeozoic j ^ ^« ^ '^^euiiuii, relative age. It will be seen that the rocks under 
 
 discussion and of all of Michigan as well, except 
 the glacial deposits, stand low in the scale and are, 
 therefore, ancient. 
 
ae ■ : ■ INLAND LAKES OF MICHIGAN 
 
 were crossed by streams, formiug gaps which served as channels 
 for the advancing ice of glacial times. Upon the peneplain and 
 about the monadnocks were deposited paleozoic sediments which 
 were largely removed by erosion before the advent of the ice. 
 
 The effect of the ice action was to scour, out the gaps, round off the 
 hills, and fill the valleys with heavy deposits of drift obscuring most 
 of the former surface. Thus, we find today an area covered for the 
 -most part with glacial material through which rock knobs, either 
 rounded or linear in form, project. The more important regions 
 where such knObs are found are the Porcupine Mountains, Gogebic 
 Eahge, Keweenaw Range, Huron Mountains, Marquette Range, Iron 
 River district, and the Menominee district. Of these, Sheridan Hill 
 in the Iron River district and some limestone capped hills in the 
 Menominee district, are composed of Paleozoic rocks. 
 
 The glacial deposits of this province were laid down for the most 
 part during the last recession of the ice by the lobes that extended 
 into Lake Superior and Michigan with their subsidiary lobes, 
 Keweenaw and Green Bay. The part first uncovered by the ice is 
 in southern Iron County, an area of till deposits with drumlins 
 or drumliu-like hills. A great moraine swings around this area, 
 formed on the north by the ice of the Superior lobe and on the east 
 by that of the Michigan lobe. The succeeding moraines show the 
 same directions. Thus, the moraines, the inter-moranic till plains, 
 and outwash run roughly parallel to Lake Superior in the western 
 X)art of this province, except where influenced by the Keweenaw 
 lobe. They have a nearly north-south trend in the ea^^teru part 
 Which was covered by the Michigan lobe. ' 
 
 The Highlands are drained to Lakes Superior and Michigan with 
 the exception of a small area in southern Gogebic County which is 
 tapped by the Wisconsin river. The drainage is controlled both 
 by the glacial formations and the pre-glacial topography. In the 
 western part a strong moraine forms the divide, and in the vicinity 
 of Watersmeet are situated the headwaters of the Wisconsin, On- 
 tonagon, and Menominee rivers. To the east the divide shifts north- 
 ward to a watershed north of Michigamme composed of thinly 
 dirift-covered crystalline rocks from which streams flow in all di- 
 rections. The drainage is incomplete and small lakes and swamps 
 are: abundant, especially in the morainic districts and the thinly 
 drift covered area north of Michigamme. There are several lakes 
 of considerable size in this province whose basins are of excep- 
 tional interest and will be discussed later. 
 
 Lowland Province. The Lowlands extend from the meridian of 
 Slarquette eastward to the Sault Ste. Marie and swlns: to 
 
ORIGIN AND CLASSIFICATION OF LAKE BASINS 17 
 
 the southwest iuto Wiscousin and Miiniesota in a broad semi- 
 circle. The greatest extension of the Lowlands along the northern 
 edge of the Highland region is found in the continuation of Ke- 
 weenaw Bay to the southwest and this is connected to the main 
 Lowland area to the east by a narrow coastal strip. 
 
 This region is on the average more than lOOO feet lower than the 
 Highlands and its general elevation does not exceed 250 feet above 
 the Great Lakes, although in places it rises considerably above 
 this. It is underlain by Paleozoic rocks which slope gently in a 
 southerly direction. The bevelling of these rocks by stream action 
 in pre-glacial times gave rise to a plain arranged in belts which 
 mark the surface exposures of the various layers of dipping rocks. 
 These belts run roughly parallel to the curve of the north shores 
 of Lakes Huron and Michigan. 
 
 The hardness of the different layers varies, and the softer were 
 worn into broad valleys whereas the harder stood in relief in forms 
 peculiar to this tj'pe of structure. They consist of low, linear 
 ridges which slope gently on the side formed by the surface of the 
 rock layers but are more abrupt on the side which cuts across the 
 layers. Such forms are known as cuestas. In this region they 
 have almost imperceptible southerly slopes but stand usually about 
 100 feet above the plain to the north, although in places bluff-like 
 escarpments with altitudes of 200.300 feet are found, for example, 
 Burnt Bluff, Big Bay de Noc. Two cuestas are present in this 
 province, one near the south shore of Lake Superior in the vicinity 
 of Munising which swings to the south in its eastward extension, 
 and another just north of the Michigan and Huron shores. Both 
 are largely obscured by glacial deposits but the southern is the 
 better developed. It begins with the Garden Peninsula at Big Bay 
 de Noc and continues across the Lowlands to Drummond Island 
 and eastward across Lake Huron as a great series of islands which 
 partially isolate the North Channel and Georgian Bay from Lake 
 Huron. 
 
 The recession of the ice in this region was from south to north 
 and there is in general an east-west trend to the deposits. As 
 shown in Fig. 3, a low moraine runs the length of this province. A 
 large morainic tract is also found at the Junction of Luce, School- 
 craft, and Mackinac counties and another important topographic 
 feature, aside from the thinly drift covered southern cuesta, is 
 the great swampy plains. The two most important are those which 
 form the major part of the drainage basins of the Manistique and 
 Tahquamenon rivers. These sandy plains have veiy gentle slopes 
 and are almost featureless, the greatest relief being the small, 
 3 ^ 
 
18 INLAND LAKES OF MICHIGAN 
 
 but frequent sand dunes each crested with a clump of pines which 
 accentuate the grass covered plain. These plains are the result of 
 stream deposition and may be referred in part to the waters escap- 
 ing from the ice border. 
 
 The drainage of the Lowlands is very, poor and a large percentage 
 of the area is covered by swamps and lakes. In the western part 
 the divide is far to the north, giving a drainage area for the Man- 
 istique river of 1,400 square miles. To the east the divide swings 
 southward and separates extensive basins on either side which 
 are drained by the Tahquamenon and Carp rivers. The portion of 
 the Peninsula east of St. Ignace and White Fish Bay is drained 
 mainly into the St. Mary's River and Lake Huron. The inland 
 lakes are found among the moraines, on the outwash plains, and 
 along the shores of the Great Lakes, the largest being situated 
 in the low morainic tract south of Seney and McMillan, and along 
 the shores of Lake Michigan near Manistique and St. Ignace. 
 
 Southern Peninsula. The altitude of the Southern Peninsula, 
 Fig. 7, is, in general, much lower than that of the Highlands of the 
 Northern Peninsula and corresponds more nearly with that of the 
 Lowland province. The highest points are found in Osceola county 
 where elevations in excess of 1700 feet have been noted. The lowest 
 altitudes are obviously determined by the level of Lake Erie, 572 
 feet, in the southeastern part of the peninsula. Ninetj^-six per 
 cent of the area stands below 1200 feet, and probably the average 
 elevation is not greatly in excess of 800 feet above the sea. The 
 highest area is situated in the northern half of the peninsula, em- 
 bracing about 1500 square miles largely within Osceola, Wexford, 
 Missaukee, Crawford, and Otsego counties, and exceeds 1200 feet 
 in elevation. -Other elevated areas are found in the southern part, 
 chiefly in Hillsdale county. 
 
 The Southern Peninsula has been divided into several physio- 
 graphic provinces which have a diagonal trend in a northeast- 
 southwesterly direction. See Fig. 7. Beginning in the southeastern 
 part is a low plain, the Erie Lowland, bordering Lake Erie and 
 to the northwest is the Thumb Upland, extending in a northeasterly 
 direction from Hillsdale county to the Thumb. This gives way to 
 the Saginaw Lowland, to the north of which lies the Northern Up- 
 land. The Michigan Lowland follows the Lake Michigan shore 
 and is an exception to the diagonal trend. 
 
 The underlying rocks are of Paleozoic age and are closely asso- 
 ciated with those of the Lowlands of the Northern Peninsula. 
 Structually the rocks dip towards the center of the State and form 
 a shallow basin which has been bevelled by erosion, distributing the 
 
ORIGIN AND CLASSIFICATION OF LAKE BASINS 
 
 19 
 
 edges of the various I'oniia lions in conccnti'ic ovals, that is, one 
 within the other. The formations are of different resistances and 
 the sandstone in particular yielded slowly to the erosive agents. 
 Tlie o-eneral relief features of this area before the advent of the 
 
 Fig. 7. Map showing physiographic provinces of the Southern Peuinsu'.a, (after 
 Leverett). Note: Erie Lowland not designated. 
 
 glacier bore a striking resemblance to the present in that the 
 larger features were similar in kind and distribution. Thus, there 
 were upland areas in the northern and the southern parts with low- 
 lands between and on either side. The low area in the northwestern 
 
20 INLAND LAKES OF MICHIGAN 
 
 part between Ludington and Frankfort would now be below sea 
 level if the glacial deposits were removed. 
 
 In contrast to the Northern Peninsula, the glacial deposits of 
 the Southern are largely those of an ice invasion just previous to 
 the Wisconsin, or last stage, and form not only the main filling of 
 the valleys but in places prominent ridges. The lobation of the 
 ice of the earlier glaciation was similar to that of the Wisconsin 
 stage and the same areas were, in general, regions of accumula- 
 tion. It is of interest to one engaged in the study of lakes to 
 know' that the present surface features are largely due to the 
 veneer of drift left by the last ice recession. 
 
 During its maximum extension the ice of the Wisconsin stage 
 covered the entire Southern Peninsula and its recession uncovered 
 first an area in the southern part of the State. See Fig. 3. In- 
 spection of the moraine numbered 2 on the map shows this area 
 to have been interlobate between the Michigan and the Huron-Erie 
 lobes. With further recession two narrow interlobate areas were 
 formed on either side of the small Saginaw Lobe, as shown by 
 moraines 3 and 4 on the map. Thus, there were areas of excessive 
 accumulation in the southeastern part, more or less coincident with 
 the Thumb Upland, and in the northern part, coincident with the 
 Northern Upland. The latter was an area of especially great accu- 
 mulation, and here are found extensive moraines, till plains, and 
 broad out\^^ash aprons, the latter constituting the great sand plains. 
 
 The apparent coincidence of the pre-glacial physiographic pro- 
 vinces of this peninsula with those of the present time would lead 
 one to infer that the pre-glacial topography controls the present 
 relief but such is not the case, since the corresponding areas do not 
 actually coincide except possibly in the Thumb Upland. It is prob- 
 able, however, that the main influence was the indirect one of de- 
 termining the positions of the ice lobes and thus the moraines, and 
 that the present topography is due, for the most part, to the dis- 
 tribution of the drift, and near the shores of the large lakes to the 
 working over of this material by the waters of Lakes Algonquin and 
 Mpissing. However, it may be that some of the topography in 
 Hillsdale County and in the region northwest of Thunder Bay is 
 referable to the underlying rocks. Stream action since glacial 
 times has modified the surface so slightly that it is negligible. 
 
 The drainage of the Southern Peninsula is determined in its 
 larger aspects by the physiographic provinces discussed above, and 
 the sources of the main streams are found in the high interior por- 
 tions. In the northern part the divide is situated near the center 
 of the Northern Upland and the streams flow outward to Lakes 
 
ORIGIN AND CLASSIFICATION OF LAKE BASINS 21 
 
 Michigan and Hnron. In this locality the headwaters of the Mus- 
 kegon, Manistee, and An Sable rivers are in close proximity. Many 
 of the smaller streams, however, have their sources on the slopes 
 of the ujjiand province and flow more or less directly into the 
 lakes. 
 
 In the southern part the long Thumb Upland forms a veritable 
 v.'atershed and on this are located the sources of the St. Joseph, 
 K:^Iama;;oO; Saginaw (south branches), Huron, and Raisin rivers. 
 As in the northern section, the minor streams head on the slopes 
 of til is })rovince and flow directly to the lakes. Between the two 
 upland areas lies the Saginaw Lowland which is drained by the 
 Grand and Saginaw Rivers. 
 
 Notable and peculiar drainage patterns are shown by some of 
 the streams, especially the Saginaw and Black (Thumb region), 
 but these are due to the distribution of glacial material, more 
 especially the moraines and the uplifted beaches of the predeces- 
 sors of the Great Lakes. In Alpena and Presque Isle counties 
 are many sink holes, and the surface drainage is interrupted by 
 these in some cases. From the distribution of the sinks and the 
 presence of "fountains" in parts of Thunder Bay, it has been 
 inferred that there is an extensive underground drainage system, 
 reaching from this region to the vicinity of Black Lake in Cheboy- 
 gan County, but this has not been carefully worked out as yet. 
 
 The abundance of lakes in the Southern Peninsula is an indica- 
 tion of the incompleteness of the drainage, and it is noteworthy 
 that many of the larger as well as the smaller lakes are to be found 
 in the morainic districts, for example, in the northern and southern 
 interlobate areas and in the morainic region within Calhoun, Barry, 
 and Kent counties. Aside from the lakes due to the irregular dis- 
 tribution of the glacial material, there are those bordering Lakes 
 Michigan and Huron which are more important in the extreme 
 northern part of the peninsula. 
 
22 
 
 INLAND LAKES OF MICHIGAN 
 
 ORIGIN OF LAKE BASINS, 
 
 A physiographic study of lakes has as its starting point the origin 
 of the basins, and studies in the past ^lave resulted in a classification 
 according to manner of formation which includes many types. 
 The list given, although not complete, will serve to illustrate the 
 diversity of types. 
 
 Diastrophism (movements of the 
 earth's crust) 
 Slow movements 
 Rapid movements (faulting) 
 
 Vulcanism 
 
 Gradation 
 Rivers 
 
 Waves and currents 
 Wind 
 Glaciers 
 Mountain 
 
 Continental 
 
 Ground water 
 Gravity 
 
 New-land lakes 
 Ponded lakes 
 Basin range lakes 
 Rift-valley lakes 
 Earthquake lakes 
 
 CouUee Lakes 
 Crater lakes 
 
 Ox-bow lakes 
 Alluvial dam lakes 
 Saucer lakes 
 Crescentic levee lakes 
 Raft lakes 
 Delta lakes 
 Side-delta lakes 
 
 Lagoons 
 Dune lakes 
 
 Rock basin lakes 
 Valley moraine lakes 
 Border lakes 
 Morainal lakes 
 
 Marginal 
 
 Ground 
 Morainal dam lakes 
 Inter-morainic lakes 
 Pit lakes 
 Glint lakes 
 Ice dam lakes 
 Glacial lobe lakes 
 Glacial scour lakes 
 Fosse lakes 
 Sink lakes • 
 Karst lakes 
 Landslide lakes 
 
ORIGIN AND CLASSIFICATION OF LAKE BASINS 23 
 
 It is seen from the list that diastropliism and the gradational 
 processes of rivers and glaciers produce the greater number of types 
 of basins, but of tliese, glaciers are productive of the greater num- 
 ber of examples, and the lakes thus formed are of greater import- 
 ance. Although it is comparatively simple in most cases to assign 
 the general cause of origin, there are many cases where it is difiScult 
 to determine the specific cause, inasmuch as several factors, each 
 of which may be sufficient to form a lake basin, have been active. 
 For example, basins due to deposition by continental glaciers are 
 easily recognized but it is often a perplexing study to determine 
 which of the various deposits plays the most important part. In 
 cases where several factors enter the most important must be de- 
 cided upon and the lake classified accordingly. Within the State 
 of Michigan lake basins of the following types have been recognized : 
 
 Graciers 
 Continental Glacial scour 
 
 Morainal 
 Morainal dam 
 Pit 
 
 Inter-morainic 
 Fosse 
 
 Waves and currents Lagoons 
 
 Diastrophism Eift-valley 
 
 Ponded 
 
 Ground water Sinks 
 
 Rivers Ox-bow 
 
 It is clear from this list that a considerable number of types 
 have been recognized but by far the greater number of examples is 
 due either wholly or in part to glacial action. In fact, there are 
 but few that can be referred to other causes but they are interest- 
 ing in that they are exceptional, for this State at least, and their 
 characteristics as well as those of glacial origin will be given. 
 
 Rivers. Ox-bows. Lakes of this class are found in the vicinity of 
 streams which have reached the limit of downward cutting and 
 are meandering. In streams of this age the adjustment between the 
 carrying power and the material in transport is so close that even 
 a slight decrease in velocity will cause deposition of some of the 
 suspended material and, on the other hand, any increase will cause 
 removal. The current of a stream that is meandering on a valley 
 flat is increased on the outside (convex) side of the bend and cor- 
 
24 
 
 INLAND LAKES OF MICHIGAN 
 
 A 
 
 Fig. S. Map and diagrams showing characteristics of oxbow lakes. Map a part. 
 of the Ypsilanti quadrangle. 
 
ORIGIN AND CLASSIFICATION OF LAKE BASINS 
 
 25 
 
 respoudiugl}' deceased on the iuside (concave). Therefore cutting 
 takes place on the outside and deposition on the inside and at the 
 same rate. This causes the meander to increase in swing, the width 
 of the stream remaining constant, and also deejjens the channel 
 next the outside bank. In addition to the increase in width, meand- 
 ers tend to work downstream, due to the fact that the stream is 
 running down a slope. The neck of the meander is gradually worn 
 away and the stream eventually straightens its course. The ends 
 of the abandoned channel are soon filled by deposition and a lake 
 formed. The characteristics of such lakes are: Its position on a 
 valley flat composed of alluvium, crescentic shape, and greater 
 depth near the convex side. See Fig. 8. An excellent example of 
 an oxbow lake is that shown on the map in Fig. 8, and lies on the 
 valley flat of the Huron River (Washtenaw County) about three 
 miles below Ypsilanti. 
 
 Ground Water. Sinks. These lakes occur in regions underlain 
 by limestone which is readily soluble in water containing carbonic 
 acid gas in solution. The water with its dissolved gas seeps through 
 the fissures and bedding planes of the rock and carries away much 
 material in solution. The underground openings thus formed 
 sometimes reach large dimensions and are called caves or caverns. 
 The continuation of the process finally causes the roof of the cave to 
 fall, forming a depression on the surface which may become filled 
 wath water. Such basins may be irregular in shape but are gen- 
 erally somewhat circular in outline and have no surface outlet, al- 
 though they may have inlets. The characteristics are shown in 
 Fig. 9. 
 
 Fig. 9. 
 
 Map and diagram to sliow manner of formation of lalies in sink-holes, 
 (after Hobbs). 
 
26 
 
 INI.AND LAKES OF MICHIGAN 
 
 Lakes of this type are rare in this State and are confined to the 
 Southern Peninsula as far as now known. Ottawa Lake in the 
 southeastern part is ascribed by Lane to this cause. Numerous 
 sinks occur in Alpena and Presque Isle counties but no lakes oc- 
 cupying true sink holes are known. Sunken Lake, a few miles 
 south of Metz in Presque Isle County, is closely related to this type 
 of lake and was visited by the writer. It has little interest except 
 as a type of basin and is now permanently dry. It appears to be a 
 stream channel about one fourth of a mile long which is cut off 
 from the Upper South, a branch of Thunder Kiver, by a dam at its 
 upper end. It ends abuptly at the lower end in a sink which stands 
 above the stream bed and ponded the water w^hich formerly entered 
 this basin at high water stage. A view of this lake bed in shown 
 in Plate I. 
 
 DiASTROPHisM. Faulting. Lake basins due to faulting are ex- 
 ceptional in Michigan and in no case known to the writer can the 
 actual faults be detected. Canyon lake in the Huron Mountain 
 group has characteristics which point so clearly to this origin that 
 it is included in this class. This lake is too small to appear on 
 the map, Fig. 84, but in reality is a most interesting and fascinating 
 body of water. It is scarcely one-fourth mile in length and does 
 not exceed one hundred feet in width. Nevertheless it possesses a 
 charm which lies not in its size but rather in its picturesque loca- 
 tion and surroundings. It lies in a narrow canyon of almost uni- 
 form width which cuts directly across a hard rock saddle. The 
 shores at the ends are low and swampy but the sides are cliffs which 
 
 '7:;;^reej.-^-i^-^^,^^ 
 
 .-^>''mvIv_ 
 
 
 Fig. 10. Diagram and map showing characteristics of rift valley lalies, (After Hobhs). 
 
 increase in height from either end to a maximum in the center. The 
 cliffs, while very steep, are not perpendicular but ascend in steps 
 consisting of high risers and narrow treads upon Avhich stand trees 
 and shrubs in precarious positions. The outline of the sides is a 
 somewhat regular zig-zag rather than a straight line, showing the 
 presence of fractures. In addition, the depth of the lake is rela- 
 tively great and the cliffs descend perpendicularly into the water. 
 
ORIGIN AND CLASSIFICATION OF LAKE BASINS 
 
 27 
 
 Such characteristics lead one to the conclusion that this basin 
 was formed by the dropping of a portion of the earth's crust, not 
 as one block but several long, narrow blocks, the amount of drop 
 or displacement being progressively greater towards the center 
 of the depression. 
 
 The general characteristics of such lakes are shown in Fig. 10 
 and may be enumerated as follows : Long and narrow, of great 
 depth, and bounded on the sides hj rock cliffs. 
 
 Warping. Along the western coast of the Southern Peninsula 
 south of Frankfort many of the streams broaden on approaching 
 Lake Michigan, and the expansion is suflficieut to warrant their 
 being classed as lakes, since they have been separated from the 
 main lake by the development of bars. Their sej)arate existence is 
 due to the work of waves and currents but the expansion of the 
 river mouths — the lake basins — :is due to a warping of the earth's 
 crust. The warping consists in an uplift of the land to the north- 
 east which is raising the outlets of the Grreat Lakes and, in the 
 
 Fig. 11. Diagrams sho«-ing characteristics of lagoons of the drowneil stream type. 
 
 case of Lake Michigan, is causing the water to encroach on the 
 land in the region referred to. This is flooding or drowning the 
 river mouths arid, in conjunction with shore action which forms the 
 bars across their mouths, forming lakes of moderate depth whose 
 beds slope gently to a channel formerly occupied by the stream. 
 The shape is usually irregular due to the flooding of main stream 
 and tributaries. In case a single stream is flooded the shape is 
 roughly triangular with the upstream point truncated by a delta- 
 
28 INLAND LAKES OF MICHIGAN 
 
 like deiDosit made by the entering stream. Fig. 11 illustrates these 
 characteristics. 
 
 Glacial Lakes. Scour. In some localities the scour of the ice is 
 responsible to a large extent for depressions, although other factors 
 may be important. The scour, was a planing action accomplished by 
 rocks held in the base of the ice and was localized according to the 
 topography, structure, and hardness of the rocks over which the 
 ice flowed. Topographic depressions, such as stream valleys run- 
 ning in the same direction as the ice, were deepened by an increase 
 in flow and a greater thickness of ice. Gaps in ridges running 
 transverse to the direction of ice movement were both widened and 
 deepened because of increased flow somwhat analogous to the 
 greater velocity in river narrows. Once through a gap, the ice 
 spread and in some cases flowed along the transverse valley on the 
 far side of the ridge, deepening it locally. It is probable also that 
 the ice in these depressions was the last to melt and outwash was 
 deposited at the ends which accentuated the basin formed by scour. 
 Such lakes are usually long and narrow and the evidence of glacial 
 scour is the presence of striations on the exposed rock surfaces 
 running parallel to the length of the lake. Another structure 
 which facilitates glacial wear is an abundance of fissures in the 
 rocks. Rocks which are easily abraded also were more rapidly worn 
 down by the ice, but regions underlain by such rocks were the 
 locations of stream courses previous to glacial times and the ice 
 merely increased the pre-existing relief. 
 
 Deposition. Morainal. This term is reserved for the small basins 
 caused by irregularities in the surface of marginal and ground 
 moraines. In marginal moraines the surface is characteristically 
 hummocky and the knobs and basins vary in size and regularity. 
 The size and shape of the lakes occupying moranic depressions 
 are determined not alone by the nature of the individual basins but 
 by the amount of water draining into them. If sufficient water is 
 available to cover several adjoining basins, a lake is formed which 
 is irregular in outline and usually of moderate size. In case a 
 single basin is flooded, the lake conforms to this basin and is often 
 oval in outline and small in extent. The smallest of these are 
 mere ponds without outlets and few, if any, inlets, and are rapidly 
 being filled with vegetation. The depth of such lakes is not great 
 and corresponds roughly with the relief of the surrounding country. 
 The shores are frequently strewn with boulders and cobbles and the 
 bottom is composed of a clay mud, where not covered with peat. 
 
 In the ground moraines the relief is less pronounced and the lakes 
 are usually very shallow, although they may be of considerable 
 
ORIGIN AND CLASSIFICATION OP LAKE BASINS 
 
 29 
 
 ■ *''' '•-' <f : v.: ;.'.>;i?/y3*jjrV-. ;'';.•'; U-'*"- ■ 
 
 ^c^/^ 
 
 \A///es 
 
 Fig. 12. Diagrams and map showing characteristics of morainal (marginal) lakes, 
 (Diagrams after Ilobbs. Map a part of the Three Rivers, Mich., Quadrangle.) 
 
30 
 
 INLAND LAKES OF MICHIGAN 
 
 extent. Houghton Lake in Roscommon County, one of the largest in- 
 land lakes in the State, lies mainly on ground moraine and the 
 depth does not exceed 25 feet. The characteristics of morainal 
 lakes are shown in Fig. 12, 
 
 Inter-Morainal. During the recession of the glacier it sometimes 
 happened that the halts of the ice border were close together, 
 leaving parallel morainic ridges with a narrow depression 'between, 
 composed of ground moraine or outwash. Often such depressions 
 are below the general drainage level and, if the ends are blocked, 
 become the sites of lakes of considerable extent. Such lakes are 
 elongate in the direction of the flanking moraines whose slopes 
 have the characteristic knob and basin topography, and the ends 
 are frequently blocked with outwash. The presence of outwash may 
 be explained by assuming a block of ice which filled the depression 
 and from whose sides outwash developed in either- direction; or 
 that the outwash aprons developed locally at various places along 
 
 Fig. 13. 
 
 Lake Antoine, a lake of the morainal dam type, (fropi the Menominee 
 Special Quadrangle). 
 
ORIGIN AND CLASSIFICATION OF LAKE BASINS 31 
 
 the ico front but did not coalesce. This type of basin is very simi- 
 lar to tluit of morainai lakes situated on ground moraine and per- 
 haps should be considered as a special case rather than a distinct 
 type. 
 
 MoRANiAL Dam. Although the general statement that continental 
 ice masses sucli as covered northeastern North America assume a 
 form that is largely independent of the relief of the land is true, 
 yet it is a significant fact that the ice border in its details was 
 extremely sensitive to the topography over which is flowed. At the 
 ice border hills served to' divert the flow, and depressions became 
 locations of increased movement especially when their direction 
 was parallel to that of the ice flow. In this way minor tongues 
 were formed in valleys which deposited a series of moranic dams 
 as they receded and above these dams lakes now lie. The valleys 
 must have been present before the final retreat of the ice and may 
 have been pre-giacial or inter-glacial. Lakes of this type are often; 
 elongated in the direction of the valley but may be circular or- 
 even run transverse to the valley if the dams are close together or- 
 low in elevation above the valley floor. They occur frequently in. 
 series in the same valley. 
 
 In the Iron Kiver district of the Northern Peninsula there are 
 some lakes which are held by drift dams thrown across the valleys 
 between drumlinoidal hills, for example, Fortune, Chieagon, Stan- 
 ley, while in the vicinity of Iron Mountain an irregular moraine 
 across a pre-glacial valley forms the west border of Lake Antoiue, 
 see Fig. 13. 
 
 Fosse. Fosse, as used in this connection, refers to a long, narrow 
 depression that is sometimes found between a moraine and an out- 
 wash plain. It is a remnant of ground moraine upon which the 
 ice stood when the out wash was being formed. The outwash was 
 built up along the steep ice front partially burying it, and when 
 the ice retreated part of the material at the inner edge of the out- 
 wash fell back on the ground moraine, forming a very steep slope. 
 A short distance back the moraine was piled up, leaving a depres- 
 sion, as shown in Fig. 14. 
 
 Such lakes may be distinguished by the attenuated form, the 
 presence of moraine on one side and outwash with steep edge on the 
 other, and the absence of outlet or inlets of importance. The water 
 of the lake seeps readily through the sand and gravel of the out- 
 wash in lieu of an outlet. Inlets may develop on the side slopes of 
 the moraine but the more important will run in the unoccupied 
 portion of the fosse. Inasmuch as the fosse is a local development^ 
 the inlets must necessarily be of little impoi-tance and the lake 
 
32 
 
 INLAND LAKES OF MICHIGAN. 
 
 fed principally by ground water. An excellent example of this 
 type of lake is Crooked lake, situated in the group of lakes a few 
 miles west of Chelsea, Washtenaw County, 
 
 Pits. The term pit, as here used, signifies a depression in an out- 
 wash plain. It was probably formed by the isolation of an ice 
 iblock which became covered with debris and melted later, allowing 
 
 *^^^^'//m////« 
 
 fui^^uvu^* 
 
 OaT^^^^^ 
 
 Outi^aeh P/a/yi 
 
 Sancf and. G^a\yeJ -=■"• 
 o 
 
 Ouryv<7eh P/o/n 
 
 > 77//; ° \:r^ roS3& 
 
 -'o o o ^ 
 
 Fig. 14. Sketch and diagrams to illustratf manner of formation of fosse lakes, 
 (diagrams after Hobbs). 
 
 the material above to settle. The important thing in the formation 
 of these depressions is the protective effect of a coating of earthy 
 material on the ice. A small rock fragment on the surface of the 
 ice absorbs enough heat from the sun's rays to become heated 
 through and melts a depression for itself in the ice. Larger frag- 
 ments or an accumulation of small ones are not able to conduct 
 the heat to the under side and, therefore, protect the ice, and the 
 greater the thickness of the earthy material the slower the melting. 
 
ORIGIN AND CLASSIFICATION OF LAKE BASINS 
 
 33 
 
 Lei^e/ 'f^/rred 
 
 P/a/n 
 
 
 Fig. 15. Map and diagrams to illustrate characteristics of pit lakes. Map a part 
 of Three River Quadrangle. Lower diagram after Hobbs.) 
 
34 INLAND LAKES OF MICHIGAN 
 
 Continental glaciers are characteristically divided into two dis- 
 tinct zones, an upper one comparatively clean and free from debris 
 and a lower which is heavily clogged with rock fragments. Under 
 the sun's rays the clear ice of the upper zone melts rapidly but, 
 when the lower zone is reached, the rock fragments protect the 
 ice directly below them. The portions of the surface not covered 
 with debris melt until earthy material is uncovered and finally a 
 complete rock cover is formed which soon becomes so thick that 
 melting proceeds at a very slow rate. This difference in the rate 
 of melting of the upper and lower zones caused the ice of the upper 
 zone to recede possibly several miles while that of the low^er remains 
 stagnant. Wherever the drainage from the ice was vigorous, the pro- 
 tective cover was removed from the stagnant ice and it melted, but 
 where sluggish streams were depositing material, the ice was deeply 
 buried in an outwash plain with an unbroken surface sloping 
 gently away from the ice. The ice blocks did not underlie all of the 
 outwash plain but were more in the nature of scattered fragments, 
 due probably to the uneven distribution of debris in the lower zone 
 of the glacier. Where the load was exceptionally heavy the debris 
 accumulated on the surface until a cover was formed which pro- 
 tected the ice beneath so effectively that it persisted until covered 
 with outwash. At some time subsequent to the formation of the 
 outwash the ice blocks melted and allowed the material above to 
 subside slowly, causing pits or depressions in the surface of the 
 outwash. This process is illustrated in Fig. 15. 
 
 The distinguishing features are : The basin is a depression in a 
 plain, the materials of the plain are water deposits and, therefore, 
 assorted and sometimes stratified, the slope from the plain to the 
 water level is steep, the outline is roughly circular, and there is 
 often no outlet and no important inlets, the lake being supplied 
 and drained by the seepage of ground water through the sandy 
 material of the outwash plain. 
 
 Waves and Currents. Lagoons. Shore action often isolates a 
 depression forming a new and usually smaller, detached body of 
 water. The original depression is due to other causes and the 
 shore action is responsible for the isolation only. The work of 
 waves and currents is fully described in Chapter II and needs no 
 discussion here. 
 
 It must be stated that some of the inland lakes studied by the 
 writer cannot be definitely included in the classes described above 
 because of the complexity of the origin of the basins and the lack of 
 data concerning them. In particular, attention is called to a series 
 of elongated lake basins along the coast of the northwestern part 
 
ORIGIN AND CLASSIFICATION OF LAKE BASINS 
 
 35 
 
 of the t^outliern reuinsula— Pine, Torchlight, Elk, Walloon, etc. 
 These lakes exist in Ion"-, narrow Valleys whose bottoms are filled 
 with thick (lei)osits of loose sand. However, on the sides of the 
 llankino- hills at elevations from 100-300 feet above the level of Lake 
 Michigan are fonnd exposnres of clay varying in thickness from a 
 few feet to more than 100. These deposits have been buried by till, 
 showing that they were fornn^d prior to the last advance of the ice 
 over this region. The clays are interpreted by Leverett as lake beds, 
 inasmuch as they are distinctly laminated, nearly free from pebbles, 
 and in places show sandy partings between the layers of clay. The 
 data are incomplete but indicate the presence of great lakes in this 
 vicinity previous to the last advance of the ice, the extent and dura- 
 tion of which we have merely a suspicion. The puzzling thing is 
 the relation of the valleys in which the present lakes lie to the 
 ancient lakes as signified by the clays. The heavy deposits of sand 
 in the valleys preclude the possibility of determining by direct 
 observation Avhether or not the sand is underlain by the lake clays 
 found higher on the hills, and it is essential to know this. If 
 underlain by the clays, the basins would be classed as enormous 
 pits since it would indicate that the clays had sunk from the higher 
 elevation. If not underlain by the clays, it may be assumed that 
 the deposits have been removed or were never formed in the present 
 lake basins. If the material has been removed, streams, ice, or both 
 may have been the eroding agents. If the clays were not deposited 
 in these depressions, we may assume the depressions to have been 
 filled with blocks of ice, probably stagnant, while the clays were 
 settling on the beds of lakes which bordered these ice masses. 
 
 REFERENCES 
 General 
 
 Salisbury, R. D., Physiography, Henry Holt and Co., N. Y. 
 HoBBS, W. H., Earth Features and their Meaning, MacMillan, N. Y. 
 Tare and Martin, College Physiography, MacMillan, N. Y. 
 Michigan 
 
 Leverett. Frank, Surface Geology of Michigan, Michigan Geological and 
 
 Biological Survey. 
 Leverett, F. and Taylor. F. B., The Pleistocene of Indiana and Michigan, 
 
 Monograph 53, U. S. Geological Survey. 
 Van Hise, C. R., and Leith. C. K., The Geology of the Lake Superior 
 
 Region, Monograph 52, U. S. Geological Survey, Chapter 4, Physiography 
 
 by Lawrence Martin. 
 Allen, R. C, The Iron River Iron-bearing District of Michigan, Michigan 
 
 Geological and Biological Survey. 
 Levterett, Frank. Folio 155 (reprint), U. S. Geological Survey. 
 Lakes, origin of basins 
 Fenneman, N. M., Lakes of Southeastern Wisconsin, Wisconsin Geological 
 
 Survey, Bulletin VIII. 
 Russell. I. C. The Surface Geology of Portions of Menominee, etc. An- 
 nual Report Mich. Geological Survey, 1906. 
 Lane. A. C, Water Supply and Irrigation Paper, No. 30, U. S. Geological 
 
 Survey. 
 
CHAPTER II 
 
 THE DEVELOPMENT OF LAKE SHORES AND THE EX- 
 TINCTION OF LAKES 
 
 CHARACTERISTICS OF WAVES AXD CURRENTS 
 
 Waves. Coincideut -with the filling of a lake basin with water 
 there begins a development of its shores and, to some extent, its bed. 
 Conditions of absolute calm are exceptional and it is seldom that 
 roughened patches do not appear, travel forward, and die out on 
 the otherwise unruffled water surface. It is a matter of common 
 knowledge that for the most part these patches are caused by the 
 wind, and examination shows them to be composed of a pattern of 
 wavelets which run with the wind and whose crests are at right 
 angles to the wind direction. See Plate II, A. If the wind fails, they 
 disappear but with a freshening breeze the area of the ruffled 
 water soon spreads over the entire lake. 
 
 From this we might conclude that waves are caused by the fric- 
 tion of the wind on the surface of the water, and further observation 
 will strengthen this conclusion. At any one point, with a freshen- 
 ing wind, it is readily noted that the waves not only increase in 
 height but in length and velocity as well. This continues until a 
 maximum development is reached which is largely dependent on the 
 strength of the wind and the time it has been blowing. However, 
 a trip along the shore discloses the fact that the waves increase in 
 development toward the lee side of the lake and, if several lakes are 
 observed, it becomes evident that the larger the lake the greater 
 are the waves formed by winds of the same velocity. Thus, in addi- 
 tion to the velocity of the wind and the length of time it has been 
 .effective, may be added the factor of the expanse of water acr.oss 
 which it blows, or reach, in the development of wind driven waves. 
 This development of the wave in height, length, and velocity is due 
 to the fact that an almost continuous push by the wind is effective 
 on the wave from one end of the lake to the other and is, therefore, 
 cumulative. It continues until the friction caused by the differen- 
 tial movement of tlie water particles is equal to the energy sup- 
 plied by the wind. Inasmuch as the storm winds usually steady 
 down at a maximum velocitv which is somewhat uniform, a maxi- 
 
38 
 
 INLAND LAKES OF MICHIGAN 
 
 mum wave developmeut wliicli is exceeded only during exceptionally 
 strong winds, may be postulated for a given lake. On the other 
 hand a decrease in velocity or cessation of the wind allows the waves 
 to flatten and gradually die out. 
 
 Other causes of water waves are possible but it is the wind driven 
 wave that is effective on the smaller bodies of water. In waves of 
 this type there are two motions to be considered, the forward mo- 
 tion of the wave itself and the motion of the particles of water af- 
 fected by the wave. Whenever a wave is running in deep water 
 where its motions will not be interfered with, it is called a /ree 
 wave and the motion of the particles is theoretically in circular, 
 vertical orbits which do not move forward with the wave, the revo- 
 lution being in the same direction as the movement of the wave. 
 Thus, in waves of oscillation, there is a forward movement of the 
 particles through the upper half of the orbit during the passage of 
 the crest of the wave and a backward semi-circle with the trough. 
 This may readily be observed by watching the movement of a floating 
 object, and is illustrated in Fig. 16. If all the particles occupied 
 
 -M7i/e Length 
 
 i^'o^e Leng-/-/! 
 
 I-Va^e Base 
 Fig. 1'6. Movement of water particles in waves of oscillation. 
 
 the same relative position in their orbits, or pliase, e. g., all at the 
 top, and continued to revolve, the result would be a raising and low- 
 ering of the entire surface of the lake accompanied by a forward 
 and backward movement. This is not the case and the wave pro- 
 gresses because each impulse is communicated from one particle to 
 the next adjoining on the lee side. It follows, then, that no two 
 particles in any wave are in the same phase but are more advanced 
 in their orbits to the windward. This is represented in Fig. 16 in 
 which the orbits of conveniently spaced particles have been drawn 
 
DEVELOPMENT OF SHORES— EXTINCTION OF LAKES 
 
 39 
 
 and must be considered as representative, tlie intervening spaces 
 being tilled with particles revolving similarlj-. The particles are 
 represented as revolving in a clockwise direction and, beginning 
 at the left, each particle is more advanced in its orbit (has a more 
 advanced phase) than the next to the right by one-eighth of the 
 circumference of the orbit. By connecting these points a curve is 
 obtained which represents the form of the wave in cross section. 
 This curve* shows the steeper crests and wide troughs, and from 
 this it will be seen that the upper half of the revolution of a particle 
 described during the passage of the crest of the wave must be ac- 
 complished in a shorter time and with greater velocity than the 
 lower half, since the forward movement of the w^ave is uniform. 
 
 Both the height and the length of the wave may vary and great 
 variations in the dimensions and velocity of waves are possible. Ob- 
 servation has shown, however, that in general there is a more or 
 less definite relation in fully developed waves between the height, 
 length, and velocity. This varies somewhat in different bodies of 
 water but may be approximately stated as follows : Wave length is 
 five and one-eighth times the square of the period (time between two 
 crests) and the length fifteen times the height. The relationship 
 is shown in Fig 17 in which the phasal difference is the same but 
 the size of the orl)its increased. 
 
 Diagram illustrating the ideal development of waves. 
 
 The agitation of the water particles at the surface is communi- 
 cated to the particles below, but the motion decreases very rapidly 
 and practically dies out at a very limited depth. See Fig. 18. It is 
 usually stated that the size of the orbits is halved for each in- 
 
 *The curve here shown is a trochoid and is obtained from the trace of a point 
 inside the circumference of a rolling circle. A cycloid is formed if the point is on the 
 circumference. 
 
40 
 
 INLAND LAKES OF MICHIGAN 
 
 crease in depth equal to one-ninth of the wave length, but this is 
 only a rough approximation. An illustration may give a better 
 idea of the rapidity of the decrease in motion with depth. A free 
 wave which has a height of three feet at the surface is approximate- 
 ly forty-five feet long and the time of passage about three seconds. 
 At a depth equal to the wave length the diameter of the orbits is less 
 than one iive-hundredtli of that at the surface, or in the case cited 
 seven-tenths of an inch and the time three seconds. The rate at 
 which the water moves in describing an orbit of this size is about 
 two-thirds of an inch per second, a very feeble current and in- 
 capable of geologic work except possibly the transportation of the 
 very finest sediment. The depth of a wave length below the sur- 
 face has, therefore, been called the wave hase, that is, the lower 
 limit of effective wave action, but observations show that this depth 
 
 \ 
 
 
 Fig. IS. Diagram illustrating rapid decrease, with depth, of motion of water 
 particles in a wave, (after Fenniman). 
 
 is excessive and probably should be placed at less than half the 
 wave length. 
 
 The relationship in the dimensions of the wave does not hold 
 under a freshening breeze. Under this condition the height in- 
 creases faster than the length, and the wave gradually becomes 
 steeper until a limit is reached which, if exceeded, would cause the 
 water to describe a loop at the crest of the wave, a condition ob- 
 viously impossible. Fig. 19, which shows the effect of increasing the 
 height while the length remains constant, illustrates this point. The 
 wave is then said to break into a tvhitecap, but it is probable that 
 the wind whicli is moving much more rapidly than the wave blows 
 the crest over into whitecaps before the theoretical limit of steep- 
 ness is reached. 
 
 As stated previously, from a theoretical standpoint there is no 
 permanent forward movement of the particles of water in waves 
 but, due to the constant forward push by the wind, it happens, in 
 reality, that each revolution of a particle finds it slightly advanced 
 from its former position and there results a drift with the wind. 
 It is probable that the wind has a greater push on the crest of the 
 
DEVELOPMENT OF SHORES— EXTINCTION OF LAKES 41 
 
 Avave than on the trough because the wind and water are moving iu 
 the same direction and the steepness of the crest allows the force 
 to be applied to better advantage. Consequently, most of the for- 
 ward movement occurs during the passage of the crest and an 
 attempt has been made to show this in Fig. 20. By following the 
 course of this continuous line it will be noted that all of the for- 
 
 Fig. 19. Diagram to illustrate the formation of wMtecaps. For convenience the 
 wave length is kept constant. 
 
 ward motion is represented as taking place in the upper half of the 
 orbit. The curve is not accurately drawn principally because there 
 are no data from which to construct it but is presented merely to 
 give some idea of the manner in which a forward drift is set up 
 bv the wind. 
 
 Fig. 20. Diagram of the path of a particle of water affected by a number of 
 
 successive waves. 
 
 In addition to the drift, there are other positive forward move- 
 ments. Of these whitecaps have already been considered. Another 
 takes place when the waves approach the shore. As the water be- 
 comes shallower the agitation extends to the bottom and the lower 
 portion of the wave is retarded by friction. The orbits of the 
 
42 INLAND LAKES OF MICHIGAN' 
 
 water particles are no longer circles but are deformed into ovals. 
 These are more nearly circular at the surface but become flattened 
 with depth until the bottom is reached where the motion is a hori- 
 zontal oscillation. The retardation of the lower part of the wave 
 is shared by the upper but not in the same degree and results in a 
 decrease in velocity which is less pronounced at the surface. The 
 velocity of the wave is continuously reduced as it approaches the 
 beach and is necessarily accompanied by a corresponding decrease 
 in wave length and a steepening of the crests. In addition, the 
 wave height is actually increased, due to the transmission of the 
 motion to a continuously smaller quantit}^ of water as the depth de- 
 creases, and the form of the wave becomes steeper on the front or 
 shore side because of the more rapid forward movement near the 
 surface. The variations are all progressive and continue until the 
 crest topples forward into foam, or hreaks, see Plate II, B, which 
 reduces the height of the wave. The storm waves break first at a 
 more or less uniform distance off shore, called the off-shore breaker 
 line, and the reduction in height by breaking is sufficient to allow 
 them to regain their true wave form. They then proceed for some 
 distance, in some cases to the shore, without further breaking. 
 It is possible to have the slope of the bottom so flat that the waves 
 may break at the oft'-shore line and continue to the shore with 
 crests a mass of foam, but this usually does not occur. In such a 
 case the waves may be dissipated before they reach the shore. 
 
 Under the conditions stated above — a very gentle off-shore slope 
 — waves of oscillation are not only modified in the manner described 
 but may in some cases change their character and become a different 
 type, called waves of trmislation. The motion of the particles, in- 
 stead of being in onbits, may become a definite advance accom- 
 panied by a lifting and sinking, together making a semi-elliptical 
 path for each particle which does not return to its former position. 
 Each particle from top to bottom starts simultaneously with the 
 approach of the wave, moves forward and upward until the crest 
 arrives, then sinks in its forward movement, and finally comes 
 to rest when the wave has passed. The forward movement is the 
 same for all particles but the verticle movement is greatest at the 
 surface and decreases downward to the bottom where it becomes 
 zero, that is the movement is horiz'ontal. Fig. 21 shows the paths 
 of the particles in a wave of this type and Fig. 22 the forward and 
 upward movement in the front portion and the forward and down- 
 ward motion of the particles as influenced by the back of the wave. 
 The velocity increases in the forward half of its path from zero 
 
Micliii;:iti ( Jcdlduiciil aiu 
 
 I'ulilicaticiii .".(I, (ii'dld-ical Scries ^'.j, 
 J'lalc n. 
 
 A. WAVES. lUUT LAKE. 
 
 i!. WlHiECAl'S AND DREAKEKS, IlIGGINS LAKE. 
 
DEVELOPMENT OF SHORES— EXTINCTION OF LAKES 43 
 
 to ii luaxiiunm at the iniddle and then decreases until it comes to 
 rest. 
 
 .Waves of translation are caused by a sudden addition of a volume 
 of water to a lake or other body of water, each wave representing 
 an addition. Obviously, they are independent of each other and 
 consist merely of crests of water moving forward at a uniform rate 
 of speed. They have neither length nor trough as the terms are 
 used with reference to waves of oscillation, the surface between 
 the crests being flat and the distance variable according to the 
 regularity of the additions of water. When formed in a lake they 
 are caused by the plunging crests at the off-shore breaker line 
 which supplies the additional volume of water and usually run in 
 to the shore in apparently related series because of the regularity 
 with which the waves from tlie lake enter the breaker zone. 
 
 All waves, whether of oscillation or translation, are eventually 
 dissipated on the shore, except in the possible case of an excessively 
 
 Fig. 21. Diagram showing paths of Fig. 22. Diagram showing upward 
 
 water particle In waves of translation, and forward movement in the front 
 
 (after Fenneman). part (right) and downward and for- 
 
 ward movement in the back part (left 
 of a wave of translation, (after Fen- 
 neman). 
 
 wide breaker zone. With the final plunge on the shore the true 
 wave motion is lost and the water rushes forward and back over the 
 shore, which acts as an inclined plane. The outgoing water running 
 down the inclined shore meets the next incoming wave but suc- 
 ceeds only in modifying its front, which becomes increasingly con- 
 vex toward the shore, and increasing its height until it wavers, 
 curls forward, and crashes on the beach. This final breaking of 
 the wave is popularly known as comber. 
 
 Currents. In the preceding pages it has been pointed out that 
 there is a forward drift of the water in wind driven waves which 
 is further increased within the breaker zone by the partial or total 
 conversion of the waves of oscillation into waves of translation, and 
 that this forward movement occurs mainly at the surface. This 
 transfer of water from the windward to the lee side of lakes necessi- 
 tates a return. Tliis is accomplished by means of currents, although 
 in large lakes, as Erie and Superior, a small amount of the transfer 
 is accommodated by a piling up of the water at the lee end. The 
 currents are set up when tlie waves strike the shore and mav be 
 
44 INLAND LAKES OF MICHIGAN 
 
 horizontal currents along the shore, sJiore currents, or a vertical 
 return into the main body of the lake along the bottom which is 
 called the undertow, or both. The nature of the current is de- 
 termined by the angle at which the waves strike the shore. 
 
 When the Avaves strike the shore at exactly 90 degrees the in- 
 coming water runs up on the shore and returns underneath with- 
 out lateral movement along the shore, forming undertow only. As 
 stated above, the agitation of the water by waves within the breaker 
 zone extends to the bottom where it is a forward and backward 
 horizontal movement having a maximum velocity midway between 
 two periods of rest. Hence the undertow must be a pulsating 
 current in which the particles move to and fro but advance slightly 
 out into the lake with each oscillation. Its strength depends on 
 the height of the waves and the steepness of the off-shore slope 
 but decreases as it advances into the lake because it is distributed 
 through larger amounts of water as the depth increases. In this 
 way it loses its indentity but may continue throughout the entire 
 length of small lakes as an inappreciable drift. 
 
 But it is impossible for waves to approach all shores at an angle 
 of 90 degrees. When the waves strike at oblique angles the paths 
 of the particles of water are as shown in Fig. 23. Instead of a to 
 
 Fig. 23. Diagram to show actual motion of water particles striking shore at an 
 
 oblique angle. 
 
 and fro motion along a straight line as in the case of the undertow, 
 the motion is an oscillation on the shore accompanied by a lateral 
 movement and results in a current along the shore which maj re- 
 verse its direction because of the variable wind directions. Shore 
 currents should reach their greatest development when the waves 
 are running nearly parallel to the shore but this seldom happens. 
 It may occur where the off-shore slope is very steep and the wind 
 direction favorable. But where the off-shore slope is gentle the 
 shore end of the wave is retarded within the breaker zone and the 
 retardation increases with nearness to the shore. There is, then, 
 a tendency for the crest of the wave to bend and swing towards a 
 direction more nearly parallel to the shore, that is strike at 90 
 degrees. This change in direction may cause oblique waves to 
 strike the shore at 90 degrees if the obliquity is slight and the off- 
 shore slope wide. 
 
DEVELOPMENT OF SHORES— EXTINCTION OF LAKES 
 
 45 
 
 Two currents, then, are set np on a shore upon which the waves 
 are pounding, the undertow and tlie shore current, and the relative 
 importance of each varies according to the angle at which the waves 
 strike. With waves running on shore perpendicularly the undei^- 
 tow only is present but any appreciable variation of this angle 
 sets up a shore current which increases at the expense of the under- 
 tow as the angle departs from perpendicularity. Considering the 
 entire shore of a lake, the development of waves and currents 
 varies at any given time due to the varying relations between the 
 directions of the shore and the wind at different parts of the lake, 
 and over a period of time, because of changes in wind direction and 
 velocity. Thus, with shifting winds all the shores of the lake may 
 be affected but not equally because the storm winds come usually 
 from a' prevailing direction. Under the influence of a wind constant 
 in direction the location of the shore, and the size and shape of the 
 lake are important. 
 
 Obviously, the windward side is least affected and the effects in- 
 crease towards the lee. The size of the lake is important in that as 
 the reach is greater the waves are better de^^eloped. As regards 
 shape, the simplest case is a lake of circular outline. The middle 
 of the lee side, A. Fig. 24, receives the strongest waves and at an 
 
 Fig. 24. Diagram to show currents set up on a circular lake. 
 
 angle of 90 degrees, therefore undertow only is developed on this 
 shore, and at the sides, B, shore currents are formed. Along the 
 intermediate stretches both shore currents and undertow^ are pres- 
 ent but the former merge into the undertow as they approach A 
 and return underneath. To the windward the waves are inactive 
 and the only current possible is the return below the surface, the 
 continuation of the undertow. On the other hand, if the lake is 
 long and narrow, the condition is not so simple. The most effective 
 
46 
 
 INLAND LAKES OF MICHIGAN 
 
 wiuds are those which blow lengthwise of the lake because of the 
 excessive reach and develop waves and currents similar to those 
 established on a lake of circular outline. Cross winds, on account 
 Qf their short reach, produce effects of minor significance but simi- 
 lar in character. However, when the wind crosses the lake diagon- 
 ally, only a short stretch of the shore can be perpendicular to the 
 wave movement and strong shore currents are developed on the 
 lee side but the undertow is of minor importance. In this case the 
 shore current turns the end of the lake and skirts the windward 
 shore as a return current which has lost its intermittent character 
 and therefore much of its effectiveness. See Fig. 25. 
 
 r^ 
 
 Pig. 25. Diagram showing conditions under which a return shore current is formed. 
 
 THE WORK OF WAVES AND CURRENTS 
 
 Waves and currents in the neighborhood of the shores are very 
 effective erosive agents. They are active not only in tearing down 
 the land but also in removing and depositing the disintegrated ma- 
 terial. The resemblance to rivers is close and is strengthened when 
 it is realized that they are, in fact, nothing but bodies of water in 
 motion, obeying the same laws but with strikingly different re- 
 sults. Thus, the transporting power of running water is involved 
 and with it the presence of tools — ^^suspended material — which are 
 important in the degrading process. In general, it may be stated 
 that the waves are the active agents of destruction, and the cur- 
 rents the agents of transportation and deposition. In the final 
 analysis both are due to the same cause, the wind, and they occur 
 together. As regards the work accomplished, they may be said to 
 
DEVELOPMENT OF SHORES— EXTINCTION OF LAKES 
 
 47 
 
 supplement each other, that is, the curreuts are dependent on the 
 waves to furnish material for transportation and, on the other hand^ 
 the waves would soon lose their force if the material were not 
 removed. 
 
 The AVork of >Vaves. When a wave strikes the shore, there is 
 agitation of the water from the plunge of the wave and at the same 
 time currents are set up, except where a verticle cliff extends into 
 the water to a depth greater than wave base. The range of the 
 direct action of the wave extends approximately from wave base 
 below water level to the greatest height that the v/aves splash 
 abo\e water level but the effectiveness varies within these limits. 
 The action is greater above water level than below and is greatest 
 at or a little above the water level. The movement of the water is 
 greatest as it makes its final plunge on the shore and during the 
 succeeding in-and-out rush along the inclined beach. Here also is 
 the source of the suspended earthy material which is broken from 
 the shores by the force of the waves with the co-operation of the 
 weathering process. The suspended material is thrown violently 
 against the shore with each incoming Avave and acts as a powerful 
 abrasive, which is limited in its action to a narrow zone near the 
 water level. Within this zone cutting is most rapid and the effects 
 are more pronounced where the shores are steep. At first a low 
 cliff develops which recedes and at the same time increases in height. 
 The clift' soon extends above the upper limit of the direct action 
 of the waves and becomes undercut. The overhanging portion 
 must sooner or later fall and, thus, extends the action of the waves 
 indirectly to the top of the cliff and allows the recession to continue. 
 See Fig. 26. 
 
 Fig. 20. Farwell's Point, an undercut notched cUff on Lake Mendota, Wisconsin. 
 After a time the overhanging portion will fall by gravity to produce blocks like 
 those in the foreground, (after Fenneuiau). 
 
48 INLAND LAKES OF MICHIGAN 
 
 Coincident with the recession of the cliff there is formed at its 
 base and below water level a terrace cnt in the material of which 
 the cliff is composed and, therefore, known as a cut terrace. The 
 terrace itself furnishes the condition necessary for currents which 
 modify its surface to some extent. The tools which are so effective 
 in the cutting process are at first too heavy to be carried out of 
 the zone of action of the surf but the continual pounding against 
 the cliff and each other gradually reduces their size until they 
 are carried lakeward by the undertow or along the shore. 
 
 The distribution of this material along the shore constitutes a 
 heach which tends to become smoothed out into straight lines or 
 almost perfect curves. See Plates XII, A and XIII. The material 
 of the beach is rounded, except for particles of extreme sizes, and 
 shows a rather close assortment. The rounding of the particles 
 is due to the mutual abrasion caused by rubbiiig. Manifestly those 
 particles too large to be moved by waves are unaffected and small 
 particles, such as sand, are so buoyed up by the water that abrasion 
 is not effective. Both waves and currents are active in assorting 
 the beach materials. All of the particles which can be moved are 
 picked up and tossed shoreward by the incoming waves but only 
 the finer material is carried away as the water runs back. There 
 is, then, a minimum size of particles to be found on a beach but 
 not necessarily a maximum because rocks too large to be moved 
 may be present. During exceptionally heavy storms coarse ma- 
 terial is sometimes built into surprisingly strong ridges which 
 stand some distance from the shore under nonnal conditions and 
 are known as storm beaches. See Plate XIX. 
 
 Other factors in determining the character of beach material 
 are nearness to the source of supply and the nature of the material. 
 Where the material is quarried from a headland, the size decreases 
 and the assortment is better with distance from the source. This is 
 due to the more effective wave action on the exposed headland and 
 reduction in size of the particles by abrasion as they progress 
 along the beach. If solid rock is quarried, structures such as frac- 
 tures and bedding are important. Thus, with thin beds closely 
 fractured, small flat blocks are supplied which are soon rounded 
 off and constitute a shingle heach. In Michigan the nature of the 
 glacial deposits is important in determining beach materials, since 
 little solid rock is exposed on the shores of the inland lakes. The 
 clay of the till is readily washed away, leaving a strand of cobbles 
 and boulders, while sand and gravel, unless clearly a current de- 
 posit, are significant of outwash. 
 
 The outward passage in the undertow is a series of backward and 
 
DEVELOPMENT OF SHORES— EXTINCTION OF LAKES 
 
 49 
 
 forward iiiovenieuts which are far more effective tlian a steady cur- 
 reut of the same average velocity would be. Abundant tools of fine 
 texture enable the undertow to wear down somewhat and smooth 
 oft' the cut terrace. The modification of the terrace in this way 
 seems to be dependent largely on the time that the process has been 
 active and, therefore, is greatest at the outer edge and decreases 
 to the shore. Consequently the surface of the terrace slopes grad- 
 ually out to its edge and then drops steeply into deep water. But 
 when the edge of the cut terrace is encountered by the undertow, the 
 current loses its effectiveness due to the sudden deepening and 
 drops its suspended material. This accumulates beyond the edge 
 of the terrace until it reaches the general level of its surface at 
 which time the undertow again becomes effective and extends the 
 terrace into the lake by deposition. Thus, there is formed a cut- 
 and-built terrace which is constantly being widened by cutting along 
 the shore and by deposition on the lake side. See Figs. 27 and 28. 
 
 Fig. 27. 
 
 Profile of a cut and built terrace on a steep rocky shore. The cliff is 
 verticle and notched, at the base, (after Hobbs). 
 
 Ur -Ji'l 11^ li ^ ^•'-/v-^v.v^:^i--;f^ 
 
 Fig. 28. Profile of cut and. built terrace on a steep shore formed of loose material. 
 Note inclination of cliff and the stranded boulders in front, (after Hobbs). 
 
 It has a gentle slope away from the shore and drops suddenly 
 into deep water at its outer edge w^hich is limited by the depth at 
 which the water, agitated by the larger waves, loses its effectiveness 
 
 7 
 
50 
 
 INLAND LAKES OF MICHIGAN 
 
 as a transporting agent — one-half wave length or less. For a given 
 loealitj^ this depth is fixed by the size of the largest waves and sub- 
 sequent widening of the terrace serves only to flatten its slope 
 Avliich, in turn, reduces the action of the waves and currents. The 
 process, then, is self limiting and the shore eventually becomes 
 adjusted^ at which time the highest waves lose their force as the 
 shore is reached. 
 
 On well developed cut-and-built terraces it is not uncommon to 
 find a perceptible shoaling of the water just before the "drop off'^ 
 is reached, indicating the presence of a sand ridge which may reach 
 almost or quite to the water level. These sand ridges are nearly 
 
 coincident with the "off-shore breaker line and their manner of 
 
 ft 
 formation has been referred to the violent agitation of the water 
 
 during the breaking of the waves followed by more quiet condi- 
 tions as the waves regain their true form. In this way the condi- 
 tions of transportation and deposition are satisfactorily fulfilled. 
 Similar forms have been described extending above the water level 
 in large bodies of water and are called harrier 'beaches but in all 
 cases a lowering of the water level (or elevation of the land) is 
 involved. It is plausible that the waves might pound such forms 
 above the water level but as ^-et no such case has ever been observed. 
 Barrier beaches are not necessarily attached to the shore, are 
 composed of sand, and have gentle slopes on the lake side while on 
 the opposite side the slopes are steeper. Fig. 29. Between the land 
 and the barrier is a narrow, shallow lagoon. 
 
 Fis 
 
 29. 
 
 Section of a barrier with characteristic steep landward and 
 seaward slope, (after Hobbs). 
 
 jentle 
 
 In the discussion above it has been assumed that the material of 
 the shore was uniform but this seldom, if ever, occurs on all shores 
 even of a small lake. The different kinds of rock vary greatly in the 
 resistance they offer to the erosive agents and the same kind of 
 rock may vary from place to place. In addition, certain structures 
 in the rock serve as lines of weakness along which- the erosion 
 proceeds more rapidly. The more important of these are divisional 
 planes between beds of stratified rock and joints which are the 
 vertical cracks so commonly present in all rocks. These variations 
 
DEVELOPMENT OF SHORES— EXTINCTION OF LAKES 51 
 
 in the rocks give rise to a lumilter of transitory forms <lurinsi- the 
 progress of erosion. 
 
 Under the conditions stated al)ove (steep slojxM. clilV foi-iuation 
 is one of the first results of wave action. The steepness of the clilf 
 depends very largely on the firmness or consolidation of the rock. 
 The great deposits of sand and gravel which are very numerous 
 and extensive in glaciated regions otfer little resistence to the waves 
 and, being but loosely held together, form cliffs of moderate steep- 
 ness which are not undercut. See Fig. 28. The same statement 
 may be applied to the cliffs cut in clay except that the cliffs are 
 somewhat more steep. However, in solid rock the cliffs may be 
 vertical and overhang the zone of undercutting at the water's edge, 
 Fig. 27. 
 
 Of the various forms which are due to wave erosion the under-cut 
 cliff is perhaps the most common wherever hard rock is encountered. 
 In massive rocks the face of the cliff may be very ragged and the 
 undercutting excessive. The same may be true of stratified rocks 
 where the layers are of different degrees of hardness and many 
 picturesque forms result. Wherever a local weakness is present 
 the cutting naturally proceeds more rapidly and a sea care, Fig. 30, 
 
 Fi£ 
 
 30. Sea-caves in process of formation along joints, (after a pliotograpli by 
 C. W. CoolO. 
 
 is the result. The weakness may be due to a difference in the rock 
 or to the presence of fractures. The fractures or joints may run 
 for long distances along a straight line and often have extremely 
 smooth walls. They apparently run in all directions but study has 
 shown that the more important are grouped in systems which cross 
 each other at right angles. Joints running approximately parallel 
 to the shore may cause smooth-faced cliffs, often without under- 
 tutting if the joints are close together and well developed. They 
 
52 
 
 INLAND LAKES OF MICHIGAN 
 
 may run oblique to the shore and in this case the cliffs have a 
 buttressed effect. When the joints running at right angles to the 
 shore are followed, caves may develop. Often the joints converge 
 back from the shore and in time a channel is cut through. As the 
 channel enlarges this form is appropriately called an arch. Fig. 31. 
 
 Fi£ 
 
 il. A sea arch and small caves on the shore of one of the Apostle Islands, Lake 
 Superior, (after a photograph by the Detroit Photographic Company). 
 
 But eventually the arch must fall as the action proceeds and a 
 stack, Fig. 32, is left standing entirely separated from the main cliff. 
 And this, too, must give way in time to^ the irresistable attack of 
 the waves. These forms are all evanescent and represent the early 
 scenes in the development of this particular type of landscape. They 
 are very common features of sea coasts and the shores of large 
 lakes and are an indication therefore, that much of the adjustment 
 of the shores is yet to be accomplished. Inasmuch as the inland 
 lakes lie largely in glacial basins, these features are seldom found. 
 Waves, as we have seen, are most active on shores with steep 
 off-shore slope but conditions along the shores of any considerable 
 body of water vary and with them the effectiveness of the waves. 
 Headlands are almost universally subject to attack by the waves 
 
DEVELOPMENT OF SHORES— EXTINCTION OF LAKES 
 
 53 
 
 and to these may be added shores with notable relief, for the sur- 
 face of the surrounding, laud is to a great extent an indication of 
 the topography of the bottom, that is steep slopes on land indicate 
 similar slopes beneath the water. 
 
 THE WORK OF CURRENTS 
 
 On lakes of irregular outline shore currents must necessarily 
 encounter numerous changes in direction of the shore. The cur- 
 rent is aible to accommodate itself to many of these but where the 
 bend is abrupt it leaves the shore in the direction it had before the 
 
 Fig. 32. 
 
 A stack on the shore of Lake Superior, (after a photograph by the 
 Detroit Photographic Company). 
 
 bend was reached. As the current leaves the shore its velocity is 
 rapidly decreased because its motive power — waves striking the 
 shore at an oblique angle — is lost and friction is present between 
 the current and the still water of the lake. Deposition, therefore, 
 takes place and is greatest near the point of departure from the 
 shore. 
 
 If the bend is caused by a small indentation, the current may 
 carry across the mouth but the velocity is decreased nevertheless 
 and deposition takes place along its path in the form of a narrow 
 submerged bar, best developed near the shore. Inasmuch as the 
 
54 INLAND LAKES OF MICHIGAN 
 
 currents ma}' reverse their directiou along the shore, due to shifting 
 winds, the bar develops from both sides. As the submerged bar de- 
 velops it first reaches the .surface of the water at its land con- 
 nections and is then pounded above the water level by the waves into 
 pointed sand ridges, called spits, on either side of the indentation. 
 (See Figs. 49 and 50.) Spits do not usually develop equally on 
 both sides because of the prevailing direction of «torm winds and 
 the varying reac-h of the waves on different parts of the lake. With 
 the formation of spits the currents are able to continue farther 
 beyond the shore and the indentation is rapidly closed by a com- 
 plete bar which is above the water level. This is called a sub-aerial 
 bar or simply a &ar. Plates IX, A, IX, B, XI and XIV. If a consid- 
 erable amount of water drains into the lagoon, as the cut off in- 
 dentation is called, a current is set up across the bar which maj" 
 maintain an open channel. It sometimes happens that an island 
 is encountered by the current after leaving the shore and it becomes 
 tied to the land by the bar which developes. Fig. G5. It is then 
 called a land-tied island and the bar is designated as a tomholo. 
 Bars are composed almost entirely of sand and have a gentle slope 
 on the lake side, due to the action of the waves and undertow, and 
 a steep slope, often as much as one to one, on the J^agoon side 
 where wave action is much less intense. 
 
 In case the (indentation is wide the currents may not persist 
 across the mouth and a hook or re-curved spit is formed. Hooks 
 are 'closely related to spits in that the material and the manner of 
 formation is the same. The difference is in form, the hook curving 
 back towards the shore. See Plate VII. Under the conditions 
 necessary for the formation of hooks, the velocity of the current 
 dies out rapidly after leaving the shore and the waves are able 
 to turn it landAvard. The change in curvature is due in part to a 
 shifting of the wind so that the waves strike the hook more directly 
 or even from the opposite quarter and thus modify its form. Also 
 the tendency of waves which are running oblique to the shore to 
 swing to a direction more nearly parallel to the shore may be im- 
 portant in this process. The writer has seen waves break along a 
 hook when the retardation of the shore end of the wave was suffi- 
 cient to swing the crests around the end of the hook and even on 
 the lee side in the opposite direction to the waves on the lake. 
 
 In the growth of spits and hooks the varying direction of the 
 winds often causes minor developments in form which are not in 
 harmony with the general growth. As has been stated previously, 
 most of the storm w^inds come from a quarter which causes cur- 
 rents in one direction along a shore and the general development of 
 
DEVELOPMENT OF SHORES— EXTINCTION OF LAKES 55 
 
 hooks and spits is in accord witli this. However, high Avinds do 
 occur from other quarters and the end of the lioolc or spit often 
 grows rapidly during a single storm at an abrupt angle to the 
 general direction. Such forms are usually very transient, often 
 being destroyed by the next storm, but may persist on hooks as 
 Ijrongs extending landward from the sheltered side. See Plate VII. 
 In reality, spits and hooks represent a balance between the cur- 
 rents set up in either direction along a shore and in the normal 
 development one current usually greatly predominates. All grada- 
 tions are possible and do occur. The clearest case is where evenly 
 balanced currents from either direction are forced to leave the 
 shore at some point. Spits develop from both directions and join 
 in a point out in the lake forming a triangular or V-har — also 
 known as cnspatc forela)id — whose base is the shore and whose 
 sides are equal. If the currents are not equally effective, one side 
 grows at the expense of the other, the better developed often being 
 a hook and the smaller a spit. The material is typical shore drift 
 and the slopes are characteristic except at the tip of the V-bar where 
 it drops suddenly into deep water. The whole embankment en- 
 closes a' shallow depression similar in shape and normally filled 
 with water. Such cuspate forelands seen by the writer were on 
 long, narrow lakes running north-south and, at present, no reason 
 for the currents leaving the shore can be assigned. 
 
 Another case of deposition by currents sometimes occurs along 
 the shores of wide, shallow embayments between headlands. In 
 such cases much of the material from the headlands is transported 
 along the beach by currents into the bay. On account of the shal- 
 lowness of the bay, the undertow is not etfective, and this material 
 accumulates about the head of the indentation. In this way a- sand 
 Hat is built which gradually widens and reduces the indentation. 
 
 GENERAL EFFECT OF WAVES AND CURRENTS 
 
 In general the tendency of both waves and currents is to make 
 the outline of the lake more regular. The Avork of waves has been 
 described as a process of cutting, which insofar as headlands are 
 reduced, removes the irregularities Avhich project into the lake. 
 On the other hand, currents are agents of removal and deposition. 
 Deposition is favored by indentations which are cut ofl! or filled 
 thereby. Hovrever, it must not be inferred that a circular or oval 
 outline is the final outcome of the work of these agents on all lakes, 
 although this would seem to be the logical conclusion. The process 
 
56 INLAND LAKES OF MICHIGAN 
 
 of cutting is self limitiug and may stop before the headlands are 
 completely reduced. Also many indentations are too wide to be 
 crossed by currents, and hooks develop rather than bars, serving 
 merely to make the bends less abrupt. V-bars seem to disturb the 
 symmetrj^ of the shores and their complete development would 
 carry them across the lake and divide it into smaller and more 
 nearly circular bodies of water. This, however, is unlikely unless 
 great quantities of material are supplied from extraneous sources. 
 Another factor is the instability of conditions under which the 
 lake exists. The development of lake shores requires long periods 
 of time and changes in conditions may cause new levels or even 
 the extinction of a lake before the work of adjustment is finished. 
 
 THE WORK OF ICE ON LAKE SHORES 
 
 Ice Raimparts. On the shores of lakes in regions having cold 
 winters, ice is an effective agent of transportation and deposition. 
 Its action is mainly that of a shoreward movement, accomplished 
 slowly but with great force, which not only carries material for- 
 ward to the shore but is able to shove heavy rocks in front. As the 
 ice disappears in the spring the material affected by it is either piled 
 up in a wall on the shores known as an ice rcbmpcvrt, Plate XVII, A, 
 Figs. 51 and 61, or is left stranded on the terrace, to be moved 
 sulbsequently. The ramparts are best described as walls of rock 
 material and are located a short distance back from the shore. They 
 are usually a single ridge but may be compound and the material 
 is of all sizes, including boulders of large size. The slopes are 
 steep on both sides with a tendency for the front slope to be the 
 steeper. 
 
 The shove on the shore b}" the ice is exerted in two ways, b}^ ex- 
 pansion and by ice jams. Expansion occurs during the winter 
 when the lake is completely frozen. Water is rather exceptional 
 in its properties and behavior under various conditions but, once 
 frozen, acts as any other solid at ordinary temperatures and j)res- 
 sures. Thus, it expands or contracts with a rise or fall in tempera- 
 ture, and this property is involved in the formation of ice ramparts. 
 When a lake first freezes it is enveloped with a layer of ice which 
 completely covers its surface. If the temperature remains constant 
 below the freezing point, the ice merely increases in thickness. But 
 temperatures do not remain constant and the changes are often 
 great and rapid. There is always a lowering at night followed by 
 a daily rise and, in addition, there are the cold and warm waves 
 which sweep over the country. With each drop in temperature the 
 
DEVEiLOPMENT OF SHORES— EXTINCTION OF LAKES 57 
 
 ice contracts but does not pull away from the shores. It cracks 
 instead and the cracks are soon healed by freezing, since the temper- 
 ature is below the freezing point. In this way the ice cover is 
 kept intact but actually contains more ice that it did at the higher 
 temperature. If such a condition is followed by a rise in tempera- 
 ture, the ice must expand and it will then be greater in extent 
 than the surface of the lake. Repeated alternations in tempera- 
 ture serve to exaggerate this condition. The expansion may be 
 accommodated either by overriding the shores or by buckling. In 
 the latter case pressure ridges in the ice are formed out from the 
 shore and usually occur in the same places year after year. Buck- 
 ling takes place when the ice is not thick enough to withstand the 
 pressures exerted by the expansion and where the shores are steep 
 enough to prevent a landward movement of the ice edge. Even under 
 this condition a slight amount of movement may be possible if 
 the water level is low when the action takes place. In glacial ma- 
 terial, at least, this results in a cliff embedded with boulders at its 
 base or a boulder lined strand. See Plate VIII, B. 
 
 On sloping shores the ice is free to move and, if conditions are 
 favorable, pushes up a rampart. The size of the material plays an 
 important part both in their formation and permanancy. As a 
 rule fine grained material such as sand does not freeze into a solid 
 mass and the ice slides over it with little accumulation of debris. 
 Also sandy shores are more commonly low and well-developed ram- 
 parts are not formed. However, if binding material, such as the 
 roots of trees and mats of grass, is present, the material is pushed 
 into ramparts, Fig. 61, but they are soon destroyed by the waves. 
 During the winter of 1914-15 the sandy shore at the north end of 
 Whitmore Lake, Washtenaw County, was pushed into a rampart 
 about four feet high but by August of the following summer the 
 rampart was less than a foot in height. Where coarse material is 
 present it offers a good purchase to the ice which carries and pushes 
 it on the shore forming ramparts which resist wave action and are 
 relatively permanent. The fact that fine material is so generally 
 mixed with coarse in ramparts indicates that the shoving action is 
 important. The maximum size of particles moved in this way is 
 not known but boulders of several tons in weight are found in ram- 
 parts. 
 
 Ramparts, formed by expansion, although very common on the 
 shores of lakes in Michigan, require rather specific conditions for 
 their formation and are formed only in winters when these condi- 
 tions are fulfilled. The conditions involve the climatic factor, 
 shore conditions, and the size of the lake. Shore conditions have 
 
58 INLAND LAKES OF MICHIGAN 
 
 already been discussed. Other conditions being equal, the size of 
 the lake determines the amount of expansion. On small lakes, 
 possibly less than a half mile in diameter, the amount of expansion 
 is so small that the ramparts, if formed, are insignificant and soon 
 destroyed. Larger lakes permit greater expansion but the ice be- 
 comes less able to transmit the thrusts caused by expansion, as the 
 diameters are greater, and it buckles. The abjlity to transmit the 
 thrusts depends on the thickness of the ice in relation to its expanse 
 and is thus controlled by the climate. In this region the maximum 
 size of lake upon which such ramparts are formed is not definitely 
 known but probably does not greatly exceed a mile and a half. 
 
 The climatic factor includes both temperature and snowfall, 
 temperature changes are essential but they must be large and rapid, 
 and quickly transmitted to the ice. The excessive temperature 
 changes of cold and warm waves which occur in this latitude during 
 winter fulfill the temperature conditions perfectly but the daily 
 rise and fall does not seem to be adequate. The rate at which the 
 air temperatures are communicated to the ice depends on the thick- 
 ness of the snow covering which may form an effective blanket. 
 Absence of snow is, therefore, the most favorable condition. 
 
 The position of the rampart with reference to the shore is deter- 
 mined by the size of the lake and by the amount and number of 
 the temperature changes. In general, it is situated just back of 
 the shore which marks the high water stage. The size depends 
 on the supply of available material and the number of times the 
 ice shoves across the shore. If we assume that the ice moves a 
 definite amount of material at each invasion, it is obvious that the 
 development of the rampart is dependent on the number of inva- 
 sions. As a rule, the ice encroaches on the shore but once each 
 winter, each drop and rise in temperature serving only to push 
 the ice farther on the shore. More than one advance is possible if 
 the ice thaws during the winter and loses its continuity, but the 
 total movement in the several advances is the same as that in a 
 single advance and, therefore, less effective for each. Within limits, 
 large particles are more readily moved by the ice than small ones 
 and a preponderance of such material assures a strong rampart. 
 However, the supply of this material on a shore may be limited and, 
 after repeated invasions, may become exhausted, thus limiting 
 the growth of the rampart. 
 
 Inasmuch as the climatic factor is relatively constant over a 
 period of years, the typical rampart is a single ridge. If the avail- 
 able material is sufficient, the ridge develops until it becomes strong 
 
DEVELOPMENT OF SHORES— EXTINCTION OF LAKES S9 
 
 enough to resist successfully the i)ush of the ice and any furlher 
 expansion is relieved by buckling of the ice sheet. The growth from 
 this time on, if any, must be in width and from the lake side. On 
 lakes whose levels are gradually lowering the overriding of the 
 shores by the ice is not materially decreased but the position of 
 the ice edge after expansion becomes less and less advanced as the 
 shores recede and a series of ramparts are formed to Avhich the 
 name ice-push terrace has been given. Another factor in their for- 
 mation is that with the lowering in level fresh areas of the bottom 
 are brought into the zone of ice action and the supply of material 
 is replenished. 
 
 Ice ramparts of the expansion type, then, are limited to regions 
 whose winters are severe and are punctuated by frequent cold and 
 warm waves and to lakes of moderate size with absence of snow 
 covering. Of these conditions the frequency and amount of the 
 effective temperature changes, the amount of snow, and even the 
 thickness of ice vary considerably, and it is only during winters 
 when all of these factors are favorable that the expansion is efCec- 
 t'lVQ in forming ramj)arts. 
 
 Ice Jam. The explanation of ice ramparts on the basis of ex- 
 pansion is only half of the story. Kamparts almost identical with 
 those known to have been formed by expansion are common on 
 lakes which do not fulfill the conditions necessary for this type. 
 These lakes may exceed the maximum diameter postulated for ex- 
 l^ansion many times and are covered for the entire winter with a 
 thicklayer of snow. In fact, it is known from observation that the 
 ice does not expand on the shores of the larger lakes, and the ram- 
 parts have been accounted for by the action of ice jams which occur 
 during the final melting of the ice in the spring. 
 
 The melting of the ice at this time proceeds most rapidly at the 
 shores, due to the more rapid heating of the land than the water 
 under the influence of the rays of the sun, and a lane of open Avater 
 of considerable width is formed next the shore. After the lane is 
 formed weather conditions are of great importance. Some springs 
 a prolonged warm spell is accompanied by calms or light breezes 
 and the ice melts rapidly with little disturbance. It may disap- 
 pear entirely in this way or become porous or ''rotten" so that sub- 
 sequent winds reduce it to slush. If, however, a storm develops 
 after a lane is formed but before the ice has becomes porous, the 
 mass of ice is blown before the wind and moves slowly to the shore 
 with an almost irresistable force which carries everything in its 
 path. See Plates X, A and XII, B. The storms in this latitude are 
 
60 INLAND LAKES OF MICHIGAN 
 
 cyclonic in nature and may best be described as vast whirlwinds 
 which revolve in a counter-clockwise direction and at the same time 
 travel from west to east across the country. They are especially 
 frequent in Michigan because most of the storms tracks unite in 
 the G-reat Lakes region. They are accompanied by winds which 
 shift from an easterly direction usually through the south to the 
 west and northwest, but occasionally in southern Michigan shift 
 through the northern half of the circle. The velocity of the winds 
 increases with the shifting and reaches its maximum with the 
 westerly and northwesterly winds at which time it often reaches 
 the intensity of a gale. These storms travel eastward at the average 
 rate of about 700' miles a day but the rate is variable and they may 
 halt for 24 hours or more. Inasmuch as they affect areas often 
 1500 and more miles across, their effects may be felt for several 
 days in a given locality. All shores of a lake may thus feel their 
 influence but the eastern and northeastern sides are in the lee of 
 the strongest winds, at least for southern Michigan. 
 
 Where shore conditions are favorable ice jams bring in and pile 
 up material in ramparts almost identical with those formed by 
 expansion. Boulders of several tons in weight are moved shore- 
 ward in this way, leaving a trench between them and the lake and 
 having a pile of rubble in front. Several such rocks were found on 
 the northeast shore of Long Lake, Alpena County, (See Figs. 77 and 
 78.) As in the case of expansion ramparts, those formed by ice 
 jam are self limiting for they finally must reach a strength which 
 is able to stop the advance in the lower portion of the ice and the 
 top shears over. Kepetition of the process may form an ice push 
 terrace, see Plate III (Athabaska), and this may possibly occur 
 during one season as the ice is buffeted back and forth by the 
 shifting winds or by a close succession of storms. 
 
 The size of the lake is also of importance. Permanent ramparts 
 of this type are not found on very small lakes because the momen- 
 tum of the small ice masses is not sufficient for large effects. On 
 very large lakes they are formed but the material of the shores 
 is usually fine and wave action excessive so they are soon destroyed, 
 e. g., on the Great Lakes. The size of lakes on which the maxi- 
 mum push is exerted is not known but it is much larger than that 
 for the expansion type. Practically all of the larger inland lakes 
 of Michigan show evidence of ice action which is roughly propor- 
 tional to the size. On many of the lakes of intermediate size it is 
 certain that the push is exerted in both ways and the ramparts are 
 thus intensified. 
 
Mioliisan <iool(isic;il and 
 I?u>l(>si('al Siiivcy 
 
 I'lllilicntiini .".(), < Jcoldiiical Scries 25 
 I'lati' III. 
 
 
 l7-»m^cr 
 
 A. ICE-PUSII TKUKACE, ATHABAiSCA. 
 
 Courtesy Canadian Geological Survey. 
 
 B. LAKE NEARIXG EXTIXCTIOX BY VBGETATIOX, SECOXD SISri'EK LAKE, 
 
 XEAK AXX AlKBUK. 
 
DEVELOPMENT OF SHORES— EXTINCTION OF LAKES 61 
 
 Another efifect of ice occurs Avheii a great number of floating 
 cakes of ice are present. These drift with the wind to the shores 
 and, i-nstead of being pushed np on the shore, some drift along 
 the shore with the waves and currents. The action of these agents 
 is increased by tlie ice especially as regards the transportation of 
 large material. Some of the material may be supplied by rocks 
 frozen into the ice which are released on melting but some shore 
 material is moved along by the ice blocks and arranged into forms 
 similar to spits or V-bars. A V-bar formed in this way was noted 
 on Long Lake, Alpena County, I»late XIV, B. The form, the depres- 
 sion in the center, and the slopes are characteristic of current action 
 but the material is angular limestone blocks having in some cases 
 a largest dimension of as much as eighteen inches. 
 
 A similar form is shown in Plate XV, A. This feature resembles 
 a spit in its position at the neck of an indentation but differs in 
 form and material, Xormally the outline of a spit is very regular 
 but in this case the curvature is serpentine in character (the main 
 body of the lake is to the left of the spit in plate) . Also the ma- 
 terial is angular and of relatively large size on its surface, although 
 much of the submerged portion is composed of sand. It is prob- 
 able that current action is largely responsible for its formation 
 but floating ice blocks have added some material and have suc- 
 ceeded in modifying its form. 
 
 THE EXTINCTION OF LAKES 
 
 With the birth of a lake the forces which were responsible for its 
 formation leave it unprotected to the action of certain agents which 
 remodel its outlines and, to some extent, its bed. In addition, an- 
 other set of agents becomes active which inevitably results in the 
 extinction of the lake if existing conditions prevail. All of these 
 agents work rapidly from a physiographic standpoint and support 
 the idea that lakes are temporary features of the earth's surface. 
 
 The processes which are working towards the extinction of lakes 
 are filling, draining, and evaporation. The filling of a lake basin 
 may be achieved by sedimentation, by animal and vegetal remains,, 
 and by chemical precipitation. Sedimentation is active in lakes 
 whose entering streams bring in great quantities of silt which is 
 deposited near the shores to be worked over later by waves and 
 currents, and to which may be added the material torn from the 
 cliffs by the waves. The working over of this material into bars 
 and barriers also favors the accumulation of vegetation. 
 
 Some of the lakes are inhabited by innumerable minute animal 
 
62 INLAND LAKES OF MICHIGAN 
 
 organisms whose hard parts are composed of calcium carbonate. 
 When the animals die the fine shells drop to the bottom and become 
 a part of beds of a white, powdery substance known as mar\, Marl 
 is usually considered to be a mixture of calcium carbonate material 
 from several sources which, in addition to animals, include vege- 
 tation and chemical precipitates. Marl beds as much as forty feet 
 in thickness have been found in Michigan lakes and are being utilized" 
 in the manufacture of Portland cement. The bed shown in Plate 
 IV, B, is from three to four feet thick and is exposed along the arti- 
 ficial channel of the Sturgeon Eiver near Indian River. Thicknesses 
 of seventeen feet obtained by borings have been reported along 
 Crooked River in the same vicinity. 
 
 Vegetation is another source of filling. Numerous water loving 
 plants analogous to the leaves and stems of deciduous trees die at 
 the close of the growing season and sink to the bottom. These dead 
 parts, being covered with water, are protected from the gases of 
 the atmosphere and only j)artially decompbse. The yearly residue 
 accumulates as dei)Osits of peat, a light-brown to black porous sub- 
 stance composed very largely of vegetal remains, many of which 
 are well preserved. This material burns readily but with poor 
 heat values and is not used to any great extent as fuel in this- 
 country at the present time. The abundance of better fuel has held 
 back the exploitation of peat but, as the supply of coal diminishes, 
 the importance of the great peat deposits will become more and 
 more appreciated. 
 
 The plants which enter into the formation of peat may or may 
 not be attached to the bottom but iii either case probably do not 
 grow in water exceeding twenty-five feet in depth, due to unfavor- 
 able conditions of heat and light, and usually are within from two 
 to six feet of the surface. The floating forms are important in 
 lakes which are protected from strong winds and may sink and 
 form a deposit over the entire lake bottom. Those attached to 
 the bottom start in the shallow water along the shores and grow 
 outward into the lake as the accumulation of their remains de- 
 creases the depth. However, these forms, growing most abundantly 
 near the surface of the water, are not entirely dependent on shallow 
 water but extend outward over the surface as a floating bog, com- 
 posed of the felted and intertwined stems and roots. These bogs, 
 often tenacious, are elastic and give under pressure, hence the name 
 quaking bogs. They may develop so rapidly as to cover the surface 
 of the lake before the basin is completely filled and are thus under- 
 lain by clear water. The development from this time on is accom- 
 
DEVELOPMENT OF SHORES— EXTINCTION OF LAKES 
 
 C3 
 
 plished by di'oppiiiiAS I'l-om the uiidei' side of the bog, and in this 
 way the water is crowded out of the space below. The bog be- 
 comes firm, first along the shore anil progressively outward. When 
 a lake is filled witli peat the growth of vegetation on its surface 
 continues for a time. But exposure to the air is unfavorable for 
 preservation of the plant remains and the accumulation ceases a 
 short distance above the former water level. 
 
 The encroachment from the shores in typical cases is quite regu- 
 lar and shows an interesting zonal relationship between the different 
 kinds of plants. Tlie constitution of the zones may vary but for 
 southern Michigan the first plants to develop are the floating forms 
 and the pond weeds. As these grow outward a zone of water lillies 
 starts at the shore and is closely followed by the floating sedges 
 which form the floating bog. With the filling of the clear water 
 
 
 Bf Bs Bm S M O-W 
 
 
 
 ivir 
 
 Fig. 33. Diagrams illustrating the filling of a lake by vegetation. The several plant 
 associations of the Bog series, displacing one another, belong to the following 
 major groups: — (1) O. W. — open water succession; (2) M. — marginal succession; 
 (.S) S. — shore succession; (4) B. — bog succession, comprising the bog meadow 
 (B. m), l)og shrub (B. s) and bog forest (B. f) ; and (5) M. F. — mesophytic forest 
 succession. (Ri-produced from Bulletin 10, Geological Sui-vey of Ohio.) 
 
64 INLAND LAKES OF MICHIGAN 
 
 «nder the floating matt, shrubs and other plants develop on the sur- 
 face of the matt which crowd out the sedges. Next come the coni- 
 -fers, usually tamarack and spruce, and the last to propagate are 
 the deciduous trees, especially poplar, willow, and maple. See 
 Fig. 33 and Plate III, B. Such is the typical succession when fully 
 developed but in the intermediate stages the later zones are absent. 
 In this manner lakes are filled by vegetation but the process 
 varies in importance in different lakes. Quiet water is essential 
 for plant growth, hence small lakes and lagoons are most affected. 
 Shore conditions also have an effect. Gentle off-shore slopes are 
 favorable because wave action is less intense and depths are suit- 
 able for vegetation, but are often composed of sand in which plants 
 take hold with difficulty. Frequently there is found a heavy growtii 
 ■of weed'^ in the mud just beyond the edge of the terrace but the 
 sandy terrace itself supports a sparse growth of reeds. Dead lakes, 
 B.S lakes filled with vegetation are frequently called, are character- 
 ized by a monotonously flat surface composed of black soil and 
 covered with a thick carpet of moss and shrubs above which Is 
 growing a thin stand of timber. See Plate XVIII. 
 
 Lakes may become filled to some extent by chemical precipitation 
 but this process is limited in its application. In this climate it 
 -may have been of importance in the formation of marl where cold 
 springs enter the lakes, but in dry regions it plays a more import- 
 ant role. In such regions the lakes are typically without outlets, 
 due to the fact that evaporation is excessive and prevents the 
 waters from rising to an avenue of escape. The loss of the water 
 hj evaporation allows the dissolved material brought in by streams 
 to accumulate and, when sufficient concentration is reached, to 
 precipitate on the shores and bottom. Among the substances de- 
 posited in this way are salt, borax, calcium carbonate, etc. 
 
 The draining of a lake is accomplished by cutting down of the 
 ■outlet. Inasmuch as lakes act as settling basins, the outlets are 
 relatively free from sediment and in general cut very slowly. The 
 size and velocity of the outlet, and the resistance that the material 
 ■over which it flows offers to abrasion, determine the rate of down 
 cutting. Certain lakes, on account of their depth reaching below 
 sea level, cannot be drained under existing conditions but, with the 
 cooperation of filling, extinction is always a possibility. 
 
 Changes in climate are necessary for the extinction of lakes by 
 evaporation and the change must be such that the supply of water 
 is decreased or the evaporation greatly increased. A more arid 
 
DEVELOPMENT OF SHORES— EXTINCTION OF LAKES 65 
 
 climate supplies both eoiuUtioiis aud is usually accompanied by an 
 increase in temperature. Many examples of partial or total ex- 
 tinction from this cause are to be found in tbe arid west but none 
 in Michigan. Great Salt Lake, which has been greatly lowered in 
 level in this way, is one of the best known examples. 
 
 The relative importance of the different methods of extinction 
 varies greatly in different regions and with individual lakes. In gen- 
 eral, the outlet is deepened rapidly in unconsolidated rocks, but 
 even in hard rock this may be true if the down-cutting is due to the 
 recession of a waterfall, such as Niagara Falls in the outlet of Lake 
 Erie. As a rule deposition is more important than draining but in 
 Michigan this is probably not the case. Down-cutting of the out- 
 let is important because the great majority of the outlet streams run 
 over unconsolidated glacial material which is readily eroded with- 
 out the help of tools. On the other hand deposition has been slight. 
 Many of the lakes are fed by springs and the drift deposits have 
 as yet been only slightly trenched by streams, in most areas the 
 original slopes being almost intact. In addition, the streams are 
 usually short and the areas draining into the lakes small. An ex- 
 ceptional case is Torch Lake near Houghton where the Sturgeon 
 river has built a large delta at the southern end of the lake. A 
 more important source of material in our lakes is the cliffs which 
 sometimes form a large part of the shores. The cliffs, composed 
 almost entirely of unconsolidated material, are easily eroded by 
 the waves and the debris is distributed along the shores and bottom. 
 It is possible that the enlargement of the lake by shore recession 
 may equal the amount of filling. Where cliffs form a considerable 
 part of the shores the filling must be greater; and the ratio in- 
 creases according to the height and preponderance of the cliffs. It 
 is probable, however, that the amount of deposition in the lakes of 
 Michigan so far has been a matter of a few inches only on the 
 bottom. 
 
 At the present time, draining is probably more important than 
 filling, but Avith future development the down-cutting of the out- 
 lets will gradually decrease as the streams approach grade, and the 
 sediment brought in by tributary streams must increase as these 
 streams extend their courses. At the same time the material de- 
 posited by waves and currents will decrease as the terraces widen. 
 The deposition of part of this material in shore-forms reduces the 
 size of the lake by cutting oft' indentations and thus facilitates 
 filling, both in the main body of the lake and in the lagoons. 
 
 Vegetal accumulation seems to be more important in the extinc- 
 
66 INLAND LAKES OF MICHIGAN 
 
 tion of lakes in Michigan than either of the two processes discussed 
 above and is especially effective in the smaller lakes and lagoons. 
 It is impossible to give an estimate of the amount of filling that has 
 been accomplished in this way, but the prevalence of "dead lakes" 
 and quaking bogs indicates that vegetal accumulations are of fre- 
 quent occurence. All peat deposits are not necessarily evidence 
 that a lake basin has been filled, and it is only by a determination 
 of the depth and distribution of the peat and in some cases a 
 recognition of plant zones that the extinction can be proven. 
 
 As regards chemical precipitation, marl is practically the only 
 deposit of any significauce in the lakes of the State and it may be 
 formed in other ways. Three factors may be active in its formation, 
 plants, animals, and chemical precipitation, and their relative im- 
 portance is not known. In general, marl is one of the first de- 
 posits to be formed on a lake bottom and is often covered with peat. 
 It may be sufficient in itself to fill a lake basin but no cases of such 
 filling have been described in the knowledge of the writer. 
 
 THE CYCLE OF SHORE DEVELOPMENT 
 
 In the preceding pages the development of lake shores under the 
 influence of waves and currents has been traced. This development 
 is gradual and systematic, and the various stages are marked by 
 definite topographic forms. In other words, the shores pass through 
 a cycle of events which begin with the birth of the lake and termi- 
 nate when the waves and currents are impotent to further modify 
 them. A change in water level, either up or down, institutes a new 
 cycle which may or may not interru]3t the previous one before it 
 is completed. Following the practice with regard to streams, the 
 stages in the cycle have been likened to the life cycle and are termed 
 youth, maturity, and old age. These terms, in a general way only, 
 indicate corresponding lengths of time during which the forces 
 have been active, but conditions, both as to the constitution of the 
 shores and the force of the waves and currents, are so variable that 
 the emphasis should be placed on the stage of development rather 
 than on the time element. 
 
 The youthful stage is a period of active erosion. The shore is 
 marked by irregularities above and below water level, and a general 
 lack of adjustment to the movements of the water. The presence of 
 frequent headlands necessitates numerous bays with sharp curves 
 and the shore currents are consequenth^ poorly defined and discon- 
 tinuous. As the headlands are reduced and irregularities of the 
 bottom filled, the currents increase in strength and continuity, and 
 
DEVELOPMENT OF SHORES— EXTINCTION OF LAKES G7 
 
 eventually siuijilily the shoivliiic by cutting otl" reentrants. Youth, 
 then, is a time of relatively rapid changes and is bronght to a close 
 Avhen all possible cut-otl's have been accomplished. 
 
 The progress from this time on is gradual in contrast to the rapid 
 changes of youtli and characterizes maturity. The shore liue as a 
 whole either shifts landward or lakeward depending on the efficiency 
 of the currents and tlie material available. Where abundant ma- 
 terial is supplied by incoming streams, the shore will advance lake- 
 ward. If little or no nuiterial is supplied, the shore must progress 
 landward bijt the recession becomes increasingly slower until an 
 end point is reached. Shores of most lakes probably never reach 
 a stage beyond maturity because of the interference of the process 
 hj extinction or by the inauguration of a new cycle. The inaugura- 
 tion of a new cycle by a rise in level gives conditions of the same 
 nature as those present when the lake basin was first flooded. In 
 case the water level sinks, sliore action will be influenced more or 
 less by the topographic forms developed during the previous stage, 
 and the development nuiy consist largely in a remodeling of these 
 features. The latter condition is of common occurrence on the in- 
 land lakes of Michigan. On practically all the lakes, at least one 
 higher level ma}' be recognized, unless the level has been raised by 
 dams, and in some cases as many as four have been found. 
 
 REFERENCES 
 
 Waves and Shore Action. 
 
 Fexnemax, N. M., Lakes of Southeastern "Wisconsin, Wisconsin Geologi- 
 cal Survey, Bulletin VIII. 
 
 GiLBEBT, G. K., Topographic Features of Lake Shores, U. S. Geological 
 Survey, Fifth Annual Report. Lake Bonneville, U. S. Geological Sur- 
 vey, Monograph I. 
 
 Stokes. C. G., On the Theory of Oscillatory Waves, Cambridge Transac- 
 tions, Volume VIII. 
 
 Russell, J. Scott. The Wave of Translation, London, 1885. 
 
 'Cornish, Vaughan, Waves of the Sea and Other Water Waves, Chicago. 
 Ice Action. 
 
 Buckley, E. R., Ice Ramparts, Transactions of the Wisconsin Academy 
 of Science, Arts and Letters, Vol. XIII. 
 
 Tyerell, J. B., Ice on Canadian Lakes, Transactions of the Canadian In- 
 stitute, Vol. IX. 
 
 Johnson, D. W., Shore Processes and Shoreline Development. N. Y., 
 1919. 
 Also general references cited at end of Chapter I. 
 
CHAPTER III 
 LAKES OF THE CHEBOYGAN RIVER BASIN 
 
 In the basin of the Cheboygan River lie several of the larger in- 
 land lakes of the State, which, on account of their grouping and 
 manner of formation, may well be discussed together. See Fig. 34. 
 About three miles upstream from the city of Cheboygan the river 
 branches, one branch leading from Black Lake on the borders of 
 Cheboygan and Presque Isle counties, and the other draining a 
 chain of lakes known as the "Inland Route." The lakes of the 
 Inland Route and their connections are navigable for boats of small 
 draught and a regular passenger service is maintained during the 
 summer months from Cheboygan on the Straits of Mackinaw to Con- 
 way on Crooked Lake about three miles from Little Traverse Bay. 
 The lakes included in this route are Mullet, Burt, and Crooked. 
 Douglass, another lake of considerable size and importance, lies 
 directly north of Burt, and is also included in this drainage system. 
 
 So far as known these basins lie entirely in glacial deposits which 
 are somewhat complicated in this interlobate region. On the north- 
 east side the moraines deposited by the ice of the Michigan and 
 Huron lobes have a northwest-southeast trend and consist of a 
 number of ridges which overlap in some cases. The best defined 
 is probably a narrow ridge which parallels the shore of Lake Huron 
 from Mackinaw to beyond Cheboygan, the only break being that 
 through which the Cheboygan River flows. On the western side 
 the moraines were deposited by the Michigan lobe and should be 
 more nearly north-south in trend but are poorly developed. Little 
 Traverse Bay caused a small lobe of the ice Avhich penetrated as 
 far as Crooked Lake and left the Aveak morainic ridges that cross 
 this lake. The puzzling topographic feature is the extensive low- 
 land area which is irregular in outline and extends from the head 
 of Little Traverse Bay nearly to Cheboygan. This depression is 
 crossed by similar depressions running northwest-southeast. The 
 latter apparently lie between the moraines but the main depres- 
 sion runs transverse from Little Traverse Bay to Cheboygan, near 
 which place the depression is terminated by the Cheboygan moraine 
 mentioned above. It seems certain that the depressions existed 
 
70 
 
 INLAND LAKES OF MICHIGAN 
 
 Fig 34. Outline map showing lakes of the Cheboygan River Basin. Note : Douglass 
 Lake, north of Burt Lake, is undesignated. 
 
LAKES OF THE CHEBOYGAN RIVER BASIN 71 
 
 prior to the last retreat of the i;hicier and may liave heeii caused 
 by stream action i)revioiis to the advance of the ice, by the sconr 
 of the ice in its advance, or by both. Dnring the retreat of the ice 
 these depressions were filled with small lobes of ice which melted 
 more slowly than the main ice slieet and prevented heavy deposi- 
 tion in or across them. Also tills rej»ion is nnderlain by a pnre 
 limestone which has been dissolved to a consideral)le extent east of 
 this locality forming nnmerons sinks, and it is ])robable that some 
 of the dee]) holes in these lakes were formed in this manner. 
 
 All of these lakes lie in i)arts of this irregnlar depression whose 
 slopes are strikingly marked by shore lines of former lake levels 
 higher than the present. One of these shores stands on the average 
 about 90 feet above Lakes Michigan and Ilnron and marks the 
 borders of Lake Algonquin which in this region may best be de- 
 scribed as a' great archipelago. This archipelago covered all of the 
 present inland lakes of this group and large areas of the adjacent 
 lowland as well, leaving a heavy veneer of sand on the slopes now 
 exposed. Below the Algonquin beaches at elevations varying from 
 thirty-five to forty-five feet above Lakes Michigan and Huron, is 
 another well defined shore line, that of Lake Nipissing. It stands 
 below the level of Douglass and Black Lakes but is present around 
 Burt, Mullet, and Crooked Lakes, a' short <listance back from the 
 shores and at elevations varying from fifteen feet above Crooked 
 Lake to twenty-five feet above the Cheboygan River at Cheboygan. 
 Thus, with the sinking of the level of Lake Algonquin, Douglass and 
 Black Lakes become isolated basins while the lower part of the de- 
 pression, in which the lakes of the ''Inland Route" lie, was stil? sub- 
 merged and separated a large island to the northwest from the 
 mainland. During Xipissing time the opening at Little Traverse 
 Bay- was partially closed by a bar the sands of which have been 
 heaped into dunes^ see Plate IV, A. These dunes rise gently on the 
 western .sides to heights of one hundred thirty to one hundred 
 forty feet and then drop steeply on the eastern sides, showing 
 clearly the predominance of westerly winds. Near the shore of the 
 bay small dunes are now in j)rocess of formation and are migrating 
 eastward. Farther inland, however, the large dunes have been 
 clothed Avith vegetation which has prevented further movement. 
 This row of dunes forms the divide which forces the water to run 
 eastAvard into Lake Huron and is narrowest and lowest at Kegomic, 
 having a Avidth of slightly more than one-fourth mile and a height 
 of thirty-four feet above Lake Michigan. 
 
 The recession of Lake Xiidssing to the present Great Lakes level 
 
72 INLAND LAKES OF MICHIGAN 
 
 isolated a large inland lake which at first occupied all of the de- 
 pression between the bar at the head of Little Traverse Bay and 
 Cheboygan. Later it w^as divided by a bar at Indian River and 
 then low^ered to the present condition as the Cheboygan River 
 deepened its channel. The variation in elevation in this part of 
 the Cheboygan River drainage is very small, the total drop being 
 less than twenty feet in thirty miles, and most of this occurs in the 
 last mile of the river. The difference in elevation between Crooked 
 and Mullet lakes is less than sixteen inches but this is in part due 
 to a ponding of the water by a dam across the river at the mill of 
 the Cheboygan Paper Co. 
 
 CROOKED LAKE 
 
 Crooked lake is the western member of the '^Inland Route" and 
 is readily reached by the G. R. & I. R. R. which skirts its northern 
 shore. The name is none too appropriate if applied to its outline, . 
 which is roughly triangular and is nearly divided by Oden Island 
 slightly east of the center. The lake is shallow for the most part 
 but contains a good sized basin which drops to sixty-one feet in 
 depth west of the island. 
 
 The irregular basin of Crooked Lake lies in a trough which 
 crosses the general trend of the morainic ridges having northwest- 
 southeasterly trend. The constriction in the outline caused by Cin- 
 cinnati Point is due also to morainic material. This persists as a 
 submerged ridge across the lake, with a maximum depth of less 
 than twelve feet, and is flanked by deeper water. A similar ridge 
 but better developed almost divides the lake at Oden Island. It 
 seems probable, then, that the main depression existed before the 
 last retreat of the glacier and may have been formed by a small 
 lobe of ice which pushed through Little Traverse Bay. As the ice 
 retreated, small morainic ridges were deposited across the trough 
 and are largely submerged at the present time. The deep basin 
 west of Oden Island was probably filled by a protected mass of ice 
 which left this depression on melting. The whole depression was 
 later covered by the waters of Lakes Algonquin and Nipissing, 
 which deposited a veneer of sand over the morainic material. In 
 fact, this sand covers the lowlands bordering the lake, and the 
 till is exposed only where the sand has been removed along the 
 headlands by wave action. 
 
 Three former levels are easily recognized along the shores of 
 Crooked Lake. The Algonquin and Nipissing lakes have already 
 been mentioned and their shores are found at levels of seventv and 
 
Fig. 35. Map of Crooked Lake, Emmet County, showing configuration of bottom, (after XJ. S. Lake Survey Chart). 
 
.MicluUMIl ( icdiiiuir;! I nllll 
 
 I'liiilouii-al Siiixi'v 
 
 riiblicjiliori .".(). iH'iiNmical Sci-icw 2'.j, 
 I'lair IV. 
 
 A. SAXD DUNES, LITTLE TKAVBRSE BAY, KOlM) LAKE IX lOKEGROUXD. 
 
 D. .^L\KL r.EI». STliaiEOX UIVER XEAK lU'KT LAKE. 
 
LAKES OF THE CHEBOYGAN RIVER BASIN 73 
 
 fourteen feet respectively above the jji-esent level. The tliird level 
 occurs between four and five feet above the present and is but moder- 
 atel}' developed, in places dropping out entirely. The faintness 
 of the shore lines of this Post-jMipissiug level may be due in part 
 to the greatly reduced size of the lake and the consequent weakened 
 shore action, but also to the fact that this level was maintained for 
 a short period of time. The drop to the present I'evel was due to 
 the cutting down of the outlet, which runs through loose sands and 
 therefore worked rapidly. 
 
 The shore adjustments of this lake are interesting and have taken 
 place largely at the lower levels. Much vi'ork was done during the 
 Post-Nipissing stage but adjustments are still taking place that 
 will make important changes in the lake if allowed to continue. 
 These mnj best be appreciated by a description of the shores in the 
 order of a traverse. 
 
 Conway, situated at the west end of the lake, see Fig. 35, lies on 
 a sand flat but slightly above the level of the lake. This flat is 
 interrupted by a shallow, swampy trench through Avhich the outlet 
 of Eound Lake reaches Crooked. The beach at this end is of sand 
 and the lake bottom slopes gently outward, making an excellent 
 and safe bathing beach. South of the outlet of Round Lake, the 
 land slopes gradually upward to the Nipissing beach which follows 
 the lake shore to Cincinnati Point. Along this shore the Nipissing 
 terrace is well developed at an elevation of fourteen feet above the 
 lake, averaging about one hundred yards in width, and above this 
 an abrupt cliff fifty to fifty-five feet in height rises to the Algonquin 
 terrace. The beach is of clear sand and even in contour except 
 where littered with drift wood. The material of this beach is 
 working eastward and is being deposited in a spit attached to the 
 west side of Cincinnati Point. 
 
 The point is caused by a till knoll which stood as an island at 
 the beginning of the Post-Nipissing stage but w^as connected w^ith 
 the mainland by a bar along the western side behind which was 
 a lagoon. This bar now stands from eight to one hundred feet back 
 from the beach. Inasmuch as the present wave action on the' east 
 side of the point is slight, in place of a bar a terrace is found. 
 Along this side the ice has pushed a strong rampart where shore 
 conditions were favorable. It is especially noticeable at the end 
 of the jDoint and near the mainland, but fails between these places. 
 East of Cincinnati Point the Nipissing terrace is narrow, but of 
 sufficient width to allow the building of a wagon road. Above this 
 terrace a steep cliff, fifty feet high, rises to the Algonquin terrace 
 
74 INLAND LAKES OF MICHIGAN 
 
 which ends iu a well defiued shore line oue hundred or more yards 
 to the south. Along the present shore there is no indication of the 
 Post-Nipissing level and the beach is of coarse material as far as 
 the blunt swell which marks the southeastern limit of the point. 
 Here the Nipissing terrace widens and slopes gently to the shore. 
 
 The bay beyond Cincinnati Point is caused by a long, narrow, 
 swamp which savings back towards Bound Lake. Beginning on the 
 point and extending about one-third the distance around the bay is 
 a continuous rampart which reaches a height of six feet in j)laces 
 and was formed during the Post-I^ipissing stage, i:\ear its south- 
 eastern extremity it encloses a lagoon, indicating that both cur- 
 rents and ice have been active in its formation. The bay terminates 
 in a broad point which is lined with a' beach of coarse material, an 
 indication that the material is still covered with sand. Along this 
 point the ice has pushed up a rampart w'hich continues around the 
 southern end of the lake as the most prominent shore feature as far 
 as the Minnehaha River. North of the river the land is low and 
 flat but not swamp3^ The rampart is present, but poorly developed 
 along this shore. A short distance inland a faint terrace and shore 
 of the Post-Nipissing stage can be distinguished. Shore action 
 has been slight here, iu sj^ite of the fact that the waves which 
 strike this shore have the longest reach on the lake and are driven 
 by the strongest winds. The explanation is that the off-shore 
 slope is very gentle around the entire southern end of the lake and 
 the force of the Avaves is largely dissipated before they reach the 
 shore. The adjustment is not complete, however, because currents 
 are actively transporting the shore material northward and have 
 built a spit more than one hundred yards in length opposite Oden 
 Island. This spit extends outward under water and meets a long- 
 slender spit which has grown from the southeastern end of the 
 island. The opening between the island and the mainland, which 
 was originally more than one-fourth mile in width, is now less than 
 two hundred feet and is so shallow that only boats of very 
 small draught can pass. The two spits are not exactly in line at 
 present, the direction of that attached to the island being almost 
 due east. They will eventually swing into line, and, once this is 
 accomplished, the tjmg of the island to the mainland to the south 
 will be a matter of a few years only. 
 
 The till of the island is largely masked with sand, but an indica- 
 tion of its presence is found in the cobble beach along the south 
 shore. This island was evidently a shoal during Nipissing time 
 but was partly above water during the Post-Nipissiug stage. At 
 
LAKES OF THE CHEBOYGAN RIVER BASIN 
 
 75 
 
 this time most of the nrea was }»laiie(l olV to a sand terrace with tlie 
 exception of a small part near thesonth side. At the Post-Nipissinti' 
 stage tlie lake ice was very active on the small island and pnshed 
 u]i a i)rominent rami)art on all sliores. Wave action has heen 
 especially active on the west shore and to a lesser extent on the 
 south shore. This resulted in the transportation of the shore ma- 
 terial around tlie north and south ends and its deposition in the 
 form of spits, of which the one at the southeast corner has already 
 been described. The counterpart of this spit occurs on the north- 
 west corner and is actively growing at the present time. It has ex- 
 tended some distance beyond the original shore of the island and 
 encloses, a lagoon to the east. The outline of the tip of this spit as 
 shown in Fig. 36, presents a sudden jog to the east. Undoubt- 
 
 Fig. 36. Spit offset near distal end. Northwest end of Oden Island, Crooked Lake. 
 
 edly this jog represents a slight elevation of the level of the lake 
 or at least a holding up of the water to a more uniform level 
 throughout the year than it naturally would have. The only ex- 
 planation the writer can offer is that the waters are ponded to some 
 extent and kept at a more constant level by the presence of the 
 dam at Cheboygan, which is but five feet lower than the level of 
 Crooked Lake. Unless the channel at this point is kept open arti- 
 ficially, the island will be tied to the mainland from this end as 
 well as the south. No data could be obtained concerning tlie 
 date of construction of the dam, and this is very unfortunate for, 
 with this at hand, some estimate of the time necessary for the com- 
 pletion of the bar might be made, if the interpretation of the break 
 in the outline of the spit is correct. 
 
76 INLAND LAKES OF MICHIGAN. 
 
 East of the island the shore is low and shows a faint beach of 
 the Post-Nipissing level some distance back from the shore. Wave 
 action is slight here, but currents are set up which have formed a 
 spit about one hundred feet long on the south side of the small bay 
 into which the outlet of Pickerel Lake enters. Beyond this bay 
 the land rises to the Nipissing terrace which is rather wide and 
 slopes gently toward the lake. The beach is of sand which is being 
 carried northward and deposited in a well developed hook at the 
 somewhat prominent projection. Along this hook the trees line the 
 shore in places, and the roots are gradually being swept free from 
 sand, which is added evidence of an abnormally high level for the 
 lake. Back of the hook just mentioned stands a lagoon which con- 
 nects with 'the north end of the lake and is in process of filling by 
 vegetation. 
 
 Crooked River, which discharges the water of this lake into Burt, 
 is a very sluggish stream, having a drop of slightly more than 
 six inches in over four miles. The valley runs between the edges of 
 the Nipissing terrace and gradually narrows until at Alanson it 
 just allows the passage of the stream. This is the only place on 
 the river where a road-crossing has been made. Below Alanson the 
 depression widens somewhat and the river expands into Hay Lake, 
 now so filled with vegetation that it has been necessary to dredge a 
 cliaimel. Leaving Hay Lake the stream takes a straight course 
 tlirongh a low sand flat, but suddenly begins to meander at the 
 Devil's Elbow. This seems to be the highest place in the depres- 
 sion between Crooked and Burt lakes, and the banks correspond 
 closely in elevation with the Post-Nipissing level as found on 
 Crooked Lake. It is evident from this that the drop to the present 
 level is due to the cutting of the outlet through these sands. The 
 stream with its present current could hardly have cut this channel, 
 but at the higher level the gradient was somewhat steeper and 
 there were no artificial obstructions in the drainage system. 
 
 Returning to the lake, the north shore presents little of interest 
 until Ponshewaing is reached. Here the Post-Nipissing terrace is 
 well shown, and upon this an ice rampart is found somewhat west 
 of the point. Currents from the west have been active along this 
 shore, but the resulting forms are obscured by docks and "made 
 ground." However, at Oden a well developed spit was formed at 
 the Post-Nipissing level, running to the east and partially enclos- 
 ing a narrow lagoon which has been dredged and is now used as a 
 harbor for small boats. The town of Oden is built on the Nipissing 
 terrace, the front slope of which has been cut into low cliffs by the 
 
LAKES OF THE CHEBOYGAN RIVER BASIN 77 
 
 waves of the present lake. Farther to the west, this terrace is rela- 
 tively narrow and the Algonquin terrace above is the more promi- 
 nent. As the west end of the lake is approached, both terraces leave 
 the lake and continue to the north side of Little Traverse Bay. 
 Along this shore the terrace of the Post-Nipissing stage stretches 
 from the foot of the Nipissing terrace to the beach and is wet and 
 swampy. 
 
 From the description above it should be clear that Crooked Lake 
 as an isolated basin has stood at a level some four feet higher than 
 at present. Considerable adjustment of the shores has taken place 
 at the higher level and is still going on. A notable change that 
 may be expected is the tying of the island to the mainland both at 
 the southeast and northwest ends. This will probably be accom- 
 plished first at the southeast and later will have to be prevented 
 artificially at the northwest end of the island if the lake continues 
 to be navigable to its western end. The ''drop off" is well defined on 
 shores exposed to the storm winds, such as the west side of the 
 island, the large embayment on the south shore, and the north shore 
 near Oden. The depth at the "drop off" is approximately four feet, 
 and in most places it is evident that the slope of the submerged ter- 
 race is very flat. This depth seems very small for a lake of this 
 size, and it is probable that this terrace is largely the result of 
 wave and current action during the Post-Pipissing stage, at which 
 time the depth over the terrace was double that at present. This 
 flat off-shore slope must greatly reduce the force of the waves, but 
 complete adjustment has not been accomplished as yet. In th(; 
 future, more is to be expected from deposition than from cutting, 
 although the slight flooding of the lake has increased the latter. 
 Ice action has been of some importance, and in several cases excel- 
 lent ramparts have been formed. Yet, as a rule, the material and 
 topography of the shores are not favorable for their development. 
 
 As to the extinction of this lake, it is certain that it cannot be 
 drained unless the level of Lake Huron is materially lowered. It 
 stands 14. G feet higher than Lake Huron and there are three "holes" 
 which have greater depths than this. Tributary streams are few 
 and deposit little sediment, so this method of extinction may be 
 considered of slight importance. Filling by vegetation is of much 
 greater importance. In many places marl is being deposited, and 
 beds of seventeen feet in thickness have been reported in the outlet 
 south of Alan son. In addition, heavy stands of reed grow each 
 season on parts of the submerged terrace, parti^cularly in the east 
 
78 INLAND LAKES OF MICHIGAN 
 
 arm and along the soutli shore where some protection from the 
 waves is afforded, and aid the process of filling. 
 
 BURT LAKE 
 
 The second menilber in this group of lakes is Burt Lake, which 
 with an area of 26.5 square miles is one of the largest inland lakes 
 of the State. This lake is oblong in shape and extends north-south. 
 Its length is slightly less than ten miles and its width reaches 
 about five miles, although the average is probably nearer three. See 
 map. Fig. 37. It is easily reached by the Michigan Central R. R. 
 which crosses the outlet at the town of Indian River, situated on 
 the outlet one-half mile from the lake. 
 
 As far as known, no hard rock outcrops on the shores of this 
 lake, the surrounding land being composed entirely of glacial de- 
 posits. In general, it is flanked with moraines which run slightly 
 oblique to the length of the lake. One of these moraines, which 
 causes Colonial Point on the west side, ends abruptly -at the point, 
 and irregular deposition of the morainic deposits on the east side 
 has given rise to Greenman point near the head of the lake. The 
 north end of the lake heads in a swamp beyond which is the out- 
 wash plain extending to the east end of Douglass Lake. On the 
 west side, Crooked River enters the lake through a low sand 
 plain and Indian River drains the lake through a similar de- 
 pression at the south end. The basin is consistently regular, 
 usually reaching depths of forty to fortj^-five feet, but is somewhat 
 deeper towards the south end. Two exceptions to the evenness of 
 the bottom are present : A small pit east of Colonial Point Avhich 
 drops to more than seventy feet in depth, and a shallow depression 
 near the south end fifteen to twenty feet below the general level. 
 This basin seems to lie in a depression between morainic ridges 
 which on the west side especially are irregular in distribution and 
 continuity, and were deposited by the ice from the Lake Huron 
 basin. The ice in this locality did little abrading, and this basin 
 probably existed before the last retreat of the ice. The complica- 
 tion of the morainic system makes it seem plausible that Burt Lake 
 was filled with ice after the general ice front had retreated, and 
 around parts of its borders outwash Avas deposited which now lies 
 well above the lake level, e. g., the outwash at the north end. The 
 "holes" in the bottom of the basin may be due to exceptional thick- 
 ness of the ice or may possibly be sink holes. 
 
 Burt Lake, on account of its size, the excellent development of 
 the Nipissing terrace and cliff, freedom from swamj)y shores, and 
 
Fig. 37. Map sLowiug outlines and coiiliguralitm oli the bottom of Burt and Mullet L; 
 
 (Atter U. S. Lake Survey Chart.) 
 
Fig. 37. Map showing outlines and coufiguration of tlie bottom of Burt and Mullet Lakes, Cheboygan County. 
 
 (Al'ter U. S. Lake Survey Chart.) 
 
LAKES OF THE CHEBOYGAN RIVER BASIN. 7&' 
 
 its accessibility shares the popularity of the lakes of the ''Inland 
 Route" as a place for recreation. At the present time the sunnnev 
 homes are largely near the south end of the lake and along the west 
 shore south of Crooked River. These locations are near Indian 
 River, the source of supplies, and along the route of the passenger 
 service from Conway to Cheboygan. There are abundant cottage 
 sites all along the shores, and the writer confidently looks for- 
 ward to a mucli greater development of this lake as a summer re- 
 sort in the Inture. 
 
 Indian River leaves the lake at the extreme southeastern corner 
 and flows through the nortli side of a break in the upland which is 
 about a mile in width and extends to Mullet Lake. This rather 
 broad channel is flanked on either side by the high cliffs of Lake 
 Nipissing. Its bed, where not trenched by Indian and Sturgeon. 
 Rivers, rises gradually to a sand bar which extends from cliff to^ 
 cliff through the town of Indian River in a regular curve concave 
 to the west. This bar grew from the west and practically separated 
 the Burt and Mullet lake basins, forcing the outlet to the north. . 
 On the gentle front slope of the bar are several minor beaches wkicki 
 were formed during the recession of Lake Nipissing and probably" 
 mark levels of short duration, since small terraces and cliffs at like 
 'elevations are found along the shores of Burt Lake. In the la- 
 goons behind these small beaches swamp conditions prevailed, and 
 beds of marl were laid down one of which is shown in Plate IV. 
 
 Lentil thirty-five j'ears ago, the Sturgeon River flowed behind 
 this bar into the Indian River and choked the channel with its 
 heavy deposits of sand. When the necessity of navigating Indian 
 River arose, the results of this deposition were recognized and an 
 artificial channel was dug which turned the waters of Sturgeon 
 River directly into Burt Lake. Some idea of the amount of ma- 
 terial deposited by the river may be obtained from the delta which 
 has been built into Burt Lake since that time. It projects fully 
 three hundred yards beyond the general curve of the shore and at 
 present has split the stream into two distributaries. The west 
 shore of the delta curves outward gently but the turn on the east 
 is abrupt, showing that westerly currents prevail now as in former 
 times. The delta extends outward under water a short distance 
 onh^ and drops rapidly from about ten feet to nearly twenty. The 
 sub-aqueous terrace continues around the south and west sides 
 of the lake to the vicinity of Saege;i\s Resort, where it is much less 
 definite, and disappears in the bay to the north. 
 
 To the west of the delta the Xipissing terrace narrows, and the 
 
80 INLAND LAKES OF MICHIGAN 
 
 €liflfs which rise above it to a height of thirty-five feet gradually ap- 
 proach the shore. At Pittsburg Landing this terrace stands sixteen 
 feet above the present lake level and is wide enough to afford an 
 ideal location for cottages, see Plate V, A. The lake side of the 
 terrace in places merges into a lagoon and marks a level of the lake 
 "between Nipissing and the present. This shore is exposed to the 
 northerly and northeasterly winds, and the elfect of the waves of 
 long reach is 'seen a short distance to the west, where the Nipisslng 
 terrace has been entirely removed and bare cliffs in excess of fifty 
 'feet in height reach from the shore to the top of the Algonquin 
 terrace. At Kingsley Beach the Nipissing terrace reappears with 
 moderate width and is backed by a low cliff rising to the Algonquin 
 terrace. This condition persists as far as Saegers with slight 
 variations in the width of the Nipissing terrace and the character 
 of the material of the present beach. For the most part, sand 
 beaches prevail, but at the Saw Mill and at Saegers where the 
 moraine comes to the shore the beach is strewn with boulders. The 
 low cliff between the present shore and the Nipissing terrace is, 
 as a rule, covered with grass, but in a few places fresh scars boldly 
 announce a renewal of Avave work. At Saegers it has been neces- 
 sary to dump boulders along the shore to prevent the encroach- 
 ment of the waves, in other places the cutting has not as yet oblit- 
 erated a small terrace between the Nipissing and the present level. 
 
 Beyond Saegers the Nipissing terrace widens and slopes gently 
 to the lake. Trees grooving to the water's edge are being under- 
 mined by wave action and thrown over on the shore either by winds 
 or ice. As the shore swings into Poverty Bay the prevailing cur- 
 rents leave the shore and have built a spit running into the bay. 
 This spit has grown at least one-hundred yards from the shore and 
 supports a row of trees on its surface. Behind the spit is an ex- 
 cellent example of the filling of a lagoon by vegetation, mainly 
 rushes and cat-tails. Here again we find evidence of flooding, for 
 the outline of the spit has not the characteristically even contour 
 and is lined with tree roots partially excavated by' the waves. 
 
 Around the bay, called by some Poverty Bay, the shores are low 
 and swampy, and show little wave action. The bottom here is 
 muddy, and in the shallow water vegetation, protected from the 
 strong winds, is growing outward from the shore, giving practically 
 no beach. Crooked River has built a small delta into the bay but 
 at present is flowing through an artificial channel. At the head 
 of the bay, however. Maple River has built a large projection 
 through which it flows in a series of distributing channels. It will 
 
Miiliii;;iii (Jt'olofiical :uul 
 Jiiologii'al Survey 
 
 I'tililicatioii .".0, (i('(il()f;i'"il .Scries 2.j, 
 Plate V. 
 
 
 
 ^^^^^^^H^^*^ %^ J^S^X '- ~.^Ji£gZL.JttliiM^S^i^Bg-,-,-^yr^i' - 
 
 A. PITTSBURGH LANDING, BURT LAKE. 
 
 B. BOULDER-PAVEiD BANK, BURT LAKE. 
 
LAKES OF THE CHEBOYGAN RIVER BASIN 81 
 
 be noted from the map. Fig. o7, that the branches of the river avoid 
 the main part of the delta and now empty into small bays on either 
 side. This will resnlt in tlie filling of these bays until the present 
 channels become so clogged that further shifting is necessitated. 
 The even shore line of the bay northeast of Maple River is quite 
 in contrast to that to the southwest, and examination shows that a 
 low sand bar has developed from the east side of Colonial Point, 
 extending to the mouth of the river and cutting off a part of the 
 low swamp about its lower course. 
 
 There is a sudden transition from the swamp of Maple River to 
 the higher ground of Colonial Point. This point is a moraiuic 
 hill whose top was planed off during Algonquin time. The 
 Mpissing terrace is very well developed along the point but becomes 
 faint inland as it converges from the shores to the northwest. It 
 was impossible to trace the shore completely around the hill, but 
 from the elevation of the land to the northwest it is safe to con- 
 clude that this point was a peninsula with a very narrow neck, or 
 possibly a land-tied island in Lake Nipissing. The terrace is wider 
 at the end of the point than at the sides, due to protection on the 
 west side and excessive wave action on the northeast which has 
 removed part of this terrace since Nipissiug time. A Post-Nipissing 
 terrace is well preserved on the bay side, forming a low, swampy 
 zone next the shore which never exceeds twenty feet in width. 
 
 Oft' the end of the point and continuing northward the subaqueous 
 •terrace is narrow and the "drop-off" sudden at about ten feet. 
 This continues, but gradually widens and loses its identity towards 
 the north end of the lake. The shore features on the east side of 
 Colonial Point are rather uniform, consisting of a well-developed 
 but narrow Nipissing terrace the outer edge of which has been cut 
 into low cliff's by waves at the present level, and. a' beach of coarse 
 material, residual from the disintegration of the till. One interest- 
 ing exception occurs at the small projection on the east side of 
 the point near the end. At this place the Nipissing shore recedes 
 from the present shore a' distance of two-hundred fifty yards in a 
 slight indentation into which the currents are able to swing. How- 
 ever, at one of the lower "intermediate levels the currents left the 
 shore and built a bar across the head of this bay, enclosing a shal- 
 low lagoon which is now dry and supports a growth of large trees. 
 This, bar may be recognized on the present beach by the change 
 from the coarse material to sand. 
 
 North of Colonial Point the low ground which runs southwest- 
 ward to the Maple River swamp comes to the shore. The trees grow 
 11 
 
82 INLAND LAKES OF MICHIGAN 
 
 to the water's edge and are being washed away at high water, giv- 
 ing alternate stretches of partly excavated tree roots and sandy 
 beaches. This low tract is somewhat over a mile in width and 
 gives way to morainic hills on whose slopes the features are so 
 similar to those found on the east side of Colonial Point as to 
 need no further description. Near the north end of the lake the 
 cliffs leave the shore which is then bordered by a swamp through 
 which Carjj Creek runs. The contour of the sand be^ich has a 
 scalloped effect, due to the prominent delta built by Carp Creek. 
 Currents are active here, coming from opposite directions in each 
 re-entrant, but have not developed distinguishable bars at the pres- 
 ent shore. It is possible that bars maj have been built at higher 
 levels, but the nature of the swamp and the heavy undergrowth 
 makes their determination an uncertain task under the conditions. 
 
 On the east side of the lake the swamp gradually narrows and 
 is replaced by a morainic ridge of hard, red till, running slightly 
 east of south and ending abruptly at Greenman Point. Along 
 this shore the Nipissing level is represented by a prominent cliff 
 but the terrace is narrow and steep, indicating a small amount of 
 wave action during Nipissing time. This is to be expected from 
 the location of the shore which precludes the possibility of waves 
 of long reach striking it except at a very oblique angle. The present 
 beach contains much coarse material which is quite generally j^ushed 
 up above the strand, and in places patches of ice ramparts are to be 
 seen, best developed at Greenman Point. Evidence of ice action ^ 
 is not common on the shores, of this lake, and its presence on the 
 northeast shore leads to the conclusion that ice jams are the cause 
 of the shove rather than expansion. In addition, the size of the 
 lake is in excess of the maximum on which expansion is considered 
 to be effective. 
 
 At the end of Greenman Point an interesting hook discloses con- 
 siderable current action along this shore. The hook, a sketch of 
 which is given in l^'ig. 38, rounds the point and doubles back on 
 itself almost parallel to the main shore, extending well into Bour- 
 asau Bay and enclosing part of the swamp into which this bay 
 heads. The material is finely graded from cobbles four to five 
 inches in diameter near its land connection to fine sand at its end, 
 and has been supplied entirely from the cliffs to the north. The 
 weak currents moving south are unable to cross the broad entrance 
 to the bay and deposit material which is subsequently worked into 
 the bay by the strong southwesterly:^ winds. At the head of the 
 bay the only effect at present of wave or current action is the under- 
 
LAKES OF THE CHEBOYGAN RIVER BASIN 
 
 S3 
 
 iniiiiug of trees which grow to the water's edge, and this is probably 
 due to a recent elevation of the water level. On the east side of the 
 bay, however, material from the south is being worked into tlie bay 
 b}' southwesterly winds, here the most powerful on account of 
 reach, forming a sand beach. 
 
 Fig. 3S. Hook at Greenman Point, Curt Lalie. 
 
 South of the bay the moraine approaches the lake and for a short 
 distance has the characteristic profile of this region, — a flat terrace 
 surface at the top referred to Algonquin time, a cliff and terrace of 
 Nipissing stage below this, and the final descent to the lake which 
 may be notched in places by the Post-Nipissing terrace (see Fig. 39). 
 
 /l/gonquin 
 
 Post- Mpissing 
 
 Lake Le\/e/ 
 
 Fig. 39. Diagrammatic profile of the exposed terraces of Burt Lake. , 
 
 For a distance of about a mile the Nipissing terrace is narrow and 
 steep but then widens considerably' beginning at the point below 
 Fresh Breeze Eesort. Along the present shore coarse material on 
 the beach occurs almost uniformh' and occasionally numbers of 
 boulders are found. The boulders invariably stand high on the 
 beach and in some cases have been forced back into the clay cliff. 
 This indicates strong ice push, but conditions are not favorable for 
 
84 INLAND LAKES OF MICHIGAN 
 
 the formation of definite ramparts. Tlie small point about midway 
 between the ends of the lake, A on map, is an almost isolated mor- 
 rainic hill which may have been an island during Mpissing time 
 since the main clifl: runs back of it. It is virtually lined with 
 boulders, large and small, which are shoved up into a wall. See 
 Plate V. It is possible that some of these have been cleared from 
 the adjoining farm and dumped here, but the regularity of the 
 wall and the presence of drift logs four and five feet above the 
 present level shows clearly that ice push is intense at this point. 
 The Nipissing terrace here is relatively wide and strewn with 
 boulders. This shows that the action of the waves was strong in 
 this locality during Mpissing time, and the terrace is cut rather 
 than built. A glance at the map shows that the strong northwest- 
 erly winds have considerable reach here and that the direction of 
 the shore is favorable for heavy pounding by the waves and for 
 efficient current action to remove the disintegrated material. South 
 of the point the land is low and was covered during Nipissing time. 
 Three definite bars below the Nipissing level were found in this 
 swampy depression, each of which cuts off a crescent shaped lagoon. 
 The lower one was formed at the present level, but the remaining 
 two stand higher and mark levels intermediate between the 
 Mpissing and present. The material in these bars was de- 
 rived from the north and was distributed as far as the point just 
 north of Tuscarora Beach. The present shore at this point shows a 
 hook-like form which is being built largely from the south. In 
 reality, this form is being built more after the fashion of V-shaped 
 embankments, for material is supplied both from the north and 
 the south, although there is no enclosed depression. The currents 
 from the south are the stronger and the hook is consequently turned 
 to the north. The point itself is caused by the projection of bould- 
 ery drift, some of which has been pushed northw^ard into a spit of 
 coarse material at a level above the present. From the size of the 
 boulders this evidently has been formed largely through the push of 
 drifting ice blocks. See Long Lake, Alpena County, for similar 
 forms. South of the point the waves are actively cutting the Nipis- 
 sing terrace and have formed a low, freshly cut cliff from which the 
 material of the hook is derived. Along Tuscarora and Wautan 
 beaches the waves are cutting the outer edge of the Nipissing ter- 
 race, in places exposing the boulder clay. The terrace here has, 
 therefore, been cut. Large boulders have been lined on the beach 
 by ice, but their paths on the lake side are obliterated. 
 
 South of Wautan Beach the Nipissing shore continues almost 
 
LAKES OF THE CHEBOYGAN RIVER BASIN 85 
 
 south instead of following the present shore, and the terrace widens 
 to some extent. At the somewhat prominent projection about one 
 mile south of Wautan Beach, both the shore and the oustide edge 
 of the terrace of the Nipissing stage are back some distance from 
 the shore, the intervening area being swamp except at the some- 
 what higher ground of the point. North of the point this swamp is 
 cut off from the lake by a complete bar at the present level, but to 
 the south two definite bars are to be found at the Post-Nipissing 
 level during which stage the point, then an island, was connected 
 to the mainland. From this point to the outlet the slopes have the 
 characteristic profile, and the shores are lined with the coarse ma- 
 terial commonly present where waves have cut into boulder clay. 
 One place of interest is a narrow swampy area a short distance 
 north of the outlet, lying adjacent to the shore and backed by a low 
 cliff rising to the Nipissing terrace which is here poorly developed. 
 This swampy area may be a terrace of the Post-Nipissing level but, 
 if so, indicates considerable wave action at this point. The poor 
 development of the Nipissing terrace here seems to show slight 
 wave action, and the swamp is more likely the bottom of Lake 
 Nipissing beyond the zone of wave action. 
 
 From the above description it will be seen that the physiographic 
 features of Burt Lake are comparatively uniform, so much so, that 
 their description is somewhat monotonous. Shore adjustments, 
 past and present, have been few and consistent for the most part in 
 the development of a cut and built terrace bordered with cliffs, and 
 a' limited amount of depossition. The most notable changes oc- 
 curred during Xipissing time when the basins were cut off from the 
 main lake and partially isolated. The change from Nipissing to 
 the present level w^as accomplished slowly and with at least two 
 intermediate levels, as shown by beaches and by slopes notched by 
 wave action. The dropping in level was due to the cutting down 
 of the Cheboygan Kiver Avhich varied in rate, due probably to var- 
 iations in the constitution of the material of the moraine near 
 Cheboygan. A closer packing of the till or heavy accumulation of 
 boulders in the channel would hold up the waters for a time, but, 
 with their removal accomplished, the downward cutting would be 
 renewed. 
 
 The intermediate Post-Nipissiug levels were of short duration 
 and little work was done. The cliffs and terraces are faint w^here 
 present and for the most part have been entirely removed by wave 
 action during the present stage. In general, deposition was active 
 in the past in the same localities as at present, but no important 
 
86 INLAND LAKES OF MICHIGAN 
 
 reduction in tlie size of the lake was accomplished by the develop- 
 ment of bars across indentations. In fact, under the present condi- 
 tions it is unlikely that any great changes in outline due to this 
 cause will occur, except possibly at Greenman Point. Here the 
 southerly winds seem to be able to distribute the material brought 
 along the point, and the bay may be filled but not cut off. The 
 only other large point, Colonial, is being attacked by the waves. 
 Possibly the greatest development will be the delta of the Maple 
 River which may fill the bay into which it flows. This depends 
 largely on the rate of deposition, for it seems likely that this stream 
 will be abandoned as an outlet of Douglass Lake (see description 
 of this lake). 
 
 Wave action has predominated on this lake and the terraces 
 are the prominent features on the slopes facing the shore. In some 
 cases the Nipissing terrace is complete but, for the most part, 
 has been cut into at lower levels. Under existing conditions, wave 
 action is also the prominent agent, and a relatively narrow terrace 
 and "drop off" is present, especially well developed along the sides 
 and at the south end. The depth of water at its outside edge is 
 between ten and eleven feet and this is much less than the wave base 
 for this lake, which reaches a depth of at least thirty feet. In this 
 case the submerged terrace has been formed mainly at the present 
 level and indicates that the depth at which effective transporta- 
 tion of sand ceases is about one-third of the wave length during the 
 greatest storms. 
 
 Burt Lake, on account of its size, depth, regular outline, and 
 slight elevation above Lake Huron will become extinct very slowly. 
 Filling by vegetation, although in progress, is not effective due 
 mainly to the lack of protected shores. It is most active in the 
 bay west of Colonial Point and in Bourasau Bay on the east side. 
 Both of these hays support a heav}' growth of rushes during the 
 summer, and in the former marl covers much of the bottom. Con- 
 siderable sediment is brought to the lake by streams but they 
 supply only the minor part of the water of this lake. The pres- 
 ence of extended areas of sand in this locality increases the im- 
 portance of ground water over surface drainage, and much of the 
 water is supplied from this source. In fact, there are but three 
 streams of any importance which enter the lake — Sturgeon, Maple, 
 and Carp — and of these Maple River may possibly be abandoned. 
 There is the possibility that these streams will develop and drain 
 much larger basins in the future, and make sedimentation a factor 
 of importance. Deepening of the outlet at the present time is at 
 
LAKES OF THE CHEBOYGAN RIVER BASIN 87 
 
 a slaiul-still, due to Iniinan iiiterlVreiice, and it seems probable that 
 this Avill eontiime indeUiiitely. However, ignoring the human side 
 and assuming that no obstruction will be present in the outlet, the 
 amount of draining must necessarily be limited by the level of Lake 
 Huron, whose present level is but fourteen feet lower than Burt 
 Lake. Lowering of Burt Lake by fourteen feet would decrease 
 its depth more than one-third, but the reduction in area would be 
 relatively slight. Still, the shores would be much lower than at 
 present and the muds of the present bottom would afford excellent 
 conditions for a heavy and rapid growth of vegetation. Another 
 possibility lies in the fact that the land is here rising very slowly, 
 and the consequent dropping in the level of the Great Lakes would 
 allow complete drainage of the lakes of the ''Inland Koute," pro- 
 vided the uplift continues for a long enough period of time, 
 
 MULLET LAKE 
 
 A short distance east of Burt Lake and connected with it by the 
 Indian River lies Mullet Lake, see map, Figure 37. These two lakes 
 are very similar in shape and size, the greatest difference being in 
 the orientation. Disregarding the extinct arm at the southwest 
 end. Mullet Lake is almost identical in length with Burt Lake, and 
 the average width and size, 26.8 square miles are not materially dif- 
 ferent. The outline of the shores shows considerable irregularity 
 especially on the southeast side. The points for the most part run 
 directly out into the lake and have about the same general direction 
 as Colonial Point and the northeast shore of Burt Lake. 
 
 The surrounding country stands well above the lake and has a 
 somewhat northwest-southeast trend, although this is none too ap- 
 parent. Across this topography the deep basin of the lake extends 
 as a part of the peculiar depression in which the "Inland Route" 
 lakes lie. Curiously, the deepest part of Mullet Lake is situated 
 in the constricted central portion, whereas the broad expanse at 
 the north end is relatively shallow. Northeast of a line connect- 
 ing Dodge and Needle points, the water rarely exceeds thirty feet 
 in depth and is, furthermore, marked by several shoals more or less 
 in line with these two points. The drop to deep water is gradual 
 towards the southwest until a deep trough which runs , along the 
 narrow part of the lake is reached. This trough is in excess of one- 
 hundred feet deep throughout and extends southwestward to a 
 steep upward slope which follows a sinuous course northwesterly 
 from McArthur Point. It is well defined and is bounded by steep 
 slopes on all but the northeast side. Several "holes" exist in the 
 
8S INLAND LAKES OF MICHIGAN 
 
 bottom of the lake, the deepest lying off Long Point and reaching 
 a depth of one hundred forty-five feet. Another is Scotts Bay 
 which drops to more than eighty-five feet. 
 
 The origin of the depression in which this chain of lakes lies has 
 been discussed at the beginning of this chapter and in Chapter I. 
 The pecularities of the shore line and the bed of the lake are ex- 
 plained by the distribution of the glacial formations. The last ice 
 sheet covered this lake from the northeast and retreated in the 
 same direction. The main depression was in existence previous to 
 the advance of the ice and caused a local advance of the ice front. 
 As the ice retreated over this country its front halted in the vicinity 
 of Eed Pine Point, building a moraine. The ice still filled the lake 
 basin at this time, therefore the moraines do not cross the lake. 
 On the west side of the lake, clay hills are present which are in 
 line with Red Pine Point and the morainic ridge near the north- 
 east end of Burt Lake, making it probable that this line marks a 
 position of the ice front. In width the moraine reaches to Round 
 Point and thus accounts for the narrow central part of the lake. 
 This moraine is not of the distinct knob and basin type for it was 
 deposited under the waters of a lake which washed the ice front 
 and, therefore, shows much less relief. Furthermore, it was covered 
 by Lake Algonquin, and, in addition to the planing off of the hill- 
 tops, a, veneer of sand or clay, depending on the proximity to the 
 shore, was deposited over much of this territory. This moraine is 
 traced with difficulty and some doubt may be expressed as to the 
 correctness of the interpretation given above. It may be that a 
 stagnant block of ice occupied the deep trough of the lake, but at 
 any rate, its border stood at one time at the steep slope running 
 northwestward from McArthur Point. Two moraines also cross the 
 course of the Cheboygan River, running in the same northwest- 
 southeasterly direction. The first moraine crosses the river soon 
 after leaving the lake and is bordered by an outwash plain which 
 extends to the southwest and accounts for the shallow lower end 
 of the lake. The Black River flows between this moraine and a 
 narrow ridge in the vicinity of Cheboygan. 
 
 Upon leaving Burt Lake via Indian River, the amount of artifi- 
 cial control of the stream is somewhat surprising and bears strong 
 evidence of the popularity of these lakes as summer resorts, for the 
 main traffic through them is by resorters and tourists. The long 
 piers, the dredged channel, and the diversion of the Sturgeon River 
 into Burt Lake from Indian River are readily explained when it is 
 realized that the drop from Burt Lake to Mullet is less than one 
 
LAKES OF THE CHEBOYGAN RIVER BASIN 89 
 
 foot, most of whic'li occurs in the lirst' mile aud a half. Also, the 
 valley spreads to a width of more than one-half mile south of 
 Indian River and becomes a swampy mud-flat througli which Indian 
 River meanders. The stream was unable to keep a channel open at 
 its mouth and in its lower course, which was undoubtedly a shal- 
 low arm of Mullet Lake now filled by the silt carried down by 
 Sturgeon River before its diversion. 
 
 At the entrance to Mullet Lake proper a striking shore feature 
 presents itself on the west side. Currents swinging along this shore 
 from the north have deposited their load of sand in a long, narrow 
 spit which extends fully half way across the opening. At the end 
 it turns back abruptly, and many of the trees which line it stand 
 in water. These facts indicate an accident in the history of the 
 lake, — the artificial raising of the water level by a dam at Cheboy- 
 gan. Under normal conditions the spit would have continued 
 straight across the indentation, forcing Indian River to the ex- 
 treme south side. Under the present conditions the spit must be- 
 come adjusted to the higher level, and the probable course of events 
 will be a slow increase in the irregularity of its contour until the 
 trees are removed. Then the work will proceed more rapidly, and 
 the bar be re-formed farther back in the swamp, probably in line 
 with the sharp point on the opposite bank, as indicated by the di- 
 rection of the hook. Mullet Lake illustrates excellently the re- 
 newal of activity on lakes whose level has been raised, and further 
 evidence may be found within sight of the bar in the freshly cut 
 cliff to the west which is pounded by the waves driven by the power- 
 ful north and northeasterly winds. 
 
 Along this shore the rolling topography is covered with sand, 
 but, where sections are exposed, boulder clay, or till, is usually ex- 
 posed underneath. The till is seen on the cliffs and is much more 
 resistant to wave action than the sand which comes to the shore in 
 the depressions, giving rise to small projections of the shore, as at 
 Cold Springs. The cobbles and boulders of the beach, indicate 
 till from which the finer particles have been removed. North of 
 Cold Springs a depression extended below lake level and has not 
 only been isolated by a bar but filled. Across this extinct lagoon 
 the Michigan Central R. R. has built an embankment. The raised 
 water level is very evident on this shore from the presence of great 
 quantities of driftwood and from the trees whose roots are washed 
 by the waves. 
 
 Near Topinabee, however, the slopes of the main depression in 
 which this lake lies approach the shore, and the banks are some- 
 
90 INLAND LAKES OF MICHIGAN 
 
 what liigher. Tlie slopes are composed of till and much coarse ma- 
 terial is found on the beach, which has been |)ushed into a feeble 
 rampart or lined along the shore by ice action. Topinabee is one 
 of the important resorts of the lakes of the ''Inland Route" and 
 there is a geographic reason for this. Along the shore are found 
 a series of terraces and cliffs which mark the higher levels at which 
 the lake formerly stood. The diagram of the terraces on Burt 
 Lake, Fig. 39, will perhaps give an adequate idea of the relations 
 of these terraces. Next the shore a low cliff is found locally which 
 is receding into a terrace about four feet above the present level. 
 This terrace reaches a width of fifty feet at Topinabee and is 
 flanked on the land side by a low grass-covered slope, the bottom 
 of which marks the shore line at the time when the lake stood at 
 this level. Above this cliff is a much wider terrace which gradually 
 rises to a height of sixteen to seventeen feet above the lake and 
 ends abruptly in a steep cliff more than forty feet high. This cliff 
 and terrace were formed iby Lake Nipissing and are continuous 
 with those found at approximately the same level on Burt and 
 Crooked lakes. Near the top of the high cliff is a slight notch, in- 
 dicating a level of short duration. Ahove this stands a broad ter- 
 race upon which much of the town is built. The highest terrace 
 terminates at the base of a cliff and was formed by the waves of 
 Lake Algonquin. Quite generally the terraces are sandy, the re- 
 sult of the action of the undertow, and the cliffs are in clay and 
 stand at a steep angle. The sandy character and the nearly level 
 surface of the terraces insures dryness and affords excellent loca- 
 tions for buildings. The Nipissing terrace is the usual choice of 
 location for summer homes, on account of its proximity to the 
 lake and the excellent water supply derived from flowing wells. 
 Another factor is that the railroad has taken advantage of this 
 level strip which persists the entire length of the lake, making the 
 resorts readily accessible. 
 
 Such are the terraces that practically surround Mullet Lake and 
 much of the interest from our viewpoint centers around them. As 
 already stated, the Algonquin and Mpissing shores are continuous 
 around Burt and Mullet lakes, the former standing well back and 
 above the present lake. That is to say, there was a continuous body 
 of water in this region, and the tracing of its shores with their 
 varied topographic forms is a profitable and pleasing study. The 
 level below the Nipissing, which we shall call the Post-Nipissing 
 stage, stands at about the same level as on Burt Lake, four feet 
 above the present, but the lakes were probably separated by a bar 
 
Michisiin Goological and 
 Biological Survey 
 
 riibliciifioii .'lO. Ocolofiiral Scries 23, 
 Plate VI. 
 
 A. NIGGER CREEK, MULLETfT LAKE. 
 
 B. STOXEY POIX'T, MULLETT LAKE. a. NI1'ISSL\<; TERRACE, b, ICE 
 RAMPART, c, POST-XIPIiSSIN'G TERRACE. 
 
LAKES OP THE CHEBOYGAN RIVER BASIN 91 
 
 at Indian Kiver. The I*ost-Nipissing stage was of relatively short 
 duration and the terraces are narrow or absent except at the ends 
 of the lake. The absence of this terrace in many places indicates 
 that it has been destroyed during the succeeding stages. 
 
 Northeast of Topinabee the Post-Nipissing terrace widens and is 
 followed by the railroad. It slopes gently to the water's edge where 
 the trees are washed by the waves. The inefficiency of the waves 
 here is due to the protection afforded by the vegetation and the 
 lee position of the shore with reference to the prevailing strong 
 winds. Conditions soon change, and south of Nigger Creek the 
 Post-Nipissing terrace has been entirely removed. In its place is 
 a cliff in the outer edge of the Nipissiug terrace. The section af- 
 forded by thi's cliff shows till covered by stratified sand and 
 furnishes the key to the nature of the terrace. At first waves were 
 active and carved a cut terrace but, as the cutting advanced land- 
 ward, more and more of the outer portion of the terrace came 
 into the zone of deposition by the undertow, hence the covering of 
 sand. Beyond the cliff the Nipissing terrace recedes from the 
 shore, following the depression in which Nigger Creek flows, and 
 a narrow terrace of the Post-Nipissing stage is present at the 
 shore. Currents are active at the entrance of Nigger Creek and 
 have built spits from both sides at the present level. The spit 
 on the north side is the better developed and has forced the stream 
 to make a sharp bend to the south before entering the lake. Near 
 the lake, Nigger Creek is. an almost stagnant pool, Plate VI, A, 
 which is being filled with hydrophytic (water loving) vegetation, 
 through which stand trees with submerged bases. This condition 
 is due to the raised water-level and will soon kill the trees, making 
 this an unattractive, mosquito-breeding swamp. For about a 
 quarter of a mile north of Nigger Creek the railroad embankment 
 interrupts the natural contour of the shore but runs farther in- 
 land at the blunt point bej'ond. 
 
 Along this point the greater efficiency of the northerly winds is 
 again apparent. On the southern side of the point the Nipissing 
 cliff' is perhaps a thousand feet from the shore, the intervening 
 space being occupied by both the Nipissing and Post-Nipissing ter- 
 races. Along the present shore a well-developed rampart has 
 been pushed up by the ice to a height of four feet, one of the 
 strongest on this lake. The northeast side of tlie point is quite in 
 contrast to this, for the waves have reduced tlie low terrace and 
 are actively cutting into the Nipissing, exposing fresh cliffs of 
 boulder clay. In the bay between tliis point and Long Point, both 
 
92 INLAND LAKES OF MICHIGAN 
 
 the iS'ipissing and Post-Nipissing terraces are present and are 
 relatively wide. A railroad embankment obscures, the conditions 
 along the present shore but this wide indentation probably was 
 never cut off by a bar. Near Long Point a small ice rampart indi- 
 cates a moderate amount of ice push. 
 
 Long Point is interesting in that it is a region where wave action 
 has been excessive with pnactically no evidence of deposition ex- 
 cept on the sub-aqueous terrace. The Nipissing and Post-Nipis- 
 sing terraces are present on both sides and contrary to expecta- 
 tion, are better developed on the north side. Combined they 
 reach a width of more than one-fourth mile, but off-shore is a sub- 
 merged terrace of almost double this width. The submerged ter- 
 race is well defined all along the northeast shore facing the deep- 
 est portion of the lake and drops off at about twelve feet, slightly 
 lower than in Burt Lake. Both lakes are similar in size and 
 shape and the greater depth of the "drop-off" in Mullett is to be 
 ascribed to a greater rise in the water level than to any consider- 
 able difference in the force of the erosive agents. This is also 
 shown by more active cutting on the present shores of Mullet. 
 
 From Long Point to Dodge Point the shore is comparatively 
 straight with the exception of a shallow indentation south of 
 Hiawatha Beach. Conditions are very uniform along this stretch, 
 the features consisting of the Nipissing and Post-Nipissing ter- 
 races which are relatively constant in width and extend more than 
 a quarter of a mile back from the shore as a rule. The blunt 
 point near Hiawatha Beach is an exception. Here a hill of resis- 
 tant material has increased the work of the waves, and, although 
 the terraces, exposed and submerged, are well developed, the pro- 
 jection of the shore line reflects the difficulties encountered. For 
 a similar reason the shore projects slightly at Silver Beach but, 
 in this case, the cause is an accumulation of large boulders. For 
 the most part the shores are sandy but are often obscured by drift- 
 wood and vegetation growing in the water. In a low cliff below 
 Hiawatha Beach and at other places along the low shore wherie 
 trees have been uprooted, accumulations of marl are present, fur- 
 nishing a hint as to one method by which the lake is being filled. 
 Ice action is effective along the shore and has piled up ramparts 
 at various places, notably near Hiawatha Beach. Even on the 
 low sandy shores small ramparts are found, but always where 
 vegetation acts as a binder. 
 
 The shores are somewhat higher and the beach is of clear sand 
 along the sharp bend in the shore line towards Dodge Point. 
 Favorable shore conditions and the protection from storm winds 
 
LAKES OF THE CHEBOYGAN RIVER BASIN 93 
 
 afforded I>y the ])oint make tliis an ideal location for the summer 
 resort of IMiillet Lake. Also the topography of the point is sucli 
 that it is not necessary for the railroad to follow the shore, and 
 the inconvenience of the tracks and danger of accidents is partial- 
 3y eliminated. One of tlie landmarks of this part of the lake is the 
 sliarp knoll above the point crowned with a clump of pines which 
 are elsewhere lacking. The Nispissing terrace surrounds this hill on 
 all sides except the northwest, where the island rises just enough 
 to make it uncertain as to whether this height was an island or 
 a narrow-necked peninsula at that time. By using this sag, the 
 railroad is able to keep its tracks straight and at the same time 
 follow the terrace. Ou the lake side, the knoll was cut into 
 a steep cliff at the foot of which lie quantities of coarse beach ma- 
 terial. The beach pebbles have been quarried to some extent but 
 their use was not ascertained. The Nipissing terrace is broad 
 and near the lake is sandy, furnishing excellent sites for the build- 
 ings of this deservedly popular summer resort. The Post-Nipis- 
 sing level is here represented by a terrace which does not exceed 
 fifteen feet in width and whose edge is pushed up into an ice ram- 
 part at the tip of the point. Little, if any, deposition by currents 
 is to be found here. 
 
 Beyond Dodge Point, the Nipissing terrace fringes the hills 
 which run to the northwest and is narrow, but the Post^Nipissing 
 terrace widens and extends around the foot of the lake. This low- 
 er terrace stretches along the course of the Cheboygan River in 
 a V — and ends in a low outwash plain immediately in front of a 
 narrow moraine, the hilltops of which were bevelled by the waves 
 of Lake Mpissing. The evidence of this is to be seen in the river 
 banks where stratified sandy material gives way to hard clay cliffs 
 about sixteen feet high before Strawberrj- Island is reached. 
 
 The shores on either side of the outlet are low and sandy but, 
 except for local patches, are covered with drowned vegetation which 
 offers passive resistance to the onslaught of the waves. This is 
 well illustrated along the shore in one locality between Dodge 
 Point and the outlet where the vegetation has been cleared from 
 the shore. The result, shown in Fig. 40, has been a recession of 
 the shore line of forty to fifty feet but unfortunately the time dur- 
 ing which this was accomplished was not learned. From this we 
 can realize what may be expected from wave action when the trees 
 bordering the shores are killed and removed. 
 
 The Post-Nipissing terrace narrows after leaving the outlet 
 and becomes a narrow strip of variable width along the north- 
 south trending shore of the east side, as far as Needle Point. Its 
 
94 
 
 INLAND LAKES OF MICHIGAN 
 
 width varies with the topography, widening in the depressions 
 and narrowing at the points, and is always flanked by the Nipis- 
 sing terrace which developed to a much greater extent on this shore 
 than on the opposite side of the lake. This is due to some extent 
 to the flatter topography on the east but also to the exposure to 
 storm winds from northerly and westerly directions. This devel- 
 opment of the Nipissing terrace is well shown north of Aloha 
 where its width reaches nearly one mile. Beneath the lake in this 
 shallow portion the submerged terrace is poorly developed, and 
 from Dodge Point to beyond Needle Point the bottom slopes grad- 
 ually to moderate depths. Along this same shore, adjustments by 
 both waves and currents are slight. The broad terraces are sand 
 covered and often are composed entirely of this material on the 
 
 
 Fig. 40. Recession of a flooded shore line due to removal of vegetation, Mullett Lake. 
 
 outer or built portions. Consequently, they are easily removed 
 by the waves and the shore is generally receding except where 
 held up by vegetation. In fa,ct, the recession of the shore is great- 
 ly retarded here both by trees still standing and large quantities 
 of driftwood which line long stretches of the beach. The projec- 
 tions of the shore line are slight and blunt, and are due to irregu- 
 larities in the original topography rather than to differences in the 
 resistance of the material. One exception to this statement occurs 
 at Point A, on the west side of Mullet Lake, see map, Fig. 37, which 
 is lined with boulders and is probably composed of till. 
 
 As in the case of the Michigan Central, on the west side the 
 Detroit and Mackinaw R. S. uses the terraces for its roadbed as 
 far as Aloha. This town is favorably located for resort purposes 
 
LAKES OF THE CHEBOYGAN RIVER BASIN 93 
 
 but is more exposed to storms tluiu locations on tlie west side of 
 the lake. 
 
 From ^'eedle Toint on, the irregulaiities of surface and dif- 
 ferences in resisteuce of the material cause a much more broken 
 shore line, in fact a narrowing of the lake. It is probable that a 
 moraine, laid under water and later covered by Lake Algonquin, 
 crosses or runs to the lake shores here. Needle Point is composed 
 of compact boulder clay which in itself is resistent to erosion au<^ 
 also furnishes many boulders to act as a breakwater. It was for- 
 merly less sharp and extended about eight hundred feet farther 
 out into the lake. The contrast between the north and south sides 
 of the point, in accordance with practically all similar features of 
 the lake, illustrates very strikingly the importance of storm winds, 
 here northerly, in the erosion of the shores. The north side is 
 rapidly being worn back and for a short distance near tlie tip a 
 storm beach has been piled up, enclosing a narrow lagoon. The ti]> 
 of the point is kept sharp by the recession of the north side, and 
 directh" in line with it is a small island which was formerly a part 
 of this point. This is clearly a remnant or outlier and was nevei- 
 a land-tied island, for the remnant of the connection is now a sub- 
 merged boulder ridge. On the south side of the point evidence of 
 cutting at present is not to be found, but instead the beach is of 
 even contour and composed of assorted material which decreases 
 in size with distance from the point, its source. The bay" south- 
 west of Needle Point is bounded by swamp and the shores lined 
 with driftwood, stumps, and standing trees. The beach, where not 
 obscured, is of sand but no indications of a bar were found. 
 
 The broad projection culminating in Round Point is due to hills 
 of resistant clay in proximity to the shore. At the Indian Reserva- 
 tion the Post-Nipissing terrace is obliterated and the waves are 
 now cutting into the Nipissing terrace, exposing boulder clay in a 
 clific eight to ten feet high, A sandy depression to the west ac- 
 counts for the smooth beach of wave-worked material which soon 
 gives way to a knob rising sixty feet above the lake. This hill is 
 flanked by the cliffs and terrace of the Nipissing stage on all but 
 'its landward side and was an island at that time separated from the 
 mainland by a shallow strait, almost duplicating the hill at 
 Mullet Lake Station. On its northern exposure, wave action is 
 excessive and is cutting a clitf in the Nipissing terrace. The tip of 
 the point is low and is a triangular remnant of the Post-Nipissing 
 terrace. It does not show the wear that takes place on either side, 
 and probable some deposition took place here when the lake level 
 
96 INLAND LAKES OF MICHIGAN 
 
 stood lower than at present. Ice action has formed a small ram- 
 part on the tip. 
 
 Along the shore between Round and Stoney points, the Nipissing 
 terrace is again in evidence and the adjustment of the shore is 
 broken only by one minor point of boulders. The Nipissing cliff 
 rises to the high Algonquin terrace a few rods back of the shore. 
 Stoney Point is merely a repetition on a smaller scale of the ma- 
 jority of the points on the lake. The clay of the Nipissing terrace 
 is cut into a cliff six to eight feet high on the north side, but around 
 the point there is little wave action, leaving intact both of the 
 lower terraces. However, the end of the point shows the relations 
 of the different levels so well that a photograph is reproduced in 
 Plate VI, B. Note the beach of coarse material with many large 
 boulders and the till cliff of varying height. Near the end of the 
 point (center of view) the cut terrace of the Post-Nipissing level 
 is present and has been cut into a low cliff by the waves at the 
 present level. This ends abruptly at the left in an ice rampart 
 which contains many large boulders and was formed during the 
 Post-Mpissing stage. Beyond the rampart is the surface of the 
 Nipissing terrace, here in the cut portion. 
 
 The bay between Stoney and Red Pine points almost exactly re- 
 peats the conditions for the bay north of Stoney Point and need not 
 be described. Red Pine Point, however, is an extended morainic hill 
 which compares favorably in height and is in line with the high- 
 land extending beyond Topinabee towards the northeast end of 
 Burt Lake. This is probably an extension of a moraine but did not 
 continue across the lake basin. It is heavily wooded and is alto- 
 gether one of the finest locations on the lake. It is one of the few 
 points that show any tendency towards growth from current action. 
 At the present level a small spit is extending to the northwest but 
 apparently very slowly. The position of the drop-off gives us some 
 idea of what has gone on in the very recent past and shows a much 
 greater deposition than at present. The growth of the spit to the 
 northwest is unique for this lake and requires explanation. The 
 wind directions which may affect this point are about equally di- 
 vided between the tAvo sides, but in violence those from the north- 
 erly quadrant are the more important. Yet the force of the waves 
 tossed by these winds is lessened by their passage across the grad- 
 ually shoaling bottom, but , on the southwest side the submerged 
 terrace is narrow and the waves strike the shore with but slightly 
 diminished intensity. Also the regular shore to the southwest with 
 its nearly continuous cliffs furnishes abundant material and allows 
 
LAKES OP THE CHEBOYGAN RIVER BASIN 97 
 
 the development of a' far more efficient curreut than is possible on 
 the irregular, low shore north of the point. 
 
 The end of the point is the key to the events that have happened 
 here. A fragment of the Post-Nipissing terrace is present whose 
 cliff has been pushed into an ice rampart. Landward from this 
 there is the distinct iXipissing terrace of moderate development. 
 On this terrace, closely paralleling the ice rampart, is a strong bar 
 which runs to the southeast, gradually crossiug the terrace and 
 merging into the Nipissiug cliff. The cliff at the present level on 
 the north side of the point cuts the bar at a sharp angle, furnish- 
 ing an excellent cross section from which the relations are easily' 
 seen. The southeastern side has therefore been a point of departure 
 of currents since the j)oint has existed as such. The Xipissing ter- 
 race is narrow but distinct along this shore, and the cliff above it 
 rises stee]>ly to remnants of the Algonquin terrace on the hill rop. 
 The Post-Nipissing terrace has for the most part been cut away and 
 the waves are now attacking the terrace above, forming cliff's five 
 to eight feet high. 
 
 Scotts Bay is a deep depression and continues to the southeast 
 as a low swamp which supports a heavy grow^th of vegetation. A 
 narrow lagoon has been formed at the present level by the forma- 
 tion of a low storm beach, but the swamp as a whole was probably 
 never cut oil", although it is possible that a bar, thoroughly hidden 
 by vegetation, may exist farther back. The Nipissing terrace swings 
 ■far back around the swamp but reappears again at McArthur Point 
 where hills of boulder clay stand near the lake. This point was orig- 
 inally of gentle slope towards the lake and the waves of Lake Nipis- 
 sing quickly reduced it to an elongated terrace fully a half-mile in 
 length. The depression in which the Pigeon Kiver flows is so 
 baflly flooded that little could be determined as to the shores ex- 
 cej)t on the south side where we leave the lake with the waves 
 cutting back into the familiar Nipissing terrace. 
 
 A reading of the above description has no doubt left the impres- 
 sion that wave cutting is the important work being done on this 
 lake at present. Current action at the present level assumes im- 
 portance onl^' on the west side of the inlet and at Jsigger Creek. The 
 latter probably will be able to maintain a channel through the bar 
 but there is a possibility of greater growth at the entrance of In- 
 dian Kiver. This bar should adjust itself to the higher level and 
 extend to the other side, leaving a gap large enough to accommo- 
 date the flow of the stream. Undoubtedly other adjustments were 
 made but have been destroyed in recent times, as may be inferred 
 from the study of Red I'ine and Stoney points. The effects of the 
 13 
 
98 INLAND LAKES OF MICHIGAN 
 
 liftiug of the water level are excellently illustrated on this lake. 
 The flooded hays and inlets, the fresh cliffs and the trees standing 
 in water, together with great quantities of driftwood which line the 
 exposed shores stand as evidence of this fact. The future develop- 
 ment of the shores of this lake must result from the increased 
 activity of the waves and will consist at first in a recession of the 
 shores. At present this is proceeding somewhat slowly as the 
 shores are protected by vegetation, but it will increase when this 
 protection is no longer available. Adjustment should occur first 
 along the low shores and indentations since here the waves are 
 working in the veneer of sand which covers the entire depression 
 in Avhich the lake lies. Still, such places are regions of deposition 
 rather than degradation, and we may confidently look for a grad- 
 ual building out of the beaches in such places, except in the limited 
 number of bays where currents may leave the shore and form bars. 
 Evidences of ice action on this lake indicate moderate effects. 
 Eamparts are found mainly on the points and are discontinuous and 
 poorly developed. In the bays the material is sand and ramparts 
 are not developed or, if so, are quickly reduced by the waves at 
 the present period of excessive activity. We are uncertain as to the 
 shore features of this lake under normal conditions, and encounter 
 difiiculties in attempting to discuss the relative importance of ex- 
 j)ansion and ice jam. In some of the bays expansion should be 
 active, but in general the lake is too large for expansion and power- 
 ful ice jams are to be expected. 
 
 The agencies working towards the extinction of this lake are 
 apparently making little headway. Filling of any sort is insignifi- 
 cant, especially since the diversion of the Sturgeon Kiver into Burt 
 Lake. Vegetation has made little progress in the main body of 
 the lake on account of the excessive wave action, and there are few 
 localities where it is likely that it can establish itself in the future. 
 Some deposits of marl are present on the shores, it is true, but we 
 can hardly look to this alone to fill such a large basin. The tribu- 
 tary streams are few and as yet have deposited little material. As 
 these streams lengthen their courses, more sediment will be brought 
 to the lake and filling from this source will increase. Aside from 
 a change in climate which cannot be foreseen, there remains the cut- 
 ting down of the outlet. With conditions as they are at present, 
 this is impossible but might succeed in lowering the level to that 
 of Lake Huron, fourteen feet lower, provided the dam at Cheboygan 
 is not maintained. This would bring the level just low enough to 
 expose the present submerged terrace and would not materially 
 change or reduce the size of the lake except at the north end. How- 
 
LAKES OF THE CHEBOYGAN RIVER BASIN 99 
 
 ever, the rising of tlie laud to tlu' iiorlheast ol" the CJreat Lakes, 
 in itself a slow; proces.s, hu-ieases the iuiportance of the incision of 
 the outlet, but an uplift of seventy-live feet or more is necessary 
 if the lake is to be drained. 
 
 BLACK LAKE 
 
 Slightly over three miles from its mouth the Cheboygan River 
 divides, one branch connecting with Mullet Lake and the other 
 taking a southeasterly direction. Some difference exists as to the 
 name of the latter and it is designated on different maps as the 
 Cheboygan and as the Black River. The same is true with reference 
 to a large lake which is drained by this river, situated some ten 
 miles above the forks. This question has been referred to the 
 United States Board on Geographic Xames and we will follow its 
 decision bj' using Black for both the river and lake. 
 
 Black Lake is somewhat elongated in a northwest-southeasterly 
 direction and has a length slightly greater than six miles. Its 
 greatest width is approximately three and three-quarters miles 
 and its area fifteen and seven-tenths square miles. The exact eleva- 
 tion of the lake is not known but is estimated at six hundred forty 
 feet above sea-level or forty-five feet above Mullet. The shores are 
 of relatively even contour, as compared with the other lakes of 
 this system, and are^ noticeably interrupted only where the Upper 
 Black River enters on the southwest side and at the quarry near 
 Bonz Resort on the south side, see map. Fig. 41. The topography 
 of the surrounding country shows a tendency towards a northwest- 
 southeast trend caused by the dej)Osits of the glacier which occu- 
 pied the basin of the northern part of Lake Huron. Much of the 
 northeast side of the lake is bounded by highland which varies in 
 height and distance from the lake and continues in the same gen- 
 eral direction beyond the southeastern end of the lake. On the op- 
 posite shore the highland runs along the south end approximately 
 parallel to the cliffs of the northeastern side but is composed to some 
 extent of hard rock which outcrops at the quarry near Bonz Resort. 
 This highland is broken by an extensive depression, through which 
 the Upper Black* River flows, and does not reappear until near the 
 outlet. If the directions assumed by the lake itself, its outlet, and 
 the inlets at the southeast end are taken into consideration, the 
 northwest-southeast trend is rather striking, and there is a tendency 
 to attribute the basirf to a sag between the fragmentary morainic 
 ridges which trend in this direction. However, the presence of hard 
 rock outcrops and the broad depressions in which the Upper Black 
 and Mud Creek flow make the problem much more complex. In 
 
100 
 
 INLAND LAKES OF MICHIGAN 
 
 addition, the lake has not been systematically sounded and, al- 
 though probably deep, little is known of the nature of the basin 
 covered by the lake. Therefore, it seems best to leave the origin 
 of the basin as unsettled until suflScient data are known. 
 
 This region was covered hj the waters of Lake Algonquin which 
 stood more than one hundred feet above the level of Black Lake. 
 The sands and clays deposited under water at this time cover 
 the land surfaces in the vicinity of the lake, and the sands, especi- 
 ally, have been worked over subsequently by waves and the wind, 
 furnishing many interesting features. Black Lake stands above the 
 
 OuHei- 
 
 Pig. 41. Outline map of Black Lake, Cheboygan and Presque Isle Counties. 
 
 level of the Xipissing beaches and, therefore, was not a part of 
 that lake. Two levels higher than the present may be clearly 
 recognized along the shores but must be referred to transitory 
 stages of the Great Lakes while the water was dropping from the 
 Algonquin to the Nipissiug level. This is certainly the case for the 
 higher level but other causes may be advocated for an intermediate 
 level about four feet above the j^resent. 
 
 Black Lake lies some distance from the railroad and compares 
 unfavorably with some of the other lakes of the system in this 
 respect. It may be reached either from Cheboygan by automobile 
 or from Onaway. The former is the more generally used, although 
 it necessitates a longer journey and lands one at the outlet. The 
 
LAKES OF THE CHEBOYGAN RIVER BASIN 101 
 
 first impression the physiogi'ai)her gets is that the lalce is flooded. 
 The virgin hardwoods on the low flat that borders the river stand 
 in Avater near the stream, and. the same conditions hold for the 
 small islands wliich rise barely above the lake jnst off the outlet. 
 This flooding is probably caused by a dam thrown across Black 
 River six miles beloAv, wliiili has a head of eighteen feet. 
 
 The flat which lines the river rises gently to low cliffs on either 
 side but is more extensive on the north side. This flat is inter- 
 preted as the terrace of a higher water level and the conclusion 
 may be verified at almost any point on the shores of the lake. It 
 will be referred to as the Upper Level. The flat continues on the 
 north side and shows little of interest except the marlj' constitu- 
 tion of the shores, indicating a method of filling. 
 
 The irregular shore which fronts this terrace ends abruptly at 
 point A, see map, Fig. 41. This point is a sharp spit which con- 
 tinues approximately a quarter of a mile outward into the lake as 
 a' distinct submerged bar in line with the shore of the east side of 
 the spit. The material on the west side of the point is pebbles, but 
 that of the east is sand.. It is evident from this that the waves are 
 especially active on the west side and that currents have built the 
 fine sand beach of even curvature on the east. It may be inferred 
 that the waves are more powerful on the west but this is not the 
 case. It is the relative strength of wave and current action which 
 determines the character of the beach. Southeasterly and southerly 
 winds, although less powerful than the westerlies, have a long fetch 
 at this end of the lake and set up strong currents on the northeast 
 shore, causing deposition to predominate over cutting on the east 
 side of the point. On the west side waves of less power have been 
 able to throw up a strong storm beach because the shore is not 
 choked b}^ debris carried by currents of very limited development 
 along the irregular shore to the west. Inasmuch as the configura- 
 tion of the bottom of the lake is not known, no cause for the current 
 leaving the shore at this jDoint is advanced. 
 
 East of the point the sandy shore sweeps in a smooth curve to 
 point B. This point seems to have thrown the currents out a short 
 distance from the shore and a low bar enclosing a narrow lagoon 
 has resulted. This swampy lagoon terminates at the foot of a steep 
 bluff, which gradually approaches the north shore of the lake from 
 the northwest, and marks the shore of the Upper Level. This level 
 shows that considerable adjustment in the outline of the lake oc- 
 curred, and in many places the drop to the present level has ex- 
 posed a terrace of considerable width which is well shown along this 
 shore. 
 
102 INLAND LAKES OF MICHIGAN 
 
 Point B is due to a local accumulation of boulders and imme- 
 diately gives waj to a sandy terrace. The eft'ects of the flooding of 
 the lake are here apparent in the increased activity of the waves 
 which have cut the edge of the terrace into a low cliff. This cliff 
 is rapidly receding, necessitating the building of breakwaters. Near 
 point B the numerous springs issuing from the high cliff cause a 
 swampy condition of the terrace, although it is well above the 
 water level. Where dr^-, the terrace proves a suitable location 
 for summer cottages which are furnished with excellent drinking 
 water from the springs. The cliff back of the terrace has its great- 
 est development in this locality and rises generally sixty feet or 
 more above the lake. At the top of the cliff is a flat terrace formed 
 during the existence of Lake Algonquin whose shores stood some 
 distance to the northeast. Above this level now stand a few sand 
 dunes which, in some cases, form j)art of the cliff, making a sheer 
 drop of nearly one hundred feet, the highest on the lake. 
 
 To the east along the lake shore the highland drops suddenly to a 
 low swamp and beyond this is a lowland composed of clay hills with 
 infrequent sags. The terrace is narrow and the shore lined with 
 small boulders, the product of selective wave action on the boulder 
 clay. Beyond Roberts, the hills recede sharply and the low terrace 
 widens into a swamp which encircles the southeastern end of the 
 lake. A stream entering the lake in this vicinity has been turned 
 to the southeast and shows the predominance of westerly winds on 
 this shore. At the Upper Level the eastward moving current left 
 the shore and built a complete bar from the mainland to a narrow 
 island which lies adjacent to the present shore near the bend to the 
 southwest around the upper end of the lake. 
 
 The northwest shore beyond point A has been the scene of intense 
 wave action throughout the history of the lake, as is shown by the 
 preponderance of cliffs which in places are prominent features of 
 the landscape. In addition, the submerged terrace is well developed 
 and was estimated at fifteen hundred or more feet in width. It 
 drops quite regularly to deep water from depths of eight feet, 
 except in one locality a short distance northwest of Roberts. The 
 soundings here showed a terrace which slopes gradually outward 
 to eight feet upon the outer edge of which was found a submerged 
 bar three feet in height. Such a ridge is probably due to the violent 
 agitation of the water where the incoming storm waves first break 
 and may be the forerunner of a barrier. The depth of the water 
 at the "drop off" is certainly less than one-half the wave length 
 of the storm waves on this shore but cannot be taken as indicative 
 of any relationship between the two factors, since uncertainty exists 
 
LAKES OF THE CHEBOYGAN RIVER BASIN 103 
 
 as to tlie coiiipleteiie.s.s of adjustment to the i)roseiit level. Iiias- 
 iiiiu'li as the adjustments were so great during the U])i)er Level, 
 it is probable that the submerged terrace conforms more nearly to 
 conditions at tliat time than to those existing now. 
 
 The low ground through which the Kainey river flows extends to 
 the southeast and is an exposed sandy terrace. At the shore a sand 
 beach curves eveuly towards the limestone blutf east of Bonz Re- 
 sort, and below water the terrace slopes gradually out to a well 
 defined "drop off" more than one-fourth mile off shore. The ex- 
 posed terrace back of this low ground and beach is poorly drained 
 except on two sand bars which are best seen where the river cuts 
 through. Both bars are attached to the shore near Bonz Resort 
 point and run to the north around the end of the lake. They di- 
 verge somewhat as they leave the point and finally play out beyond 
 Rainey river. The bar nearer the lake stands at a level which 
 corresponds to the Upper Level along the northeast shore. Back 
 of it is a narrow swamp above which rises the second bar at a 
 slightly higher level. This bar clearly indicates a water level inter- 
 mediate between that of Lake Algonquin and the Upper Level. This 
 probabl}"^ was a transitory stage and, owing to the lack of accurate 
 elevations, it seems best not to attempt to show its relation to any 
 of the Great Lakes stages. 
 
 In the early stages of the highest level the lake was of much 
 greater extent with large bays at the southeast end, west of Bonz 
 Resort, at the Upper Black basin, and at the outlet. Adjustments 
 of the shore were few and incomplete. Shore action was of great- 
 est intensity in the Rainey River bay but here the bar was not com- 
 pleted. ]S^o such features were found in the other bays. It is in- 
 teresting to note that the development of the bars in the Rainey 
 River bay was from the west. The reason for this lies in the con- 
 figuration of the shore rather than in the difference in exposure to 
 storm winds. At this time the northeast shore was irregular and 
 gave little opportunity for the development of currents. On the 
 other hand, Bonz Point was the scene of great wave action and 
 furnished abundant material. 
 
 As the point near Bonz Resort is approached, the material on the 
 beach rapidly increases in size and the shore becomes rocky. 
 The exposed terraces gradually reduce in width and are very 
 definite. The Upper Level shore is here surmounted by a rock cliff 
 ten to fifteen feet high above which stands the flat terrace and 
 cliff" of the highest level. The point is caused by the only exposure 
 of hard rock found on the lake sliores, a closely fractured limestone. 
 It stands in a bold cliff thirtv or more feet high, which does not 
 
104 INLAND LAKES OF MICHIGAN 
 
 come to the water's edge, inclicatiug its formation at the higher 
 levels of the lake. The cliff shows no indication of the highest level, 
 but the weathering of the closely fractured rock would have quickly 
 obscured the poorly developed terrace that may have been formed 
 during this short-lived stage. Along this shore evidence of ice push 
 is seen in the line of large rocks along the present strand. This 
 point is in reality a hard rock ridge which runs northwest-southeast 
 and formed a promontory during the higher levels. To the west 
 stood a broad haj which was separated from the lake by a complete 
 bar during the Upper Level. This shallow lagoon was soon filled 
 with vegetation and now exists as a flat swamp above which rues 
 Ihe bar near the present shore. This bar is the only dry ground 
 Jiear the lake in this vicinity and upon it are built the cottages of 
 Bonz Kesort. The maximum Avidth of the swamp is about one- 
 fourth mile and it is bounded by cliffs on the land side. To the 
 west it narrows and the cliffs stand nearer the shore. At locality C 
 (see map) they come within one hundred feet of the shore and 
 form the vs'-estern attachment of the Bonz Kesort bar. The cliffs 
 again recede and another bar continues towards the outlet of the 
 Upper Black. This bar also stands at an elevation corresponding 
 to the shore of the Upper Level along the northeast shore and the 
 lower bar in the Rainey River embayment. It lies some distance 
 back from the shore and splits to the north, assuming the form of a 
 large hook rather than that of the simple spit. 
 
 The bars run almost to the present channel of the Upper Black 
 on the surface of a broad delta. This delta causes a large projection 
 of the shore line and is one of the best examples of this feature 
 to be found in Michigan lakes. The river reached the lake by a. 
 series of distributaries, some of which still flow during the flood 
 season. The effective currents along this shore are northerly, 
 formed by the easterly and southeasterly winds of long sweep, and 
 have caused the unsymmetrical development of the delta towards 
 the northwest. At present, the shores are being cut away and the 
 material shifted towards the west, turning the present channel of 
 the river in this direction. The movement of the material is so 
 rapid that it is necessary to keep the channel of the river open 
 artificially. The submerged portion of the delta is correspondingly 
 large, and the submerged terrace consequently reaches its greatest 
 development here, fully a half mile in width. 
 
 Beyond the delta the low ground persists nearly to the outlet as 
 a swamp. When the lake stood at the higher level this was a 
 locality of great current action, and the results are to be seen in a 
 series of bars standing near the shore on the exposed terrace. Near 
 
LAKES OF THE CHEBOYGAN RIVER BASIN 105 
 
 the delta a single bar cut oti" the swamp to the west, but this bar 
 splits three times in its course to the uorth, forming four distinct 
 bars. They are especially well developed north of the mouth of Mud 
 Creek and lie within forty rods of the shore. The direction of the 
 currents, as shown by the bars and the deflection of Mud Creek 
 to the north, conformed to the general direction along the south- 
 west shore. None of the bars along this shore were complete, but 
 the best developed is that standing next the present shore and 
 reaches to within a few yards of the high ground near the outlet. 
 The end of this bar forms the small hook in the present shore line 
 just south of the islands. 
 
 The half mile of shore south of the outlet is bordered by a low, 
 swampy terrace above which a cliff rises to high land. Wave action 
 was active here at the time when the bars to the south were being 
 deposited, and a terrace of moderate width was formed. The small 
 islands at this end of the lake are all flat-topped and stand at a 
 level slightly above the lake. It seems probable that they were 
 small islands during the early stages of the Upper Level but were 
 completely bevelled during that stage. 
 
 In brief, we may state that Black Lake first came into existence 
 as a separate body of water during the recession of Lake Algonquin. 
 A high level is recorded in one locality only, the land form being the 
 higher spit attached to Bonz Resort point and extending east around 
 the Rainey River bay. The lake at that time was much more irregu- 
 lar in outline and larger in size than at present. The lack of ad- 
 justments of the shoreline indicates that this level was of short 
 duration. Following this the lake halted at a level a few feet 
 higher than the present, which we have called the Upper Level. This 
 level was probably the most important one for the lake, and the 
 shores were maturely adjusted to the waves and currents. Great 
 bays were separated from the main body of the lake by currents and 
 bold cliffs cut by waves. In addition, the Upper Black deposited 
 great quantities of silt forming a large delta at its mouth. In 
 general, the present outline of the lake was determined at this 
 time. 
 
 Only minor adjustments have been accomplished since the lake 
 receded from the Upper Level; in fact, there remains little to be 
 done. Of recent years, the ponding of the waters has increased wave 
 action and some readjustment will be the result. It should be 
 largely in the form of cutting back the exposed terrace of the Upper 
 Level. - This will continue until equilibrium is established and 
 should progress rapidly, since the material is largely sand. The 
 small points will be reduced more and more by wave action, but 
 
106 INLAND LAKES OF MICHIGAN 
 
 the delta of the Upper Black will continue to increase to the north. 
 The growth of the delta must eventually fill the lower end of the 
 lake, after which two possibilities arise: A shifting of the distribu- 
 taries may pour the silt into the main body of the lake and the 
 filling proceed without interruption, or the Upper Black may con- 
 nect directly with the outlet. The latter seems the more probable 
 to the writer, since the delta is growing in this direction. Filling 
 •by marl or vegetation must be limited to the shallower portions 
 of the lake and will not be important until a great amount of filling 
 by other means has taken place. Another factor in the extinction 
 of the lake is the cutting downward of the outlet. This has prob- 
 ably caused the drop from the Upper Level to the present, and, if 
 unimpeded artificially, would eventually lower the lake level to 
 that of Lake Huron. Inasmuch as the depths are not known, it is 
 impossible to state whether this would completely drain the lake. 
 At present, the filling of the lake b}^ the silts of the Upper Black 
 River is the most important. 
 
 DOUGLASS LAKE 
 
 Douglass Lake lies on the western border of Cheboygan County 
 in Munroe township, T. 37 N., R. 3 W., and is about fifteen miles 
 due south of Mackinaw City. A mile and a half to the south is 
 the north shore of Burt Lake whose level lies one hundred eighteen 
 feet below that of Douglass. It is reached from Topinabee on the 
 Michigan Central R. R. or from Pellston on the G. R. & I. R. R. 
 by a drive of several miles over a pine '^slashing", now grown up to 
 poplar and associated trees. At present, it is the home of the sum- 
 mer stations maintained by the department of Surveying and of 
 the Biological Sciences of the University of Michigan, whose camps 
 are located on South Fish Tail Bay. In addition to the University 
 camps, there are several resorts, so that the lake is fairly well popu- 
 lated during the summer months but less so than some of the more 
 accessible lakes in the vicinity. 
 
 Douglass Lake stands at an elevation of seven hundred thirteen 
 feet above sea level and one hundred thirty-two feet above Lake 
 Michigan, into which it drains. Its greatest length is somewhat 
 less than four miles, and greatest width does not exceed two and 
 one-half miles, the area totaling 6.2 square miles. Two constrictions 
 appear in the outline of the lake which divide it into three basins 
 united by broad connections. However, if the configuration of the 
 bottom is considered these basins are not so evident. The western 
 end is a true basin which draps to a depth of eighty feet, but the 
 central portion is less than thirty feet in depth and would hardly 
 
LAKES OF THE CHEBOYGAN RIVER BASIN 
 
 107 
 
 be called a basin. Tlie eastern arm is peculiar in slia]>e and con- 
 tains two deep holes in North and South Fish Tail Bays which 
 connect with a pit otT Grape ^'ine l*oint and are separated from 
 each other by a broad shoal which extends to the eastern end of the 
 lake. The greatest depth is eighty-five feet and occurs in South 
 Fish Tail Bay. 
 
 The material surrounding the lake is all of glacial origin and 
 is composed of sand, except at the headlands. These headlands are 
 caused by till, which is much less readily attacked by the waves, 
 and it will be seen from the map, Fig. 42, that, in general, they are 
 
 DiO'^enzs PoJnf 
 
 
 J Milt 
 
 Fig. 42. Outline map of Douglass Lake, Cheboygan County. Broken line indicates 
 approximately the edge of the off-shore terrace. (After U. of M. Surveying De- 
 partment map.) 
 
 opposite each other. There seem to be two small till ridges here 
 which cause the constrictions in the outline of the lake but do not 
 persist across the basin unless possibly in the case of the more 
 westerly. On either side and between the ridges are heavy deposits 
 of sand which partially filled the depressions except where the 
 lake now lies. The eastern end of the lake is surrounded by outwash 
 but the sands of the central and western basins, although possibly 
 outwash, were deposited, in part at least, on the bed of Lake Al- 
 gonquin which formerly covered this region. 
 
 The basin of Douglass Lake lies in the region covered bv the ice 
 
108 INLAND LAKES OF MICHIGAN 
 
 from Lake Huron wliich moved in a soutliwesterly direction in tliis 
 localitj^, as shown by the northwest-southeasterly trend of the 
 moraines. A large morainic tract lies to the northeast of the lake 
 and on the southwest the Colonial Point moraine of Burt Lake 
 extends to the vicinity of Fairy Island, so that, in a way, 
 the basin is situated between moraines. Yet the peculiarities of the 
 lake, both as to form and basin, cannot be accounted for in the 
 simple inter-morainic type of basin which is usually more or less 
 regular in outline and shallow in depth. The proximity of moraine 
 on the northeast and the presence of outwash at the eastern end of 
 the lake, however, lead to the conclusion that the basin was caused 
 by the burial and subsequent melting of one large but very irregu- 
 lar block of ice or three separate blocks of which that occupying the 
 central position was relatively thin. 
 
 One of the striking physiographic forms to be seen near this 
 lake is a well-defined cliff and terrace about twenty feet above the 
 present level. It is not continuous but appears at varying distances 
 from the shore on the higher elevations which have been planed 
 off in some cases and have fiat tops. The elevation of the base of 
 the cliffs corresponds with that of Lake Algonquin in this region, 
 and the general distribution of the beaches of that lake shows 
 that Douglass Lake was at that time a depression in the bottom of 
 one of the inlets of a great archipelago. The Nipissing beach which 
 appears commonly on the shores of neighboring lakes is not present 
 here, Douglass Lake being more than a hundred feet above its 
 level. This lake, then, must have come into existence with the 
 subsidence of the waters of Lake Algonquin and the present shore 
 features are due to the forces which have been acting since that 
 time. 
 
 In South Fish Tail Bay the material of the outwash plain is easily 
 eroded, and a clearly marked cliff and terrace stand back of the 
 present shore at the Algonquin level. The present beach is gravelly, 
 indicating rather strong current action, and beneath the water a 
 built terrace of sand extends outward a short distance, dropping 
 suddenly into deep water at a depth of four feet. Wave lengths 
 of three and four times this figure are common on the lake, thus 
 making the rather low value of one-third to one-fourth the wave 
 length as the limit of effective transportation by the undertow. 
 There is, however, a probability that this terrace was formed large- 
 ly when the lake level stood about four feet higher than at present, 
 as shown by elevated beaches, and has not yet been adjusted to the 
 changed conditions. 
 
 Grrape Vine Point to the west is caused by morainic material 
 
LAKES OF THE CHEBOYGAN RIVER BASIN 109 
 
 which is less icadily at lacked by the waves, but this position is 
 exposed to the westerly and northwesterly storm winds and shows 
 considerable cutting. The Algonquin cliff and cut terrace is well 
 developed on the lieadland. The submerged terrace widens at the 
 point and the shoreward portion is clearly formed by wave cutting. 
 Some deposition has taken place on the east side of the point, but 
 the waves are able to swing around the headland for the most part 
 and the point is therefore blunt. Westward the material again be- 
 comes sandy, and the recession of the Algonquin cliff indicates ex- 
 cessive cutting in Algonquin time. The submerged terrace continues 
 wide, and its limits are sharply defined by a line of demarcation 
 where the yellow of the sand gives w^ay to the dark blue of the 
 deep water. 
 
 Along this shore at the present level deposition is taking place, 
 and the shore has been straightened in some instances. Just west 
 of Grape Vine Point is a narrow lagoon a hundred yards or more 
 in length, which has been separated from the lake by a sand bar 
 between one and two feet high. In general, the south shore as far 
 as Bryants is a succession of small projections and indentations. 
 The projections are caused by local accumulations of coarser ma- 
 terial and are marked by cliffs in close proximity to the shore and 
 by gravel or pebble beaches. The beaches show coarser material 
 on the north and northeast sides of the points but change rapidly 
 to sand on the sides facing the west. The north and easterly winds 
 are the more important on this shore because of the protection 
 from the westerlies offered by Fairy Island. The indentations are 
 lined with sand beaches and the cliffs recede from the shore. The 
 bases of the cliffs referred to here are probably washed by the waves 
 at high water stage in the spring along the projections, but in the 
 indentations they stand about four feet above the present water level 
 marking a higher level of the lake in the past. That this former 
 level, which we may refer to as the Upper Level, was maintained 
 for a considerable time, is shown in the rather large indentation 
 a short distance east of Douglass Lake Resort. This indentation 
 v/as entirely cut oft' by a bar at the Upper Level and, w^ith a lowering 
 of the water, dried up and grew up to forest w-hich has since been 
 cut. 
 
 At the Eesort the cliffs run close to the lake and the terrace of 
 the L^pper Level has been cut away. In the small bay to the west 
 they again recede. Along the shore of this bay was noted a small 
 sand spit, rather blunt in shape, which is being built by currents pro- 
 pelled by northeast winds since the "drop off" runs close to the 
 shore on its northwestern side. Fairy Island is a narrow strip of 
 
110 INLAND LAKES OF MICHIGAN 
 
 morainic material which is tied to the mainland by a bar from the 
 west side at low water but j)robabl3'' is not completely attached at 
 high water. At the Upper Level, the connection was less pro- 
 nounced or was not present, for on the projection of the mainland 
 opposite the island a blunt spit- has been built about four feet 
 above the present level which does not extend to the present shore. 
 This sjDit is more in the nature of a V-bar and encloses Bryants Bog. 
 
 The Island presents an interesting profile. The essential feature 
 is a flat top surrounded by a cliff from the base of which a terrace 
 slopes gently to the water's edge. This terrace shows a much greater 
 width at the ends than along the sides. The flat top was planed off 
 by the waves of Lake Algonquin, and the cliffs and terrace below 
 were cut after the water had subsided to the Upper Level. At this 
 level, ice action built strong ramparts which begin at the base of the 
 cliff and extend out on the terrace. The submerged terrace off the 
 north end of the island is wide and was formed largely by wave 
 cutting, as is shown by the large boulders scattered on its surface. 
 However, to the southeast the bench swings outward in a broad 
 curve and is built of sand transported by currents set up by west- 
 erly winds. West of the island the bay is shallow and the bench 
 not well marked. 
 
 Westward from the island the shore is sandy and of perfect curva- 
 ture for perhaps a quarter of a mile. It is in fact a bar which ends 
 in a small hooked spit and behind which is a lagoon supporting a 
 heavy growth of rushes. From the map, Fig. 42, it will be noted 
 that this shore conforms in curvature with the west shore of the 
 island, from which most of the material composing the bar has 
 been derived. The land west of the lake is low and sandy and was 
 covered with water during Algonquin time. During the early stage 
 of the Upper Level this end was considerably greater in extent, but 
 before the water receded to the present level a strong bar developed, 
 forming a large lagoon to the west which is still wet. East of Maple 
 River the bar was built by shore-drift from Fairy Island under 
 northerly winds, or by return currents when the winds were from 
 a more westerly quadrant. North of the river the bar continues 
 around the entire west end of the lake but was built by southerly 
 drifting currents, clearly shown by the spit just north of the river. 
 It is poorly developed around the bay at the northeastern extremity 
 of the lake and cut through by small streams but persists as far 
 as lugleside where cliffs line the shore. The "drop off" in the west 
 arm of the lake is sharply defined, except in the shallow water of 
 the south side, and the terrace is wide. The sandy material in this 
 locality obscures the manner of formation of this terrace but the 
 
LAKES OF THE CHEBOYGAN RIVER BASIN 111 
 
 presence of the great bars along the shore seeins to indicate that it 
 has been bnilt rather than cnt. 
 
 At Ingleside the moraine comes to the shore, forming a point 
 whicli is now being attacked by the waves. The Upper Level ter- 
 race has been obliterated and the liner material of the till carried 
 away, leaving a beach of rather coarse material. The bay between 
 Ingleside and Bentley Point is caused by a sag in the moraine which 
 has been partly tilled with sand. The head of this indentation was 
 completely cut off by a strong bar at the same level as those on the 
 west arm. At the present level a small ice rampart has been formed 
 twent}^ to thirty feet in front of the bar. The shore of this bay is 
 rather irregular at the low water stage of mid-summer, a condition 
 not to be expected in front of a bar. Examination of the materials 
 of the shore, however, discloses the fact that the waves have entirely 
 stripped the sand covering in places from the hard clay of the 
 moraine which holds up wave action, causing minor projections. 
 
 Some peculiar forms built of sand, called cusps, were noted in this 
 bay which, although similar to spits, differ materially and cannot 
 be explained in the same way. They consisted of sharp points of 
 sand built out from the shore at an oblique angle, extending above 
 the surface at low water and continuing outward below the water 
 level as bars. In some cases they turned back to the shore abruptly 
 similar to V-shaped bars, but in all cases their direction was to- 
 wards the median line of the bay ; that is, if extended outward, those 
 on opposite sides of the bay would meet approximately'^ along a 
 line drawn from the head of the bay out into the lake. A possible 
 explanation is that during moderate storms- at Ioav water stage 
 small storm beaches are thrown up over which the waves break. 
 At first the storm beach is continuous and the water collects be- 
 hind it in a narrow lagoon. If more water is supplied to the lagoon 
 by the waves than can seep back through the sand, the level of the 
 lagoon rises and eventually the water flows back to the lake over 
 low places in the beach. Thus, channels are cut through the storm 
 beach, each channel draining a portion of the lagoon. Such chan- 
 nels will be maintained only where the streams are able to over- 
 come the tendency of the waves and currents to obstruct them. The 
 power of a stream is dependent on its velocity and in this case is 
 determined by the amount of water in the lagoon; that is, the size. 
 Since the width of the lagoon is practically uniform, the size is 
 direct!}' proi)ortional to the length. During the early stages of a 
 storm, many such channels may be formed and obliterated, but 
 eventually the lagoon is divided into sections of more or less uni- 
 form length, which are able to maintain an open channel to the 
 
112 
 
 INLAND LAKES OF MICHIGAN. 
 
 lake. The streams in maiutaining these channels are ^constantly 
 carrying out and depositing sand which is worked over by the 
 waves and currents into spit-like forms. See Figs. 43 and 44. 
 
 Fig. 43. Sand cusp, Douglass Lake. 
 
 Fig. 44. Sand cusp, I>ouglass Lake. 
 
 These features are formed when the waves run directly into the bay. 
 At such times the waves enter the bay with straight crests but are 
 retarded at both ends as they progress, causing a curvature of the 
 
LAKES OF THE CHEBOYGAN RIVER BASIN 113 
 
 crests, see Fig-. 45. Ou the sides of the bay the waves strilie the 
 shore obliquely and set up currents running towards the head of the 
 bay, where they merge into the undertow. Thus the cusps point 
 towards the center of the baj-. The fact that such forms occur on 
 flat shores and at low water suggests that the forces of degradation 
 and deposition are so evenly balanced that once the balance is over- 
 come, the predominant force will continue its work. In this case, 
 the deposits made b}' the outlets of the lagoon are able to force the 
 currents from the shore, but at the same time are remodelled into 
 cusps, whose directions conform to the course of the currents. 
 However, at high water and during heavy storms the balance is 
 destroyed and the forms are obliterated. 
 
 ■Fig. 45. Conventional diagram illustrating the increase in curvature of waves 
 within an embayment. Broken line indicates edge of off-shore terrace. 
 
 At Bentley Point wave action is heavy, as is shown by the coarse 
 material of the beach. Currents are also active and have built a 
 spit running to the southwest, which is about one hundred feet 
 long at low water. The reach of the waves is here more important 
 than the strength of the Aviud and the currents from the east are 
 stronger. The broad bay east of Bentley ^practically duplicates 
 the shore features found between Ingleside and Bentley, except 
 that the cuspate forms are not present. The ice rampart here is 
 somewhat better developed and reaches a height of four feet in 
 places. 
 
 The blunt headland opposite Grape Vine Point is caused by a 
 projections of the same moraine and has similar features in general. 
 Cliffs rise from the shore to a flat topped area, somewhat less than 
 twenty feet above the jiresent level, which extends nearly a mile 
 to the north, terminating in the Algonquin beach. This flat topped 
 area is the cut-terrace of that time. On either side of the cliff at 
 Stony Point the terrace and cliff of the Upper Level are present, 
 and upon the terrace spits, formed during this stage, run both to 
 the east and the west. These forms are steep and narrow near the 
 cliffs and are composed of coarse material, including boulders up to 
 a foot in diameter. Farther from the cliffs, the material decreases 
 in size and the spits broaden, reaching widths of nearly' one hundred 
 1.5 
 
114 INLAND LAKES OF MICHIGAN 
 
 feet at the ends Avhere the material is saud. The elevation near the 
 cliffs is in excess of five feet above the terrace but drops to less 
 than three feet at the ends. Currents have been largely instru- 
 mental in the formation of these spits, but ice action has aided near 
 their land attachment where the forms are more characteristic of 
 ramparts than of current deposits. 
 
 Farther to the east, at Sedge Point, the currents again left the 
 shore at the Upper Level and built a recurved spit that cut off low 
 ground to the north, which is still swampy. At the present level 
 the currents are depositing in front of the fossil spit and have 
 built a series of recurved spits which enclose triangular lagoons, 
 as shown in Fig. 46. At Pine Point, a short distance east of Sedge 
 
 / 
 
 Fig. 46. Diagrammatic plan of bars and lagoon at Sedge Point, Douglass Lake. 
 
 Point, a spit similar to that just described apf)ears. Entering 
 North Fish Tail Bay the shore turns abruptly to the northwest and 
 the currents, being unable to follow the shore, have formed a per- 
 fect example of a compound hook which is reproduced in Plate VII. 
 This bay is a deep pool, showing depths in excess of fifty feet, but 
 presents little of interest along its shores until Diogenes Point on 
 the east side is reached. At this point the currents swinging into 
 the bay from the south have deposited a complex series of recurved 
 spits at the present level, which enclose irregular shaped lagoons 
 now being filled with vegetation. A sketch of these is presented in 
 Fig 47, the lagoons being numbered in the probable order of their 
 formation. 
 
 The eastern end of the lake is a long sand beach above which 
 small sand dunes have been piled by the wind. The peculiar 
 widening of the submerged terrace along this shore is of consid- 
 erable interest. It projects lakeward suddenly just below Dio- 
 genes Point and gradually widens until the deep hole in South 
 Fish Tail Bay causes it to double back and run close to the shore. 
 The entrance to North Fish Tail Bay is wider than that of its 
 
Mu'lii,i;-nn Ccolo.uicnl aiul 
 Biologicnl Survry 
 
 rulilicaticm .'lO. <;ciil(i<;iciil Series 25, 
 I'latu VII. 
 
 nn 
 
 Mil lin' iiiii iH nil nil ■ ■■■I 
 
 / 
 
 i/i 1 ll! 
 
 A. HOOK. DOl'GIvASS LAKE. 
 
 B. RAISED BEACHES, PINE LAKE. 
 
LAKES OF THE CHEBOYGAN RIVER BASIN 
 
 115 
 
 c()un(evi);irt to the south, wliicli may aceonnt for the better de- 
 veloped forms along its shores. The wide terrace in this part of 
 the lake is composed of lar<i;e roeks at its outer (Nl^e but slioreward 
 these give ]thu'e to clear sand. The explanation is that an island 
 or at least a shoal, similar to Fairy Island in sha])e and material, 
 but larger in size, existed formerly at this ])laee and has been de- 
 stroyed by wave action which was able to transport the finer ma- 
 terial only. Thus, an accumulation of boulders was left under 
 water at a depth which marks the lower limit of effective wave 
 action, and this part may be considered a cut terrace. The liner 
 material was washed shoreward and completely tilled the depres- 
 sion, making the terrace continuous to the shore. Clearly, the 
 
 Fig. 47. Diagrammatic plan of bars and lagonns at Diogenes Point, Douglass Lake. 
 
 westerly winds both ou account of their strength aud reach have 
 plaj^ed the prominent part in the formation of thvs exceptionally 
 wide cnt-and-built terrace. 
 
 History. During Algonquin time the lake did not exist as a 
 separate basin but rather as a depression in an arm of a great 
 archipelago. With the recession of the w^aters to the Nipissing 
 level, Douglass Lake became an isolated basin and stood at a level 
 approximately four feet higher than at present. At this level 
 most of the shore adjustments Avere made, the most notable being 
 the development of bars along the west end, across the indenta- 
 tions on the north side, aud also near Bryants on the south side. 
 Inasmuch as the adjustments were so largely made at the higher 
 level, it is felt that the submerged terrace, which is so well devel- 
 oped on this lake, was formed at that time with a depth of water 
 
116 INLAND LAKES OF MICHIGAN 
 
 at its outer edge of from seven to eight feet, or about one-half the 
 wave Tbase during the greatest storms, rather than the low value 
 as shown at the present level. The adjustments now in progress 
 are minor in importance and consist mainly in cutting back the 
 headlands and in some current action. Wave action on the head- 
 lands has succeeded in obliterating the terrace of the Upper Level 
 in most places. Current action is slight because much of the ad- 
 justment of the shore had been completed at the higher level, and 
 the amount of material supplied by waves and tributary streams 
 is small. No large indentations are now in the process of being 
 cut off except possibly a portion of North Fish Tail Bay. The 
 other current deposits are small in size and formed mainly during 
 the low water stage to be re-formed or obliterated during the 
 flood stage. The most significant of these is the bar which con- 
 nects the island to the mainland. Evidence of ice action is present 
 but shows no exceptional development of ramparts, due for the 
 most part to unfavorable shore conditions both as to topography 
 and material rather than insufiicient ice push. 
 
 With the shore adjustments largely completed, the interest in 
 the future development lies mainly in the possibilities of extinc- 
 tion. Up to the present, vegetation has played little part except 
 in the filling of the lagoons which has reduced the area of the lake 
 considerably. The vegetation in the main lake is principally 
 rushes and is limited to the submerged terrace and mainly to that 
 part which is exposed at low water. This filling is most import- 
 ant in the shallow water betAveen Fairy Island and the west shore. 
 Filling by pediment is so small that it may well be neglected. 
 Cutting down of the outlet has been of some importance in the 
 past and accounts for the dropping in level from the Upper Level 
 to the present. Since the outlet flows through unconsolidated 
 sand, this method of extinction may continue to be effective, but un- 
 derground drainage may greatly interfere. 
 
 Somewhat less than a mile southeast of South Fish Tail Bay is 
 located Big Springs, the source of Carp Creek which drains into 
 Burt Lake. The lower portion of this stream flows through a 
 swamp but the upper course heads in a gorge cut to a depth of 
 sixty to seventy feet in sand. Near its head the gorge ramifies 
 and at the end of each ramification is a spring. The supply of 
 water from these springs is large and constant, but unfortunately 
 the writer had no means of comparing the amount with that dis- 
 charged by Maple River, the surface outlet of the lake. Between 
 Big Springs and the central basin of the lake are several sipks in 
 the outwash plain which may be interpreted as indicating an un- 
 derground seepage line rather than the result of the melting of 
 
LAKES OF THE CHEBOYGAN RIVER BASIN 117 
 
 buried ice blocks. Fiirtbei" evidence is supplied by a well record 
 at Bogardus Camp which shows a dropping of tl»e ground water 
 level to the south. From this it seems reasonable that a consider- 
 able portion of tlie water of Douglass Lake drains southward under- 
 ground and issues at Big Springs. Also it is evident that the 
 gorge has been formed by sapping at the springs and is gradually 
 working backward towards the lake. If this is correct, the lake 
 probably will be tapped and the outlet will be shifted to a point 
 just west of Grape Vine Point. The gradient of the new outlet 
 will be much steeper than that of Maple River and down-cutting 
 will proceed at a more rapid rate than at present. The level of 
 the lake will then lower with minor changes in outline until the 
 outlet has cut down twenty feet. The east and west basins will 
 then exist as isolated basins sixty to sixty-five feet in depth separat- 
 ed by the dry bed of the central part. These lakes will still drain 
 through the new outlet and may be completely drained since the 
 greatest depths are above the level of Burt Lake. Yet, the process 
 becomes progressively slower as the gradient of the outlet flattens 
 and vegetation will probably accomplish the final extinction. 
 
CHAPTER IV 
 
 LAKES OF THE GKAND TRAVERSE REGION 
 
 In wliat is known as the Grand Traverse region, situated in 
 the northwestern part of the Southern Peninsula, are a number 
 of most excellent lakes of considerable size. Most of the more popu- 
 lar of these lakes border Lake Michigan and, in fact, were once a 
 part of it, having been isolated by great bars which developed in 
 either Algonquin or Nipissing time. The only exception among 
 the lakes visited in this region is AValloon which became an inde- 
 pendent basin when the Great Lakes subsided to the Algonquin 
 level. The popularity of these lakes is due not only to their nat- 
 ural beauty and adaptability for summer resorts but as well to 
 the proximity of Lake Michigan to the west, which considerably 
 tempers the summer heat. 
 
 The lakes included in this chapter — Walloon, Pine, Torchlight, 
 Elk, and Crystal — are typical for the region and are all attenuat- 
 ed in form, in which respect they resemble the famous "finger 
 lakes" of central New York. In addition to their attractiveness 
 las summer resorts, the situation of these lakes in an excellent 
 fruit-growing region makes them all the more important. In such 
 regions transportation is always a problem and, in this case, may 
 be solved partially by navigation. Pine Lake has for some time 
 been connected with Lake Michigan by an artificial channel 
 through which boats of considerable draught may pass without 
 difficulty. In fact, Charlevoix is a regular stop for some lines of 
 navigation during the summer months. The lake itself is navig- 
 able for boats of heavy draught for its entire length, and this 
 cheap means of transportation should lead to an increased devel- 
 opment of the agricultural possibilities of the region, already well 
 started. Of the other lakes, Elk and Torchlight off'er similar pos- 
 sibilities but at greater cost, since locks at Elk Rapids and con- 
 siderable dredging between the two lakes would be necessary. An 
 illfated attempt was made to make a navigable waterway from 
 Ci-ystal Lake to Frankfort but the result was merely to lower the 
 level of the lake. This proved so serious that a dam was built at 
 the outlet to hold the water at somewhere near its natural level. 
 
 With the possible exception of Crystal, these lakes are also simi- 
 
120 INLAND LAKES OF MICHIGAN 
 
 lar in the nature of the basins which they occupy. As discussed 
 at the close of Chapter II, the basins are large troughs running 
 more or less parallel to the direction of ice movement during the 
 last glaciation, but present difficulties of explanation as to man- 
 ner of formation which have not yet been solved. 
 
 WALLOON LAKE 
 
 Walloon Lake is the most easterly in position of the lakes of 
 this group and is situated in north-central Charlevoix County, a 
 few miles east of Pine Lake. It is easily reached by the Grand 
 Eapids & Indiana E. R,, which follows the broad valley of Bear 
 Creek south from Petoskey and runs a short spur from the main 
 line to Walloon Lake Station at the south-eastern end of the lake. 
 
 Walloon Lake is one of the most popular in the State. It is 
 of sufficient size to warrant a large fleet of motor boats, and the 
 irregular shore line lessens the fetch of the waves that would other- 
 wise become of dangerous size during storms and sudden "blows". 
 The abundance of high ground along the shores insures excellent 
 locations for cottages and its nearness to the railroad makes it easy 
 of access. The fishing is also an attractive feature. Unfortunate- 
 ly, from the standpoint of the resorters, the level of the lake has 
 been subjected to serious fluctuations by the use of the water for 
 power. A dam was constructed to regulate the flow of water 
 throughout the year, and the result has been a serious lowering of 
 the level during the summer months. This has been done since the 
 lake developed into a summer resort, causing great inconvenience 
 and loss of property to the cottagers, and has been the subject of 
 long litigation. The height of the dam has been fixed by law but 
 the lake has not been visited by the writer since that time. 
 
 The outline of this lake is very irregular and, although over nine 
 miles in length, has an area of only 8.35 square miles. In the 
 figure given for the area is included the ]^orth Arm which covers 
 slightly more than one square mile. Thus, the width on the 
 average is about three-fourths of a mile and rarely exceeds one 
 and a half miles. 
 
 From the map the idea may be gained that the lake has a gener- 
 al northwest-southeast trend which is interrupted by the North 
 Arm. However, from the physiographic standpoint, it may be 
 better described as occupying parts of two elongated basins of the 
 type mentioned earlier in this chapter. The trough occupied by 
 the main lake has a northwest-southeast direction for the northern 
 half of its extent. It then swings more nearly eastward and con- 
 juects with the Bear Creek valley a mile or more beyond the lake. 
 
LAKES OF THE GRAND TRAVERSE REGION 
 
 121 
 
 Near the southeastern end of the lake the second trough, in which 
 the North Arm lies, crosses the main depression and causes the deep 
 haj on the south side of the lake opposite the North Arm. South 
 of the lake it turns to the southeast and unites with the Bear 
 Creek depression some three miles below the main trouuh. 
 
 7- J J A/ ^ 6 fV 
 
 Sco/e / /n//e 
 
 T33M /S.SiV 
 
 Fig. 48. Outline map of Walloon Lake, Charlevoix County, 
 
 The surface features of the region are relatively simple with the 
 exception of the depressions mentioned above. The greater part 
 of the lake is surrounded by morainic deposits composed of a 
 rather sandy till. To the north lies a till plain which borders 
 about three miles of the north end of the lake and a much smaller 
 
122 INLAND LAKES OF MICHIGAN 
 
 part of tlie extremity of the North Arm, The surface, then, is 
 composed of knobs and basins, or sags and swells, except at the 
 continuations of the intersecting troughs. Both of these types are 
 descriptive of a rolling topography, the chief difference being in 
 the amount rather than in the character of the relief. Naturally, 
 a very irregular shore line of minor headlands and embayments 
 is the result. 
 
 The northern end of the lake lies within a mile of Little Traverse 
 Bay and is separated by a low divide about one fourth mile north 
 of the lake. Nearby and but a few feet below the crest of the 
 divide stands the shore of former Lake Algonquin. Thus, the 
 Walloon lake basin was not connected with this predecefesor of 
 Lake Michigan at the north end, and, if any connection existed, it 
 must have been at the present outlet. The latter is uncertain but, 
 at any rate, the lake w^as practically isolated at this time. 
 
 According to the original land survey, the outlet of Walloon 
 Lake was at the southeastern end on the north side of the valley 
 leading eastward to the Bear Creek valley. At present, the lake 
 drains over a dam and through a newly cut channel at the south 
 side of the flat. The town of Walloon Lake is built on the flat 
 and much of the topographj;- is thereby obs,cured. It seems reason- 
 able, however, that the present channel is artificial and that the 
 outlet, as shown on the early maps, represents the conditions as 
 regards shore action in this locality. The shore along the valley 
 floor is an adjusted sand beach where not interfered with by struc- 
 tures. Under the i^resent conditions no trace of a bar could be 
 found and it is probable none was formed. Material is carried to 
 this shore from both sides during westerily "blows" and is largely 
 redistributed by undertow. If we accept the position of the outlet 
 as shown on the early maps as correct, shore currents have afi'ect- 
 ed a transfer of material to the north along this shore, in spite of 
 the fact that the irregular south shore is not favorable to the 
 formation of strong currents. 
 
 Along the north shore to the entrance of the North Arm the 
 morainic knobs and basins drop gently to the lake. Thus, the 
 shore is a succession of flats and lens-shaped clifi's which reach a 
 maximum height of nearly forty feet at Three Pines. No distinct 
 submerged terrace is present, and the only deposltional form not- 
 ed extends eastward from the clifl's at A, see map, Fig. 48. This 
 form is a blunt hook about forty feet in length and is composed 
 largely of shingle. The coarse material has been shoved into a 
 distinct rampart near the attachment to the clifi's by the expan- 
 sion of the ice during the winter. Further to the east the ram- 
 
LAKES OF THE GRAND TRAVERSE REGION 123 
 
 part splits into three distinct ri(lj>es wliieh decrease in lieiglit and 
 play out as the end of the hook is reached. 'Shore action is con- 
 stantly sujiplying material to the hook, which is being reworked by 
 the periodic ice shove into a series of ramparts, that is, a local 
 ice-push terrace. Ice push is also in evidence at B, where boulders 
 have been forced into the clitf. In general, the shore forms are 
 the result of wave action in this locality, due largely to the lack 
 of suflSciently large embayments. The effective winds are from 
 the west, and strong eastward moving shore currents develop al- 
 most to the exclusion of undertow. These currents are virtually 
 uninterrupted for more than two miles and are able to transport 
 relatively coarse material which is ground to smaller sizes as it 
 travels along the beach. Thus, there is a noticeable grading of 
 the beach material which decreases in size to the east and becomes 
 sand at the lower end of the lake. It is here distributed by the 
 undertow into which the shore current merges. See Chapter III. 
 
 Another interesting feature found along this shore is the com- 
 bination of narrow terrace and low cliff which borders the low 
 parts of the shore. The terrace supports a heavy growth of vege- 
 tation, including trees of considerable size, and may, therefore, 
 be taken as an indication of a stage of the lake which stood two 
 feet above the level of the water in the summer of 1913, and not 
 merely a higli water mark. It must have been continuous when 
 the water subsided and has been removed since by Avave action 
 except along the low, protected parts of the shore. 
 
 At the narrow entrance to the North Arm, conditions are rather 
 abruptly changed. Current action here assumes the prominent 
 role, and the wind directions which were so important on the 
 shore just described are secondary. It will be noted from 
 the map that the shores of the approach to this bay gradual- 
 ly converge to two opposite points, forming a channel a quarter 
 of a mile in width. Within the bay, the shores recede rapidly and 
 increase the prominence of the points. The significant fact is that 
 the currents on both sides are not only forced to leave the shores 
 at these points but are able to maintain their courses across the 
 channel. Furthermore, the winds from both the northerly and 
 southerly quadrant are effective and have about the same reach. 
 Therefore, the spits which developed from these points are not un- 
 expected. 
 
 These spits, shown in Figs. 40 and 50, restrict the channel more 
 than one-fourth of its original width and are connected by a sub- 
 merged bar which is within eight feet of the surface at its low- 
 est part. Within the memory of settlers this depth was as great as 
 
124 
 
 INLAND LAKES OF MICHIGAN 
 
 eighteen feet, therefore, the bar is developing rapidly and the 
 channel will soon have to be kept open artificially, if it is to be 
 maintained. As seen in the sketches, these spits are triangular in 
 shape and are of regular curvature on both sides, indicating that 
 currents from both the main lake and the North Arm have been 
 instrumental in their formation. Yet, if the attachments and 
 
 
 Fig. 49. Spit at the west side of the entrance to the North Arm, Walloon Lake, 
 (iSketch from photograph). 
 
 'ir^:',^"^ * :>>ii' 
 
 Fig. 50. Spit at east side of the entrance to the North Arm, Walloon Lake, 
 (ISketch from photograph). 
 
 curvature of the spits with reference to the adjoining shores are 
 considered, it is clear that the greatest development has been from 
 the main lake. This may be due to several causes : The pre- 
 valence of storm winds, the depth of the water affected, and the 
 nature of the shores. Of the first two, we are not certain, but it 
 is probable that the main lake is the deeper, and that the storm 
 
LAKES OF THE GRAND TRAVERSE REGION 125 
 
 winds shift more frequently tlirongli the southerly quadrant than 
 the northerly. As to the nature of the shores, we find a greater 
 prevalence of cliffs along the converging approach than in the bay, 
 although the shore affected is shorter. This is probably due to the 
 intensification of the waves and, therefore, current action in the 
 narrowing approach. It appears, then, that all three factors are 
 favorable to growth from the main lake, but detailed study is neces- 
 sary for a decision. 
 
 Furthermore, it is apparent that the spit on the east side is 
 the better developed. This is clearly due to the greater shore line 
 affected and the unquestionable prevalence of storm winds having 
 a westerly component. It is also apparent from the sketches that 
 a considerable part of these spits developed during the higher 
 stage. The drop to the present level was so slight that conditions 
 were unchanged and the present growth is a continuation of that 
 of the previous stage. 
 
 As regards shore action, the North Arm acts as an isolated basin. 
 Shore conditions are similar on both sides and resemble those of 
 the north shore between the east end of the lake and the entrance 
 to this bay. Moraine borders the southern part of this embayment 
 but drops to a till plain which skirts the shores of the northern 
 half. The characteristic shore features, therefore, are the now 
 familiar cliflPs and terrace in front of which runs a sandy beach 
 interspersed at the small points with boulders. AVave action is 
 prominent but the cliffs are considerably lower than on the main 
 lake. Current action, however, has not been productive of any 
 decided effects unless it be a gradual building out of the flats to 
 the line of the cliffs. This could not be determined on account of 
 the heavy growth of vegetation wliicli obscures the surface of the 
 lowlands. In but one locality, aside from the spits at the entrance, 
 are currents actively depositing and this occurs on the south side 
 of Birch Point. Here a small spit composed of well assorted 
 pebbles extends southward from the point and continues under 
 water as a sand terrace. The relief is much less near the north 
 end, and much of the shore is swampy. Continuous swamp 
 fringes the north end with the exception of a low but conspicuous 
 swell west of the inlet. The vegetation of this swamp is creep- 
 ing outward over the marl-covered bottom, indicating the inception 
 of the final stage in the deveiopment of this embayment — its ex- 
 tinction by vegetation. Ice shove of the expansion type is active 
 here, but shore conditions are not favorable for decided results. 
 Ramparts of local extent are present on the low cliffs near the 
 
126 INLAND LAKES OF MICHIGAN 
 
 Walloon Lake Country Club but are much inferior to those found 
 on the main lake. 
 
 In general, it is evident that shore action within the North 
 Arm is much less intense than on the main lake and the adjust- 
 ments are correspondingly weaker, as may have been inferred 
 from the discussion of the spits at the entrance. 
 
 Outside the North Arm the increased activity of shore forces is 
 apparent. Interest centers first at point C, where the shore makes 
 a right-angled bend to the west. Ordinari'lj^, one might expect the 
 currents set up by westerly winds to leave the shore at tMs point 
 and deposit their suspended material in alignment with the north 
 shore. Instead, however, we find bold cliffs below which stands 
 a narrow terrace of the higher level. Around the point this ex- 
 jDOsed terrace widens and upon it a well-developed spit, which 
 has been modified by ice-shove, swings from the shore, enclosing 
 a narrow lagoon, now drained. From this it is evident that the 
 south and southeasterly winds which sweep without interruption 
 across the widest part of the lake are the most effective. It is 
 probable that the currents flowing eastward along the north shore 
 deposit material at the point, since lagoons are found further to 
 the west, but any such deposits are subsequently worked around 
 the point by the southerly winds. 
 
 The lagoons referred to above are found between C and Bacon 
 farm. The depressions are sags in the moraine, closed by bars 
 at the higher stage of the lake. These bars have been remodelled 
 by ice-push to such an extent that somewhat close observation is 
 necessary to detect current action. The assortment and gradation 
 in size of the material along the bars are the deciding characteris- 
 tics. The first bar encountered from the east shows two distinct 
 ramparts which rise in steps away from the lake. The elevations 
 of the ramparts correspond with the present and higher levels of 
 the lake, and the ramparts were, therefore, formed during these 
 stages. Farther west the bars across the mouths of the small 
 indentations have been remodelled into single ramparts. 
 
 Beyond Bacon farm an almost continuous cliff faces the lake 
 and extends back of Eyan Point to the north part of the lake. 
 The principal break occurs at the slight recession of the shore 
 north of Illinois Point. The gentle slopes which comg to the 
 shore here were carved into a distinct terrace at the former level 
 and are now heavil}' wooded. The smaller initial adjustments by 
 both waves and currents have taken place along this shore, and 
 the shore-line, although sinuous, extends with little variation from 
 the cliffs to flats. The adjustments are far from complete, how- 
 
LAKES OF THE GRAND TRAVERSE REGION 127 
 
 ever, since the submerged terrace is almost entirely lacking at the 
 present levd and was poorly developed at the higher level. 'Near 
 Eyan Point an outcrop of black shale rock was found in the cliff 
 above the lake. This rock offers little resistance to eroding agents, 
 in fact, less than the adjacent boulder clay, and weathering is dis- 
 integrating it so rapidly that it exhiMts none of the characteris- 
 tics of rocky shores. Eock ontcrops are so infrequent on the 
 shores of Michigan lakes that it is mentioned in i)assing. The 
 most noticeable shore activity is due to ice-shove. Expansion 
 must be very active on this shore for its effects are seen on virtual- 
 ly every cliff and flat where conditions are at all favorable, and 
 the enumeration of eaeh rampart and boulder-paved cliff' would 
 become monotonous. 
 
 At the narrows formed by Eagle Island and Eyan Point the ad- 
 justments are of striking proportions. Currents have left the 
 shore on both sides of the lake and have developed spits which 
 have reduced the width of the narrows relatively more than those 
 at the entrance to the North Arm, although the channel is not so 
 restricted nor so shallow. Naturally, we compare these two lo- 
 calities and find that the spits on the east sides show the greater 
 development in both cases, due to the same cause — the greater 
 strength and the prevalence of westerly winds. Eyan Point, how- 
 ever, whose north side has a curvature in conformity with the 
 shore to the north, has been bnilt to a large extent by currents 
 from the north, a fact readily accounted for by its position near 
 the south end of the extended west shore. This spit, which ex- 
 tends almost half way across the narrows, is of clear sand and 
 w^as built mainly at the abandoned level of the lake. Thus, its 
 surface stands two feet or more above the present level. As the 
 spit developed, grasses and, later, trees took root, forming a mat 
 over its surface. Ice action was then able to form a series of low 
 ramparts parallel to both shores but better developed on the north 
 side. When seen by the writer, this point was being eroded on 
 the north side and built up on the south, a process Avhich, if it 
 continues, will shift the position of the entire spit to the south. 
 This shifting was well shown at the tip of the point by a sudden jog 
 in the shore-line which occurs at the attachment of a recent ex- 
 tension of the spit. 
 
 On the opposite side of the lake conditions are similar but the 
 results are on a much smaller scale. The curi'ents leave the shore 
 at the extremity of Eagle Island but are relatively feeble, due to 
 the infrequency of strong east Avinds and the irregular shore to 
 the north. The blunt sand spit which reaches southward from 
 
128 INLAND LAKES OF MICHIGAN 
 
 the end of tliis point consists of two parts : A swampy, grass-cov- 
 ered flat next the cliffs and a bare outer zone bordering the lake. 
 There is practically no difference in the elevation of the two parts, 
 1but the failure of the grass cover forms a sharp line of division. 
 It is possible, of course, that the vegetation is gradually creeping 
 outward and that the development of the spit has 'ibeen continuous. 
 Yet, from conditions found elsewhere on the lake, it seems more 
 probable that a broad bar, for the most part submerged, developed 
 during the higher stage, and that the slight lowering of the level 
 exposed a portion of this bar upon which vegetation soon took 
 iiold. If this is correct, the bare portion must be an extension of 
 this bar formed under the present conditions. At any rate, it is 
 evident that the currents from the north are the more potent. 
 
 On the east side north of Ryan Point, the more or less regular 
 alternation of cliffs and flats again appears. Shore conditions, 
 even to the frequent evidences of ice shove, are very similar to those 
 below the point, but show, in general, greater activity of waves and 
 currents. Thus, a persistent submerged terrace is present which 
 reaches a width of one hundred feet or more on the southern 
 stretches of this shore and drops into deep water at a depth of four 
 or four and a half feet. Opposite the St. Louis Club two small 
 hooks which extend southward from minor projections indicate 
 the prevailing movement of the shore currents. It is evident that 
 here wind direction is more important than reach in the develop- 
 ment of currents. 
 
 The conspicuous embayment, D on map, in the rather regular 
 «hore along the northeast side, is caused by a large amphitheatre- 
 shaped basin, a sag in the moraine. North of this the even slopes 
 of the till plain dip gently to the lake, forming shores which are 
 low but not swampy. 
 
 The north end of the lake is called the Mud Hole. True to its 
 name the bottom is covered with an ooze of marl upon which is 
 accumulating the yearly residue of a heavy growth of rushes. It 
 Is a distinct basin with a shallow, narrow entrance, which is furth- 
 er constricted by the development of a spit on the west side. 
 Within the Mud Hole shore action is limited to the expansion of 
 the ice, and this is not important at the present level. However, 
 a distinct rampart, containing boulders of considerable size, stands 
 a,t a higher level near the north end where the width of the bay 
 was not greater than one fourth mile. See Pig. 51. This observa- 
 tion is interesting in view of the rather prevalent opinion that 
 ramparts are not formed on lakes of much less than a half mile 
 in diameter. The north shore of the Mud Hole is fringed by a 
 
LAKES OF THE GRAND TRAVERSE REGION 
 
 129 
 
 swani]> whicli extends noi-tlnvard to the low divide wliicli stood 
 between this lake and J^ake Al<;onqnin. Tt seems certain that Wal- 
 loon Lake stood at the hitihei- level at that time and covered this 
 swamp. Tliis l)ein<>- tlie case, the divide was but a few i*ods in 
 width. 
 
 Fig. 51. Ice rampart. Mud Hole. Walloon Lake. (Sketch from Protograph.) 
 
 The west shore above Eagle Island is much more broken than 
 that of the opposite side and, although shore action is relatively 
 feeble, more deposits are found. The first to be encountered is 
 the spit at the entrance to the Mud Hole, already mentioned. This 
 spit is turned to the northeast and, therefore, is being built by 
 currents from the south. The material for this spit is quarried 
 from the short stretch of shore in tlie bay to the south and is 
 limited in amount. Nevertheless, the spit is developing rapidly, 
 on account of the small amount of filling necessary to close the 
 channel and the fact tliat the currents are quickly brought to a 
 halt. Ice push is strong in this locality and has formed a ram- 
 part on the south side of the spit which merges into a boulder 
 paved cliflf at its attachment. 
 
 Again at the north end of the blunt point upon which the St. 
 Louis Club is located, the currents have held to their course at 
 the present level, even though the bend iii the shore is not pro- 
 nounced and the hook, thus formed, shuts off a narrow lagoon 
 which is open at the north end. This lagoon supports a heavy 
 growth of lily pads, rushes, and grass and will soon become filled. 
 Ice action does not seem to be effective at tlie present level but its 
 effects are evident at the ramparts along the old shore. However, 
 17 
 
130 INL)AND LAKES OF MICHIGAN. 
 
 in the bay to the south of the St. Louis Club ramparts are found 
 at both levels, but that at the present shore is of moderate devel- 
 opment and is not continuous. 
 
 Below this bay the waves are working on the lower slopes of the 
 hills and have formed low cliffs along a stretch of shore a half 
 mile in length. Currents are also of considerable force and have 
 carried away the finer particles, leaving coarse material on the 
 beach. The effective drift is to the south and much of the debris 
 has been deposited in a hook and an extended submerged terrace, 
 E on map, which are detaching a narrow lagoon to the rear. 
 
 Before Eagle Island is reached the slopes drop to a narrow 
 swamp which runs directly south across the neck of this projec- 
 tion to the decided embayment on the south side. This swamp is 
 barely above the present water level and was evidently covered 
 during the higher stage of the lake, therefore the name, Eagle 
 Island. The swamp borders the north shore of the bay partially 
 ■ enclosed by Eagle Island, and dense vegetation has obscured the 
 beach.' A small spit on the east shore of this bay at the edge of 
 the swamp is interesting in that it is an index of the power of the 
 winds from the southwestern quadrant. The spit is turned to 
 the northwest and derives its material, therefore, from the short 
 stretch of shore between it and the end of the point. The fetch 
 of the waves is short and must be driven by strong winds to be of 
 any significance. The shallowness of the bay is, however, an im- 
 portant factor in the formation of this spit, on account of the 
 small amount of filling necessary and the rapid decrease in in- 
 tensity of the waves as they progress towards the beach. As in 
 other shallow parts of the lake, heavy deposits of marl cover the 
 bottom upon which reeds are now taking hold. 
 
 The protected west side of the bay is bordered by gentle slopes 
 which have been carved into a low terrace at the higher level but 
 show little evidence of wave action at the present level. As a 
 matter of fact, conditions along this shore are reversed for the 
 higher and present levels, and currents are now the important 
 agent of adjustment. They leave the shore in two places along this 
 side, due probably to the shoaling of the water, see Crystal Lake, 
 and have formed small sand spits which point northward. East- 
 erly winds are, of course, the most eff-ective, since the bay is well 
 protected on the north by Eagle Island. 
 
 Zs'ear Harpers, ice action is well shown by ice ramparts across 
 the mouth of a ravine at the present and higher levels. As usual, 
 the older rampart is the better developed. The embayment south of 
 
LAKES OF THE GRAND TRAVERSE REGION. 131 
 
 Harpers is liued by a saud beach of even curvature iu spite of tluv 
 alternation of cliffs and sags, and the effects of ice action are evi- 
 dent as ramparts or boulder-paved cliffs. From F to G wave 
 action has predominated and cliff's of variable height face the lake, 
 with the exception of a wide depression on the west side of point G. 
 During the higher stage a bar developed across this depression, and 
 the elongated lagoon was filled with vegetation. The growth of the 
 bar must have been from the west under the influence of northerly 
 winds. Ice action piled up a rampart along the old shore previous 
 to the development of the bar and later has been active along the 
 bar. The sandy character of the bar is not favorable for decided 
 effects and the present rampart is inferior in development. 
 
 Again at G conditions were reversed with the sinking of the water 
 to the present level, and a' broad terrace is being built at the foot 
 of the cliffs by currents from the west. The point of departure of 
 the currents is not definite and a blunt point is the result. A large 
 amount of material is dropped at this point since it extends three 
 hundred feet or more into the lake as a submerged terrace, dropping 
 into deep water at four feet. The turning of the currents from the 
 shore at G and the yerj shoal j\'ater between G and Air Castle Point 
 have effectively prevented adjustments along the intervening shore. 
 
 But at Air Castle Point, the constructive work of shore agents is 
 shown on a scale comparable with that at the entrance to the North 
 Arm and at Kyan Point. This great spit is irregular iu outline on 
 the west side but has an even curvature iu accordance with the 
 trend of the east shore. Clearly it has been built by currents from 
 the south. Deposition is still taking place, and a submerged por- 
 tion is growing into the lake as a relatively narrow bar with a 
 somewhat greater curvature than the subaerial part, a form in 
 striking contrast to the re-curved spits or hooks which are usually 
 formed when currents are dissipated in deep water. The increase 
 in curvature occurs along the part of the bar which has grown into 
 deep water and is exposed to the force of the waves from the west, 
 undiminished by the projecting point G and the intervening shoal. 
 Under such conditions the spit will increase in curvature as it 
 grows, and its position will represent the relative strength of the 
 forces acting on either side. It is probable that this spit will de- 
 velop to the west of point C and, therefore, will not divide the lake, 
 but it is impossible to make a definite statement on this point 
 considering the present development of the spit. The greater part 
 of the spit was built during the higher level and, after the establish- 
 ment of vegetation, a continuous ice rampart was pushed up on the 
 east side. 
 
132 INLAND LAKES OF MICHIGAN. 
 
 Below Air Castle Point cliffs, unbroken save for a narrow valley 
 which is blocked by ice ramparts, line the shore to the head of the 
 large embayment which forms the southern extremity of the lake. 
 As already stated, this bay is caused by the continuation of the 
 North Arm trough which extends several miles to the south and 
 southeast. The shore is, therefore, low and swampy.' The notice- 
 able features are the well-developed submerged terrace and the ice 
 ramj)arts. The ramparts are three in number and are especially 
 well developed and distinct. They increase in elevation, size, and 
 continuity with distance from the shore. The best developed, the 
 one farthest inland, stands three to five feet above the adjacent 
 land and encircles the bay with but a single break where a small 
 stream crosses. The middle rampart is inferior in development but 
 still is a decided ridge; the lowest is discontinuous and poorly 
 defined in places. Two ramparts are common and may be correlated 
 with the two stages of the lake. The presence of the third rampart 
 in this one locality is, however, somewhat puzzling. 
 
 The lake attains its greatest width opposite this bay but still does 
 not exceed the maximum limit for ice expansion. The expansion, 
 then, is greatest in this locality, and the ramparts are exceptional- 
 ly well developed. From its elevation, the rampart farthest inland 
 may be considered the equivalent of the higher rampart formed in 
 other favorable localities where two are present. Also it is evident 
 that the ramparts nearest the shore are in process of formation at 
 the present time and are correlatives. But a lake stage, corres- 
 ponding in level with the intermediate rampart, cannot be assumed 
 since corroborative evidence at other localities on the lake is en- 
 tirely lacking. To the writer this series of ramparts seems to have 
 been formed in a manner similar to that of an ice-push terrace but 
 on a shore of such flat slope that the ramparts are separate and 
 distinct ridges. The cause of the lowering of the lake level was the 
 gradual deepening of the outlet by natural processes. The earliest 
 and largest rampart was formed at the higliest stage after vegeta- 
 tion had become well established and served to bind the loose sands. 
 The size indicates that the higher stage must have been of relatively 
 long duration. During the lowering of the level the vegetation 
 slowly encroached on the emerging lake bottom and was not dis- 
 turbed by the ice which, under normal conditions, expanded to 
 positions less and less advanced as the shore receded. But under 
 especially favorable conditions such as high water, light snowfall, 
 and numerous alternations of temperature during the winter, ex- 
 cessive expansion took place, and the ice advanced into the zone of 
 vegetation and pushed up the rampart. The slope seems to be the 
 
LAKES OF THE GRAND TRAVERSE REGION. 133 
 
 most inipoiiaiit factor in this considei-atioii since tlic latc oi' reces- 
 sion of tlie sliore and, tlierefore, the advanced i)ositiou of tlie ex- 
 panding ice is dependent on tlie flatness of tlie slope. Another 
 possible cause for the intermediate rampart is that the lowering 
 of the level was temporarily halted by an obstruction in the outlet, 
 which Avas sufficient for the formation of a rampart but not for 
 distinguishable eflt'ects of the other shore agents. 
 
 The shores from this bay to the outlet are of the clilf and sag 
 type, modified locally by ice action. This type has been so fre- 
 quently mentioned in connection with other localities that repeti- 
 tion is not necessar}' and the description of the shores may be left 
 at this point. 
 
 In resume, the episodes in the history of this lake are but two, 
 the present level and a stage a few feet higher. This may seem 
 somewhat meager when compared with the numerous stages of 
 several of the nearby lakes, but the many adjustments of the shores, 
 begun at the higher level and continuing at the present, are of suf- 
 ficient interest to compensate for the deficiency. 
 
 Walloon Lake stands well above the levels of the predecessors 
 of the Great Lakes since Algonquin. There is a possibility of a 
 connection with Algonquin at the southeastern end and a certainty 
 that the higher stage was in existence at this time. Whichever may 
 have been the case, the agencies aft'ecting the shores of the higher 
 level were similar in intensity to those active today, since the reduc- 
 tion in area has been slight and the lake Avas practically enclosed. 
 
 The irregularity of the basin and the adjacent sloi)es afforded 
 many opportunities for large and significant adjustments of the 
 shores, but on the other hand, eft'ectively reduced the intensity of 
 the forces by which such adjustments are accomplished. The 
 limited reach of the winds and the irregularities of the shores per- 
 mit a moderate development of waves and currents and the results, 
 in general, correspond. Waves have cut back many of the minor 
 salients, reducing the smaller sinuosities of the shore line, but the 
 very limited development of a submerged terrace shows a relatively 
 small amount of wave action. And the currents can neither be 
 of great power nor continuity on account of the short stretches of 
 even shore. The striking thing, however, is the localization of their 
 effects at critical points, which greatly increases their importance. 
 Thus, a continuation of their activitj^ at the entrance to the North 
 Arm, at Kyan Point, at the channel to the Mud Hole, and possibly 
 at Air Castle Point will lead to a division of the lake into smaller 
 members, and this, in turn, will greatly hasteif its extinction by the 
 l)rocesses already well started. Ice expansion is active to a re- 
 
134 INLAND LAKES OF MICHIGAN. 
 
 markable degree ou the shores of this lake: The series of ramparts, 
 the boulder-paved cliffs, and the ice-push terraces are unequalled on 
 Michigan lakes. 
 
 By far the greater development of the shores occurred at the 
 higher stage. The slight lowering of the level has caused a reduc- 
 tion in the activity of the waves which in some localities has been 
 sufficient to reverse conditions from cutting to deposition. The 
 decrease in wave action furnishes less material to the currents, and 
 all of the shore adjustments are necessarily taking place more 
 slowly. Nevertheless, it is evident that the Mud Hole and the 
 North Arm will soon be separated from the main lake, to be fol- 
 lowed later by the division of the remainder of the lake into two 
 basins by the growth of Ryan Point. Shore activities will be 
 further reduced in the separate basins and in the meantime vegeta- 
 tion, which has already accomplished considerable filling in the 
 partiallj^ enclosed baj^s, will increase. Thus, it may be suggested 
 with some confidence that this lake will become extinct before the 
 completion of the adjustment of its shores takes place. 
 
 PINE LAKE 
 
 Pine Lake, called Long Lake on the earlier maps, is elongated 
 in a northwest-southeasterly direction and at Charlevoix lies with- 
 in a mile of Lake Michigan. See map Fig. 52. The main body of 
 the lake is slightly over thirteen miles in length and probably 
 does not exceed one and one-half miles in average width. Where 
 greatest, the width is but little more than two and one-half miles, 
 and in one place only, near the upper end, does it contract to less 
 than a mile. Thus, the main lake may be considered rather uni- 
 form in its dimensions, with only minor bays and projections re- 
 lieving the regularity of its shores. However, an important ex- 
 ception is found in the narrow South Arm which extends nine 
 miles in a direction slightly east of south. The South Arm is 
 much narrower than the main lake, its average width being esti- 
 mated at less than a half-mile, and is constricted to five hundred 
 feet in the narrows near its entrance. The total area of the lake 
 is 26.7 square miles. On account of its peculiar shape and navi- 
 gability, the lake has influenced to some extent the grouping of 
 population about it, and we find the cities of Charlevoix, Boyne 
 City, and East Jordan at its extremities. 
 
 The region in which Pine Lake lies is one of the few localities 
 in Michigan where drumllns are found. More than half of the main 
 lake and virtually all of the South Arm lie in longitudinal depres- 
 sions which are surrounded by these peculiar hills of hard boulder 
 
Fig. 52. Map showing outline and configuration of tlae basin of Pine Lalce, Cliarlevoix County. (After U. S. Lake Survey Chart.) 
 
LAKES OF THE GRAND TRAVERSE REGION. 135 
 
 olav. Tlio reinaiiuler is moraiuic material of sandy character. 
 The drumlius are cliaracterized by smootli slopes and a general 
 parallelism of their longer axes and were formed nnder the glacier, 
 the longer axes indicating the direction of the ice movement. The 
 smooth slopes adjoining the lake are high, rising in places to heights 
 of three hundred feet above the lake. In many cases, the drumlins 
 are ronghly parallel to the basin of the lake, bn-t considerable dis- 
 cordance is found, especially in the South Arm, and the basin is 
 considered to have been independent of the ice movement at the 
 time when the drumlins were formed. This idea is further strength- 
 ened by the fact that the basin of the lake is exceptionally free from 
 islands, shoals, and deep holes. 
 
 The iiliysiographic history of Pine Lake shows four distinct levels, 
 in which respect it is not exceeded by any of the inland lakes of the 
 State. The shore lines of these levels, especially the two highest, are 
 conspicuous on the slopes above the lake and stand usually at mod- 
 erate distances from the present shore, see Plate VII. The highest 
 terrace stands about eighty-five feet above the lake at Charlevoix 
 and was formed by Lake Algonquin. In contrast to the lakes of 
 the Cheboygan basin, Pine Lake was not greatly extended in area 
 during this stage except at Horton Bay, Boyne City, and in the 
 South Arm. The latter was connected with the series of narrow 
 troughs which lead to Grand Traverse Bay and in which lie Inter- 
 mediate, Torchlight and Elk Lakes. 
 
 The next lower level is that of the Nipissing Lakes, and the beach 
 stands twenty-seven feet above Pine Lake at Charlevoix. The sink- 
 ing of the water to this level was accompanied by considerable con- 
 striction of the lake at the extremities, and the basin was isolated 
 except for a narrow strait at Charlevoix. Following the Nipissing 
 stage, a drop of eighteen feet brought the level to nine feet above 
 the present. This level may be designated as the Post-Nipissing, and 
 the basin was completely isolated for the first time. The drop from 
 the Post-Nipissing stage to a level four feet above the present, which 
 we shall call the Upper Level, probably accompanied the downward 
 cutting of the outlet. This level was abandoned when the lake was, 
 connected with Lake Michigan by an artificial channel in 1873, the 
 amount of lowering being 3.6'2 feet, according to the United States 
 Engineer. 
 
 The interesting history of the lake may be profitably supple- 
 mented by a study of the shore features at the various levels. In 
 general, it may be stated that the regularity of the basin and the 
 smoothness of the surrounding slopes have not furnished conditions 
 for large adjustments of the shore lines. Also, on account of the 
 
136 INLAND LAKES OF MICHIGAN. 
 
 greater power of the waves and possibly longer periods of action, the 
 adjustments at the 'higher levels were of greater magnitude. 
 
 The narrow neck of land which separates Pine Lake from Lake 
 Michigan is less than a mile in width. Its surface is composed of a 
 flat terrace, the Nipissing, standing at an elevation of slightly less 
 than thirty feet above Lake Michigan, and above this on either side 
 rises a steep cliff to the Algonquin terrace, more than eighty feet 
 above the lake. A rather deep depression, occupied by Round Lake, 
 breaks the monotony of the Nipissing terrace and connects with 
 Lake Michigan by a narrow channel, artificially deepened and 
 widened. The connection with Pine Lake is now by an artificial 
 channel which isolated Park Island the north side of which is formed 
 by the Old Elver, the former outlet of Pine Lake. The channels are 
 dug to a minimum depth of twelve feet and allow the entrance of 
 large vessels into Round Lake, which makes an excellent harbor. 
 The town of Charlevoix is built on the higher terraces but largely 
 on the Nipissing terrace west of Round Lake. This thriving town 
 is one of the most popular summer resorts of the State and is by 
 far the best location on the lake. The extension of the lake towards 
 the southeast is almost directly away from Lake Michigan, and the 
 cooling effect of the lake breezes is slight at the farther extremities. 
 Consequently, the lake as a whole is not so extensively patronized 
 by summer visitors as are some others in th6 State. 
 
 On the Lake Michigan side of the iNipissing terrace at Charlevoix 
 stands a narrow sand bar which developed from the southwest and 
 must have crowded the outlet to the north, although it probably did 
 not completely close it. Towards Pine Lake the Nipissing terrace 
 is bordered by a cliff which drops to the Post-'Nipissing level. In the 
 early part of this stage Round and Pine lakes were connected by a 
 strait about five hundred feet in width, but this connection was 
 gradually narrowed by a bar which developed in a northeasterly 
 direction from the cliffs on the south side of the present channel to 
 the large bend in the Old River. Sufficient water passed through 
 the channel to keep it open, but the bar was able to force the stream 
 to the cliffs on the north side. Above the shore of Pine Lake in this 
 vicinity, the Upper and Post-Nipissing shores are well defined, the 
 former having an especially fresh appearance. Thus, in the vicinitj^ 
 of Charlevoix the four main stages of the lake may be readily dis- 
 tinguished. 
 
 , Along the north shore the Nipissing terrace narrows and soon 
 disappears. At the present shore are found storm beaches, indicating 
 the existence of powerful waves which develop with a fetch of several 
 miles when the wind is from the southeast. At the most northern 
 tip of the lake is a lowland extension which was the scene of strong 
 
LAKES OF THE GRAND TRAVERSE REGION. 137 
 
 cuiTeut actiou in former times. Tlie i)re.seiit shore swings in an even 
 curve towards Pine Point and is bordered by a flat wliicli slopes 
 gently upward to a distinct sand bar about one hundred yards from 
 the shore. Behind the bar, the crest of which stands four feet above 
 the lake, is a swampy lagoon, which, gradually becomes drier to the 
 north. The depression is again interrupted beyond by a strong bar 
 which stands twenty feet above Pine Lake. In both cases, the bars 
 extend completely across the depression and, since their elevations 
 correspond to the Nipissing and Upper Levels, we may conclude that 
 portions of this bay were cut oft' from the- main lake at those levels. 
 The present conditions at Pine Point show moderate wave and cur- 
 rent action from the west, thus enabling us to determine the direc- 
 tion of the development of these bars. 
 
 Pine Point was formerly much more prominent than at present 
 because, in addition to the bay just discussed, there existed a narrow 
 indentation east of the point, which extended fully a quarter of a 
 mile inland. This bay persisted until tlie Upper Level, during which 
 a bar was built near the present shore. An attempt at draining the 
 lagoon, thus formed, was made by digging a ditch through the bar, 
 but with mediocre success. At the present level the only shore 
 action is accomplished by waves which have thrown up a storm beach 
 of sufiicient height to enclose a narrow, crescent-shaped lagoon. 
 Along this shore the waves have laid bare considerable marl whicli 
 was deposited during earlier stages. 
 
 On either side of Oyster Bay drumlin-like projections of nearly 
 north-south trend reach the lake shore and are responsible for its 
 irregularity. On the lake side of the promonotory west of the bay 
 the waves are actively cutting as they have been in the past. The ef- 
 fects are seen in a beach of rather coarse material and in distinct, but 
 narrow, terraces of the Upper, Post-Nipissing, and Nipissing levels. 
 Oyster Bay occujiies a shallow sag that continues northward forty, 
 rods or more beyond the present shore. During the Upper Level, the 
 greater part of this depression was covered, but a distinct bar wathin 
 two hundred feet of the present shore indicates the formation of a 
 lagoon at the head. The extinction of this lagoon was casued by ac- 
 cumulation of marl, heavy deposits of which may now be seen. 
 
 The long stretch between Oyster Bay and Horton Bay is noticeable 
 mainly for the perfection of the terraces of the former levels of the 
 lake, shown diagrammatically in Figure 53. The least developed of 
 the terraces is the Post-Nipissing which has been cut away in places. 
 The minor projections, for example Wilson I'oint, all show active 
 cutting by waves on the west side and a tendency towards deposition 
 on the east. Thus the beaches are stonv to the west but of fine ma- 
 
13S INLAND LAKES OF MICHIGAN. 
 
 terial or of marl on the opposite sides. The coarser beaches are 
 in places pushed into low ice ramxjarts. One mile beyond Wilson 
 Point is the Shale Dock which marks the location of the only outcrop 
 of rock near the j)resent shore of the lake. It is a dark-colored, soft 
 shale and was formerly used by the Bay Shore Lime Co., which 
 operated tlie quarry and shipped the rock by water to its plant at 
 Bay Shore. This outcrop does not reach the ]3resent shore but was 
 undoubtedly carved by waves in the past. However, the rock 
 weathers so easily that the characteristic shore forms have been 
 destroyed. 
 
 P/ne La fie 
 
 Fig. 53'. Diagram showing exposed terraces along the shores of Pine Lake. 
 
 The shore west of Horton Bay shows an extension of the Upper 
 Level terrace which continues to the bay. This bay lies in a depres- 
 sion which caused a large 'expansion of the lake during the stages 
 previous to the Upper Level. This was especially marked during 
 Lake Algonquin and extended several miles up the Horton Creek 
 valley. The rounded projection of land on the west side of the bay 
 w^as covered until the lake stood at the Post-Nipissing level. During 
 this stage it was a narrow promontory and was severely pounded 
 by the waves on the lake side, forming a beach of coarse material. 
 This rubble was pushed into a decided ice rampart which still stands 
 four to live feet above the beach and contains many good sized 
 boulders. The Upper Level is well shown by a definite beach which 
 swings around the point and into the bay as a bar, cutting off the 
 swampy hinterland. On the east side of the bay the high ground 
 lies from one-eighth to one-fourth of a mile from the shore, and a 
 similar bar at the same level cut off the lowland at the foot of the 
 hills. This bar developed from the east, as is shown by the westward 
 turning of Horton Creek before entering the bay. From Horton Bay 
 to Boyne City and beyond the surrounding topography is morainic 
 rather than drumlinoidal, and the material is sandy till. This is 
 readily detected by the change in the beach material from rubble to 
 sand about one-half mile beyond the bay. This shore is exposed to 
 strong westerly winds and the wave action is intense. The former 
 levels are well shown with the exception of the Post-Nipissing which 
 
LAKES OF THE GRAND TRAVERSE REGION. 139 
 
 lias been out awa}'. The absence of this terrace in a locality of 
 strong- wave action, while those above and below it are well de- 
 veloped, may be interpreted to mean that this stage was of short 
 duration. 
 
 The configuration of the lake is such that the fetch of the waves 
 driven bj' westerly winds, the most important here, decreases to the 
 south and currents become relatively more effective. Therefore de- 
 positioii^is to be expected where shore conditions permit. This is 
 well illustrated as Horse Point is approached from the north, where 
 a broad indentation formerly existed. At present, the beach is 
 of fine sand and of even curvature to the end of the point. Inland 
 at the Upper Level is a narrow spit which almost parallels the 
 present shore and encloses an elongated lagoon. Similarly on the 
 landward side of this lagoon is another spit at the Post-Nipissing 
 level, which likewise has cut off a lagoon but of smaller proportions. 
 At the tip of the point, however, wave action again predominates, 
 and the Upper Level terrace has a width of from one hundred to two 
 hundred feet. In addition, the coarse material has been forced up 
 into a decided rampart at the present shore. Ramparts on a lake 
 of this size must be largely of the ice jam type and are a further 
 indication of the power of the waves. South of the point the hills 
 recede from the shore, and broad terraces of the L^pper and Post- 
 Nipissing levels are present. The width of the terraces is due to 
 the flatness of the slope rather than excessive shore action. At the 
 present shore wave action is cutting into the Upper Level terrace. 
 This cutting is slight as a rule, but at A (see map) the waves have 
 reduced the L'^pper Level terrace and are cutting into the sand of 
 the Post-Nipissing. As Boyne City is approached, the upland stands 
 nearer the lake and the terraces are narrow but distinct. 
 
 At Boyne City the lower levels follow the present shore but are 
 obscured by buildings. The Algonquin shores run to the southeast 
 as far as Boyne Falls and it is on the terraces of this and the 
 Nipissing stages that most of the city is built. Along the south- 
 west shore from Boyne City to the entrance of the South Arm the 
 upland slopes somewhat steeply to the shore, and consequently the 
 terraces are narrow. The Algonquin and Upper Level terraces are 
 well defined, but the intermediate ones vary in development and 
 are obliterated locally. Shore action is limited in this part of the. 
 late and is mainly by waves which have made a stony beach. In- 
 ferior local ice ramparts on the beaches of the Upper and jjresent 
 levels are evidence of an ice shove of moderate force. The moderate 
 wave action is somewhat intensified on the southeast sides of the 
 points which, with the exception of that at Flatten Dock, are due 
 
140 INLAND LAKES OF MICHIGAN. 
 
 to the glacial topography. On the northwest sides of points, cnr- 
 rent action has enclosed small lagoons in some cases, as at B. At 
 Flatten Dock a stream entering the lake has built out a delta of con- 
 siderable proportions. This delta was formed during Nipissing 
 times and since then has been degraded by wave action on its 
 borders, forming terraces at the Post-Nipissing and Upper Levels. 
 The terrace of the latter is now swampy, due to the presence of a 
 storm beach at the present shore. In places along this shore wave 
 action at the present level has cut low cliffs, which often expose 
 marl. Such localities are easily recognized when waves are run- 
 ning by the milky appearance of the shallow water. 
 
 The broad point between the main lake and the South Arm is 
 exposed to the heavy seas of the northeasterly winds and wave action 
 is powerful. In addition, the original slopes were low, conse- 
 quently all the terraces are well developed and wide. Ice has also 
 been active in this locality, having formed distinct ramparts on the 
 beaches of the present and Upper levels on the west side of the 
 point. The intensive work of the waves on this point may be appre- 
 ciated by a trip of somewhat over one-fourth mile eastward from 
 the narrows of the South Arm to a drumlin, the top of which was 
 completely bevelled during Algonquin time. The drumlins are, 
 perhaps, best developed along the South Arm but as a rule lie be- 
 yond the borders of Lake Algonquin. 
 
 The entrance to the South Arm is almost closed by an abrupt 
 projection of the shore from the east side. The south end of this 
 projection makes a sharp re-entrant and affords at its tip an ex- 
 cellent index of current action through the narrows. An incipient 
 spit shows that weak southward moving currents pass through. It 
 will be seen from the map that waves of considerable size may be 
 formed from the north. The drive of these waves tends to pile up 
 the water on this shore and the only outlet is through this nar- 
 rowing channel, even though it is tortuous. 
 
 The South Arm nowhere reaches a mile in width and the ad- 
 joining slopes are consistently steep. Consequently, wave action 
 is moderate and shore adjustments are much less striking than 
 along the shores of the main lake. However, the Algonquin and 
 Nipis'sing terraces, shown in Plate VII, are well developed and en- 
 circle this arm of the lake. This is to be expected since the lake 
 was larger during these stages and also received some of the swells 
 of the main lake. For the most part, the lower beaches are distinct 
 but only locally are they well developed. The best development is 
 found on the numerous small headlands, illustrated by point C. 
 All four terraces are found here, and, in addition, a small spit is 
 
LAKES OF THE GRAND TRAVERSE REGION. 141 
 
 growing to the nortli at llie picsciit h'vcl. The only winds ettective 
 on this shore are those from northerly or southerly quadrants, 
 and the latter are the more imi)ortant on account of their long 
 reach. Farther south, point 1) shows a wide Nipissing terrace into 
 which the waves have cut a low clitf. The lower terraces seem to 
 have been poorly developed here and were quickly obliterated. . 
 South of this blunt point the slopes are flatter and all the ter- 
 races are present. The effects of ice action are seen in the frag- 
 mentary ramparts at the Post-Nipissing and Upper levels, caused 
 probably by expansion here. 
 
 ;With slight variations, the conditions just described continue to 
 East Jordan where the narrow, lower terraces encircle the end of 
 the lake with some extension up the valley of the Jordan river. 
 This stream has deposited large quantities of material ami tilled 
 in a considerable area at the head of the lake during the lower 
 levels. It enters from the west and has constricted the end of the 
 lake by deposition from this side. The development of the higher 
 terraces is almost identical witl* that at Boyne City, except that 
 the Algonquin was even more extended. 
 
 On the west side of the South Arm the features are so similar 
 to those opposite that a detailed description seems unnecessary. 
 The shore from point E northward to the next prominent projection 
 is worthy of mention on account of the prominence of the Post- 
 Xipissing terrace, which is usually poorly developed. The adjust- 
 ments at present are mainly due to cutting and this on the north 
 sides of the points. Again, at Holy Island the terraces are well 
 developed. This island first stood above water during the Post- 
 ]^ipis'sing stage, and this and succeeding levels are distinctly shown. 
 On the sides, the levels are indicated by notches and low cliffs, 
 a.nd at the ends terraces thirty or more feet in width are present. 
 At the north end a bar has nearly bridged the shallow water be- 
 tween the island and the mainland, and artificial filling with brush 
 has sulficed for a rude roadway to the island. 
 
 The west side of the narrows is flanked by narrow terraces which 
 broaden somewhat as the main lake is reached, and on these ter- 
 races Sequanota is built. Beyond Sequanota the terraces are again 
 narrow and the lower ones are indistinct in places, as at F, the 
 Upper Level having been entirely removed. The broad embayment 
 west of F is a depression between drumlins and was formerly 
 much more extensive. When the water dro|)ped to the Upper 
 Level, this was reduced to a shallow bay, and during this time a 
 bar and lagoon Avere formed. The bar stands near the present 
 shore and the lagoon is now a swampy lowland. At Two Mile 
 
142 INLAND LAKES OF MICHIGAN. 
 
 Point, a drumlin, the Post-Nipissing and Upper Level terraces are 
 again well developed with a low ice rampart of rock on the shore 
 of the latter. This rampart runs around the point into Newman 
 Bay where it changes from coarse material to sand and has the 
 characteristics of a storm beach. It is readily recognized by a row 
 of pines growing on its surface. Back of it stands the sandy ter- 
 race of the Post-Nipissing level, and in front the broad, sandy ter- 
 race of the Upper Level extends to the present shore. The loose 
 sands on this terrace are being blown into small, irregular dunes. 
 Beyond the bay to Charlevoix the upland stands near the shore 
 and the terraces are very distinct. 
 
 As may be inferred from the description above, the adjustments 
 of the shores of Pine Lake, although moderate in effects, have been 
 numerous and were made at the higher levels. Probably the greatest 
 changes took place at the Upper Level, and the most noticeable of 
 these was the development of bars across the more prominent embay- 
 ments. This level was abandoned less than fifty years ago and only 
 minor adjustments have occurred ^nce that time. The main work at 
 present is the cutting by waves and this has seldom advanced to 
 the limits of the Upper Level. Incipient current forms are present 
 in a few localities but are almost negligible when compared with 
 those formed at the previous levels. A well-developed, but narrow, 
 submerged terrace is continuous around the lake and varies in 
 depth at its outer edge from three feet off sheltered shores to ten 
 feet, the latter being more nearly representative for the lake as a 
 whole. The interpretation of this terrace in terms of wave lengths 
 developed on the lake is uncertain on account of the small differ- 
 ences in elevation between the Post-Nipissing, Upper and present 
 levels, in all nine feet. Nowhere was the undisturbed outer slope 
 of the Post-Nipissing terrace seen, while, as a rule, the Upper 
 Level terrace is continuous with that of the present. Thus, the 
 difficulty arises of determining at which level the present submerged 
 terrace developed. The Post-Nipissing adjustments are much in- 
 ferior to those of the Upper Level and, since we know that the 
 present conditions have existed for an insignificant period of time, 
 it seems safe to conclude that the present submerged terrace is the 
 unexposed portion of that formed during the Upper Level. Neg- 
 lecting the small amount of lowering of the surface of the terrace 
 which may have taken place since that time, the depth was about 
 fourteen feet at the outer edge for most parts of the lake. This is 
 probably somewhat less than one-half of the wave length of storm 
 waves on this lake. 
 
 The lack of adjustment at the present level is due in part to 
 
LAKES OF THE GRAND TRAVERSE REGION. 143 
 
 its short i)eriod of existence and also to tlie fact that adjnstments 
 were of an advanced stage during the preceding level. Thus, the 
 embayments were largely reduced and the submerged terrace was 
 very Hat. The latter must materially reduce the force of the 
 breakers and will have to be lowered from the outer edge before 
 the waves can strike the shores with normal force. The planation 
 of this terrace will be slow because it is covered with a deposit of 
 marl on most shores, which is very compact when wet and also 
 furnishes no tools to aid in the work of the waves. When the 
 waves are tinally able to work effectively, the headlands will be 
 rapidly cut away and the broad embayments built out rather than 
 cut off by currents, making a mature shoreline. 
 
 The extinction of a large and deep lake such as Pine is an ex- 
 tremely slow process. Filling is proceeding at a very slow rate on 
 account of the small and infrequent influents. This will increase 
 as the streams enlarge their basins but, up to the present time, has, 
 had little effect except at the end of the South Arm. Filling by marl 
 and peat near the shores may have some importance but, as a rule,, 
 vegetation finds difficulty in getting started on a wave-swept shore.- 
 It seems probable then that the future of Pine Lake is linked] 
 with that of Lake Michigan. If the tilting of the Great Lakes basin 
 shall be sufficient to lower the level of Lake Michigan more than 
 one hundred feet, Pine Lake will be drained. Otherwise, the slow 
 process of filling will cause extinction. 
 
 TORCHLIGHT LAKE 
 
 Bordering Grand Traverse Bay on its eastern side are two deep 
 troughs in which Torchlight and Elk Lakes lie. These troughs 
 are depressions in the drumlin area mentioned in the discussion 
 of Pine Lake and run almost north-south, conforming very closely 
 to the trend of the drumlins. As may be seen from a map of the 
 Traverse region, these basins lie oblique to the eastern side of 
 Grand Traverse Bay, approaching it at the northern ends. The 
 narrow strips of land separating the lakes from the bay are at the 
 north ends in both cases a series of bars which developed early in 
 the history of the lakes and have been blown into dunes. Thus, the 
 lakes, from one viewpoint, may be classed as lagoons, but the de- 
 pressions themselves are similar in formation to that of Pine Lake 
 and have been briefly discussed in Chapter I. 
 
 The larger and more easterly basin is occupied by Torchlight 
 Lake which is connected with Elk Lake tlirough Pound Lake, Fig. 
 54. Torchlight Lake is one of the larger lakes of the State, its 
 area being 28.5, and is known to be deep, lalthough systematic 
 
144 
 
 INLAND LAKES OF MICHIGAN. 
 
 Fig. 54. Map of Torchlight Lake 
 
LAKES OF THE GRAND TRAVERSE REGION. 145 
 
 soiiiuliiigs liave not been made. Its U'ii<>tli is slightly less than 
 eighteen miles, the longest of the inland lakes of our state, and its 
 width nowhere exceeds two and one-half miles, the average being 
 cousiderabh' less than this figure. So nearly is the lake oriented 
 north and south that only at the northwestern end does it cross a 
 range line. Furthermore, the outline of the lake is consistently 
 regular. As a consequence of its regular configuration, its size, 
 and orientation, the lake becomes dangerously rough during the 
 ^'blows'' from the north or south. Although this may be disad- 
 vantageous for navigation by small boats, it is productive of in- 
 tense wave and current action, and numerous and imjiortant adjust- 
 ments of the shores may be anticipated. 
 
 As may be inferred from the description of Pine Lake basin, the 
 history of Torchlight has much in common. The Algonquin and 
 Nipissing shore lines stand out ]>rominently on the smooth slopes 
 of the drumlins and are counterparts of those on Pine Lake. Also 
 a Post-]Mpissing stage is to be found in favorable localities and, 
 where present, is but slightly above the present level. The present 
 stage is artificially maintained by a dam at Elk Kapids which has 
 held the water above its normal level for about seventy years. Tlie 
 head of water at the dam is seven and one-half feet, but the amount 
 of flooding of the lake cannot well be determined, due to lack of 
 lioth physiographic evidence and human records, although it is 
 believed to be much less than might be inferred from the height of 
 the dam. It is well to keep in mind, however, that an apparently 
 insignificant raising of a lake level may attain considerable im- 
 portance as the shores develop under the new conditions. As a 
 matter of fact, this lake is an excellent example of the effects pro- 
 duced under such circumstances. 
 
 This lake is readily reached by the Pere Marquette R. R. at Alden 
 near the southern end, and is patronized annually by numerous 
 visitors seeking recreation and relief from the summer heat. Its 
 proximity to Grand Traverse Bay mitigates temperatures and its 
 almost parallel trend with the hax makes this condition uniform 
 over the entire lake. Many excellent locations for cottages are to 
 be found, but as j-et they are largely limited to the numerous points 
 and the south end. The size of the lake and its flooded condition 
 'make storms especially severe, therefore a sheltered location is 
 essential if boating is to be enjoyed. Such locations are found at 
 Clam River and on the south sides of the points. Where shores 
 exposed to wave action have been utilized, it has been necessary to 
 build breakwaters of some kind to prevent the rapid recession of 
 the cliffs. 
 
 19 
 
146 INLAND LAKES OF MICHIGAN. 
 
 Navigation is possible on the lake and its connecting waters 
 through Elk Lake to Elk Rapids, and when visited by the writer 
 daily service was maintained by boat from Elk Rapids as far as 
 Clam River. The surrounding country- is a rich agricultural section 
 and the feasibility of connecting this chain of lakes with Grand 
 Traverse Bay is a problem for future development. 
 
 For the visitor with physiographic bent, Alden is a convenient 
 starting point. See map, Fig. 54. Characteristic morainic topog- 
 raphy borders the east side of the lake along this shore and extends 
 northward beyond Clam River, a distance of about six miles. From 
 this locality to the north end of the lake and along the entire west 
 shore, the smooth and rather steep side-slopes of drumlins rise a 
 hundred feet or more above the lake. 
 
 The view across the lake from Alden is most pleasing, but to 
 the practiced eye the significant observation is the terracing of the 
 slopes. An inspection of the immediate surroundings discloses 
 these terraces at hand for closer study. Two distinct terraces, 
 separated by a grass-covered slope, are easily discernible, and 
 measurement places their elevations at thirty-eight and fourteen 
 feet above the lake level. The higher is the Algonquin beach and 
 the lower indicates the level at which Lake Nipissing stood. 
 
 Along this shore northerly winds have full sweep of the lake and 
 waves of great power are developed. As a result, the intense cut- 
 ting has formed low cliffs in the Nipissing terrace which, in places, 
 is composed of stratified sand and gravel, see Plate VII, A. This 
 represents the outer or built portion of the terrace, but farther 
 inland the boulders which are scattered over the sand covered sur- 
 face of the terrace are an indication that here the terrace was cut 
 in boulder clay and a veneer of sand was later deposited on its 
 surface. Further evidence of a cut portion of the terrace is found 
 in the numerous springs issuing from the base of the cliff which 
 rises from this terrace to the Algonquin, These springs are caused 
 by the surface water seeping down through the built portion of the 
 Algonquin terrace to the impervious underlying till, which was ex- 
 posed by waves during Nipissing time. The ground water which is 
 unable to tlow in the compact boulder clay seeps laterally along its 
 upper surface and issues as springs where the claj^ is exposed. The 
 accompanying diagram. Fig. 55, illustrates the conditions described. 
 In addition to the exposed terraces, there is a well defined sub- 
 merged terrace of var;^T.ng width and depth that virtually surrounds 
 the lake, and its outer slope is known locally as the "channel bank." 
 The description above holds, in general, for the east shore of the 
 lake south of Clam River. Locally, conditions have varied and a 
 
Michigan Oculouiciil mihI 
 Biolotjical Survey 
 
 I'nliliciilion ;ui. (Ji'iilouical Sorirs 2."!, 
 
 riiUu VIII. 
 
 ^t:*^7 .'*<*, ^^I', 
 
 *- *»(, 
 
 >. t 
 
 --.ii.V.T' 
 
 A. STRATIFIED EDGE OF BUILT-TEKKACE. TORCHLIGHT LAKE. 
 
 B. RArSED BOULDER STRAND. TOR€IILIGIIT LAKE. 
 
LAKES OF THE GRAND TRAVERSE REGION. 
 
 147 
 
 diversity iu the ilevelonmeiit of the .shore features i.s found at tlie 
 various levels, including the present. Much of the adjustment of 
 the shores has taken place at the liiglier levels, thus deterniinilig 
 to a large extent adjustment at the present level. 
 
 At the present shore the cliff, which is receding into the Xipis- 
 sing terrace, is almost continuous and stands, usually, six to eight 
 feet high. On the longer reaches, this cliff is composed^of sand and 
 gravel, which is often stratified, and at its base are smooth sand 
 beaches. , At the minor projections the material is resistant boulder 
 clay and the beaches are of coarser material. These points are 
 caused by morainic knobs which were formerly more prominent 
 but have been worn back by wave action. The intervening embay- 
 ments, however, were never pronounced, and the material derived 
 from intense wave action on the north sides of the salients was 
 deposited in these bays in comparatively wide built terraces rather 
 than in distinct bars. Thus, when the lake level subsided and the 
 terrace was exposed, the shore line was made more regular. In one 
 
 /}/gonguin terrace A//p/ss/ng terrace 
 
 1 I ' ■ I 
 
 . 1 
 
 Cof Aa//t ' ci/t 
 
 _ bai/t 
 
 
 J 
 
 
 \ 1 
 
 *>■■■ 
 
 1 ' 1 ■ r 
 1 1 
 
 
 
 
 
 O: 
 
 » . -P , ■ •, o • ■■■ ■ . ' ; ■ ° " '^-"^^ 
 
 ___&u/^ 
 
 
 
 
 « ■ 
 
 "' -L .' .'. -■ ' h ou/der c/au - ' '9 
 
 ■ ■ a ^-^ 
 
 v"!" 
 
 H^ 
 
 /.o/ivr-f 
 
 
 
 j 
 
 • ■•• "■.".' ■. : :■' ' "" .•".•■•«"..•'.■■ ° 
 
 
 -" ■' 
 
 c» ., 
 
 -^~r-^^I?\ 
 
 
 - . ■ • •' . ' ■ ■^ . '• . ■ ■ ' ■ •■".'•'■'''•. 
 
 
 , 
 
 
 ■ o r'*^ "^ -=^"i,\ — 
 
 
 
 ■ ■ . ^ 
 
 H 
 
 . '' 
 
 .«. ■ .*-^^' 
 
 Fig. 55. Diagram showing terraces about the sliorc-s of Torchlight Lake. Note the 
 location of springs at the base of the Nipissiug cliff which has receded into the 
 built portion of the Algonquin terrace. 
 
 locality, point A, the process was aided by the formation of a small 
 delta at the Algonquin level by a stream, now dry. The blunt 
 projection north of locality A shows an excessive amount of cutting 
 on the north side, where the waves have removed the Nipissing 
 shore and are now attacking the Algonquin terrace, forming cliffs 
 of considerable height. 
 
 The Algonquin terrace bevels the neck of land between Thayers 
 Lake and Torchlight, which shows that the two basins were con- 
 nected during Algonquin time. At first, the connection was re- 
 stricted to the outlet of Thayers Lake but widened as wave action 
 reduced the narrow headland which separated them. During 
 Nipissing time, the lakes were entirely separated, since the Xipis- 
 sing terrace does not enter Thavers Lake basin. 
 
148 INLAND LAKES OP MICHIGAN. 
 
 In the wide embaymeut one-half mile south of Lone Tree Point, 
 concentric sand bars are found on the Nipissing* terrace. At the 
 north end two such bars are present but they split and double in 
 number to the south. These do not appear to be the typical bars 
 which are built by currents across the neck of an indentation, since 
 there is no indication of lagoons and no abrupt change in direction 
 of the shore to cause the currents to swing cut. They are better 
 interpreted either as storm beaches or a series of submerged sand 
 bars, sometimes found under similar conditions, which are probably 
 formed simultaneously by breakers during a storm. The close 
 assortment and fineness of the material makes the latter interj)re- 
 tation the more probable. 
 
 Lone Tree Point is the most prominent projection in this part 
 of the lake and was originally due to a morainic knob near the 
 lake. This knob was bevelled by the waves of Lake Algonquin, but 
 the waters of Nipissing succeeded only in notching its lakeward 
 side and forming a terrace of considerable width. When the water 
 level dropped from the Nipissing stage, this projection was suffi- 
 cient to turn the currents out into the lake, and deposition rather 
 than cutting became the predominant process. Near the present 
 level, a spit was built at the end of the jjoint, the main portion of 
 which stands two to three feet above the water and represents a 
 former level of the lake. This we shall call the Upper Level. 
 
 Points, such as Lone Tree, which are the result of currents leav- 
 ing the shore, are interesting and instructive because they serve as 
 indices to the effectiveness of the forces acting. Currents may run 
 in opposite directions along a shore, depending on the direction of 
 the winds, and their strength is determined by the force and direc- 
 tion of the waves. Not only do the stronger currents transport 
 proportionally greater amounts of -material than the weaker, but 
 they deposit it in forms which are more nearly in line with the 
 shore at the point of departure, i. e., have a lesser curvature. In 
 the case under consideration, the curvature of the north side of the 
 spit is much less than the south, indicating a much stronger cur- 
 rent from the northerly direction. There is little or no difference 
 in the resistance of the material upon which the waves are working 
 along the shores of this lake and the strength attained by the 
 currents must be determined by the force of the winds. Winds 
 whose directions have no easterly component are effective on this 
 shore, but there is a preponderance, both as to velocity and fre- 
 quency, of those from the northwesterly quadrant rather than from 
 the southwest. When we consider the added advantage of a reach 
 twice as great for northerly winds at this point, the unsymmetrical 
 
LAKES OF THE GRAND TRAVERSE REGION. 149 
 
 form of tlie spit is readily ap]>r(H-iate(l. Yel waves and cnrreiits of 
 considerable force are active on the sontli side, as is shown by the 
 even, alth()n<;li sharp, cnrvatnre of the shoreline and the presence 
 of storm beaches at the Upper level. The superior strength of the 
 forces at Avork on the north side is shown at the present time by the 
 decided contrast in the work accomplished under the tiooded con- 
 dition. Current action is still effective on the south side, but on the 
 north the waves are cutting- back the point and have necessitated 
 some form of breakwater. In this connection it may be stated that 
 the name Lone Tree is no longer appropriate, for the solitary sen- 
 tinel has long since succumbed to the force of the waves. 
 
 Northward from Lone Tree I'oint the shore swings to the north- 
 east and is exposed to the full sweep of northwesterly winds. Much 
 cutting by the waves is taking place and the cliff's in the Nipissing 
 terrace are rapidly receding, causing considerable anxiety to the 
 cottage owners in the locality. Breakwaters of brush, placed with 
 twig ends outward, seem to prove temporarily effective. The blunt 
 point presents a decided contrast to Lone Tree Point in that no 
 deposition has taken place here. The Algonquin and Nipissing 
 beaches swing back into the narrow depression in which Clam 
 Lake lies but reappear on* the north side. The smooth beach with 
 cliff, above which stand the Xipissing and Algonquin terraces in 
 turn, are present as far as Balls Point. This point is clearly the 
 result of deposition by shore currents of considerable power, as 
 shown by the rather coarse, but assorted, material. It is difficult 
 to assign reasons for the currents leaving the shore at this point, 
 inasmuch as the configuration of the bottom is not known, but, 
 nevertheless, those from both directions do so. The currents from 
 the north, however, are the more powerful and have laid down a 
 much heavier deposit on the north side. In general, a submerged 
 terrace of considerable width is present along the shore, but it 
 widens to nearly one-fourth mile off the point and drops into deep 
 water at eight feet. The great quantity of deposited material shows 
 intensive action at this point, and a widening of Nipissing and 
 present terraces shows also that this has occurred since Algonquin 
 time. Most of the work was done during Nipissing time, with con- 
 siderable addition at the i»resent level during which the wide sub- 
 merged terrace has been formed. A small but interesting ice 
 rampart was found at the very tip of the point, off which the lake 
 has a width of nearly two miles. This width' is too great for ice 
 expansion and, since ice jams are known to have been effective at 
 one locality on the lake, it is probable that this agent has been 
 effective here. 
 
150 INLAND LAKES OF MICHIGAN. 
 
 North of Balls Point the shores offer nothing of additional in- 
 terest until point B (see map) is reached. This is one of the best 
 developed points on the lake, having a length of five to six hundred 
 feet, and is similar to Balls, although much sharper. As it is ap- 
 proached from the south, there first appears along the shore a cliff 
 rising to the full height of the Nipissing terrace, but as the point 
 is. reached the cliff drops to a height of eight feet and later gives 
 way to a lower terrace fronted by a storm beach. The explanation 
 is furnished by the topography of the point which is shown in a 
 conventional sketch. Fig. 56. From the sketch it will be noted that 
 
 Fig. 56. 'Conventional sketch of point designated as B on map, Torcliliglit Lake. 
 
 the highest or Algonquin terrace does not widen at the point but 
 that below this are three roughly triangular terraces which stand 
 at successively lower elevations and are separated by low cliffs. 
 These cliffs diverge somewhat as they cross the point in a north- 
 westerly direction and end abruptly at the cliff' on the north side. 
 The highest of these terraces is the Nipissing and the rise from the 
 lake level to this is accomplished in three steps. The surface of 
 the lowest terrace is somewhat irregular, but those above are of 
 characteristic slope and surface, except for a depression in the 
 Mpissing terrace which, although slight, is quite noticeable. The 
 point started to develop during the Nipissing stage and was ex- 
 tended beyond the present limits of this terrace. The depression 
 in this terrace shows that currents from both directions left the 
 shore in this vicinity and developed spits which met some distance 
 oft' shore and formed a point, a V-bar, see Chapter II. Its position 
 was slightly north of the present point. Then the water dropped 
 to the next lower level, called the Post-Nipissing, of which we know 
 little except that the waves were active on the south side of this 
 point and cut a well defined cliff in the Nipissing terrace. The 
 
LAKES OF THE GRAND TRAVERSE REGION. 151 
 
 Upper LeveJ is represented by the terrace ■which stands below the 
 Post-Nipissing and extends to the present shore ot the lake. The 
 uneven surface of this terrace suggests a somewhat different manner 
 of formation than for the smooth surface of the typical cut-aud- 
 built terrace which develops under water. Close examination dis- 
 closes the presence of indistinct ridges, and the deposit may, there- 
 fore, be interpreted as a series of poorly defined storm beaches 
 modified by ice shove. 
 
 At present the point is being cut back on the north side, and the 
 material is either being transferred to the south side or carried 
 out into the lake. On the south side deposition predominates, al- 
 though waves of considerable power are active, as shown by a re- 
 cent storm beach composed of pebbles up to three inches in diameter. 
 In general, it may be stated that the point is gradually shifting 
 southward and possibly being diminished in size. 
 
 North of the point the two lower terraces are absent but the 
 Nipissing and Algonquin are well developed, especially at j^oint C 
 where the waves are now cutting into a hill of boulder clay. The 
 submerged terrace is here scattered with boulders and is, therefore, 
 formed by cutting rather than by deposition, a condition infrequent- 
 ly met at present on this lake. Beyond C shore action decreases 
 and the Upper terrace reappears on the grass covered slopes. 
 
 At point D an interesting variation of the general shore condi- 
 tions is to be found. At the present level an ice rampart lines the 
 shore and causes a swampy condition on the gently sloping surface 
 of the Upper Level terrace back of it. This terrace rises gradually 
 to the low front slope of the Nipissing terrace, which apparently 
 was not attacked by waves during the Upper Level stage and stands 
 at its original width. Similarly, the wide Nipissing terrace is 
 bounded inland by a cliff of such gentle slope that it may be con- 
 sidered the original front slope of the Algonquin terrace. 
 
 Xorthward from this locality two small streams flow through a 
 sag in the hills and cross the terraces. Singularly, these streams 
 have been able to deepen their channels onh^ in the Algonquin ter- 
 race which here has the characteristics of a small delta. If this 
 sudden change in the activity of the streams were due solely to 
 the fact that the older terraces have been exposed to their action 
 for a longer period of time, one might expect a gradational decrease 
 in the amount of cutting in the lower terraces. Btit the change is 
 abrupt and it is probable that a large decrease in the volume of 
 the streams occurred as the water dropped to the Nipissiug level, 
 indicating a climatic change. 
 
 Northward towards point E the shore swings to the northwest 
 
152 
 
 INLAND LAKES OF MICHIGAN. 
 
 with no unusual variation in the shore. The slopes have not been 
 cleared, and the trees which grow to the water's edge have effi- 
 ciently protected the shores so that cliffs are rare. The monotony 
 is relieved at point E near which is located the State Y. M. C. A. 
 camp. The bend in the shore line at the point caused the currents 
 to leave the shore and deposit their suspended material in a form 
 which is almost perfectly preserved. The point had its inception 
 during Nipissing time and developed into a perfect hook from 
 the south, see Fig. 57. The curvature of the hook was greater 
 
 'IMf^-^ ^'lll-r^^l^' Wy 
 
 X- ^^^-^^' ^^^^^^' 
 
 Fig. 57. Conventional sketch of point near Y. M. C. A. camp, Torchlight Lake. 
 
 than that of the main shore and the re-curved portion was grow- 
 ing almost directly toward the mainland. The narrow channel 
 which connected the lagoon, thus formed, with the lake was partial- 
 ly filled b}^ the development of a submerged bar from the north 
 side. This form will be recognized as an unsymmetrical V-bar in 
 process of formation. 
 
 Below the Nipissing level is a narrow bench at the Upper Level 
 and at the present shore there is an accumulation of fresh gravel. 
 The Post-Nipissing is absent. During the Upper Level the shores 
 were continuous on both sides of the point and the development 
 was outward into the lake with less tendency towards growth to 
 the north. This is also true at the present time, as shown by the 
 relatively larger accumulation at the end of the point. 
 
 The points on this shore previously discussed indicate strong 
 currents from the north but here conditions are reversed. It will 
 be seen from the location of the point near the north end of the 
 lake that the reach of the waves is greater from the south and is 
 the controlling factor rather than the prevalence of storm winds. 
 
LAKES OF THE GRAND TRAVERSE REGION. 153 
 
 As might be expected from the study of point E, the evidence 
 of wave action is noticeably less along the shore from this paint 
 to Eastport, situated at the north end of the lake. At the present 
 shore fresh cliffs are mnch less frequent and the Nipissing terrace 
 is relatively narrow, the intermediate levels being absent. About 
 one mile north of point E a line of boulders forced into the base of 
 a cliff is evidence of strong ice push along this shore. 
 
 At Eastport a sand beach curves around the north end of the 
 lake and the hinterland rises very gradually to a well defined bar 
 at the Nipissing level, upon which much of the town is located. 
 Back of the bar there is a sandy depression dotted with small sand 
 dunes. The narrow neck which here separates Torchlight Lake 
 from Lake Michigan is of sand and stands at the Algonquin level 
 except for a zone of dunes which rise to a maximum of twenty-five 
 feet. Beyond the dunes a series of parallel bars with intervening 
 lagoons extends to the cliff overlooking Lake Michigan. From this 
 description it is evident that the north end of Torchlight lake was 
 connected with Lake Michigan and was cut off during Algonquin 
 time by a series of bars. As the water receded to the Nipissing 
 level, the earlier bars were blown into dunes and the somewhat ir- 
 regular outline of the north end of Torchlight lake was straightened 
 by the development of the bar at Eastport. The lower levels were 
 not productive of adjustments here. This end of the lake is sub- 
 jected to strong ice jams in the spring, but the sandy beach with its 
 scant vegetation is not favorable for the formation of permanent 
 ramparts. However, east of the dock there is a row of poplars a 
 few feet back from the shore. The roots of the trees have served 
 as binding material for the sand, and a well preserved rampart has 
 been formed in front of the row, but disappears abruptly at each 
 end. Observation of this rampart in process of formation, by in- 
 habitants of the locality, makes it certain that ice jams exerted the 
 shove are, therefore, effective on the shores of this lake. 
 
 Along the west side low ground borders the lake and has been 
 converted into a lagoon by the formation of a bar at the present 
 or Upper Level. This continues to point F where the divide be- 
 tween this lake and Lake Michigan rises above the Algonquin level 
 and both the Algonquin and Nipissing terraces are present. Off 
 this point the "'channel bank" is very decided and drops into deep 
 water from a depth of six feet at a rate of almost one to one, the 
 slope of the bottom being from thirty-five to forty degrees. 
 
 At Torch Lake the divide narrows and stands at an elevation 
 which is below the Algonquin beach. Wells in the vicinity penetrate 
 clay before reaching a water-bearing layer and, since no bar is to 
 
154 INLAND LAKES OF MICHIGAN. 
 
 be found on the crest of the divide, we must conclude that a con- 
 nection with Lake Algonquin was open in this locality, although 
 closed at Eastport. This connection was nearly two miles in width, 
 reaching from point F to point G. Further evidence of an open con- 
 nection was found in the vicinity of point G, where a strong spit 
 at the Algonquin level runs in a southeasterly direction into the 
 Torchlight basin from a point of the upland on the south side of 
 the strait. 
 
 , It appears, then, that the north end of the Torchlight basin was 
 connected with that of Grand Traverse Bay by a double connection 
 during Algonquin time. The adjustments of the shores of Lake x4.1- 
 gonquin were numerous and diverse, and, in general, it miay be stated 
 that virtually all indentations, such as those occupied b}^ the border 
 lakes, were isolated b}^ the development of bars. In this case, it is 
 exceptional that only the northerly connection was closed, and the 
 most reasonable explanation involves the factors of effective winds 
 and available material. The westerly to northerly winds were the 
 most effective and bars developed from the north along the shore 
 of Lake Algonquin. The long stretch of shore below Charlevoix 
 furnished sufficient material for the bar across the north channel 
 but the limited amount of land between the two channels, F, was 
 inadequate for a similar development across the sotith channel. 
 In addition, the westerly winds were able to turn the limited de- 
 posits on the south side of this channel almost directly into the 
 Torchlight basin, as shown by the bar back of Point G already 
 mentioned, and so kept the chajinel open. 
 
 Along the west shore wave action is less intense than on the 
 opposite shore and the terraces are somewhat better preserved, al- 
 though not so well developed. Below the Algonquin level, the 
 Nipissing terrace is alwa5^s well developed but that of the Post- 
 Nipissing stage is very poorly defined. The points were as a rule 
 started during Nipissing time but considerable additions were 
 made during the Upper Level stage. This level may also be recog- 
 nized in some of the bays, for example, that north of point H, either 
 as a terrace or as ice ramparts. 
 
 Point H started its development in Nipissing time as a V-bar 
 which now stands slightly south of the present point. The two 
 bars, enclosing a depression more than ten feet deep, are excellently 
 preserved. The greatest deposition occurred during the Upper Level 
 and formed the main portion of the point. At present it is being 
 cut away on the north side but is increasing on the south and at 
 the end of the point. The deposition is further shown by the broad 
 submerged terrace on the south side. Comparing conditions at this 
 
Jlicliisjin (iiMilduical and 
 Bioliijiical Siuvry 
 
 rul)licati(m :'.(». ( H'olniii.-al S('n( s 2."), 
 I'latu IX. 
 
 A. AL(;()X(,)rix r.AK, torciili<;iit lake. 
 
 
 B. ALGOXQUIX BAR, ELK LAKE. 
 
LAKES OF THE GRAND TRAVERSE REGION. 155 
 
 point with tliose at E across tlic lake. \\(' liiid tlioin reversed, tliat is. 
 E has been bnilt mainly by southerly currents and II by northerly 
 currents. 
 
 (South of point H the sl()i)es are gentle and rise to the Nipissing 
 level Avith slight indications of intermediate stages. At point I 
 a V-bar with charaeleristics and history almost identical with those 
 already described l)reaks the rather even shoreline. The chief in- 
 terest lies iu the amount of deposition that has takeu place in re- 
 cent times. A long spit extends fully one hundred yards beyond the 
 portion built at the Upper Level in a direction somewhat south of 
 east but is being cut away on the north side under the present 
 flooded condition of the lake. 
 
 South of I the usual shore conditions prevail except where druni- 
 lins approach the lake. Here the currents have left the shore and 
 two such points occur before point J is reached. At J a large 
 drumlin caused the original projection in the shoreline and became 
 a locality of intensified wave action. During the Algonquin and 
 Nipissing stages well defined terraces and cliffs were cut and the 
 side of the drumlin was steepened considerably. However, late iu 
 Nipissing time conditions changed and two small V-bars developed, 
 which have been enlarged at the lower levels, making a double 
 point. It is interesting to note that the more northerly point is 
 being added to under the present conditions and especially on the 
 north side, while the southerly one is being worn away. On the 
 latter numerous large boulders have been lined on the shore by ice 
 action. 
 
 In general, on this lake the adjustments of the shores at levels 
 below the Algonquin are the more important, due largely to the 
 fact that cutting by weaves predominated almost to the exclusion of 
 currents on the relatively smooth shores in this embayment of 
 Lake Algonquin. Thus the rather monotonous description of the 
 terrace and cliffs of this shore has not been dwelt upon. On the 
 west side of the lake the topography is less regular, and several 
 small indentations were encountered at this level within a distance 
 of six miles south of point J. The result was a straightening of 
 the shore by the development of completed bars across the mouths 
 of these bays, Plate IX, A. After the drop to the Xipissing level, 
 the waves worked back towards the bars and replaced the gentle 
 front slopes by a steep cliff. In no case was the bar entirely re- 
 moved and the remnants, with their flat tops and steep side 
 slopes, now resemble railroad embankments. Ice also was active 
 during the Xipissing stage but shore conditions were such that the 
 
156 INLAND LAKES OF MICHIGAN. 
 
 most noticeable result was the lining of boulders on the beach, il- 
 lustrated in Plate VIII, B. 
 
 Immediately south of point J the first of the bars is encountered, 
 but here the Nipissing shore did not advance far inland and the 
 embankment effect is not so pronounced. Between this point and 
 Parks five similar features are to be found. The indentations were 
 all small and at present furnish limited drainage basins, so that 
 the bars are, with one exception, intact and the low ground adjacent 
 to the bars is swampy. In the south part of section 7, T. 29 N., R. S 
 W, such an indentation of larger size furnished sufficient surface 
 water to cut a drainage channel through the bar. The intervening 
 stretches are marked by relatively narrow terraces of the Algonquin 
 and Nipissing levels which widen at the well developed points 
 indicated on the map as K and L. The higher terraces are complete, 
 except for a short stretch south of Parks where the Nipissing shore 
 has been removed. The waves are working into the front slope of 
 the Algonquin terrace but have not as yet extended the cliff to its 
 full height. 
 
 The feebleness of the shores of the exposed levels and the presence 
 of a narrow submerged terrace at the present level mark this shore 
 as one subjected to relatively light wave action. Winds from the 
 northeasterly quadrant are the most effective, having an excessive 
 reach which embraces the whole length of the lake. This is clearly 
 shown by the points K and L which are being built to the south. 
 Also the ''channel bank" in each case is wide on the north side (the 
 side of stronger wave action) but becomes much narrower on the 
 lee, or south, side. Although this shore has been subjected to 
 light wave action, on the average, it must not be inferred that the 
 waves are of meager development. On the contrary, as study of 
 points K and L show^s, the Avaves during occasional severe 
 storms beat with great power against this shore. Both points 
 started their growth during Lake Nipissing and developed into 
 small V-bars. These bars have since been added to mainly at the 
 present level, and in each case the deposit is in the form of storm 
 beaches. Thus, at K are found two well developed storm beaches 
 on the north side of the point, which merge into a single one on the 
 south side and enclose a triangular cedar swamp. At L the storm 
 beaches are not so pronounced and are confined to a series of soutli- 
 ward extending loops at the tip of the point. The activity seems to 
 be less on the more southerly point L, an observation readily con- 
 firmed at M, the last point on the lake. The latter is a simple 
 broad point trending southward with no indication of storm 
 beaches. 
 
LAKES OF THE GRAND TRAVERSE REGION. 157 
 
 The southward extension of the lake basin is a swampy flat hnt 
 slightly above the lake level, leading directly to Bound Lake. This 
 was formerly a part of the much larger Algonquin and Nipissing 
 lakes, which included Round and Elk in addition to Torchlight 
 Lake. Torch River leaves the lake at the extreme southwest corner, 
 a fact readily explained if the south shore is traversed. Starting 
 at the river one soon notices a low sand ridge which gradually in- 
 creases in strength and elevation and swings with even curvature 
 to the east side of the lake. This is unmistakably a bar, although 
 its profile is somewhat obscured by an ice rampart on the front slope^ 
 and may be easily traced to the Nipissing shore. Thus, during Nipis- 
 sing time a bar was developing which would have eventually iso- 
 lated this lake basin, but a drop in the water level accomplished this 
 result before its completion. The submerged terrace is very wide 
 at this end, due to the exceptionally strong undertow developed. 
 This wide terrace together with the clean sand beach fronting the 
 bar, affords excellent bathing facilities and makes the location, 
 Crystal Beach, a favorite with summer visitors. The presence of 
 two sand bars on the submerged terrace, parallel to each other and 
 the shore, is an interesting development. These bars shift during 
 storms and may vary in number and height but, as far as is known, 
 never reach above the water level. Since they are formed by break- 
 ers, their growth into true barriers is a possibility but it seems more 
 probable that a depression of the water level suflftcieiit to expose 
 a portion of them is necessary for further development. Many 
 examples of such forms, composed of a series of parallel bars, are 
 to be found in the Great Lakes and on their former shores now ex- 
 posed, e. g., at the north end of Torchlight Lake, but this is the only 
 inland lake in which more than a single bar was found and is there- 
 fore noteworthy. 
 
 Having traversed the shores of the lake, Ave may now attempt a 
 resume of its history and conditions. During Algonquin time the 
 basin was part of an archipelago which included Elk, Round and 
 Torchlight Lakes and extended into the depression now occupied 
 by Clam, Grass and Intermediate Lakes, which in turn was con- 
 nected with the South Arm of Pine Lake. The connection between 
 the basin of this lake and Grand Traverse Bay was a double one at 
 the north end but was partially closed at this time. The develop- 
 ment of the shores was largely by waves and the terraces are now 
 continuous on both sides of the lake. Currents were effective locally 
 on the west side and succeeded in throwing bars across some minor 
 embayments and in. forming a large spit on the south side of the 
 open connection with the main lake at the north end. 
 
158 INLAND LAKES OF MICHIGAN. 
 
 With the recession of the water to the Nij)issing level the basin 
 was definitely separated from the main lake at the north end and, 
 later, partially so at the south by the growth of a bar which was 
 largely submerged. The Nipissing Level was such that wave action 
 during this time encroached on the Algonquin Terrace, forming a 
 steep cliff: and terrace somewhat inferior in development to that 
 of the Algonquin stage. However, currents assumed a more im- 
 portant role and started the development of the present points. 
 With one or two exceptions, the deposits were V-bars which varied 
 in sjanmetry according to their position on the lake shore. Also 
 in some of the broader embayments they aided in increasing the 
 width of the submerged terrace, thus straightening the shore line 
 when the water receded from this level. The most pronounced ad- 
 justment of the shores was accomplished at this time by the de- 
 velopment of bars at both the north and south ends of the lake. 
 
 The Post-Nipissing level was of short duration and the forms 
 were of inferior development. In fact, were it not for the distinct 
 terrace at point B, its recognition would be difficult. 
 
 The Upper Level was of considerable duration and is recogniz- 
 able largely by the depositional forms existing. Nearly all of the 
 points show considerable growth at this level and these fornis 
 gradually merge into those being formed at present. In addition 
 to the growth of the points, considerable low ground was cut off 
 by the development of a bar at the northeast end. 
 
 As has already been stated, the lake is now in a flooded condi- 
 tion. Wave action is very active on all parts of the shore exposed 
 to strong winds, and cliffs are common. These cliffs are receding 
 rapidly and in a few places have removed the Nipissing terrace. 
 The points also show the effect of the increased activity and are 
 being eroded on one side at least. Erosion will continue until 
 equilibrium is established and will result in continued cliff re- 
 cession and in a reduction or shifting of the positions of the points. 
 Also the abundant wave-worked material will add greatly to the 
 submerged terrace. As the activity of the waves decreases some- 
 what, currents may be more effective, causing 'sl greater growth of 
 the points. 
 
 The final limit of point expansion would divide the lake into 
 several smaller bodies but wave action can hardly be expected to 
 furnish enough material on such a deep lake. Tributary streams 
 are few and short and the only large one. Clam Kiver, drains a 
 nearby lake. Therefore little sediment can be supplied in this way. 
 Little reduction in size by the formation of bars is to be expected 
 
LAKES OF THE GRAND TRAVERSE REGION. 159 
 
 since this was accomplished at the higher levels in the few locali- 
 ties where conditions were favorable. 
 
 Vegetation has hardly made a beginning and cannot take hold as 
 long as the waves continue to actively erode. This lake shows a 
 revival of activity and presents problems of shore development 
 rather than of extinction. 
 
 ELK LAKE 
 
 Elk Lake is another member of the series of lakes which occupy 
 similar basins east of Grand Traverse Bay. These basins were 
 briefly discussed in Chapter I and need no fnrther discussion here.. 
 Elk Lake is slightly over nine miles in length and averages less than^ 
 one and one-balf miles in width, the maximum width nowhere ex- 
 ceeding two miles. See Fig. 58. Its surface covers thirteen square- 
 miles and is, thus, less than half the size of its neighbor, Torchlight. 
 We compare it with Torchlight Lake purposely because of the very 
 striking similarity between these two bodies of water. They occupy 
 similar narrow, regular basins which follow the trend of the flanking 
 drumlins, are oriented nearly north-south, are deep, have many fea- 
 tures in common in the adjustment of their shores, and have passed 
 through the same succession of events in their past. In fact, it 
 vrould be difficult to find two lakes in such close proximity so nearly 
 alike. The same winds and storms have whipped the waters into 
 waves and developed the currents which have adjusted the shores 
 during the same period of time. The variable factor is, then, the 
 size. Differences in shore adjustments, both as to kind and amount, 
 are attributable to this cause. 
 
 Elk Lake is reached by a spur of the Pere Marquette Railroad, 
 which terminated at Elk Eapids, situated on Grand Traverse Bay at 
 the outlet of the lake. As the name indicates the drop in level from 
 Elk Lake to Michigan occurs rather suddenh' near the latter lake, 
 causing a rapids in the outlet. Advantage has been taken of the 
 steeper gradient and the river has been dammed at this point. The 
 history of these operations could not be traced back by the writer, 
 but it is known that a dam on the present site was built prior to 
 185G and has been maintained since that time with a fall of seven 
 or seven and one-half feet. 
 
 Beginning our study at Elk Rapids, we fiud the flooding of the out- 
 let above the dam very noticeable. The current is very slack and tree 
 trunks stand in the water. As the lake is approached, the outlet 
 widens and some wave action is evident in the low cliffs. Above the 
 cliffs at an elevation of fiften feet stands the Nipissing terrace which 
 terminates landward in a cliff reaching up to the Algonquin terrace 
 
160 
 
 INLAND LAKES OF MICHIGAN. 
 
 Fig. 58. Outline map of Elk Lake, Grand Traverse and Antrim Counties. 
 
LAKES OF THE GRAND TRAVERSE REGION. 161 
 
 forty feet above the water. Meguzee Point is caused by a low <lnini- 
 liii which is placed s!i<;htly obli(Hie to the lake and runs to the shore 
 about a mile to the north. On the west side of the point a low 
 terrace skirts the shore, marking a former level which probably is 
 the equivalent of the Upper Level on Torchlight and may be so desig- 
 nated. The end of this blunt point is bounded by cliffs, showing 
 strong wave action, and is fronted by a well-defined submerged ter- 
 race which "drops off" at seven feet about one-hundred yards from 
 the shore. 
 
 The southeast side of the ]>oint slopes gently to the shore where 
 it is being cut into low cliffs. Ice jams have formed a boulder strand, 
 but only patches of ramparts of feeble development are present, 
 although conditions for their formation are favorable. The small 
 amount of cutting on this unprotected shore seems out of proportion 
 to that found on other parts of the lake, and a XJlausilDle explanation, 
 but one w%ich cannot be proven, is that a rampart was formed here 
 under normal conditions of level. Under the j)resent flooded condi- 
 tion, the waves have expended their energy in its removal and are 
 just beginning the process of cliff formation. 
 
 At A, see map, conditions change and the waves have cut a more 
 decided cliff in stratified sand and gravel. This is the built portion 
 of the Nipissing terrace which is not well developed on the point 
 below. The submerged terrace is very definite off A and drops into 
 deep water from a depth of eight feet. The soundings sliow a de- 
 crease in depth just before the "drop off," indicating the presence of 
 a low sand bar. This is probably formed by the violent agitation of 
 the water where the waves first break during severe storms. 
 
 The hill which forms Meguzee Point gradually low^ers north of A 
 and gives wa}' to a swamp opposite Bass Lake. During Nipissing 
 time Elk Lake connected with Grand Traverse Bay through this de- 
 pression, but at present the connection stands slightly above the lake 
 level and is further separated by a low ice rampart along the shore 
 of Elk Lake. 
 
 To the north the upland again appears and forms the broad double 
 point B. This projection, in reality, consists of two headlands sep- 
 arated by a sag. On the headlands the waves have cut cliffs in sand, 
 which are uniformly ten to twelve feet in height, while along the 
 intervening bay the smooth shore is an indication of some current 
 action. However, the most prominent feature is a low ice rampart 
 and boulder strand. At the north end of the point currents leave 
 the vshore and are building a small spit which may eventually enclose 
 the rather deep bay to the north. In this protected bay the weak 
 Upper Level terrace is again found. Soundings off the north part of 
 point B disclose a double submerged terrace. The bottom slopes 
 21 
 
162 INLAND LAKES OF MICHIGAN. 
 
 graclually outward to a depth of about four feet where it drops sud- 
 denly two feet or more to a second terrace which continues outward 
 until a depth of eight feet is reached before it drops into deep water. 
 This double terrace is probably due to the abrupt change in condi- 
 tions coincident with the damming of the outlet. The deeper offshore 
 portion was formed at the lower level previous to the flooding of the 
 lake and upon it has been built the shallow part adjacent to the 
 shore under the conditions existing at present. We may designate 
 the parts as the younger and older but, in reality, they are two dis- 
 tinct terraces. Neither is proportional in development at the pres- 
 ent time to the waves which were instrumental in its formation. The 
 depth of eight feet over the outer edge of the older part is too great 
 inasmuch as we know that the lake has been lifted an undetermined 
 but appreciable amount, and the younger part has begun its develop- 
 ment only in the sixty or more years since the present conditions 
 were inaugurated. 
 
 The highland encircles the north end some distance from the lake, 
 and the lakeward slopes are interrupted by the Nipissing terrace and 
 cliff. The wet, grass-covered terrace of the Upper Level appears near 
 Kewadin and fringes the shore around the narrow arm of the lake 
 at this end. This terrace is so low that the waves have little to 
 work on. In fact this has not been a locality of intense wave action 
 at any time since the isolation of the lake basin. This statement is 
 based on the presence of a strong bar at the Nipissing level which 
 starts in the locality of Kewadin and runs to the uplands on the east 
 side of the lake in" a broad, swinging curve, enclosing a crescent- 
 shaped lagoon. The submerged terrace at this end is exceptionally 
 wide and nearly meets from the opposite sides of the lake otT point 
 C. Here again it has the double character as described for B, 
 the inner part dropping from a depth of three feet and tlie outer at 
 seven. The water is shallower over the terrace thrai at B and this 
 is due to moderate wave action from the southerly winds, although 
 they are of great reach. 
 
 Continuing southward along the east shore, conditions at locality 
 C first attract attention. A narrow knoll not over eight feet in 
 height caused the broad projection of the shore line. Back of this 
 knoll, i. e., east, a strip of swamp runs from the^aiortheastern extrem- 
 ity of the lake south to the eastw^ard bend of the shore and separates 
 the knoll from the upland. During the Upper Level this knoll at 
 first stool as a low island but later was connected to the mainland by 
 a bar which developed at the south end. 
 
 Along this shore the Algonquin and Xipissing shores are much 
 more distinct than on the west side and are well developed in the 
 broad embayment south of C. The Xipissing consists of a cliff and 
 
LAKES OF THE GRAND TRAVERSE REGIOX. 1G3 
 
 narrow terrace Avhich does not reach the present slioi-e. The front 
 slope of this terrace is soniewliat confnsing bnt slu)nhl not be inter- 
 preted as the cliff of a former level of the lake. At D the Nipissing 
 shore is poorly defined but the Algonquin is very strong. A climb up 
 the forty foot rise to the Algonquin level is well worth the effort, 
 for in this vicinity two excellent examples of the straightening of 
 the shoreline by the development of bars across the mouth of an in- 
 dentation may be seen. The first to be encountered is shown in 
 I*Mte IX, B. From the slight sag of the top of the bar it may be 
 inferred that spits developed from both sides of the embayment and 
 that the bar was not quite completed before the subsidence of the 
 lake to the Xipissing level. A similar bar is located about one-half 
 mile to the south. 
 
 Along the present shore to locality E the waves are cutting actively 
 and low cliffs are being formed. The material is largely boulder 
 clay, and the line of boulders on the shore shows that a moderate 
 ice-shove occurs. The "drop off" is very distinct at eight to nine feet, 
 and the submerged terrace has a width of more than one hundred 
 yards in places. Similar conditions extend to the inlet from Eound 
 Lake, except that the Nipissing terrace is being cut away b}' the 
 Araves, and cliffs which reach a maximum height of twelve feet are 
 prevalent along the shore. At F the south side of a drumlin forms 
 the point, upon which the Upper Level is shown by a terrace. Most 
 of the re-curved portion of the point was probably formed by current 
 action but is now being worn away, due to the revived activity of 
 the waves. 
 
 Eound Lake is well protected by highland on all but the west side 
 and the Torchlight depression on the northeast, and shows relatively 
 little shore activity. It supports a heavy growth of water-loving 
 vegetation which is, without doubt, rapidly filling this basin. Skege- 
 mog point, on the south side of the inlet, shows very clearly the 
 contrast in the activity along the shores of the two lakes. On the 
 Round Lake side the shores are low and the Upper Level terrace is 
 well preserved. In addition, the beginning of a spit runs into Round 
 Lake from the end of the point. The material for this spit is derived 
 from the low cliffs along the Elk Lake shore, where no evidence of 
 the Upper Level is to be found. 
 
 Below G the high ground recedes, this recession forming an identa- 
 tion during the Upper Level. The curvature of the shore was suffi- 
 cient to cause the currents to leave the shore, and a submerged bar 
 was formed across the embayment about one hundred feet back 
 from the present shore. At Many I'ines Point a hill causes the projec- 
 tion and the Algonquin and Nipissing terraces again appear. The 
 
164 INLAND LAKES OF MICHIGAN. 
 
 south side of the point is protected from the strong winds and the 
 Upper Level terrace has been preserved locally. 
 
 The submerged terrace reaches its maximum development at the 
 south end of the lake where its width exceeds one-half mile. This 
 shore is exposed to the strongest storm winds which often blow 
 the full length of the lake, and under this condition the undertow 
 attains its strongest development. Similar conditions may confi- 
 dently be assumed during the former lake levels, and a wide com- 
 pound terrace was formed in this locality, which is now the wide 
 swamp extending from the south end of the lake, to the hills a half 
 mile away. This should consist of a series of three steps but it is 
 difficult to determine on account of the mask of vegetation. At the 
 present shore the low sand bar which skirts the entire swamp is 
 being thrown back by the renewed wave activity and is somewhat 
 irregular in outline. The writer is inclined to consider this a storm 
 beach which developed under the conditions previous to the present 
 flooded stage but realizes that it may w^ell be a bar built at the 
 Upper Level. Conclusive evidence, however, is lacking since the form 
 is being rapidly remodeled. 
 
 The west shore south of the outlet rises with much gentler slopes 
 on the average than the east side but has similar forms. In general, 
 however, wave action is weaker on this vshore. The Algonquin and 
 Nipissing terraces are present but are less decided. This is due 
 largely to the protection of this shore from the storm winds which 
 usually blow from a westerly quarter. The Upper Level terrace is 
 found only in embayments and usually on the north side. A sub- 
 merged terrace is present but is relatively narrow and drops into 
 deep water at seven to eight feet on the average. The exposed ter- 
 races are poorly drained in many places anrl support a growth of 
 swamp trees. Such a condition is found near the mouth of the 
 stream which enters the lake on the west side near the south end. 
 During the Upper Level this stream built a small delta which is 
 now being removed. The low swamp bordering this stream extends 
 northward and around a narrow hill which must have been an 
 island at the Mpissing stage, if not at the Upper Level. 
 
 Northward, Cam's Hill causes a projection in the lake which is 
 accentuated by a depression on the south side, forming a muddy bay. 
 Currents from the north left the shore at this point during the 
 Upper Level stage but were too weak to carry across the indentation 
 and a hook was formed. At present this is being forced back into 
 the bay by the waves, leaving tree trunks standing in the water. 
 North of this hook, cliffs are working back into the Nipissing terrace 
 which developed here at the expense of the Algonquin. In fact, the 
 latter was entirely removed and a steep cliff ris.es from the Nipissing 
 
LAKES OF THE GRAND TRAVERSE REGION. 
 
 165 
 
 shore to the top of the hill, a height of sixty to eighty feet. North of 
 Cam's Hill no current deposits were fouud in a number of embay- 
 ments either at the present or former levels, with the exception of 
 two small indentations at the Algonquin above locality H. The 
 forms found here are duplicates of those found at D on a smaller 
 scale and need no further description. 
 
 Northward to the outlet the shores need no special consideration. 
 The Algonquin and Nipissing terraces are universally present and 
 the Upper Level is preserved wherever the tree growth lias not been 
 removed, except at the headlands. Ice ramparts are found locally 
 and are more noticeable than on the east side of the lake. This may 
 seem strange, since the ice-shove has been attributed to jams which 
 are usually more powerful on the east side, but is due to the gentler 
 slope of shores which offers more favorable conditions for ramparts. 
 It may also be considered that the ramparts are in the process of 
 destruction by waves on this side of the lake, but this has already 
 been accomplished to a large extent on the east side. 
 
 The south bend of the outlet is accounted for by the topography. 
 Another drumlin in line with that forming Meguzee Point causes 
 the low, heavily wooded point on the opposite side of the outlet, 
 Hunter's Point. Spencer Bay is due to a sag between the hills. 
 Another line of drumlins lies to the west and has forced the outlet to 
 the north before it crosses to Lake Michigan. The final bend to the 
 southwest is due in part to the encroachment of dunes, formed from 
 the sands deposited in a great bar which cut off this lake from 
 Michigan during Nipissing times. An excellent example of the move- 
 ment of dunes may be seen at the tenement of the blast furnace, 
 where a large dune is slowly advancing on the building from the 
 west and will soon cause its complete abandonment. Fig. 59. 
 
 Fig. 59. Tenement house in process of burial by a moving dune, Ellv Rapids. 
 (Sketch from photograph.) 
 
166 INLAND LAKES OF MICHIGAN. 
 
 Comparison of tliis lake with TorcUight reveals almost identical 
 conditions, history and characteristics. Both basins are similar in 
 shape and manner of formation. They were connected during the 
 Algonquin and Nipissing stages at least and show similar develop- 
 ment of the shores at these levels. There seems to be little variation 
 in the development of the two basins during Algonquin time, except 
 in the strength of the shore features, the stronger being on Torchlight 
 on account of the larger size. Neither lake was entirely separated 
 from Lake Michigan at this time but the connections w^ere greatly 
 restricted on Torcihlight. As on Torchlight, much of the shore 
 adjustment was accomplished on Elk Lake during this stage and 
 consisted both in reducing the headlands and tlie closing of indenta- 
 tions. In both of these particulars Elk Lake shows less shore action. 
 
 The Nipissing shore is much less prominent on Elk Lake but is 
 generally present. Current action was not important and none of 
 the interesting V-bars found on Torchlight were discovered. The 
 decrease in activity may be ascribed to the restriction of the basin 
 by the development of bars tending to separate it from Lake Michi- 
 gan. 
 
 No evidence was found on Elk Lake of the Post-Nipissing Level of 
 Torchlight. This stage is considered of very short duration and 
 must have been pres^ent on Elk Lake, but the shore features, weaker 
 even than on Torchlight, have been completely destroyed. 
 
 The Upper Level is present in favorable locations but is much more 
 evident on the w^est side, especially in the embayments. The flooding 
 of the lake raises the present level very close to that of the Upper 
 and is causing its rapid destruction wherever the shores are exposed 
 to strong wave action. The demolition of ice ramparts and the 
 building of a double terrace in places is interesting in this connec- 
 tion. 
 
 The development of the shores by wave action is the predominant 
 factor at present and is causing a general recesi^ion both of cliffs 
 and of current deposits. The lake is in a youthful stage at present, 
 although complicated by former levels. The causes of extinction 
 are, therefore, not important. The only case of filling by sediment 
 of any importance was found in the southw^est corner and at the 
 Nipissing level. Most of the water is derived from Eound and 
 Torchlight Lakes, which are efficient settling basins. Vegetation has 
 not taken hold as yet and no marl deposits were found. Complete 
 draining is impossible unless the level of Lake Michigan lowers. The 
 presence of the dam prevents the deepening of the outlet but with 
 this removed the result would be merely a lowering of the level of 
 about fifteen feet. As in the case of Torchlight, the problems are of 
 shore adjustment rather than extinction. 
 
LAKES OF THE GRAND TRAVERSE REGION. 1(17 
 
 CRYSTAL LAKE 
 
 Crystal Lake, the last of tliis jii-oiip to be described, is situated in 
 tlie central-western part of Benzie County and at its west end lies 
 within lialf a mile of Lake Michigan. The Toledo & Ann Arbor R. R. 
 skirts its southeastern shore and has a station stop at Beulah. See 
 map, Fig. GO. The lake is more than eight miles in length and 
 has an average width of two miles, making the area 16 square miles. 
 Its width nowhere exceeds two and one-quarter miles and in no 
 place is it much less than one and one-half miles. Thus the lake is 
 ver}^ uniform in width. The major irregularities of the present 
 shore occur mainly on the south side and consist in a broad projec- 
 tion near Robinsons and a narrowing of the lake from Outlet Bay 
 eastward. Another embayment, now closed, occurs on the north 
 shore and is occupied by Round Lake. 
 
 The topograph}^ of the bottom of this lake is not known but it is 
 stated that its depth is as great as two hundred feet. A well- 
 developed submerged terrace is uniformly present about all shores, 
 and the drop into deep water is clearly marked by a sudden change 
 in color from the light yellow of the shallow water to a deep blue 
 where the depths are greater. This change in color is due in part 
 to the clearness of the water, and the name of the lake has appro- 
 priately been changed from Cap, as found on the old maps, to 
 Crystal. 
 
 As regards the basin, it may be stated that it is relatively old. 
 In fact, it is certain that it was in existence before the ice made its 
 final advance, for it was filled with a small lobe, an offshoot from 
 the Michigan lobe, which pushed through the opening lat the west 
 end, now^ closed AAith sand. This lobe deposited a strong morainic 
 loop around this basin, which is continuous except a't the outlet and 
 a depression on the north side which runs northward into the Platte 
 Lake depression, in the viciuit}' of Round Lake. At present the 
 lake shores do not reach the morainic hills but are separated from 
 them by a rather broad zone of sandy terrace. This widens greatly 
 at the east end and extends nearly two miles before it is interrupted 
 by the moraine. 
 
 The striking physiographic characters are the predominating high 
 cliffs from whose base the sandy terrace mentioned above extends 
 to the water's edge. The first surmise is that this lake has stood at 
 a higher level and further observations prove this to be correct. 
 
 The most convenient starting place for a study of the shores is 
 at Beulah. The town is built on a flat terrace somewhat more 
 than ten feet above the level of the lake and we might almost say 
 nestles at the foot of the clitt's carved in tlie morainic slopes which 
 
168 
 
 INLAND LAKES OF MICHIGAN. 
 
 /Vt/^/HO/W 
 
LAKES OF THE GRAND TRAVERSE REGION. 169 
 
 rise rapidly to the south. Proceeding northward along the shore, 
 one may note the cliffs continuing to the east, but with swampland 
 instead of the lake at their base. A short walk allows an uninter- 
 rupted view and the physiographic significance of this end of the 
 lake becomes evident. One cannot fail to note the sandy character 
 of the soil, the distinct ridges, three in number, stretching in a broad 
 curve to the limiting cliffs on either side of the lake, and the shallow 
 depression to the east. These bars stand from fifty to one hundred 
 feet from the present shore and about ten feet above the lake level. 
 Curiously, the middle bar is not so Avell developed as those on either 
 side and stands two feet lower in elevation. The bars are clearly 
 shore features of the same high level of the lake which formed the 
 cliffs and wliich corresponds in level to Lake Algonquin. During 
 this time wave action was intense wherever the water reached the 
 morainic hills and cut strong cliffs. The quarried material ^Vas 
 carried outward by the undertow to a large extent and contributed 
 to a wide submerged terrace. Discussion of the conditions at this 
 end of the lake is reserved until later. 
 
 Near Windermere was found a small spit which indicates a level 
 between four and five feet above the present. Although so small as 
 to be easily overlooked, it is, nevertheless, of interest since it is 
 almost the only indication of what may be termed the natural level 
 of the lake. Crystal Lake stands only a few feet above Lake Mich- 
 igan, the Ann Arbor tracks at Beulah are fourteen feet higher, and 
 a project to make a waterway from Frankfort through the Betsie 
 Eiver to Crystal Lake and thence through Round to Long and Platte 
 lakes was attempted in the early seventies. Operations began and 
 ended with the making of a cut at the present outlet which lowered 
 the lake level considerably. After the project was abandoned a 
 dam was constructed at the outlet but not of sufficient height to 
 raise the lake to its former level, the natural level of the lake. 
 Conditions are now relatively stable and the previous level may be 
 convenient!}^ termed the Upper Level. 
 
 Along the north shore as far as the Round Lake depression steep 
 cliffs and a broad terrace, partly exposed, are the predominant 
 shore features. The terrace, largely formed during Algonquin 
 time and exposed by the subsidence of the water to the L^pper Level, 
 was further widened by the artificial lowering and in places is 
 swampy and foul near the present shore. The wet condition of the 
 terrace is due to the seepage of ground water from the cliffs and 
 the presence at the shore of ice ramparts which have been worked 
 over by the waves. The submerged portion of the terrace has a 
 rather uniform width in excess of one hundred yards and drops into 
 
170 
 
 INLAND LAKES OF MICHIGAN. 
 
 deep water from a depth of seven feet. The sharpness of the edge 
 of the terrace may be best observed from the top of the clitTs which 
 rise from the Algonquin shore, attaining heights of eightj^ feet or 
 more. 
 
 Even though the exposed terrace is not suitable here for summer 
 cottages on account of its wet condition, this shore abounds in 
 picturesque locations at the frequent sags in the cliffs, caused by 
 the morainic basins which appear from the lake as rounded valleys 
 abruptly truncated by the cliffs. Their resemblance to the famous 
 hanging valleys of Switzerland has been appreciated in one case, 
 at least, where a cottage built in chalet style hangs on the edge of 
 the cliff in one of these depressions. In a few cases the sags are 
 deeper and reach to the lake level or below. Thus, at Clark's cot- 
 tage, at A, and just east of Round Lake small lagoons were cut off 
 by bars in Algonquin time. iS^earer Round Lake the exposed ter- 
 race is dry and covered with sand which has been heaped into 
 small dunes. 
 
 Ice action is plainly evident here. During Algonquin time ice 
 jams swept the terrace free from boulders which Avere lined on the 
 shore and at the present level a low, but sharp, sand rampart. Fig. 
 61, bound together by dune grasses, was found by the writer. The 
 
 Fig. &1. SmaU ice rampart of sand. Crystal Lake. ('Drawn from photograph.) 
 
 ice-push is asserted with considerable confidence to have been 
 caused by jams since the lake is somewhat large for expansion and 
 is subject to frequent jams during the spring thaws. Copies of 
 photographs of an ice jam which occurred on this lake a few 
 years ago were obtained, one 'of which is ehown in Plate X, A. 
 
 The depression in which the miniature Round Lake lies is w^ell 
 below the Algonquin level and extends northward into the large de- 
 pression in which Platte and several other smaller lakes lie. The 
 Platte Lake depression was open in the early stages of Lake Al- 
 
Micliisan Oonlosical and 
 Bi(iU)sifal Survey 
 
 rnlilicatioii :',(). (icolosical Sorios 2.j, 
 IMatL' X. 
 
 A. ICE-JAM, CRYSTAL LAKE. 
 
 Photograph l)y Donald Cilibs 
 
 B. "DROP-OFF," COREY LAKE. 
 
LAKES OF THE GRAND TRAVERSE REGION. 171 
 
 Jionquin on tlie west, as weie Crystal Lake ami the Hetsie River de- 
 pressious to the south, making- a rather irref>ular eo^ast line with an 
 inside passage. Across the Crystal Lake side of the Kound Lake 
 dei>ression there now stands a strong bar more than tAventy feet 
 above the lake and in alignment with the cliffs on the north side. 
 The height of the bar which is somewhat above the Algonquin 
 level indicates that the bar was well above the water level and. 
 therefore, nearly if not quite complete throughout its extent. How- 
 ever, as the level dropped the water from the depression was able 
 to channel the bar and maintain an outlet to Crystal Lake. The 
 position of the channel nearer the eastern attachment of the bar 
 and the presence of sand dunes at the western end, indicating 
 greater age, show that the prevailing currents came from the 
 west. The maintenance of this channel seems almost prodigious 
 considering the small amount of water in the depression at present 
 and the strength of the bar throngh which it was cut. It is prob- 
 able, however, that current action became much less powerful with 
 the dropping of the water level and most of its energies were con- 
 sumed in building a broad spit-like extension of the submerged 
 terrace eastward from the point at Herdmau's landing. 
 
 The terrace narrows considerably in front of the Round Lake bar 
 and this, together with the i:>resence of Round Lake on the opposite 
 side, makes its certain that the depression is a portion of the Crystal 
 Lake basin which was isolated by a bar. However, it was not com- 
 pletely separated from the Platte Lake depression until the water 
 dropped to the Upper Level. 
 
 The prominent boulder wall at the Algonquin level on the west 
 side of the depression near Crystal Lake is indicative of strong ice 
 action which probably was caused by ice jams from the main lake 
 before the development of the bar, although the possibility of ex- 
 pansion cannot be excluded. In the same locality but at a level 
 corresponding to the LTpper Level of Crystal Lake, a well developed 
 bar follows the outline of the w^est side of Round Lake and joins 
 the back slope of the Algonquin bar near its west end. This bar ex- 
 tends more than^one-fourth the circumference of the lake and ap- 
 pears much too large to be accounted for by shore action on a cir- 
 cular lake of less than a half mile in diameter. The possibilities 
 suggest themselves that the bar may have been subaqueous during 
 the late Algonquin stage or developed subsequently to the formation 
 of the Algonquin bar but while the depression was still connected 
 with the Platte Lake area to the north. 
 
 From the Round Lake depression to point B on map, cliffs line 
 the Algonquin shore below which a' sandy terrace heaped into low 
 
172 INLAND LAKES OF MICHIGAN. 
 
 dunes extends to the present lake level. At the shore these dunes 
 have been eroded by the waves, forming the only cliffs on the north 
 side of the present shore of the lake. Since the change in level 
 has been recent and the dunes are but sparsely covered with vegeta- 
 tion, they must be in process of formation, 
 
 A study of the west end of the lake discloses the fact that Crystal 
 Lake is a lagoon. The material of the land forms is nothing but 
 sand. Adjacent to the Crystal Lake shore the subsidence in level 
 exposed a portion of the terrace three to four hundred feet in 
 width which, in general, .slopes gently towards the lake but is 
 modified to some extent by low dunes of recent formation. Beyond 
 are the steep lee slopes of the great dunes between which, near their 
 eastern limit, may be distinguished portions of a double bar at 
 the Algonquin level. The dunes, heaped in confusion to heights of 
 one hundred feet or more, extend to the Lake Michigan shore, three 
 fourths of a mile to the west, and the zone stretches in a nearly 
 north-south direction between the two morainic boundaries of the 
 Crystal Lake depression, a distance of about two miles. Most of 
 the dunes are fixed in position, due to a vegetal covering, except 
 near the Michigan shore where they are moving landward. In 
 several locations the vegetation has been removed either by cutting 
 or fire, and extensive "blow outs" in the dunes are evidence of 
 renewed movement. This great zone of sand is clearly a bar formed 
 during Algonquin times, since the Nipissing beach has been located 
 in places on its front slope, but the usual concave outline is re- 
 versed along the Michigan shore. The explanation is that the 
 limiting morainic ridges formerly extended farther into Lake Mich- 
 igan as headlands and a normal bar of concave outline developed 
 between them. However, subsequent erosion has caused a general 
 recession of this shore, as shown by the extensive cliffs, but greater 
 in amount at the northern headland, causing a convex curvature 
 and somewhat irregular outline of the bar. 
 
 Accompanying the development of the Algonquin bar on the 
 Michigan shore was an adjustment by currents along its inner 
 margin, the Crystal Lake side. The result was the formation of 
 long twin bars which extend from the southern morainic ridge in 
 a broad curve north and northwestward to the vicinity of point C 
 (see miap) and account for the regularity of the shore along this 
 end of the lake. A narrow lagoon, somewhat irregular in outline 
 on its western shore, was thus formed which was inclosed by bars 
 on either side along the west shore but stood between a bar and the 
 Algonquin cliffs along the northwest shore of Crystal Lake. Along 
 the west side the eastward misfration of the dunes has filled the la- 
 
LAKES OP THE GRAND TRAVERSE REGION. 
 
 173 
 
 goon aiul partially covered the bar. Fortunately, however, the uu- 
 buried portions are sufficient for its recognition. To the northwest 
 the dunes have not encroached on the lagoon to so great an extent, 
 and the bar stands out prominently above the dry lagoon on the one 
 side and the exposed terrace on the other. 
 
 As far as may be determined the bars consist of two parallel 
 ridges along the west shore but these coalesce' and again divide 
 into separate ridges before their attachment to the cliffs at A is 
 reached. In development, elevation, and characteristics they are 
 practically identical and are clearly the result of current action. 
 Why then the two bars? 
 
 The key to the explanation is to be found at the attachment of 
 bars to the mainland, that is, the extremities. On a lake of the size 
 and orientation of Crystal, the effective work in the formation of 
 bars at the west end is accomplished by currents driven by north- 
 easterlj^ or southeasterly winds, affecting the south and north 
 shores respectively. The fact that the paths of the larger part of 
 the great storm centers cross or lie above this locality causes a 
 preponderance of storm winds from the southeast over those from 
 the northeast. Consequently, the northern attachment of the bars, 
 B on map, is the critical locality. The shore conditions at B are 
 shown in the accompanying sketch. Fig. G2, 
 
 Fig. G2. 
 
 Diagram showing the attachment of the bars at the west end cf Crystal 
 Lake to the north shore cliflfs. 
 
 Attention is called to the recession of the cliffs in progressive 
 steps or jogs, each jog serving as a place of attachment of a bar. 
 Inasmuch as the cliff continues westward beyond the attachment 
 of the bars, wave action played the important role in this locality 
 during the early stages of the lake. Below the v^^ater level this re- 
 sulted in the formation of a terrace which developed much more 
 rapidly in the loose sand along the west side than in the morainic 
 material of the north shore. As the terrace widened the waves be- 
 came progressively reduced in size as they crossed the terrace and 
 reached the beach with less and less force, diminishing the force of 
 the shore currents in the same ratio. Currents moving Avestward 
 along the north shore east of A were relatively strong, since the 
 
174 INLAND LAKES OF MICHIGAN. 
 
 submerged terrace was narrovv'^, but upon reaching A they were 
 obliged to accommodate themselves both to bends in the shoreline 
 and to a reduction in velocity. This combination of factors caused 
 the currents to leave the shore first at the more westerly bend in 
 the shore, and formed the outer bar. The development of this bar 
 hastened the construction of the built terrace which shifted the 
 point of departure of the currents from the shore eastward to the 
 eastern bend in the shore (Fig. 62), from which the inner bar de- 
 veloped. The coalescence of the bars west of A may be accounted for 
 by a slight obstruction which modified the curvature of the outer 
 bar locally. 
 
 In this connection the triple bars of the east end of the lake 
 demand consideration. The question naturally arises concerning 
 the variation in number at the opposite ends of the lake. Several 
 ways of accounting for the three bars at the east end suggest them- 
 selves but, in order to deduce the most plausible, it is necessary to 
 consider their characteristics somewhat more carefully.- In gen- 
 eral, they stand at a lower elevation than those at the west end, 
 and below the Algonquin level. The curvature of these bars ma}^ 
 be seen from the outline of the east shore (see map), and the 
 rather abrupt angle at which they leave the cliffs is not character- 
 istic of current deposits. Finally, the lagoon extends nearly two 
 miles to the east. ]S"ow, if the same conditions are assumed for 
 the development of these bars as were found for those on the west 
 shore, the irregular curvature of the bars and the development of 
 a built-terrace of nearly two miles in width must be accounted for. 
 The latter alone is sufficient to force us to seek a different explana- 
 tion. 
 
 It seems more likely that the eastern end of the basin was origin- 
 ally shallow and that its bed was remodelled or built up into a 
 terrace in the early stages of the lake. The outer edge of this ex- 
 ceptionally fiat terrace stood near the east shore of the present 
 lake and determined the outer breaker line of the incoming waves. 
 As frequently happens in shallow hajs, a series of submerged bars, 
 three in number, progressivelj^ lower in elevation towards the lake, 
 was formed by the breakers. These bars v^^ere exposed by a lowering 
 of the water level and the inner bar, now forming the beach, was 
 subjected to storm waves on this exposed shore and was built up 
 above the level of the intermediate bar. 
 
 Along the south shore cutting has been the predominant factor 
 and the Algonquin cliffs are almost continuous to Outlet Bay. One 
 sag in the hills at C, see map, extended below lake level and was 
 cut off by a bar which developed from the west, showing the preval- 
 
LAKES OF THE GRAND TRAVERSE REGION. 175 
 
 once of winds from the westerly quarter over tliose from the east. 
 The submerged terrace is very well detined aloug this shore and 
 drops into deep water quite uniformly at seven feet. 
 
 East of Kobinsons the cliffs are exceptionally high, and the ex- 
 posed terrace is heaj^ed with low dunes which extend to the })resent 
 shore of the lake and are cut into low cliiTs by the waves. Along the 
 west side of Outlet Bay the most distinct development of the sub- 
 merged terrace on the lake is seen but this may be due to the shal- 
 lowness of the water which drops at three feet instead of seven, 
 making the effect more pronounced. The depression which caused 
 the bay was one of the channels of the inside passage after Crystal 
 Lake basin was cut otf from the main lake. Currents were active 
 here and not only cut otf small indentations on the west side but 
 built a great bar in the vicinity of the outlet which connected with 
 the cliffs on the east side. From its elevation it is apparent that 
 the bar was not exposed for its entire length but was sufiEicient to 
 hold back the water after the subsidence to the Upper Level. From 
 the bay to Beulah the Algonquin cliffs are again the prominent 
 feature and are interrupted only by two minor embayments which 
 were cut off by bars at the Algonquin level. 
 
 In conclusion we may summarize as follows: Crystal Lake existed 
 as a fjord-like bay of early Lake Algonquin. This depression was 
 crossed by a much smaller one which connected the bay with the 
 depressions to the north and south, which in turn were open to 
 the main lake. The development of bars isolated all three of these 
 basins but left the inside passages free. Wave and current action 
 were excessive in the Crystal Lake depression, after its separation 
 from the main lake, and resulted in the carving of prominent cliffs 
 in the morainic borders, the formation of a broad terrace, and the 
 development of strong bars in front of the depressions and at the 
 west end. iVt this time tlie passage to the north was closed and 
 that to the south partially so. The fornuition of triple barrier 
 ridges at the east end caused a great reduction in size by cutting 
 off a large lagoon when the level was lowered. In fact, it may be 
 stated that virtually all of the adjustments took place and the out- 
 line of the lake was fixed at this time. The waters receded from the 
 Algonquin level to the Upper, a drop of twelve to fifteen feet, and 
 left a broad exposed terrace, the sands of which have been heaped 
 into low dunes. This level persisted until about forty-five years 
 ago when the lake was lowered artificially. At present the shore 
 action consists mainly in removing portions of sand dunes and 
 the formation of low ice ramparts of sand which are remodeled and 
 obliterated by waves. 
 
CHAl'TEK V 
 
 INTERIOR LAKES OP THE SOUTHERN I»ENINSULA 
 WESTERN INTERLOBATE AREA 
 
 The lakes discusised so far in this report lie near the borders of 
 Lakes Michigan and Huron, and the majority have been connected 
 with them at some time during the past. Furthermore, a rather 
 simple grouping of these lakes suggested itself. The interior lakes 
 are much more numerous and are smaller in size, therefore a rela- 
 tively smaller nundjer was selected for detailed study. Undoutedly 
 a more extensive study of our lakes would bring to light relations 
 which would serve as a basis for grouping but for the present this 
 is not at hand. Fortunately we are able to determine more or less 
 accurately the relative time of formation of these lakes and may use 
 this as a natural order for discussion, even though other relation- 
 ships may be very remote. 
 
 The fact that the glacier receded from the Southern Peninsula 
 in a general northeasterly direction and that great , interlobate 
 areas were first uncovered between the Michigan-Huron-Saginaw 
 lobes and the vSaginaw-Erie lobes has been mentioned previously. 
 For convenience we may term the interlobate areas as the eastern 
 and western, although they merge in the southern part of the Pen- 
 insula. In general, for each of these areas the relative age of the 
 lakes is determined by their position, the more southerly being the 
 older. However, Avithin the moraines which border the interlobate 
 areas the older lakes necessarily lie nearer the interlobate and it is 
 only by a consideration of both of these factors that the chronolog- 
 ical order may be determined. Laicertainties exist as to the correla- 
 tion of the glacial deposits and, therefore, the order here presented 
 is tentative and open to revision. 
 
 LAKES OF ST. JOSErH COUNTY 
 
 Probably the first part of the Southern Peninsula to be uncovered 
 by the ice is a small triangular area included largely in St. Joseph 
 County. Within tliis county the glacial formations consist of two 
 morainic tracts, one in the eastern part and the other crossing the 
 northern part of the western boundary, and a large area of outwash 
 2:i 
 
178 
 
 INLAND LAKES OF MICHIGAN 
 
 which stands between and around the northern borders of the mo- 
 rainic areas. A number of small lakes occupy pits in the outwash 
 and basins in the moraine in this region. The more numerous, how- 
 ever, lie in the morainic basins and several of these w^ere examined 
 by the A\Titer, namely, Corey, Clear and Long lakes in the western 
 morainic area and Klinger in the southeastern district. No careful 
 attempt was made to determine the relative age of these lakes and 
 the order of their discussion is of no significance. 
 
 Klingp>r Lake. Klinger Lake is the largest of a group of small 
 lakes which are aligned in a northeast-southw^esterly direction in the 
 south-central part of St. Joseph Count3\ These basins lare moraiiiic 
 in character but that of Klinger Lake is much larger than is usual 
 for this type of basin. It appears to be a depression that is almost 
 surrounded by a narrow zone of moraine, which extends as a spur in 
 a southAvesterly direction from the main moraine, and is probabh' 
 more closely related to lakes situated, in basins of till plains than to 
 
 /V//^: 
 
 Fig. 03. 'Topographic map of Klinger Lalve and surroundings. (From "U. S. G. S. 
 Three Rivers Quadrangle.) 
 
 those found in simple morainic basins. The morainic topography 
 about its shores is the cause of varied shore conditions and the 
 sandy nature of the glacial deposits has been conducive to shore ad- 
 
INTERIOR LAKES OF THE SOUTHERN PExMNSULA 179 
 
 juslinonts Jo a dcjii-oo nnusiial on a lake of less than two square 
 miles in area. Tlie many blurt's and clean sand beaches have been 
 instrumental, lo some extent at least, in making this lake one of 
 the most po})uhir in this section of Ihe State. 
 
 From the map. Fig. (>?,, it will be noted that the shores of the 
 j>resent lake are relatively free from prominent irregularities. The 
 elongate form is rather noticeable, and the direction of the longer 
 axis is such that some of the strongest winds, notably those from 
 the southwest, have full sweep of the lake. The adjustments, although 
 stronger along the east and northeast shores, are, nevertheless, dis- 
 tributed about the shores rather consistently. 
 
 The lake is easily reached by the Toledo, Adrian, Hillsdale and 
 Elkhart branch of the New York Central Lines which runs within a 
 half mile of the soulh shore of the lake. Arriving by train one ap- 
 proaches the lake at Oakwood. This thriving summer resort is beau- 
 tifully located on high ground that commands a view of the entire 
 lake. Fresh clift's rise from the excellent sand beach which lines 
 the shore. Westward, cliffs face the shore for about one-half mile 
 with the single exception of a marshy lowland just west of Oak- 
 wood. This lowland stands less than three feet above the present 
 lake level and is very suggestive of a higher level. If such were the 
 case, one might expect to find this depression bridged by a bar but 
 none can be definitely recognized. Yet the presence of a road along 
 the natural course of such a bar is at least significant. This lowland 
 ])roves to l)e a narrow neck connecting with the flat marsh which 
 borders the southwestern end of the lake and extends westward to 
 the Fawn Kiver. A single depression drops below the level of the 
 marsh, forming a small, unnamed pond west of Klinger Lake, and 
 several island-like hills stand above its surface. Such a hill causes 
 the broad projection of the shore west of Oakwood. In the amphi- 
 theater-shaped northwestern side of this hill an exposed terrace of a 
 former level is easily recognized and, in addition, a bar extends west- 
 ward from the Avestern point of the hill. Thus, it seems clear that 
 the lake has stood at a higher level. Also since the bar and the 
 terrace stand above the marsh, it is evident that the lake flooded 
 this lowland, including the unnamed pond and possibly extending to 
 the Fawn River. Another hill stands near the present shore at the 
 head of the bay. A small spit runs eastward from this "island" but 
 does not connect with that extending towards it from the hill to 
 the east, so that the shore line was not completely straightened in 
 this locality. However, on the west side a complete bar extends to 
 Turtle Hill which was also an island during the higher level. The 
 greater development of the bars towards the west is readily ac- 
 
ISO INLAND LAKES OF MICHIGAN 
 
 con II ted for by the great reach of the waves and cnrveiits from the 
 ea.st. 
 
 Turtle Hill comes to the lake with gentle slopes which have been 
 cut into low cliffs. The material derived from these cliffs was shifted 
 in both directions along the sliore, but the greater part seems to have 
 been carried northward and deposited in a well-defined spit which 
 extends from Turtle Hill towards Breezy Beach. The development 
 of the spit was interrupted by the lowering of the water level before 
 the opening was completely closed, and the form changed somewhat 
 by recent ice action. Much of the Avest end of the lake is shallow and 
 the ibottom covered with marl near the shore. In places the sub- 
 merged terrace of the present level is clearly definable and drops 
 into deep water at about five feet. 
 
 At Breezy Beach the land rises slightly above the level of the 
 former lake bottom and is dry. The shore line was somewhat more 
 irregular during the former stage but currents do not appear to 
 have been effective along this shore. Northeast of Breezy Beach low 
 hills slope gently to the shore and a low cliff and narrow exposed 
 terrace of the higher level are present. The cliffs vary in height and 
 where highest show fresh cutting. Around the broad point the sub- 
 merged terrace is very decided and drops at thirty inches within 
 one hundred feet of the shore, indicating much less powerful wave 
 action here than on the south shore west of Oakwood. A well-de- 
 veloped, but local, ice rampart was noted near the end of the point. 
 
 To the north, wave activity seems to be greater and the exposed 
 terrace has been completely removed, the cliffs rising directly from 
 the beach. Also the submerged terrace widens gradually towards 
 tlie outlet where it reaches a width of more than one hundred yards. 
 The lowland through which the outlet flows was nearly closed by a 
 bar which developed from the south and forced the stream to the 
 slopes of the hills at the north end of tlie lake. This bar is obscured 
 by the road but the smooth curvature, the tree growth rather than 
 swamp vegetation, and the dry ground make its presence certain. 
 
 From the outlet to Bluff Beach continuous cliffs are evidence of 
 strong wave activity. Between Bluff Beach and Sandy Beach a 
 swamp borders the lake. This SAvamp was clearly a part of the lake 
 during the former level, and from the elevations we may deduce that 
 both Tauiiarack and Thompson lakes were also included. It Avas 
 stated earlier that the form of the lake was conducive to strong 
 activity at the northeastern end of the lake. The almost continuous 
 cliffs and Avell-defined submerged terrace along the shore southwest 
 of Sandy Beach are evidence of poAverful wave action due to westerly 
 AA^inds, even though their reach is less than those from the southwest, 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA |S1 
 
 and in this locality the slioie adjustnient.s are of the j>i*e'atest ina«;iii- 
 tude on the lake. The most notewortliy change took place dui-ing 
 the former stage across the lowland between Sandy and Blntl" 
 beaches and was brought about by tlie development of a great bar 
 of more than a Iialf mile in length, wliich practically sei)aratcd the 
 northeasteni jjoitinn of the l;ikc (Thompson and Tamarack basins) 
 from the i)resent Klinger ]>iake basin. The position of tlie inlet from 
 Tamarack Lake near Blntf Beach shows clearly that the bar <le- 
 veloped from Ihe elites near Sandy Beach towards the northwest 
 and that the material mnst have been derived almost exclusively from 
 the clitt's below Sandy Beach. Ice action exerts a strong push along 
 this bar but the ramparts are of moderate strength on account of the 
 sandy character of the material. 
 
 East of Oakwood the bluffs drop' to a low marsh above which rises 
 a small hill near the present shore. A bar at the level of the former 
 stage extends from this hill to the bluffs to the west but no such 
 form wa's noted on .the east side. From this it is clear that the 
 marsh Avas formerly an arm of the lake and that the hill was a land- 
 tied island. Also westerh' winds were the more effective in this 
 locality since the material for the bar must have been derived from 
 the bluffs at Oakwood. 
 
 From the i)hysiogTaphic viewi)oint Klinger Lake is most iutei-est- 
 ing in its past. Clearlj' it has stood at a level approximately three 
 feet above the present and at that time was part of a lake of much 
 greater area, although shallow for much of its extent. No attempt 
 Avas made to trace the old shore lines Avhere they deviate greatly from 
 the shore of Klinger Lake, but from the topographic map some con- 
 ception of the former extent of the lake may be deduced. The eleva- 
 tion of Klinger Lake is given as 807 feet above sea level. At the 
 former level, then, it must have stood at about 810 feet. At this 
 level, marked extensions of the basin existed at the southwestern 
 end, at the northeastern end, including Tamarack (SOS elevation) 
 and Thompson (80!) elevation i lakes, and possibly at the outlet. 
 The latter is moist interesting, since there is a possibility that the 
 lake spread over an area of more than five square miles to the north. 
 and is well worth the effort necessarv to trace this out. 
 
 The adjustments of the Klinger Lake basin were accomplished 
 largely during the former stage and, although not comnleted in manv 
 cases, they determined to a large extent the present outline of the 
 lake. The activity of all the forces acting on shores excepting that 
 of ice, is excellently shown. Ram])arts are present but are not exce])- 
 tional in develo])ment. This is due to the prevalence of sandy ma- 
 terial which is unfavorable to the development of strong ramparts. 
 
182 
 
 INLAND LAKES OF MICHIGAN 
 
 This lake is within the limits necessary for ice expansion but the 
 testimony of observers is that the push is exerted by ice jams. TKey 
 are especially effective on the east and northeast sides, because of the 
 presence of an open zone of water which is maintained throughout 
 the winter by the many springs near the ishore. 
 
 As regards extinction, the greatest effect has been produced by the 
 lowering of the water level. Little has been accomplished within 
 the present lake basin either by vegetation or sedimentation. 
 
 At the present time the adjustment of the shores of the lake is 
 not marked. There is little work for currents except the distribu- 
 tion of material derived in small amount from the cliflfs which show 
 fresh cutting in a few localities only. Probably the principal 
 adjustment taking place is the remodeling of the submerged terrace 
 to conform to present conditions. 
 
 Corby Lake. Within the morainic tract in the southwestern 
 part of St. Jos'eph County are a number of lakes occupying morainic 
 basins of which Corey, Clear, Long, Kaiser, and Mud Lakes were 
 
 Fig. G4. Topograjihic map of Corey, Clear and Long Lakes and surroundings. 
 (From proofslieet of Three liivcrs quadrangle U. S. G. S.) 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA is:? 
 
 examined. The latter two were dry at the time of the writer's 
 visit and are too small to show decided shore features. 
 
 Corey Lake, Fi«>'. 04, is the largest of those examined and lies at 
 the edge of the moraine. Its shores are bounded hj morainie material 
 with the exception of abont one-half mile on the south side. It is 
 the only one of those mentioned whicli is at all popular as a summer 
 resort and may be reached by a short drive from Three Elvers. On 
 first impression one might expect insignificant sliore adjustments 
 on this small lake of hardly more than one square mile in area, 
 but a study of its shores quickly dispels the notion. The lake lies 
 in several connected baisins and the ishores are not only irregular 
 in contour but varied in relief as well. Furthermore, the material 
 is sandy and easily worked, so tliat the shore features are excep- 
 tional for a lake of this size. 
 
 At present, the lake flows into Kaiser Lake which has no outlet. 
 Formerly when the lake stood at a higher level the water escaped 
 southward through an extended swamp, eventually reaching Mill 
 Greek. An exposed terrace of the higher level stands below the 
 cliffs at the southeastern shore of the lake but is somewhat obscured 
 by the numerous cottages. Further west, a flat caused by the reces- 
 sion of the high ground furnishes the key to the former condition 
 of the lake. The most noticeable feature is a vsand bar of smooth 
 curvature, which stands nearly four feet above and about thirty 
 feet back of tlie present shore. This bar at present incloses a lagoon 
 of approximately three hundred feet in greatest width. Closer ex- 
 amination reveals the presence of a spit attached to the east side 
 of this indentation farther inland and at a level still higher than 
 that of the bar. From evidence found on other parts of the shore 
 the highest level stood seven feet above the present and the spit 
 just mentioned was in an early stage of development. 
 
 West of the lagoon clifl:s again line the shore and the terraces of 
 both of the former levels are readily distinguished. The terrace of 
 the four foot level reaches a width of twenty feet and supports ice 
 ramparts locally. An especially well-developed rampart has been 
 formed under present conditions on this terrace just west of the 
 lagoon. Similar conditions persist to the entrance of Little Corey 
 with the exception of a sag at Shore Acres. This sag dropped nearly 
 to the level of the four foot stage and wais artifieially deepened some 
 sixty or seventy years ago to allow the water to flow southward. 
 The writer's infornuition concerning this channel is none too reliable 
 but it seems certain that the lake stood at the four foot level at that 
 time, and it is ]>resumed that the channel was dug to accommodate 
 the water under flood conditions. Since that time the ground water 
 
184 INLAND LAKES OF MICHIGAN 
 
 table has sunk; Corey Lake has di^opped to its present level; and 
 Kaiser and Mnd lakes have dried up. 
 
 The point on the south side of the entrance to Little Corey is 
 composed of sand and swings northward from the cliffs in an even 
 curve. It is clearly a current deposit which developed largely dur- 
 ing the highest level, and the lagoon on the west side is therefore 
 dry at present. Indications point towards a continuation of its 
 growith but at a very slow rate. 
 
 In addition to the forms described, a well-defined submerged ter- 
 race follows the south shore. As a rule it drops into deep water 
 within one hundred and fifty feet of the shore but swings outward 
 to double this width in front of the lagoon. The change in color 
 at the ''drop off" was very marked near the entrance to Little Corey 
 and the writer liad the good fortune to succeed in registering this 
 on a photographic plate, a reproduction of which is shown, in 
 Plate X, B. 
 
 A narrow moraiuic depression extends westward from the south- 
 western part of the main lake and is composed of two basins, both 
 of which are filled with water. The more westerly, not shown on 
 the map, iforms a small pond which drains into Little Corey, situ- 
 ated in the easterly basin. The two basins are iseparated by a 
 swamp which rises scarcely above the present lake level. Both 
 exposed terraces are continuous below the cliffs around the depres- 
 sion, ishowing that the entire dej)ression was connecited with Corey 
 Lake, as represented on the earlier maps. A sharply defined sub- 
 merged terrace follows the present shore of Little Corey and sup- 
 ports a heavy growth of vegetation and a deposit of marl as Avell. 
 Within the depression the waves have been the most active agent 
 and have accomplished the most work on the south side, due to 
 the greater strength of winds having a northerly component. Some 
 deposition by currents may have taken place at the west end of 
 Little Corey but the presence of a road across the flat makes this 
 uncertain. 
 
 However, an easterly drift along the north shore of Little Corey 
 cooiDerated with a southerly drift along the west shore of the main 
 lake to form a V-bar with characteristic central depression at the 
 north side of the entrance to Little Corey, the point occupied by 
 the Y. M. C. A. Camp. This V-bar stands fully eight feet above 
 the present level and is considered to have been a fully developed 
 form during the highest stage of the lake, that is, stood slightly 
 above the water level. At present, the currents on the main lake 
 predominate and the V-bar is extending southeastward as a spit. 
 This extension is not in line with the spit on the south side of the 
 entrance but there is, nevertheless, a likelihood that the channel 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA 
 
 LS.- 
 
 ^\•l\\ he closed, as iiidicaled by the very shallow water between the 
 Iwo spits, wliicli, in addition, receives each year the deposits from a 
 lieavy growtli of rushes. 
 
 North of the Y. INI. C. A. caniji the terraces of both of the former 
 levels are present on the clilT-lined shore either side of the point 
 opposite the island. This was the scene of stroiij* wnve action which 
 accomplisluMl most dnrino- the highest stage, since that terrace is 
 the better developed. l*art of the material derived from the cliffs 
 was distribnt(Ml on the terrace but a considerable portion was 
 carrie<l along the shore in both directions. Thns, at the point o])- 
 posite the island, currents from both the north and the south left 
 the shore but. instead of foi-miug a sim])le spit or ^'-bar, tied a small 
 island to the mainland by spits which developed in accordance with 
 the sliores on both sides. See Fig. 05. The currents Avere unable to 
 
 ^^^ 
 
 Fig Co Sketch nf small island tied to the mainland by two bars or tomliolos. 
 
 fontinue to the island or else were entirely depleted of their load, 
 for the shallow water between the island and the mainland is due 
 to the natural contiguration of the bottom, as shown by the clay 
 bottom ui»on which rest many large boulders. 
 
 The island is an oblong "sugar loaf" with steep cliffs on all sides 
 but the north. The shores show the effects of the activity of waves 
 rather than currents, and the ex[)osed terraces of the former level 
 ;ire the predominant features. >>light activity is manifested on the 
 north side but the well-developed terrace, formed at the highest level 
 on the remainder of the shoi-e. has been largcdv cut away on the 
 <'xposed south and east sides. 
 
186 INLAND LAKES OF MICHIGAN 
 
 Cliffs line the greater part of the shores of the bay north of the 
 island and usually are notched hj the terraces of the higher levels. 
 Near the head of the hay and on the north shore are indentations 
 which were completely cut off by bars at the highest level. At the 
 sharp headland on the west side of the entrance to the north arm 
 of the lake, the effects of strong current action are again evident. 
 This headland is caused by the projection of the hills into the lake 
 but is ISO sharp that the currents were unable to follow the shore. 
 The current-borne material was deposited in the form of a spit 
 which runs to the northeast and has a length in excess of three 
 hundred feet. The spit began its growth during the highest level 
 and has continued to develop during the succeeding stages, including 
 the present, but at less rapid rate. Clearly the southerly winds 
 are the most effective here on account of their long reach and the 
 deep water over which they pass. 
 
 The west shore of the north arm is lined by a continuous low cliff 
 below which stands the four foot terrace. The bay is shallow and 
 vegetation is encroaching along the shore. This is especially notice- 
 able at the north end where rushes extend from two hundred to 
 three hundred yards off shore. On the east shore wave action pro- 
 ceeded to an advanced stage of adjustment, as shown by the current 
 deposits about midway between the head of the bay and Perch 
 Point. In this locality a slight bend in the shore line of the highest 
 level was sufficient to throw the currents off shore, forming a com- 
 plete bar across a narrow lagoon. Likewise at the four foot level 
 the currents swung away from the shore and built a hook from the 
 east in front of the bar of the highest level. The lagoon in the latter 
 case does not exceed two hundred feet in width. The lower bar 
 shows clearly that the effective currents drift northward along this 
 shore 'and that reach of the wind is again the determining factor. 
 
 Between the bars just described and Perch Point the cliffs are 
 less prominent and the exposed terraces are better developed. Perch 
 Point is almost -a duplicate on a larger scale of the point opposite 
 the island. The bars which connect the low island to the main- 
 land were built at the highest level. The bar on the south side con- 
 nects with a well-defined terrace at the seven foot level, which soon 
 merges into high cliffs. The absence of the highest terrace along-^ 
 these cliffs, which are continuous to the outlet, is due to its removal 
 during the four foot stage. This shore is exposed to the strongest 
 winds of maximum reach, and the obliteration of the shore features 
 of former levels is to be expected here if ^anywhere on the lake. The 
 material quarried from these cliffs has drifted in large part to'^ards 
 the south and was deposited across the low ground at the outlet 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA 187 
 
 which was loiincrly Hooded. The bar develoi»ed diirinfi' the hi<>host 
 stage and nearly separated Corey Lake from Kaiser. 
 
 It remains to point ont briefly some of the more important epLsode.s 
 in the history of tliis lake. At its inception the lake stood at a 
 level approximately seven feet above the present. At this time it 
 Hooded all of the depression at the sonthwestern end of the lake and 
 connected with Ivaiser, Mnd, and ]>ossibly Clear Lake. The adjnst- 
 meiits of the shores during this level were carried to an adv^anced 
 stage and in some cases the currents completed their Avork before 
 the level dropped. The four fool level was in existence a relatively 
 short time ago and the adjustments followed along lines detennined 
 during the i)revi(»us stage. In general, however, they were less ex- 
 tensive. Within I'ecent years the levels of all the lakes mentioned 
 have dropped several feet, which has caused the extinction of Mud 
 and Kaiser Lakes. The adjustments at present are slight, the one 
 of greatest consequence being at the entrance to Little Core}^ which 
 may become closed. 
 
 Clear Lake. Clear Lake lies within a half mile of the north 
 end of Corey Lake, see Fig. 61, and runs parallel to the northeast- 
 ern shore. As in the case of Corey Lake, this small body of water, 
 which has a length of somewhat more than a mile and a width of 
 less than one-third of its length, is surprising in the number and 
 extent of the adjustments of its shores. The surrounding land 
 rises well above the level of the lake except at the south end. Inas- 
 much as the Mud Lake basin, now dry, lies but a few hundred feet 
 to the south, this low divide is a favorable locality for beginning 
 our study. At this locality it is evident that Clear Lake has stood 
 at a higher level than the present, and from the elevation of the 
 divide it seems probable that the two lakes w^ere connected with 
 each other and with Corey Lake through Kaiser Lake. Conditions 
 are favorable here for the formation of a bar along the south shore 
 of Clear Lake l)ut the presence of a road along the logical position 
 of such a bar makes its identity uncertain. Very often, however, 
 advantage has been taken of the higher ground along the course of 
 a bar in the building of roads and such an occurrence may serve as 
 indirect evidence of the existence of the bar. 
 
 Northward along the west side of the lake, the slopes are gentle 
 and little adjustment of the shore has taken place. Proceeding 
 along the shore one notices evidence of an increase in the activity 
 of the shore forces. The first definite feature to be found is a small 
 spit which stands three feet above the present level and points 
 southward from the point designated A on the map, Fig. 64. This 
 spit protects a narrow lagoon which is merelv a continuation of 
 
188 INLAND LAKES OP MICHIGAN 
 
 the shallow, mud-covered lake bed adjacent to this shore. Vegeta- 
 tion has established itself off shore but is especially abundant in 
 the lagoon which is rapidly being filled. North of the spit the slopes 
 increase in steepness and the activity of the waves becomes evident. 
 The wave cut cliffs increase in height to a maximum just south of 
 point B and, where well developed, rise from an exposed terrace 
 rather than from the present beach. This terrace corresponds in 
 elevation with the spit at A and is further evidence of a former 
 level of the lake. The cliffs and hills are wooded but thre terrace 
 supports only a limited growth of bushes and young trees; conse- 
 quently one may conclude that the drop in the water level has 
 been of recent occurrence. The material derived from the cliffs be- 
 tween points A and B drifted in both directions, inasmuch as a 
 current deposit Avas formed at B as well as at A. 
 
 The deposit at B is a typical V-bar with characteristic central 
 depression. Its greatest development was from the northwest, due to 
 the greater power and reach of the winds from that direction and to 
 the abundajit material quarried from the cliffs which are more 
 prominent on this side of the point. During the former stage 
 of the lake, the north side of the V-bar attained the greater develop- 
 mient, 'and, in addition, a considerable part of the current-borne 
 material was distributed oft' shore, forming a submerged terrace 
 of greater width and gentler slope than on the south side. At the 
 present time, a small hook pointing northward indicates that con- 
 ditions are reversed and that the greater growth is from the south 
 side. This reversal was brought about by the general reduction of 
 the activity of waves and currents, caused by the lowering of the 
 water level. Under the present condition of reduced activity, the 
 submerged terraces are more effective than formerly in reducing the 
 size, and therefore the power of the waves i)assing over them, and 
 the wider the terrace the greater its effect. In this case the broader 
 terrace on the north side so reduces the power of the incoming 
 waves, and therefore the currents, as to render them less effective 
 than those passing over its narrower counterpart on the south side. 
 
 North of point B the cliff's which face the slioi^e indicate the source 
 of material forming the north side of the V-bar just described. 
 The presence of an alluvial fan and a fossil delta along this shore 
 are of interest although not formed hj the agents active on the 
 lake shores. The delta Avas built hj a small stream Avhich entered 
 the lake at the time of the higher level, but the waves and currents 
 were able to distribute the material as fast as supplied, so that the 
 shore line was not affected. The exposed terrace of the former level 
 is a consistent shore feature as far as the north end of the lake. Its 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA IX'.) 
 
 elevation willi reference to the present level was nieasnred in sev- 
 eral j)Iaces and found to be within a few inches of four feet. This 
 means tliat many of the deposits formed duriu;;' lliis si aye did not 
 extend a.bove the water level. Near the uorth end of the lake I lie 
 side s]o])es are ji,eutle and the exposed terrace is wide. An 
 i n teres t in ii,- cliauiic has been ])roduced at this end of the lake by 
 the lowering- of the water level. Areas formerly covered l)y open 
 water are uoav swamps tilled with rank growth of vegetation. One 
 such wa.s a bay just soutli of tlie end of the lake. Currents fr(Mii the 
 south nuide some headway in cutting off this indentation bur suc- 
 ceeded in building only a submerged bar, now clearly outlined ))y a 
 row of willows. 
 
 The muddy north slioves <d' the lake show little adjustment aside 
 from the exi)osed terrace. The absence of any indication of a bar 
 1)etween the snmll island, not shown on the map, and the mainland 
 is evidence of the feebleness of the currents. 
 
 The east shore is lined with cliffs which are continuous along 
 the northern half of the lake. Below the cliffs stands the exposed 
 terrace of the higher level, and so uniform are the shore features 
 that any deviation is very noticeable. Thus, the current deposits 
 at either end of the cliifs are readily detected. At the sharp bend 
 near the north end of the lake a small hook is indicative of cur- 
 rents driven by southwesterly winds, and at C a sj)it shows the 
 effectiveness of the winds from the northwesterly quarter. The 
 reach from both quarters is approximately the same, so that the 
 size of the deposit is an index of the strength and frequency of the 
 winds. The greater development of the spit at C shows conclusively 
 that the northwesterly winds are the more effective. 1 u each case 
 the deposits were Ituilt during the former level and are not growing 
 at present. 
 
 The turning of the currents from the shore at (' lias ])reveuted 
 the fornmtion of a spit on the Avest side of the nearby point to the 
 southeast but some such activity might be expected on the east side. 
 This, however, is not the case and it is due to the topographic fea- 
 tures of the southern half of the east shore of tlie lake. Insteail of 
 the continuous cliff's of the shore north of C, which furnished a max- 
 imum of material, there is an alternation of cliff's and lowlamls. 
 In addition to the smaller amount of nmterial furnished by the 
 cliff's, many of the lowlands extended below the water level during 
 the former stage of the lake and were areas of de])osition. Thus, 
 the material, relatively small in anmunt to start with, was further 
 depleted by deposition at the mouths of indentations. T!ie absence 
 of any deposit at the end of the ixtint indicates that the currents 
 
190 INLAND LAKES OF MICHIGAN 
 
 not only were weak but practically without load by the time they 
 reached this locality. 
 
 The first indication of the Avork of currents along the southeastern 
 shore was found in the bay east of the point just discussed. The 
 head of this bay consists of a hooked spit which came within thirty 
 feet of completely cutting ofif a narrow lagoon. This spit developed 
 from the south and at the former level of the lake. A hundred yards 
 or more to the southwest the cliffs are again interrupted by a small 
 triangular indentation which was completely isolated by a bar. 
 Beyond this bar cliffs again line the shore as far as the depression at 
 D. The mouth of this indentation was broad but, nevertheless, was 
 completely bridged by a bar, forming a lagoon of several acres in 
 extent, which is still wet. To the south, the cliff's are less promi- 
 nent and gradually give way to gentle slopes. Yet currents, prob- 
 ably from the north, were active and succeeded in cutting off two 
 small indentations between D and the south end of the lake. 
 
 From the discussion above it should be evident that Clear Lake 
 has stood at a level some four feet higher than that found by the 
 writer in the summer of 1914. The presence of only very young vege- 
 tation on the exposed terrace is evidence that the lowering of the 
 level was of recent occurrence. This was fully corroborated by 
 information obtained from residents of the locality. Practically 
 all of the adjustments of the shore have taken place at the higher 
 level and the indications are few indeed that they are continuing 
 at the present time. 
 
 The adjustments are very pronounced for a lake of this size and 
 include those due both to wave and to current action. In general, 
 the eastern shore was most affected, and this is shown by a less 
 abundant growth of vegetation along the shores as well as by more 
 prominent shore features than on the opposite side of th^ lake. 
 
 At present the lake has no surface outlet and no inlets of im- 
 portance. Since there is little adjustment of the shores, extinction 
 is the active physiographic process and this is being accomplished 
 by vegetal accumulation. Little has been said of the submerged 
 terrace in the description above. The reason for this is that it is 
 impossible to determine its limits on account of the heavy growth of 
 vegetation, not only on the surface of the terrace but on the lake 
 bottom as well. Over much of the bottom a complete carpet of vege- 
 tation may be seen through the transparent water which gives the 
 lake its name. 
 
 Long Lake. Long Lake lies less than a half mile northwest of 
 the north end of Clear Lake and resembles the latter closely in size 
 and form. This lake is also surrounded by moraine and is without 
 

 0-1-' 
 
 O OS 
 
 of: 
 
 ^ >» 
 
 002 
 
 tuo 
 
 3S 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA. 191 
 
 outlet 01- inlets of eoiisequeiu-e. Tlu' j^roatost difl'ereiu-c is in the di- 
 rection of the longer axis which runs at right angles to that of Clear 
 Lake. Inasninch as the contoni* of the bottom is not known, it is 
 not possible to determine whether this basin is a simple elongated 
 depression or a series of connected nujrainic basins. The morainic 
 material in this locality contains a large percentage of sand and the 
 shore adjustments are of the same order as those found on the other 
 lakes of the group. 
 
 In Plate XI is shown the adjustment of the northeastern end of 
 the lake, which is not only the most pronounced on this lake but is 
 one of the most perfect of its kind found on the inland lakes of the 
 State. As nmj be seen from the plate, the north end of the lake 
 has been cut off by a complete sand bar, perfect in development and 
 preservation. This bar stretches from the cliffs on either side of 
 the lake in a beautiful curve and stands slightly more than live 
 feet above the present level. Inasmuch as there is no sag in its 
 crest, this bar was fully developed and extended above the water 
 level which prevailed at the time of its formation. This bar, then, 
 establishes a former level of the lake which probably stood some- 
 wliat less than five feet above the present stage. The lagoon of 
 about ten acres in extent is still wet and in late summer is lit- 
 erally crowded with lily pads which are rapidly converting it into 
 a peat bog. 
 
 The continuous cliffs which face the shores of the southeastern 
 side of the lake rise from a well-defined exposed terrace which 
 varies in width from forty to sixty feet. The lake level indicated 
 by this terrace stood more than four feet above the present level and, 
 thus, is in agreement with the level deduced from the bar at the end 
 of the lake. The presence of coarse material on the terrace is an 
 indication that it was formed largely by wave cutting, furnishing 
 the material for the bar above. Along the northwest shore the 
 cliffs are less prominent west of the blunt projection and become 
 insignificant to the south, where the slopes are gentle. A wide 
 exposed terrace is the predominant shore feature, although cur- 
 rents were effective in closing two small indentations in the vicinity 
 of A, see map. Fig. 01, and another near the southAvestern end of 
 the lake. 
 
 West of the broad point on the south shore, the side slopes are 
 gentle and the effects of wave action are not prominent. However, 
 currents from the north were effective and built bars across the 
 mouths of two small indentations which existed during the higher 
 level. The more westerly embayment was cut off" by as perfect a 
 bar as thiat described at the northeastern end of the lake but is 
 
192 INLAND LAKES OF MICHIGAN 
 
 less prominent on account of its smaller dimensions. Apparently 
 the current-borne material was limited in amount and the greater 
 pa'rt was deposited on this bar, since the deposit at the mouth of 
 the embayment nearer the point is a spit attached to the west side. 
 
 The broad projection on the south shore is caused by a low 
 moraiuic knoll whose lakeward slopes were carved into the charact- 
 eristic cliff and terrace profile during the former stage of the lake. 
 The material from the cliffs drifted westward and was deposited 
 on the east side of the knoll, forming a small spit. This spit would 
 have cut off a large part of the swamp which borders the shore to 
 the east had its growth not been interrupted by the sinking of the 
 water level. 
 
 The west side of the prominent projection on the southeastern 
 shore of the lake is exposed to the action of strong waves due to 
 westerly winds, and the cliff and terrace of the higher level are con- 
 tinuous to the vicinity of point B. Here the velocity of the along- 
 shore currents was reduced at a small indentation south of B, which 
 was completely cut off. At B an unsymmetrical V-bar, elongated on 
 the south side, indicates that the currents from both directions left 
 the shore but the more effective currents were from the south. This 
 bar whose longer side has a length of nearly one hundred yards 
 is an exceptionally strong shore feature for a lake of this size. 
 
 Between B and C cliff's are again prominent but at C another 
 V-bar, which is a duplicate of that at B on a smaller scale, indicates 
 similar conditions. Wave action predominated along the shore 
 from C to the northeastern end of the lake as shown by the well- 
 developed cliff and terrace. An interesting break in the cliff's 
 Avas found in the bay east of C in the form of a dry gully, at the 
 foot of which stands an excellent example of an alluvial fan. 
 
 Long Lake also is interesting chiefly on account of its past. The 
 water level formerly stood between four and five feet above the 
 present and at this time prominent adjustments of the shore took 
 place. Strong wave activity carved steep cliffs where conditions 
 Avere favorable. Currents were also active in like degree, so that 
 there are few localities where adjustments are not readily detected. 
 Not only did the currents cut off indentations but also swung away 
 from the lake shores, forming the interesting Y-bars commonly 
 found on the long and narrow lakes. Little activity is manifested 
 at the present level and the important physiographic process of 
 the future is that of extinction. Vegetation has become firmly 
 established in the lagoons but the process is not so active on the 
 main lake as was found to be the case on Clear Lake. Since there 
 is no outlet the lake could not have been lowered bv downcutting 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA 193 
 
 and the dvoj) in level, therefove, must be referred to some other 
 lause. 
 
 The lowering in level is shared by all of the lakes in this group, 
 and the idea that it is connected with the pumping of water for the 
 city of Three Kivers from the outwash plain six miles east of the 
 lakes seems to exist. Therefore some pains have been taken to 
 obtain information concerning the lake levels and the pumping op- 
 erations since the installation of the plant. In particular the writer 
 is indebted to Mr. Eugene A. Schall, City Clerk of Three Rivers, 
 Mich., for collecting this information. 
 
 The levels of these lakes and also all other water-ways in the 
 vicinity have fluctuated somewhat periodically during the last forty 
 years at least, conditions previous to that time being unknown to 
 the writer. Previous low water periods occurred in 1882-83 and 
 181)5-96, and during intervening years the water came back to 
 normal. The levels of the lakes during the low periods are not 
 known to the writer and, therefore, cannot be compared with 
 the present low stage. 
 
 The city of Three Rivers has been obtaining its water supply 
 from nearby wells since 1876, the amount used increasing to '70,000 
 cubic feet (approximately 500,000 gallons) daily at the present 
 time. Previous to 1896 a few small, flowing wells were utilized but 
 since that time twelve six-inch wells have been installed, five in 
 1896 and seven in 1910. In 1915 six of these wells were reset about 
 one hundred yards away and, although drilled to the same depth, 
 tapped an entirely difl'erent source of supply. In some respects the 
 lowering of the lake levels 'and the pumping operations are related, 
 but it must be kept in mind that our record of lake levels does not 
 antedate the installation of the water system. Also during the 
 first known period of low water, which occurred soon after the 
 installation of the water works, the wells were not pumped and 
 the amount of Avater used was insignificant. The low period of 
 1895-96 occurred slightly previous to the driving of additional wells 
 and could not have been caused thereby. The present low water 
 stage follows the change in position of some of the wells, which 
 tapped a new Avater-bearing layer, and shows a closer relationship 
 than the iDrevious stages. In addition to the discrepancies Avhich 
 appear in the statements above, the pumping operations have stead- 
 ily increased and there should have been a gradual lowering of the 
 levels rather than a periodic fluctuation. Also it is probable that 
 the water pumped from the wells comes from a large surrounding 
 catchment area rather than the local area occupied by the lakes, 
 since all other water ways in the vicinity have been similarly affect- 
 
194 
 
 INLAND LAKES OF MICHIGAN 
 
 "gagpsF 
 
 Fig. GG. Outliue map of Gun Lake, Barry County. 
 
 ^Miks 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA 195 
 
 ed. Thus, it appears from the data at hand that the wells have had 
 little or no effect on the fluctuation of the levels of the lakes. 
 
 The cause is probably a natural one and may be related to periodic 
 changes in climate which are none too well established as yet. 
 Whatever ma}' be the cause, the effects are very pronounced and a 
 careful study of the problem promises interesting results. 
 
 Gun Lake. Of the lakes included in this group Gun was the next 
 uncovered by the retreating glacier. This lake is one of the largest 
 of the numerous small, interior lakes which lie in the southwestern 
 part of the State. It lies directly in the center of the western 
 border of Barry County and crosses into Allegan County at the 
 extreme western end of the lake. The lake is not readily accessible 
 but may be reached either from Middleville or Hastings, both of 
 which are on the Grand Rapids branch of the Michigan Central 
 R. R. Nevertheless, the lake is a popular resort and draws a large 
 number of summer visitors from both nearby and distant points. 
 
 This lake is almost rectangular in shape, see Fig. 66, but is 
 nearly cut into two basins by Englands and Murphy points, whicli 
 are attached to the south and north shores respectively. Numerous 
 other points and bays make the lake most irregular in outline. This 
 is better appreciated when it is known that, although the area is 
 but slightly over four square miles, the shoreline measures more 
 than seventeen miles. If Mud Lake is included, the area is 4.4 
 square miles and the shoreline nineteen miles (a perfectly rec- 
 tangular lake of the same area would have a shoreline of eight 
 miles in length). The reason for the complicated shoreline is 
 apparent when the origin of the basin is understood. This basin 
 lies on the western border of a strong morainic ridge, trending 
 north-south. To the west is a triangular outwash plain which de- 
 veloped from the north and west and is very thin and incomplete 
 near the border of the taoraine. In the vicinity of the lake frag- 
 ments only of the surface of the outwash are present due to the 
 fact that large ice blocks of irregular shape were present at the 
 time of its formation and prevented deposition of the outwash ma- 
 terial. It is doubtful if these blocks were entirely covered by tho 
 outwash material since till plain borders parts of the shore of tne 
 lake, even Ihough its surface stands below the general levs-l cl the 
 fragments of the outwash plain (see Chapter II on the formation 
 of pit lakes). Thus, the lake may be classified as a pit of shallow 
 depth but very complicated in outline. The basins represent the 
 locations of the ice blocks and the points are spurs of the moraine, 
 fragments of the outwash plain, or swells of the undulating surface 
 of ilie till plain not covered by outwash. The varied conditions 
 
196 INLAND LAKES OF MICHIGAN 
 
 aloug tlio shore, both as to contour and material, are favorable to 
 adjiisfniejits but, on the other hand, the shallowness of the lake 
 and Hs partition into partially enclosed basins greatly interfere 
 with noruaj. wave development. 
 
 The outlet of this lake, the Gun River, flows from the southwest- 
 ern end and takes a southwesterly course to the Kalamazoo. The 
 gradient of Gun River is very flat, and the stream has cut a very 
 shallow trench in the sand flat. Shore features are now exposed 
 along certain parts of the lake which indicate a former level in 
 accordance with the down-cutting of the outlet. There also exists 
 a probability that at the higher level Gun Lake formed a part of a 
 glacial lake known as Lake Dowagiac, a matter readily determined 
 when the exact level of Gun Lake is known. Furthermore, the outlet 
 gives us a clew to the small amount of variation in the level of this 
 lake during the year. The writer was surprised to see the outlet 
 almost filled with a heavy growth of lilies and rushes and realized 
 at once that they must offer considerable resistance to the flow of 
 water which might hold the lake at a higher level. An obstruction, 
 natural or otherwise, in the course of the stream would cause a 
 ponding of the water and the encroachment of vegetation in its 
 channel would follow as a result, but the writer knows of no ob- 
 struction in Gun River, although a positive statement to this 
 effect cannot be made. Assuming the absence of an obstruction, 
 one might well conceive the encroachment of vegetation in a stream 
 channel such as Gun River which has a very flat gradient and 
 carries little sediment as it leaves the lake. In addition, it flows 
 in an old river channel which is merely a veneer of alluvium over 
 till and, after cutting through the sand, would pick up very little 
 solid material from its bed. Thus, with few tools to work with, the 
 abrasive power would be reduced and vegetation could take hold. 
 It is also interesting to note that such interference with the outflow 
 would take place mainly during the growing season and would 
 tend to hold the lake level higher during the summer months, norm- 
 ally a time of low water. At other times, the dead parts of the 
 plants would offer much less resistance and might be removed dur- 
 ing the spring floods. Another possibility is that the dead parts 
 may accumulate in the channel and gradually raise the lake level 
 until a new outlet is found or the old one cut off during exceptional 
 floods. The latter is known to have happened on at least one lake 
 in our State. 
 
 Near the outlet the shores are bordered by a low, grass-covered 
 swamp and have no definite beaches. This condition persists east- 
 ward until the low swells of the ground moraine which causes 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA 197 
 
 Englaiids Point are encountered. One of the smaller swells forms 
 Brush Island, which rises barely above the surface of Pickerel Cove, 
 and i)ossibl3^ is the cause of the blunt projection of the main shore 
 at the head of the cove. The absence of wave action along this 
 shore is probably due to the shallowness of West Lake which, al- 
 though the reach is large, prevents the formation of large waves 
 and also reduces materially the power of those formed as they 
 approach the shore. Engiands Point, however, runs out into the 
 deeper water and has been carved into low cliffs on the north and 
 northwest sides. Elsewhere the shores are low and the slopes are 
 in places carved into a low terrace hardly two feet above the present 
 level. Locally the shores have been pushed into low ice ramparts 
 by the expansion of the ice during the winter. East of Engiands 
 Point another low swell projects above the surface of the lake and 
 is known as Engiands Island. Some planation of its surface by 
 waves may have taken place in the past, but at present it is cov- 
 ered with a black muck, an indication of vegetal accumulation. 
 
 The south shore of the large embay ment, formed by Engiands 
 and Hastings points, is lined by' a definite sand beach, and the 
 shore features consist of alternating low cliffs and swales. The 
 clififs increase in height as the moraine is approached at the extreme 
 southern end of the bay. At the swales a narrow but definite ter- 
 race of the upper level is present, fronted very commonly by low 
 ice ramparts. Considerable adjustment of this shore has taken 
 place, but the low areas were not cut off by definite bars. A single 
 exception to the last statement was found in the embayment below 
 Chicago Point where a sand bar developed across a triangular 
 depression. The lagoon has been filled by vegetation so that the 
 features are none too evident. 
 
 The broad projection which includes Chicago and Rogers points 
 is due to morainic hills, and the shore features are those due to 
 cutting. The boulder-lined strand is an indication of strong shove 
 by the ice. Off Eogers Point another small island is appropriately 
 named Stony, inasmuch as the finer material of the till has been 
 removed by waves, leaving a mass of rocks standing slightly above 
 the water level. It will be noted that this island is exposed to much 
 stronger winds from the west and northwest than is Engiands 
 Island on the opposite side of tlie bay, therefore the greater amount 
 of wave action. 
 
 At the head of the embayment north of Eogers Point, a sand 
 beach of even curvature lines the shore and has been pushed up into 
 a low rampart. The terrace of the upper level is distinct on this 
 lowland but no evidence of a lagoon was found. Deposition has 
 
198 INLAND LAKES OF MICHIGAN 
 
 taken place, however, but has succeeded in building out a terrace 
 rather than forming a definite bar. On the north shore of this 
 bay wave action is slight, and the swamp vegetation of the swale 
 grows to the water's edge. Northward, the shores are high but 
 drop to a low swamp just south of Hastings Point. The swamp 
 runs across the neck of this peninsula and may have been open 
 water during the higher stage, thus making Hastings Point an 
 island at that time. The point is a swell of the ground moraine 
 rising ten to fifteen feet above the lake level. It is exposed to 
 wave action from the west especially and considerable cutting has 
 been accomplished. Thus the shores are faced by cliffs, and a rocky 
 cut-terrace runs off the end of the point. The rocks on the terrace 
 are an accumulation resulting from the removal of the finer ma- 
 terial and many have been transported to the beach by ice action. 
 The ice shove is strong on this point, for boulders up to five feet 
 in diameter are now lined along the shore. 
 
 The division of the lake into east and west arms by Engiands 
 and Murphy points has already been mentioned. The east arm is 
 similarly divided by Stony and Hastings points and the northern 
 basin of this arm is likewise constricted by Bairds Point and a 
 projection on the east side of Murphy Point. The bays on either 
 side of Bairds Point are both shallow and the point itself is low 
 with the exception of two knolls, one forming the end of the f>oint 
 and the other the expansion near the middle. The terrace of the 
 higher level is distinct on all but the west side and, inasmuch as it 
 extends between the two knolls and between the central knoll 
 and the main shore, it is evident that Bairds Point was originally 
 two distinct islands. Considerable adjustment took place along 
 the island shores and also on the north shore of Hastings Point. 
 Not only was the terrace well developed on the low slopes but, in 
 addition, a bar developed across the head of the bay between Hast- 
 ings and Bairds Points which connected the more southerly of the 
 islands with the mainland. The material of this bar was derived 
 mainly from the north side of Hastings Point, although some may 
 have been added from the Bairds Point islands which are bounded 
 by cliffs on the west side. The higher level terrace is well exposed 
 on the east side of the point and in places reaches a width of one 
 hundred feet. The development of the terrace on this side was due 
 to favorable conditions of shore topography and material rather 
 than to excessive wave action. The latter is much more effective 
 on the west side because the winds from this direction have the ad- 
 vantage of greater velocity and reach and blow across deeper 
 water. When the water level dropped Avave activity was decreased 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA 199 
 
 in greater proportion ou the east side, and sufficiently so that tlie 
 terrace was preserved on this side but was cut away on the west. 
 In fact, the south end of Bairds Cove is effectively protected from 
 wave action and is being rapidly filled with vegetation. It is an 
 excellent example of the encroachment of vegetation from the shore 
 in the form of a floating bog, and the present area of open water is 
 but a small remnant of its former extent. 
 
 The slight projection, A, on the east side of Bairds Cove is a 
 small spit pointing southwestward and is the southerly limit of 
 sand beach Avhich lines the shore to the entrance into Mud Lake. 
 At the northern end of this beach there is also a spit which is re- 
 sponsible for the isolation of Mud Lake. It is probable that the 
 spit was complete, but conditions are somewhat obscured by the 
 roadway built upon it. The entire east shore, including Mud Lake, 
 is lined by moraine whose slopes were washed by the waves at tlie 
 former level. Considerable adjustment took place at this time and, 
 in addition to the spits mentioned, a terrace was formed which 
 varied in width according to the slopes encountered. In general, 
 the terrace was wider in the embayments and has effected a 
 straightening of the shore line with the subsidence of the lake to 
 its present level. At present, the morainic slopes come to the shore 
 in one locality only, the attachment of the Mud Lake spit, and little 
 material is being quarried from the cliffs at this point. It seems 
 reasonable, then, that the active period of adjustment of this shore 
 was limited, to a large extent, to the former level. 
 
 Mud Lake is well named and is a most uninviting place. Shore 
 adjustments on this small lake were much less extensive than on 
 the main lake, and the exposed terrace is not clearly defined. In 
 its place are found bogs and muck shores, in fact, there is scarcely 
 a foot of firm beach on the lake. The lake is not over five feet in 
 depth and the bottom is composed of soft muck. As may have been 
 already inferred, the lake is in an advanced stage of filling by vege- 
 tation and this is being accomplished both by plants encroaching 
 from the shores and by floating forms. On the west side the pecu- 
 liarly shaped cove is caused by a narrow strip of boggy lowland 
 which continues southwestward to the north shore of the east arm 
 of Gun Lake. In the early stages of these lakes an open but shallow 
 connection existed at this lowland as well as at the present outlet, 
 forming an island of the present large point designated B on the 
 map. This connection was not cut off by current action but was 
 filled with vegetation and abandoned when the water level sank to 
 its present position. 
 
 The shores of Gun Lake west of the outlet of Mud Lake are lined 
 
200 INLAND LAKES OF MICHIGAN 
 
 with low sand cliffs, alternating w^ith patches of the terrace of the 
 higher level. The cliffs are being washed to some extent under the 
 present conditions, but the more prominent shore features are the ice 
 ramparts which form in front of the fragmentary terrace. This 
 condition holds until the former connection with Mud Lake is 
 reached. Across this lowland the shores are mucky and have no 
 definite beach. Beyond the lowlands the shores lie on the slopes 
 of ground moraine whose southward extension forms Murphy Point. 
 The swells of this moraine rise barely above the surface of the 
 water and have been carved into low cliffs by the waves, while the 
 sags form swampy shores. On the low shores the terrace of the 
 former level is generally present, and in such localities ice ramparts 
 are usually found. One of the most conspicuous of the morainic 
 sags extends across the neck of the point near its attachment to the 
 mainland. It stands slightly below the level of the higher stage of 
 the lake and, therefore, was an open water channel separating the 
 main portion of Murphy Point from the mainland at that time. 
 This island was later tied to the mainland by a bar which developed 
 on the east side and may now be readily traced across the swampy 
 lowland. 
 
 The adjustments of the east shore of Murphy Point are the result 
 of wave activity and consist of alternating low cliffs and flats which 
 are characteristic of the shores in morainic material. Stony Point 
 is a swell of the ground moraine that juts out into the lake and 
 in this exposed position has suffered considerable dissection by 
 waves. On the north shore the terrace of the former level is well 
 developed but on the south this has been removed by the present 
 activity of the waves. Also the waves have reduced the low neck 
 which connected this point with Murphy Point, and it now stands 
 as an island save for an ice rampart nearly three feet in height 
 which was built during the former stage by material shoved up from 
 both the north and south sides. The push from the south was the 
 more powerful and the stronger rampart follows the south shore. 
 Nearer the end of the point many of the boulders which line the 
 shore have been forced into the base of the low cliffs by the ex- 
 pansion of the ice. 
 
 The south end and west side of Murphy Point are exposed to the 
 strongest winds, and the shore features are at first due to wave cut- 
 ting. The material quarried from the low cliffs has been moved 
 largely in a northerly direction along the west shore and deposited 
 in a small spit at C, which has efficiently protected the partially en- 
 closed lagoon to the rear so that it is now filled with vegetation. 
 North of C the smooth shore is largely the result of adjustments at 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA 201 
 
 the foiTiier level, but uo bar was noted crossing the swamp which 
 forms the neck of the point. The absence of a bar in this locality is 
 ]»uzzliug in view of the fact that a small indentation at D was com- 
 jdetely cut otl", evidently from the south. 
 
 The shores are low along the north side of West Lake, and the 
 shore features are due to wave action of very moderate intensity on 
 account of the shallowness of this arm of the lake. However, the 
 .sandy material was very easily worked, and a terrace was formed 
 at the upper level Avhich is distinct in places. It is best developed 
 in Miles Bay Avhich is a pit of considerably greater depth than West 
 Lake. If currents were active we should expect results at the 
 entrance to this bay, but there seems to be little or no indication of 
 bars. The entrance is very shallow and is becoming more so, but 
 this is due to filling by vegetation rather than by currents. This 
 end of the lake supports a heavy growth of rushes and other water 
 plants and much filling in the future may be expected in this way. 
 
 Farther south along the west shore the sand plain gives way to 
 ground moraine and the land sloj^es gently to the shore. Where 
 not removed tlie woods grow to the shore and have aided in the 
 formation of the excellent ramparts which are the most conspicuous 
 shore features in this locality. 
 
 From this description it should be clear that in the early stages 
 of its existence Gun Lake stood at a higher level and covered a 
 larger area than at present, including some of the nearby lakes. 
 The shores of the upper level are well preserved and show adjust- 
 ments which are less jDronounced than might be expected from the 
 size of the lake. This is due probably to the shallowness of a large 
 part of the basin and to the irregular shoreline. Much of the work 
 was accomplished by waves which succeeded in straightening some 
 of the minor irregularities and forming a terrace of flat slope. Cur- 
 rent activity- played a subsidiary part but succeeded, nevertheless, 
 in producing some important changes, such as the blocking of the 
 channels to Mud Lake and the tying of some of the islands to the 
 mainland. Ice was also very effective and the shove exerted pro- 
 duced many well-defined ramparts and boulder strands. The sinking 
 of the level left some of these features within reach of the waves 
 which have since removed them at exposed places. The activity 
 at present, liowever, is very much less tlian at the liigher level. 
 
 Vegetation has filled, or is in process of filling, the indentations 
 find is now taking hold in the shallower parts of the lake. The 
 outlet of the lake is especially interesting in this regard, since the 
 growth of vegetation appears to hold the level of the lake more 
 nearly constant throughout the season. This, in itself, has to some 
 extent the same effect as though the level were raised and may 
 
202 
 
 INLAND LAKES OF MICHIGAN 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA 2(« 
 
 account for the slight recession of some of the shores. A continua- 
 tion of this ])rocoss will accentuate present conditions and nia^' even 
 raise the level of the lake. 
 
 LAKES CADILLAC AND :MITCIIELL 
 
 Lakes Cadillac and Mitchell followed Gun in order of appearance 
 and lie in a region which is m'ost interesting from a physiographic 
 viewpoint. The location of the city of Cadillac is especially fortu- 
 nate. It is built at the junction of a large niorainic Iract on the east 
 and an extended outwash plain on the west. The surface of the 
 outwash plain is extensively pitted, but few of the pits hold water. 
 However, two large ones are located just west of Cadillac and form 
 the basins of Little and Big Clam lakes, or Cadillac and Mitchell 
 lakes, as Ihey are now called. Tliose lakes, altliongli not large, are 
 well adapted to physiographic studj^ on account of the adjustments 
 in the easily worked material and, furthermore, are readily acces- 
 sible, Cadillac being the junction of the Grand TJapids and Indiana, 
 and Ann Arbor Railroads. It is but a few steps from the railroad 
 station to the shore of Cadillac Lake which extends for nearly three 
 miles in a direction south of west. In width it is rather uniform 
 and nowhere reaches one mile. Its area is 1.9 square miles, making 
 the average width very close to three-fourths of a mile. Compared 
 with its neighbor, Mitchell, it is very regular in outline, the main 
 exceptions being the large projection on tlie north shore and a 
 small baj' at the southwestern end. 
 
 The outlet. Clam IJiver, leaves the lake at the northeastern end 
 and flows through a depression in the outwash in a broad cun^e 
 convex northward to the Muskegon Kiver in northwestern Clare 
 County. It flows in a cliannel which has been cut a few feet below 
 the level of the plain but has been obstructed by a dam of low 
 head built by one of the numerous manufacturing plants in the city 
 of Cadillac. The obstruction of the outlet leads at once to the 
 expectation of flooded shores, which is fulfllled. 
 
 The vshore facing the city is obscured by buildings and lumber 
 docks and shore conditions are not well represented. Around the 
 north shore, point A, see map. Fig. 67, projects into the lake and con- 
 tinues under water as a shoal of less than four feet in depth for a 
 distance of three hundred yards or more. Little evidence of a distinct 
 submerged terrace is present, and this point is considered as one of 
 the less deep portions of the pit which has been exposed by a sinking 
 of the water level. The swampy condition of the point furnishes 
 early evidence of the flooded condition of the lake. The swamp con- 
 tinues along the shore westward to the vicinitv of localitv B on 
 
204 INLAND LAKES OF MICHIGAN 
 
 map, wliei'e the edge of the outwash comes to the lake and stands 
 in a steep cliff. Between the foot of the cliff and the beach a narrow 
 terrace stands ahout one foot above the present level. This terrace 
 merges into the swamp just described and with its surmounting 
 clifif may be traced to the northeast. 
 
 West of B the terrace widens and has been pushed into an ice 
 rampart near the present shore, forming a foul lagoon which is rap- 
 idly being filled with vegetation. The rampart persists to the west 
 end of the lake but the lagoon widens and extends through to Mitch- 
 ell Lake along the course of the natural outlet of this lake. A canal, 
 navigable for boats of light draught, was dug about 1870 through 
 the narrow neck of land w^hich separates the two lakes and has 
 caused the abandonment of the outlet. This neck of land stands 
 below the general level of the outwash plain with the exception of 
 a small fiat-topped mound, probably a remnant of the outwash, on 
 the south side of the canal midway between the two lakes. On the 
 Cadillac Lake side the neck has been carved into the persistent low 
 terrace which widens south oif the canal and runs to the embayment 
 at the southwestern end of the lake. On this low flat the footp'ath 
 follows a well-defined bar of the higher level from the low cliffs 
 just south of the canal to the bridge which crosses the bay at the 
 southwest end. Across the bridge the outwash and narrow terrace 
 soon appear, and the latter is made use of for the roadway as far 
 as locality C. East of this point the land is low and the terrace is 
 not distinct for a short distance. East of the broad bend of the 
 shore line the outwash again comes to the lake with the usual nar- 
 row terrace and continues thus until the shores are obscured by 
 buildings or docks. Along this shore the effects of the artificial 
 raising of the lake level are very evident. The terrace, wliich prac- 
 tically surrounds this lake, has been utilized for a boulevard but is 
 quite generally being attacked by the waves, so that some protection 
 is necessary. The pine stumps, so abundant in this region, have 
 been used for a breakwater but have proved none too satisfactory, 
 so that in addition to an unsightly shore, much repair work must be 
 done. 
 
 From the above, it is evident that Cadillac Lake has stood at a 
 somewhat higher level than at present and that the cause of the 
 lowering of the former level was the deepening of the outlet. During 
 the higher level the main adjustment of the shores was the formation 
 of a narrow terrace which is well preserved where natural condi- 
 tions have not been disturbed. Current action accomplished little 
 during this stage, the only characteristic fonn recognized being at 
 the west end south of the canal. The condition of affairs previous 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA 20") 
 
 to the artificial elevation of the water level lias been obscured by 
 the recent increased activity of the waves which are rapidly remov- 
 ing the former terrace, except where protected. During stages of 
 high water this terrace is covered, and it may be considered that it 
 merely represents the activity of the shore agents during such 
 times. This argument would have more force were not the lake held 
 abnormally high. Also the rounded, grass-covered slopes above the 
 wide portions of the exposed bottom are an indication of a definite 
 water level, long since abandoned. 
 
 Ice action is very effective wherever shore conditions permit and 
 some well-defined ramparts are to be found. The process of extinc- 
 tion is making slow progress at present and little has been done in 
 the past. At the abandoned outlet and in the bay at the southwest 
 end, considerable filling has been accomplished by vegetation but the 
 main body is still relatively clear. Filling by sedimentation is in- 
 significant for the water comes either from seepage or flows from 
 Lake Mitchell, which acts as a perfect settling basin. Undoubtedly 
 an unobstructed outlet would continue to deepen but, since the 
 depths of the lake were not taken, the final effect cannot be ventured. 
 
 LAKE ailTCHELL 
 
 North of the canal on Lake Mitchell, the shore, as shown on 
 the map, is particularly regular, but inferences as to the probable 
 shore conditions based on this Avould be greatly misleading. As a 
 matter of fact, the outwash })lain is excessively pitted and the shore 
 consists of a succession of flats interrupted by island-like mounds 
 of outwash. Proceeding northward from the canal, evidence of a 
 higher level for this lake is soon at hand. A low sand bar rises 
 above the flat at a distance varying from twenty to fifty feet back 
 from the shore and encloses a poorly drained lagoon behind it. 
 On the gentle slope in front, faint beaches may be discerned which 
 in places enclose small narrow depressions, now dry. These bars 
 were probably formed as storm beaches during the gradual lowering 
 of the level. Farther north, the waves at the higher level succeeded 
 in cutting back the "islands" of outwash, while the currents dis- 
 tributed the material in a series of connecting bars which have 
 been pushed up into ramparts in many places. These bars stan.l 
 at a somewhat lower elevation than those found on other parts of 
 the lake and it is probable that they did not extend above the water 
 level. But, with the dropping of the water level, the bars were ex- 
 posed and the shore line assumed its present regular contour. 
 
 Northward the lowland areas decrease in size and at the north 
 end disappear, leaving a continuous cliff in the outwash. This, how- 
 
206 INLAND LAKES OF MICHIGAN 
 
 ever, stands some distance back from the shore and the exposed 
 flat below is a portion of the former shallow bottom. In order to 
 get the full view of the flat it is necessary to stand on the embank- 
 ment of the Ann Arbor Eailroad which crosses in this locality. 
 The beach of the present level, called the White Sands, swings 
 around the north end in a perfect curve, see Plate XII, A, and is an 
 exceptional!}^ fine example of an adjusted sand beach. The flat 
 adjacent to this shore swings far to the west and forms the so-called 
 "Thousand Acre Swamp," above which rise ouitvN'ash remnants simi- 
 lar to those on the opposite side of the lake. The iDerfection of the 
 adjustments along this shore is again detected in the excellent bar 
 which stands a short distance inland between the lake and the 
 swamp. Ice has been particularly active along this shore and has 
 piled up two distinct ramparts in front of the bar. At the present 
 level a low rampart is in process of formation, making three in all, 
 a rare occurrence. The continuation of the bar to the south forms 
 Saw Dust Point Avhich shows the direction of the prevailing cur- 
 rents at the time of formation. Thus, the material must have been 
 derived from the cliff's at the north end of the lake. 
 
 Ice action on this lake is most interesting, not only on account 
 of its effects but also because of the manner of its occurrence. In- 
 formation from reliable sources makes it certain that both ice jams 
 and expansion take place. At Doxsie's ice jams have shoved twenty 
 feet on the shore and have piled up ten feet in height, a statement 
 well within reason as may be seen from Plate XII, B, w^hich is repro- 
 duced from a photograph of an ice jam on the north shore of this 
 lake taken by Mr. W. E. Sours of Cadillac. Also at Doxsie's, the 
 ice advances by expansion between four and five feet a season. The 
 width of the lake at this locality is one and five-eighths miles and 
 approaches the maximum limit for ice shove of this type. 
 
 Below the shallow bay back of Saw Dust Point the outwash, with 
 cliff and terrace of the higher level, again appears and extends to 
 the entrance of Mud Cove. From this shore the depths of the lake 
 are readily distinguished when light conditions are favorable. The 
 northern part is very shallow with the exception of a narrow chan- 
 nel in the middle running nearly north-south and appears as though 
 it were almost possible to wade across. The southern lobe lies in a 
 much deeper j)ortion of the pit and is sharply contrasted in color 
 with that on the north. 
 
 Mud Cove is a partially detached portion of the pit which con- 
 tinues to the west as an extended swamp. Currents have accom- 
 plished little or nothing at the entrance and there is little likelihood 
 of its being isolated in this manner. Nevertheless, its existence as 
 
Michigan Geological and 
 Biological Survey 
 
 I'ul)lication 30, Geological Series 25, 
 Plate XII. 
 
 A. WHITE SAXD,S, LAKE MITCHELL. 
 
 B. ICE-JAM, LAKE MITCHELL. 
 
 (Photography l)y W. H. Saurs.) 
 
cV, 
 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA 2U7 
 
 i\ body of open water is very liiuitod on account of the rapid fiUinj;' 
 ^W vegetation. On the soutli side of tlie Cove, a case of complete 
 extinction of a sliallow part of this irregular pit is found. In this 
 case thei indentation was separated from the lake by a narrow strip 
 of outwash which, at first sight, appears to be a bar. The extinct 
 part has an uninterrupted, flat surface composed of a black, peaty 
 soil upon which a few shrubs are growing. An attempt was made 
 to drain this for agricultural use but evidently was not an unquali- 
 tied success. 
 
 East of Mud Cove tlie outwash, with cliff and terrace, lines the 
 shore but soon drops to a small swamp which was cut off by an 
 ice rampart about twenty feet inland from the present shore. The 
 outwash again appears at the shore but finally gives way to a low- 
 land which extends to the double bay at E. This bay is caused by 
 two small pits which drop below the level of the swamp. It is set 
 off from the main lake by very shallow water at the entrance. Cur- 
 rents undoubtedly flow across the entrance of the more southerly 
 lobe but no bar is present, although the water is scarcely deep enough 
 for rowboats to cross. This bay will eventually be isolated, how- 
 ever, by the deposits from the heavy growth of vegetation in the 
 shallow water across the mouth. 
 
 From this locality eastward, the outwash cliffs face the shore 
 but the terrace of the higher level is fragmentary. The northwest 
 winds strike this shore with full force, and the waves and currents 
 have accomplished much in the adjustment of the shore. This is 
 shown b}^ the regularity of the beach, see Plate XIII, the well-de- 
 fined submerged terrace, and the presence of cusps at low water. 
 (See Douglas Lake on cusps.) Near the east side of the lake the 
 outwash cliffs continue to the south shore of Cadillac Lake, and 
 the shore of the higher level is well shown by a bar which follows 
 the curvature of the southeastern shore towards the canal. The 
 land separation between the two lakes stood somewhat above the 
 water level of the higher stage and the connection was restricted 
 to a narrow strait in the vicinity of the abandoned outlet. 
 
 Thus, in times past, a large lake of very irregular outline stood 
 in this localitj'. The greatest variation from the present condition 
 occurred in the Lake Mitchell basin which then included a great 
 part of the swamp areas to the west. In the early stages, the shores 
 were much more irregular than at present, due to the numerous 
 shallow pits above which stood "islands" of the original outwash 
 plain. At the higher level, important adjustments of the shores 
 took place which resulted in the formation of a definite submerged 
 terrace and the straigliteniiig of the sliore line in many places, due 
 
208 INLAND LAKES OF MICHIGAN 
 
 both to cliff recession and to the development of connecting bars. 
 It is probable that many of the bars Avere not bnilt to the water 
 level but it seems clear that the ''Thousand Acre Swamp" was par- 
 tially cut off. The downcutting of the Clam Elver in the loose 
 sands of the outwash lowered the level of this lake and formed ts\'o 
 definite basins connected by an outlet, now abandoned. After the 
 separation of the basins the further deepening of Clam Eiver was 
 felt directly by Cadillac Lake which stood at a lower level, while 
 Mitchell was held up by its outlet. At this stage the adjustment 
 of the shores of Lake Mitchell, accomplished during the upper level, 
 became fully effective and are the cause of the long stretches of 
 regular shore line. . 
 
 Ice exerts a powerful shove on the shores of these lakes and is 
 especially active on Lake Mitchell. Wherever vegetation served to 
 bind the sand, ramparts have been formed and the multiple ramparts 
 on the front slope of the bar at "Thousand Acre Swamp" are of 
 exceptional development. The manner of the ice push ma}- be 
 .studied to advantage on Lake Mitchell since both expansion and 
 jams are active. We are very fortunate in having a photograph 
 of the latter in action. 
 
 The cutting of the canal between the two lakes has caused but 
 little lowering of the level of Lake Mitchell, and its shores remain 
 practically in their natural condition. However, on Cadillac Lake 
 the interference of man has obscured a large part of the shore 
 topography and initiated far reaching changes. Aside from the 
 buildings and lumber docks along the east shore, the raising of the 
 level by a dam should be mentioned. The natural level of the 
 lake has been covered and obliterated by subsequent wave action. 
 Also a new cycle of shore activity has been brought about which is 
 working havoc and has necessitated the novel but unsightly break- 
 waters. 
 
 The lakes are fed to a large extent by ground water and the few 
 entering streams flow from swamps, so that filling by sediment is 
 not of importance. Vegetation has made little headway except in 
 the protected bays, and this may be due to some extent to the 
 sweeping effect of the ice jams near the shore from which the plants 
 usually encroach. This type of filling, however, may be expected 
 to increase in the future and eventually will fill the basins. Little 
 may be expected from deepening of the outlet as long as the dam 
 is maintained, a condition likely to prevail. 
 
CHAPTER VI 
 
 INTERIOR LAKES OF THE SOUTHERN PENINSULA, CON'T. 
 
 LAKES OF THE WESTERN INTERLOBATE AREA 
 AND OF ALPENA COUNTY 
 
 HOUGHTON LAKE 
 
 Of the interior counties of the Southern Peninsula, Roscommon 
 is perhaps the most fortunate as regards lakes. Others there are 
 which surpass it in number but none in respect to size. Houghton 
 and Higgius lakes, the most important, are both of large dimen- 
 sions as our inland lakes go, the former with an area of 30.8 square 
 miles taking first rank in the State. They lie in the western and 
 northwestern part of the county and, although not so readily ac- 
 cessible as many, are nevertheless very popular during the summer 
 and fall mouths. Roscommon on the Michigan Central Railroad is 
 the most convenient point of departure for these lakes, although 
 Houghton Lake is frequently reached from the west. 
 
 A i^hysiographic study of these lakes is most interesting and, in 
 fact, surprising. Obviously, for such a study all the advance in- 
 formation possible proves of great service and much information is 
 always obtainable. In this case, one may find that the glacial geology 
 has been thoroughly worked out and described.* Tliis region is a 
 great interlobate area between the lobes of ice which filled simultan- 
 eously the Michigan and Huron basins. The outline of the Michigan 
 lobe wias relatively ismooth at this stage and ran northenst-southwest 
 in the northwestern part of the State. The Huron lobe, however, was 
 noticeably irregular and extended far to the southwest into the 
 Saginaw Lowland, forming a large subsidiaiy lobe, the Saginaw 
 lobe. Thus, between the northeast side of the Saginaw lobe and 
 the Michigan lobe is a broad interlobate area which extends north- 
 east-southwest, see Fig. 3. The recession of the ice in this area 
 was to the northeast, and somewhat fragmentary cross-moraiuic 
 ridges were deposited at intervals in the region extending northeast 
 from Cadillac to Roscommon. Three such ridges cross Roscommon 
 County and in the troughs between them are "located Houghton and 
 
 *'Monograph 53, U. S. Geol. Survey. 
 27 
 
210 
 
 INLAND LAKES OF MICHIGAN 
 
 Higgins lakes, see Fig. G8. This accounts for their northwest- 
 southeast trend and similarity in size. With such information at 
 hand one naturally visits the lakes prepared to find other points of 
 similarity but herein is the surprise, which may well be deferred 
 until the lakes have been described. 
 
 Morp/ne 
 
 77// P/o'/n 
 
 Ocr/-iA/as^ 
 
 Fig. €8. Map showing the distribution of th-e glacial formations in the vicinity of 
 Houghton and (Higgins Lakes. (After Leverett) 
 
 Houghton Lake is not a summer resort, strictly speaking. Its 
 popularity is due to the excellent fishing and hunting and for this 
 there is a geographic reason. This lake, although more than eight 
 miles long and over four wide, does not anywhere exceed twenty-five 
 feet in depth and is filled with an almost continuous weed bed, except 
 in a zone about the shore- This is the lair of countless fish of many 
 kinds and accounts in part for the attraction of this lake to sports- 
 men. Interesting and instructive though a study of the plants and 
 animals of these waters might be to one capable of undertaking it, 
 the writer must dismiss it with the mere mention of wild rice. This 
 furnishes food supply to migrating birds which flock here in great 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA. 211 
 
 inimbers to feed and rest in the fall of the year, thns the attraction 
 to the hnnters. 
 
 As already mentioned, Uonghton Lake lies in a trongli between 
 two roug-hly parallel moraiuic belts trending uortliwest-sontheast 
 but does not comi)letelY fill the depression. In reaching this lake 
 from Eoscommon, one crosses the northerly moraine just south of 
 Higgins Lake, see Fig. G8, and drops down to a Avooded lowland 
 which becomes swamp within a half mile of the lake. This till 
 plain borders the northeastern shore of the lake. Outwash plains 
 swing around tlie ends and border tlie southwest shore of the lake 
 along the constricted portions at either end, with the exception of a 
 small area of till plain near the northwest end. The main body of 
 the lake, however, is retained on the west by moraine. From the 
 shallowness of the basin and the presence of broad shoals in the 
 main body of the lake, it seems clear that this body of water lies 
 mainly on till plain, and is retained at the ends by outwash and 
 on the southwestern side by a moraine. Such basins were classed 
 as inter-morainic in Chapter I. 
 
 The first impressions of this lake depend very largely on the point 
 of view. From the heights of the moraine on the southwest side, 
 the broad expanse of water, fringed with forest on the opposite shore 
 and with highland in the background, presents a pleasing landscape. 
 From the east shore, however, conditions are very different. After 
 a trip of a half hour across a low swamp, the first glimpse is caught 
 through an opening in the trees near the water level and, while not 
 so fortunate perhaps, gives the more accurate impression. The hori- 
 zontal dimension is exaggerated at the expense of the vertical and 
 things appear flat. The familiar fishing boats, which dot the surface 
 during favorable weather, are enlarged by optical illusion and sug- 
 gest at once the favorite recreation on this lake. 
 
 The road from the north reaches the lake at the inlet, or Cut, as 
 it is called, see Fig. CO, and the initial observations here lead to the 
 conclusion that shore action due to waves and currents is feeble, 
 due obviously to the shallowness of the lake. The inlet, a rather 
 sluggish stream in this part of its course, has, nevertheless, brought 
 down considerable silt which has been carried out and deposited 
 in parallel submerged bars on either side of the current, extending 
 from the shore directly out into the lake for a distance of about 
 one hundred yards. A cousiderable portion of the water of the 
 Cut comes from Higgins Lake which serves as a settling basin, and 
 the stream is not nornuilly heavily laden with sediment. Since this 
 material is not distributed along the shore, it is evident that the 
 waves and currents are not only relatively, but actually, of little 
 power. 
 
212 
 
 INLAND LAKES OP MICHIGAN 
 
 The presence of a third bar, similar in form but located more to 
 the south and in front of a recently abandoned stream channel on 
 the flat adjacent to the inlet, shows that a northward shifting of 
 the channel of the stream has taken place. Those qualified to know 
 state that this was caused by a large, temporary increase in volume 
 of the Cut, due to the removal of a dam at the outlet of Higgius 
 Lake a few years ago, and that the more northerly of the two bars 
 which formerly existed in front of the abandoned channel was re- 
 
 Fig. 69. Outline map of Hou.gliton Lake, Roscommon County. 
 
 moved as the channel shifted. Other evidence of the weakness of 
 the waves and currents is the absence of a well-defined submerged 
 terrace and the broad sinuosities of the shore line. 
 
 -Some action has, of course, taken place and has formed a rather 
 broad zone of sand off-shore in lieu of a terrace, which does not 
 support the heavy growth of vegetation found farther out in the 
 lake, Also there is an uneven beach of clear sand, which is reced- 
 ing at the present time and laying bare the roots of the trees grow- 
 ing at the high water mark. The recession of the shores is due to the 
 obstruction of the outlet. In the lumbering days a dam of about 
 four feet in height was constructed three-fourths of a mile below 
 the lake and was used to retain the water in the spring. It is prob- 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA. 213 
 
 able that the level of the lake was raised ouly a small amount, if 
 at all, but it was held at the liif»;h water marlv for a longer period 
 than normally at least and the waves became relatively more power- 
 ful. The dam is no longer maintained but the same effect is some- 
 times produced by log jams. 
 
 Quite in contrast to the work of waves and currents is the effect 
 of the ice on these shores. Ice jams, which reached a height of fifteen 
 to eigliteen feet at the Cut in the spring of IDlo, are of frequent oc- 
 currence on this shore. The gentle slope and the sandy character 
 of tlie shore are not favorable to the formation of ramparts, and 
 the chief effect is exerted on the trees which often stand in water 
 along the shore at this time of the year. DriftT\'Ood is piled high 
 on the beach, live trees snapped off and overturned, and the bark 
 is scrubbed from the trunks of trees still standing. 
 
 Shore conditions are so uniform on the northeast side of the 
 lake from East Bay to Long Point that no detailed description 
 seems necessary. Cusps, which are low water forms, are found on 
 other parts of the lake and might well occur on this shore where 
 not littered with driftwood. They are, however, very transitory 
 and their occurrence in definite localities would be of little signifi- 
 cance. 
 
 Long Point, although it extends more than half way across the 
 lake, is, nevertheless, an extension of the outwash and not a current 
 deposit. However, shore action is somewhat more in evidence on 
 the north -side of this point than along the shore just mentioned. 
 When this lake was visited by the writer some exceptionally well- 
 developed cusps were present near the tip of the point and a definite 
 sand bar free from vegetation extended a short distance into the 
 lake in line with the shore. Also, a very definite zone of sand 
 off the entire north side of the point indicates an ineffectual attempt 
 at terrace formation. 
 
 The land surface bordering the east side of North Bay has very 
 low relief but portions of outwash, standing approximately three 
 feet above a swamp which is at lake level, may be detected. In 
 front of the swamp areas, bars have developed but at a level slightly 
 higher than the present. These bars are probably a single discon- 
 tinuous bar which becomes less distinct at the northeastern end 
 of the bay but reappears along the swamp at the north end of the 
 lake, the flat of the Muskegon Kaver. It follows the present shore and 
 ■swings in a gentle curve to the outlet. Across the outlet it continues 
 to the low swells of the ground moraine nearly a mile to the south 
 but stands a hundred yards back from the shore for most of its 
 length. Along the present beach is a similar bar which ends abrupt- 
 ly at the outlet and stands at a slightly lower elevation than the 
 
214 INLAND LAKES OF MICHIGAN 
 
 bar farther inland. In general, the bars on the west and north 
 shores are better developed than those in the vicinity of Long Point. 
 The short stretch of low shore along the north side of Long Point 
 furnished a ver}^ limited amount of material for the currents flow- 
 ing eastward with the winds from the northwesterly quadrant. On 
 the Avest side, however, a relatively large amount of material was 
 quarried by the waves from the cliffs to the south and was deposited 
 by currents first in the bar whicli stands farthest inhrnd. This bar 
 swung around the north end of the lake (the outlet was farther to 
 the west at this time) and turned a small inlet to the east before 
 merging into the undertow. 
 
 The altitude of the bar shows that the lake must have stood at an 
 elevation at least two feet higher than the present high water 
 mark and possibly more, since the bar may have been submerged. 
 This higher level must have covered the flat adjacent to the outlet 
 and extended some distance down stream. Also the low swamj) 
 bordering the northeast side of the lake was flooded at this time. 
 The drop in level was due to the deepening of the outlet and was 
 gradual. As the level lowered a passage was maintained by the out- 
 flowing water through the bar which terminates farther to the east. 
 Eventually the currents were unable to continue along the bar, on 
 account of the decreasing depth of the water, and took a course in 
 the deeper water near the present shore. In this way the second 
 bar was started but, with continued lowering of the water level, its 
 length became fixed by the establishment of a definite channel by 
 the outlet. Its subsequent growth was small but sufficient to force 
 the outlet slightly to the north at its debouchure. 
 
 South of the grass-covered outwash lie the gentle swells and sags 
 of the till plain which rises to the moraine near Houghton Lake 
 village. The moraine borders the shore, as shown in Fig. 68, and 
 then drops to the outwash plain which nearly surrounds the south- 
 western end of the lake. The prominent features along this shore 
 are the wave-cut cliff'S which rise from the high water mark and 
 show no evidence of the higher level found near the outlet. Where 
 the till plain and the moraine come to the shore, the cliffs are com- 
 posed of boulder clay and are steep. The outwash is composed of 
 unconsolidated sand but is covered with a close mat of grass so that 
 the cliffs are somewhat steeper than is usual for this material. The 
 different formations are also expressed in the character of the beach 
 material which contains cobbles and boulders in front of the mo- 
 raine but is of clear sand along the outwash plain. 
 
 In conclusion, it may be stated that shore agencies on Houghton 
 Lake have been productive of meager results. Waves and currents, 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA. 215 
 
 although rejuvenated to some extent at the present time, are of 
 little power and this inactivity is due to the shallowness of the 
 lake and the heavy growth of vegetation which effectively interfere 
 with wave and current development. Shore adjustments have been 
 of somewhat greater importance in the past when the lake stood at 
 a higher level. At this time, currents shoAved their greatest activity 
 in North Bay and succeeded in decreasing the size of this arm to a 
 considerable extent. The greatest activity of the waves has taken 
 place along the southwestern shore, as shown by the prominent 
 cliffs. An index to the combined activity of waves and currents 
 is the development of the submerged terrace and this is nowhere 
 well-defined. Ice jams are of frequent occurrence and great in- 
 teusitj", especially on the northeast side of the lake, but shore con- 
 ditions are not favorable for decided effects. 
 
 The lake is, however, an interesting example of one physiographic 
 process, — that of extinction. Little sediment is brought in by 
 streams and the outlet is deepening at a very slow rate. Yet the 
 lake is filling very rapidly due to vegetation. Geologic processes 
 are, as a rule, acting very slowly according to human standards 
 of time but one ma}- look forward to the extinction of this lake in 
 the course of a few generations. 
 
 HIGGINS LAKE 
 
 Higgins Lake is reached conveniently from Roscommon on the 
 Michigan Central R. R. by stage and is well worth a visit. With an 
 area of nearly fifteen and one-half square miles and a length of 
 seven miles, this lake ranks among the larger of our inland lakes. 
 Also it is one of the most beautiful. As seen from the summit of 
 the moraine to the south, the lake consists of two broad arms 
 which stretch west and south from a central constriction. The 
 interesting island, the green slopes and the clear blue water blend 
 into a most attractive landscape. From most points on the shore, 
 however, the view is restricted but is sufficiently inviting to induce 
 one to explore further. The clarity of the water is remarkable and 
 might well have inspired a less prosaic name than Higgins. It is 
 so clear that it looks cold and such is the case, for the lake is deep 
 and is fed by springs. The change from the light color of the shal- 
 low water to the blue of the depths is sudden. So sudden, in fact, 
 that to float across this zone on a peaceful day gives one the in- 
 describable sensations of sailing into space. 
 
 Likewise from a physiographic viewpoint Higgins Lake is most 
 interesting. The general features of the region are quite simple as 
 regards geology, see Fig. GS. Moraines border the north shore of 
 
216 
 
 INLAND LAKES OF MICHIGAN 
 
 the west arm and parallel the southwest shore of the entire lake. 
 The moraiuic slopes on the southwest side do not reach the lake 
 but are separated by a narrow zone of sand which widens into broad 
 outwash plains at both ends of the lake. The butwash was formed 
 when the ice border stood at the northern moraine, and covered the 
 till plain between the two moraines, with the possible exception of 
 the lake basin itself. Kearer the moraines the outwash merges into 
 a terrace, and till is exposed beneath the sands on the southwest 
 side of the lake. 
 
 Fig. 70. Outline map of Higgins Lake, Roscommon County. 
 
 Ideas are prevalent concerning remarkable depths of the water 
 in this lake but the greatest depth reported from reliable sources is 
 one hundred twenty-five feet. This depth makes it certain that the 
 lake does not cover a till plain, as does its neighbor Houghton, or 
 else the outwash is excessively thick. From the shape of the lake 
 it seems probable that the basin is a large, irregular pit in the out- 
 wash formed by buried blocks of ice, as explained in Chapter I. 
 
 In addition, the history of the lake is as complicated as that of 
 some of the lakes which were connected with the Great Lakes and is 
 in itself an interesting story which could not be fully deciphered at 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA. 217 
 
 » 
 
 the time this study was made. The former strands are very distinct 
 aud show a hike of much greater size at its maximum hei.ulit. A 
 careful mapping of these shores and the adjustments wliicli took 
 place at the various levels promises a most interesting and profit- 
 able study. 
 
 At the outlet of the lake natural conditions are somewhat dis- 
 turbed by the activity of man. In the lumbering days a dam was 
 constructed to retain the water for the log runs in the spring. The 
 level was raised about forty inches but lowered rapidly during the 
 runs, remaining low during the greater part of the year. With the 
 passing of the forests the dam lost it§ usefulness and the lake 
 assumed its normal level. The outlet was again obstructed in 1911 
 for the purpose of permanently raising the level of the lake. This 
 caused a renewal of the activity of the waves, and the effects were 
 sufficient in one year to convince those concerned of the inadvisabil- 
 ity of such a condition, so that the dam was blown out. ^>uch dis- 
 turbances make the interpretation of conditions at the outlet some- 
 what uncertain but the significant facts may be ascertained. 
 
 The outlet, see Fig. 70, flows in a channel which has been cut 
 approximately eight feet below the level of the sand plain. It flows 
 in a southeasterly direction to the nearby Marl Lake and thence 
 eastward for more than two miles before turning south through a 
 low sag in the moraine which stands between Higgins and Hough- 
 ton lakes. In the bed of the outlet and along the shore to the 
 southwest a large number of boulders of considerable size are 
 found. These boulders lie at the base of the sand plain which is 
 relatively thin here and are washed free from the finer material by 
 the movement of the water. Considering the outlet, it is obvious 
 that down-cutting has taken place rapidly while the stream was 
 running over the sands but was accomplished much more slowly as 
 the boulders were encountered. Thus, from this limited locality 
 one may hypothecate higher levels for the lake and be prepared to 
 find conclusive evidence on the slopes adjacent to the present shores. 
 
 At the south end of the lake, the outwash soon gives way to a 
 narrow flat which borders the shore to the west side. The even 
 curvature of the shore and the clear sand beach are suggestive of 
 current action and further evidence is found in two sand bars a 
 short distance inland. These bars stand about two feet above the 
 present lake level and parallel the shore from one side of the lake 
 to the other. The lagoon, thus formed, was very shallow and is now 
 dry. If these bars were formed by currents, their growth must 
 have been from the west because shore drift on the east side of the 
 lake could not well have passed the outlet. Undoubtedly currents 
 
218 INLAND LAKES OF MICHIGAN 
 
 brought mucli material to this shore from the cliffs along the west 
 side of the South Arm, where au uninterrupted shore of nearly 
 three miles in length is exposed to the northeasterly winds. In the 
 early stages of this level, the currents did not leave the shore in 
 this locality but, as the submerged terrace developed and widened, 
 both the waves and the currents were retarded by its influence. 
 Finally, the retardation was sufficient to exert a back pressure on 
 the strong currents moving southward along the west side and 
 they were forced from the shore. Thus the first bar (farther in- 
 land) was formed and a slight lowering of the level as the outlet 
 deepened may have caused the second bar somewhat farther out in 
 the lake. (See Crystal Lake for a similar occurrence.) 
 
 At the present level the submerged terrace is wide but not sharply 
 defined. Soundings showed an exceptionally wide terrace near the 
 west side which extends into water nearly twenty-five feet in 
 depth before dropping into deep water. The zone of sand, hoAvever, 
 stopped at depths of ten to twelve feet, and this probably repre- 
 sents the greatest depth of the terrace, since the finer material is 
 usually swept from the terraces in lakes of this size and may 
 possibly be carried from one end of the lake to the other. 
 
 The west shore of the South Arm is bounded by low cliffs for most 
 of its extent. In two localities only were the effects of current 
 action detected. At locality A, see map, a small point breaks the 
 regularity of the shore line, and currents have formed a small spit 
 pointing southward, a fact in harmony with the view that the bars 
 at the south end were formed by currents from this shore. Again at 
 'Chicago Point currents are active but this will be reserved for dis- 
 cussion later. The cliffs along this shore do not rise above six to 
 eight feet and have been carved in the sands of the adjacent plain 
 which is a terrace of a higher level. The material is homogeneous 
 and unconsolidated and has been very uniformly worn by the waves, 
 forming an exceptionallj^ smooth shore line. A distinct submerged 
 terrace drops at eight to ten feet but is of relatively narrow width. 
 Off locality B a rock shoal comes to within six or eight feet of the 
 surface. It is thus within the zone of action of the waves which 
 have removed the finer material from the north end and deposited 
 it in a bar at the opposite end. Clearly the northerly winds are the 
 most effective in this locality. 
 
 Chicago Point has much of interest for the physiographer. In 
 this locality, evidence of levels higher than the present is at hand 
 and, in addition, changes are now taking place which are of great 
 significance. Back from the shore stands a broad sand terrace 
 which slopes very gently upward to the morainic hills in the dis- 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA. 219 
 
 tance. Its li(M<2,lit wns not iiieasnvod in this locality but a similar 
 terrace at the State Forest marks a level of the lake sixteen feet 
 above the present. This terrace, whidi borders tlie shore of the 
 south arm. drops to a lower one standing at an elevation of eiiihl 
 feet above lake level, formed probably at the time when the outlet 
 flowed over the surface of tlie sand flat. The spit began its develop- 
 ment at this time but, inasmuch as the outlet cut down through 
 the loose sand rapidly, the growth was small. The outlet soon cut 
 through the sand ami encountered the boulders at its base which 
 held the lake at a level two feet above the present for a considerable 
 time. During this stage the spit increased in length more than two 
 hundred feet. At the present level the spit is continuing its growth 
 and, including the submerged portion, extends fully a quarter of a 
 mile out into the lake. On the opposite side of the lake the sub- 
 merge terrace is fully as wide as the length of the spit, so that the 
 narrows is in reality much more restricted than appears from tlie 
 map. An ice rampart borders the south side of the spit at the 
 present level but its counterpart is not present on the opposite 
 side. This decrepancy may be due to the destruction of the rampart 
 on the north side. The material of the spit is not suitable for the 
 formation of a permanent rampart and, if formed, it would be 
 subjected to strong wave action, due to northerly winds of consid- 
 erable reach. 
 
 West of Chicago Point the eight-foot terrace becomes faint and 
 drops to the two-foot level in the vicinity of point C, see map. Off 
 this shore the lake is shallow to the Island, due probably to the 
 nature of the basin. The condition of the Island during the stages 
 above the two-foot level cannot be determined but it is probable 
 that its top was bevelled to some extent at the eight-foot level. 
 During the two-foot level a much smaller area than the present 
 stood above the water and was subjected to considerable adjust- 
 ment by waves and by ice. Terraces and strong beaches were 
 formed all around the island but were better developed on the north- 
 west side throughout. In addition, the bouldery material was 
 pushed up into ice ramparts which are very pronounced on the 
 northwest shore. The expanse from the Island to the west shore, 
 a' distance of more than two miles, is generally considered excessive 
 for the formation of ramparts by expansion, so that ice jams ap- 
 pear to be effective here. However, the maximum size of lake upon 
 which expansion is effective is as yet uncertain and it is unfor- 
 tunate that little could be learned concerning the manner of tlie 
 ice push on this shore. At the present level the waves are active 
 on the northwest side and much of the material quarried is trans- 
 
220 INLAND LAKES OF MICHIGAN . 
 
 ported to the point at the northeastern end of the island, where it 
 is being deposited in a spit extending almost due east. Another 
 spit was found at the sharp projection on the southeast side, which 
 was partially removed at the time of the writer's visit, due probably 
 to the temporary rise in level a year previous. 
 
 Along the south shore of the West Arm the moraine lies nearer 
 the present shore and the terrace of the sixteen-foot level is rela- 
 tively narrow. The small terrace of the two-foot level is quite 
 generall}^ present but that of the eight-foot stage occurs infre- 
 quently. At the two-foot and present levels, the waves have re- 
 moved the gravel of the higher terraces, and many of the boulders 
 of the underljdng till are concentrated on the beach at the projec- 
 tions. Ice action is of little importance on this shore. At point D, 
 which is caused by an outlying morainic knob, the eight-foot ter- 
 race is well preserved but that of the two-foot level has been re- 
 moved. The submerged terrace is well developed along this shore. 
 The line of the "drop off" swings in a southward bending loop from 
 the Island to point D where it narrows to a width of one to two 
 hundred yards and continues thus around the west end of the lake 
 with minor variations only. 
 
 About one-half mile west of point D, an indentation not manifest 
 in the contour of the present beach is to be found. This was an em- 
 bayment of considerable extent during the eight-foot level but was 
 cut off by a low bar during the two-foot stage. The lagoon was a 
 shallow portion of the submerged terrace and was readily drained 
 by seepage through the sand of the bar when the lake level lowered. 
 At E the waves have demolished both the two and eight-foot levels 
 and are working on the sixteen-foot terrace, having formed a cliff 
 twelve to fifteen feet in height. 
 
 North of point E the moraine drops to the sand plains and there 
 is a possibility that the lake extended considerablj^ farther to 
 the west when it stood at the level sixteen feet above the present. 
 The two-foot terrace is well developd and in places reaches a width 
 of over two hundred feet. This shore was badly cut during the 
 temporary high water due to the recent dam at the outlet. Most of 
 the beach has been w^orn away to the tree roots and the effects of a 
 strong ice push are evident in places. The latter is said to have been 
 caused by ice jams in the spring of 1913. 
 
 At the State Forest the moraine which borders the northern end 
 of the lake approaches the shore and the levels are poorly defined 
 with the exception of that at the two-foot stage. However, the 
 sixteen-foot terrace is definable and has been accurately determined 
 at sixteen and one-half feet above the present level of the lake. 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA. 221 
 
 At the Saginaw Grounds the highest terrace is reached by a con- 
 tinuous clitl' of about fifteen feet in lieight. Fartlier to the east at 
 the Bay City Grounds niorainic liills are being carved by the waves, 
 forming cliffs up to heights of thirty-hve feet. These two locations 
 are well suited for resort purposes and afford an excellent view of 
 tlie entire length of the lake. Tlie only disadvantage is the exces- 
 sive width of tlie present submerged terrace, the greatest on the 
 lake. Inasmuch as it drops at six feet, the water is very shallow 
 for a long distance off shore. 
 
 Sand plains again border the lake in the vicinity of Sovereign 
 . Park and continue along the eastern side of the lake to the outlet. 
 The clump of pines at the Park gives some idea of the wonderful 
 forests that formerly covered this region and adds to 'the attractive- 
 ness of the location. At the two-foot level the shore stood con- 
 siderably farther back than at present. So much so, that the cur-. 
 rents left the shore and deposited a bar nearly a mile in length in 
 front of a very narrow lagoon. As is frequently the case on Higgins 
 Lake, the lagoons of this level stood on the sandy terrace and were 
 quickly drained when the water level lowered. Along the shore iu, 
 the vicinity of Cole low hills come to the shore and have been 
 carved into cliffs which step down to the two-foot terrace. This 
 higher ground leaves the shore about one-half mile north of Parker 
 Point, F on map, and swings back from the lake iu a broad curve, 
 reaching the lake again at Oak Grove. The lowland between this 
 and the lake was flooded at the higher levels and was thus a shallow 
 embayment whose bed was w^orked into a submerged terrace of 
 gentle slope. The change in direction of the shore line at the sides 
 of this embayment was abrupt and especially so at the northerii 
 end, F, where the angle was approximately ninety degrees. Cur- 
 rents must have left the shore at each side, but, strange as it may 
 seem, accomplished little at the north side. However, on the south 
 side a series of bars converging on the low cliffs above Oak Grove, 
 stand at progressively higher elevations inland, or in other Avords, 
 are arranged in steps. They sweep around the bay in broad, swing- 
 ing curves and are truncated by the somewhat irregular shore north 
 of Parker Point. 
 
 Apparently the bars did not begin to develop until the eight- 
 foot stage, during which the outlet was being cut down rather 
 rapidly. As the water level lowered at least four bars were formed, 
 which may bereadily distinguished along the road running east 
 from the head of tlie bay. The individual bars were not traced 
 through the swamp back of Parkers Hotel and there is the possibil- 
 ity that the number is increased in this locality by splitting. As 
 
222 INLAND LAKES OP MICHIGAN 
 
 the water subsided to the two-foot level, the embayment was drained 
 with the exception of a narrow strip adjacent to the present shore, 
 which was isolated by a low bar during this stage. The submerged 
 terrace is very well defined along this shore but gradually decreases 
 to a depth of about four feet and a width of less than fifty on the 
 south side of Parker Point. Beyond the end of the point it makes 
 a broad loop to the southward and also widens in conformity with 
 .the terrace fronting the east shore above the point. 
 
 In the development of the shore features in this locality several 
 factors have been of importance. Obviously, winds with a westerly 
 component are the only ones effective on these shores and those 
 from the northwest are probably the more important both as to 
 strength and frequency. That the currents caused by these winds 
 have been powerful is shown by the development of the shore be- 
 tween Parker Point and the bluffs at F in accordance with that, to 
 the north, and by the southward extension of the submerged ter- 
 race off the east side of the point. Nevertheless, the deposits in 
 the embayment have been made almost exclusively by northward 
 moving currents in the South Arm and something other than the 
 force and prevalence of the wind has been the determining factor. 
 
 From either direction the material is derived from sources which 
 are practically equivalent but the waves have a slight advantage 
 at Oak G-rove on account of the narrower submerged terrace. This, 
 however, is not sufficient to account for the great preponderance of 
 the work of the currents in the South Arm and the explanation must 
 be found in the form of the lake basin. The lake lies for the most 
 part in a deep pit, but during the higher levels overflowed on the 
 adjacent slopes, forming shallow embayments. Such an embayment 
 is the one under discussion but off the point lies the deep water of 
 the main basin. The currents from the West Arm left the shore 
 at the abrupt turn, F, and ran directly into deep water, causing a 
 rapid reduction in velocity and therefore deposition of the sus- 
 pended load. This material was utilized in filling the deep basin and 
 consequently was not localized in characteristic shore forms. How- 
 ever, on the south side of the embayment the currents flowed over 
 shallow water after leaving the shore at Oak Grove and deposited 
 their load in bars which crossed the bay and terminated at the 
 shore north of Parker Point where the stronger cross currents of 
 the West Arm were encountered. 
 
 The protuberance of the shore in the vicinity of Oak Grove is 
 of higher ground and was the source of the materials of the bars in 
 the embayment just discussed. Here the sixteen-foot level is well 
 defined and the terrace is utilized as a building site for several 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA. 223 
 
 .suiuiiioi' cottages. The eight-foot level is not well deUued, much of 
 it liaving been removed at the lower levels. Kecent cutting due 
 to the temporary raising of the lake level is much in evidence along 
 the low cliffs which line the shore. At Highland Point the land 
 drops in elevation and the two-foot level is prominent as a wide 
 terrace. Ice action has built a low rampart at the present shore 
 but is of moderate intensity. Between Highland Point and the out- 
 let a low sand flat extends several miles to the eastward. This 
 flat was covered by all of the higher levels but the borders of the 
 sixteen and eight foot stages were not traced. During the two-foot 
 level a low bar, conforming to the present contour of the shore, 
 developed and definitely cut off this lowland from the lake basin. 
 
 In conclusion, it seems fair to state that Higgins Lake is one of 
 our most interesting and instructive lakes. The history of this 
 interior lake is exceptional and is merely sketched in this account. 
 This phase of the work is left with reluctance for it is felt that a 
 detailed study of the various levels would be productive of most 
 interesting results. We may be certain of a much larger lake 
 during the highest or sixteen-foot level. At this time, the ends were 
 extended in broad, shallow embayments and the southwestern 
 border reached to the morainic slopes. Much work was done by 
 the waves, and broad terraces were formed which now appear as a 
 veneer of sand on till adjacent to the shores. AVith the lowering 
 to the eight-foot level, accomplished probably by deepening of the 
 outlet, a new cycle was inaugurated. At this time the waves were 
 working on the unconsolidated sands of the sixteen-foot terrace 
 and quarried great quantities of material which was distributed by 
 currents on the submerged terrace and in the embayments. The 
 rapid deepening of the outlet through the sand caused a gradual 
 depression of the level to two feet above the present stage. The 
 adjustments at this level were mainly by currents which cut off 
 numerous minor embayments, notably at the south and west ends. 
 The slight drop to the present level served mainly to reduce the 
 activity of the shore agents, a condition emphasized by contrast 
 Avheu the lake was held temporarily at a high level. A re-adjust- 
 ment to present conditions is undoubtedly taking place but the 
 results are not striking. However, a continuation of the develop- 
 ment of the spit at Chicago Point will have far-reaching results, 
 since it is growing across the narrowest part of the lake. Eventu- 
 ally, the lake must be divided in this locality, but much filling is 
 necessary on account of the deep water and the progress will be 
 slow. 
 
 The process of extinction has hardly started. Filling by vegeta- 
 
224 IiNLAND LAKES OF MICHIGAN 
 
 tion and sediment is of slight importance at tlie present time on 
 account of the great depth of the lake and the absence of entering 
 streams. The deepening of the outlet will eventually drain the 
 lake, but this process is making slow progress now that the stream 
 is flowing over boulders. It seems reasonable, then, that the 
 shores may become adjusted to a late stage of development before 
 the cycle is interrupted. Ice push is not of great importance on 
 this lake as regards effects but observational work on this phase of 
 our study should produce interesting results, especially in the West 
 Arm which is near the maximum limit for expansion. 
 
 Before leaving these lakes the writer wishes to point out the 
 unexpected (to him at least) contrast between the two lakes in 
 this region. Higgins Lake is deep ; its limpid, blue water is de- 
 rived from underground sources ; the adjustments of its shores 
 are numerous and varied; and its history is punctuated with 
 interesting episodes. A few miles to the southwest Houghton 
 Lake occupies a similar position with reference to the glacial 
 formations, but is almost diametrically opposite in characteristics. 
 Thus, it is shallow, its turbid waters, derived from surface drainage, 
 ' are agitated throughout by every storm ; it is choked with vegeta- 
 tion; the adjustments of its shores are infrequent and of little 
 significance ; and conditions have varied little in the past from those 
 existent today. 
 
 PORTAGE LAKE^ CRAWFORD COUNTY 
 
 Portage is a rather popular name for lakes and it seems neces- 
 sary to state that the one under consideration is situated in the cen- 
 tral-western part of Crawford County. It lies about eight miles 
 almost due north of Higgins Lake, and, thus, came into existence 
 at a later time. It is reached from Grayling on the Michigan Cen- 
 tral Eailroad by a short drive. 
 
 The best idea of the outline of this lake may be gained from 
 the accompanying map. Fig. 71, drawn from the U. S. Land Survey 
 map of the region. It is approximately three miles long, less than 
 a mile in width except across the broad embayment at the north 
 end, and has an area of slightly less than three square miles, 2.96. 
 It lies in one of the deeper pits of an out^^ash plain but at the 
 border of a moraine which ^stands not far back from the south and 
 west shores. Although within three miles of the An Sable River it, 
 nevertheless, is the source of one of the tributaries of the Manistee 
 which leaves the lake at the northwestern end. The lake is best 
 known on account of the location of the state camp of the National 
 Guard on the west side. 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA. 
 
 225 
 
 Tlio entire east side of tlie lake presents little of pliysiograpliic 
 interest. Shallow depressions which sink below tlie level of the 
 oiitwash plain but stand above the lake level cause swampy shores 
 for the greater part of the distance. The swamp condition is fur- 
 ther accentuated by low ice ramparts of sand through which occa- 
 sional small streams break. Along the north shore the adjustments 
 are better defined and the broad salient, Eagle Point, is the index 
 of this work. A well-defined bar extends out into the lake from 
 
 Ta(B A/ -R 4- IV 
 
 .^/\A//s-rc£r 
 
 Fig. 71. Outliuo map of Portage Lake, Crawford County. 
 
 the end of" the point and connects with the beaches on either side 
 in unbroken curves. The low cliifs in the vicinity of Mclntyres' 
 west of tbe point have furnished a considerable j)ortion of the ma- 
 terial for this spit. On the opposite side some material drifts in 
 from the northeast, as shown by the turning of the stream courses 
 towards the point before they cross the beach. The west side of the 
 point is exposed to the strong westerly winds and the adjustments 
 of this shore should, therefore, be more pronounced. This, however, 
 is not striking and may be due to the formation of return currents 
 29 
 
226 INLAND LAKES ,0'F MiICHIGAX 
 
 OH the east side during ''blows" from the southwest in addition to 
 the direct effect of the easterly winds. 
 
 At the northAvestern end of the lake the land drops to a low 
 swamj) through which the outlet runs. A strong ice rampart follows 
 the shore, but fails at the outlet and near the southern edge of the 
 swamp. This rampart totally otoscures any possible current de- 
 posits but, nevertheless, the fact that it plays out to the south sug- 
 gests the possibility of its be'lng a remodeled bar which developed 
 from the north. 
 
 South of the swamp low, outlying knobs o'f the moraine relieve 
 the swamp conditions. These knobs determined the position of 
 currents along this shore and are connected in some cases by bars. 
 In one locality a definite bar with a well-developed rampart on its 
 front slope stands above the present water level. These forms^ 
 which show a higher level of the lake, are made more prominent 
 by the presence of a second rampart at the present water level. 
 Northwest of Bear Point the moraine comes to the lake shores and 
 the slopes are cut into steep cliffs which rise from the exposed ter- 
 race of the higher level. At Bear Point the currents left the hills 
 during the former level and built a strong bar in a southeasterly 
 direction which now ends abruptly at the present shore on the south 
 side of the point. The lagoon behind this bar is filled with vegeta- 
 tion. A rampart has been pushed up on the front slope of the bar 
 but is being removed by waves under the present conditions. This 
 material is being carried outward and deposited on a submerged 
 bar in line with the northeast shore of the point, one of the few 
 adjustments in process. 
 
 South of Bear Point the morainic topography has given rise to 
 an alternation of cliffs and small swamps. The activity of the waves 
 is unmistakably evident 'but the currents seem to have carried little 
 of the material along the • shore, for the swamps are not cut off 
 except at the small indentation near the rifle range and this by an 
 ice rampart. Near the south end the land drops to an extensive 
 swamp which is separated from the lake by ice ramparts rather 
 than by typical bars. Beginning at the western side two ramparts 
 fringe the shore, the inner rampart reaching a maximum height of 
 five feet. They decrease in strength towards the east and die out 
 before reaching the higher ground of the moraine at the south end 
 of the lake. In this locality the slopes are gentle and an exposed 
 terrace of the former level is present locally. Wave action seems 
 to have been very slight for the cliffs are very low or are entirely 
 lacking. 
 
 The shores of Fortage Lake show a relatively small amount of 
 adjustment and this was accomplished at a level somewhat higher 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA 227 
 
 than the present. Tlie lack of well-defined sliore features makes an 
 accurate determination of the elevation of this former level some- 
 what difficult but it probably stood between two and three feet 
 above the level of tlie lake at tlie time of the Avriter's study. The 
 original lake was much more irregular in outline than the present 
 and the opportunities for adjustments were numerous. Neverthe- 
 less, the greatest change was brouglit about by the draining of many 
 of the shallow indentations when the level lowered. At the present 
 level the adjustments in process are few in number and progressing 
 very slowly. 
 
 It is interesting to compare the development of the shores of 
 this lake with that of such lakes as Corey, Clear and Long in St. 
 Joseph County. These lakes are much smaller than Portage and 
 are surrounded by material which, although sandy, is probably 
 more consolidated than the outwash sands along the east and north 
 shores of Portage Lake, Nevertheless, the shore features are much 
 more decided than those of Portage Lake and have been carried to 
 a more advanced stage of development. In the absence of complete 
 data one is inclined to infer that the greater age of the more south- 
 erly located lakes is the determining factor. 
 
 OTSEGO LAKE 
 
 Further retreat of the ice in the western interlobate area uncov- 
 ered the region centering about Otsego County, in the southwestern 
 part of which Otsego Lake is situated. This lake is nearly five miles 
 in length but is scarcely a mile in greatest width. Its area is three 
 and one-quarter square miles, so that the average width is approxi- 
 mately three-fifths of a mile. This lake is hemmed in on both sides 
 by moraines for most of its extent, but the material at the shores is 
 outwash sands, so that the lake may be classed as a pit. It is, 
 however, somewhat exceptional in form and topographic location, 
 and, therefore, some discussion of its manner of formation may not 
 be out of place. 
 
 The glacial deposits of fhis region lie south of a well-defined 
 moraine which was formed by the ice of the Michigan lobe on the 
 northwest and by that of the Huron lobe on the northeast, thus 
 the right-angled bend. See Fig. 72. An extensive outwash plain 
 stands south of this moraine and surrounds a number of morainic 
 tracts which rise above the plain as island-like forms. The morainic 
 tracts have a general north-south trend and are separated by nar- 
 row, sand-filled troughs, three of which, designated I, II, and III 
 
228 
 
 INLAND LAKEiS OF MICHIGAN 
 
 on map, are well defined. The central trough, II, branches north- 
 ward and Otsego Lake lies at the north end of the eastern branch. 
 Attention is called in particular to the eastern troughs, I and the 
 Otsego Lake branch of II, which lie in the drainage system of the 
 Au Sable River. 
 
 The drainage of these troughs divides near the north end, the 
 major part flowing southward. The upper parts, however, flow 
 southeastward around the morainic tracts and contain lake basins 
 which are pits in the outwash. Thus, in accounting for these 
 basins, and in particular that of Otsego Lake, the elongated form 
 
 Fig. 72. Map of the glacial formations in the yicinitj' of Otsego Lake. Solid bJack 
 indicates moraine. Lined areas are outwash. (After Leverett.) 
 
 and the presence of divides in the sand-filled troughs must be con- 
 sidered. 
 
 The recession of the ice border in this locality was to the north- 
 west within the Michigan lobe and to the northeast within the 
 Huron. Thus, it is very probable that at one stage the ice front 
 stood near the northern border of the morainic tracts and extended 
 in narrow tongues into the inter-morainic valleys to approximately 
 the position of the present divides. During this time, the drainage 
 flowed southward and the valleys were partially filled with outwash 
 material. As the ice melted the higher elevations were completely 
 uncovered, but parts of the tongues of ice in the depressions became 
 detached and were buried by outwash from the ice front which was 
 receding northward. At the same time a lower channel, parallel to 
 the ice border, was opened to the southeastward which in the case 
 of Otsego Lake crossed the northern tip of the moraine on the east 
 side. Finally the melting of the buried ice block formed the depres- 
 sion in which Otsego Lake stands. 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA 
 
 229 
 
 The loose surface material, the irregularity of tlie shoreline, and 
 the elongated form of the lake, presenting a broad expanse of water 
 
 "aZ/in^i 
 
 
 1' ^ ^\ . 
 
 yMrhc/^L/s deefch 
 
 ^ M/'/et 
 
 Fig. 73. Outliue map of Otsego Lake, Otsego County. 
 
 to winds from certain directions, have caused numerous adjust- 
 ments of the shores which are exceptional for a lake of this size. 
 Many excellent beaches and cottage sites are, therefore, available 
 
230 INLAND LAKES 01^ MICHIGAN 
 
 Tint are most appreciated on the east side in proximity to the 
 railroad. 
 
 On a map one looks in vain for an outlet of Otsego Lake (Fig. 73) 
 Tjut may notice that the North Branch of the An Sable River heads 
 straight for the east side of the lake somewhat north of the center. 
 It does not extend to the lake, however, and one interested in drain- 
 age problems naturally wonders what the conditions at the head 
 of this stream may be. In this case a trip is worth while, so we may 
 advantageously begin our study at this place. 
 
 Back from the lake a dry channel, with even floor and gentle 
 slope away from the lake, runs through a break in the hills and 
 connects directly with the North Branch of the An Sable. Nearer 
 the lake the channel broadens to a low marsh in the sands of the 
 trough and splits before reaching the lake shore, having formed 
 an island in past times. At present the water stands four to five 
 feet below the level of the outlet, long since abandoned. 
 
 The shores along the head of the broad embayment from which 
 the outlet leads are bounded by cliffs of stsatified sand which reach 
 to the surface of the outwash about ten feet above the lake. This 
 unconsolidated material was easily quarried and was distributed 
 in both directions by shore currents and offshore by the undertow. 
 The relative importance of these agents is dependent very largely 
 on the reach of the winds which may be rated in the following order: 
 Southwest, northwest and west. Thus, shore currents developed at 
 the expense of the undertow and the submerged terrace is poorly 
 defined. In most places it is merely a zone of sand which shows a 
 sharp line of demarcation from the muddy bottom of the deeper 
 part of the lake. Of the shore currents those which develop under 
 southwest winds have the advantage and the northward drift along 
 this shore is the more powerful. Therefore, much of the material 
 has been deposited along the south side of the broad point north of 
 the outlet. The current action is first apparent at the outlet where 
 a bar started to develop at a level four feet above the present and 
 partially enclosed the small marsh bordering'the outlet, which must 
 have been a shallow arm of the lake at that time. At the present 
 level ice-shove, probably of the expansion type, has formed a poorly 
 defined rampart, although shore conditions are unfavorable. North 
 •of the outlet this upper beach extends to the broad point and is 
 known as Wah-Wah-Soo Beach. The point is caused by an original 
 irregularity of the basin and was almost obliterated by wave action 
 during the early stages of the lake. Thus, in the background stands 
 a cliff' in the outwash and from the foot of this stretches the now 
 exposed submerged terrace of that time. But with the subsidence 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA 23l 
 
 of the water, caused by the (lownciittino: of the outlet, the activity 
 of the shore agencies was reversed and currents from the north <Ie- 
 posited a bar almost in line with that shore, enclosing a very narrow 
 lagoon just in front of the cliffs. Either during or following this 
 period the currents from the south became effective and carried the 
 Wali-Wah^Soo Beach beyond the bar from the north and are at 
 the present time flowing to the end of the point. This condition is 
 expressed by a well-defined submerged terrace which drops into 
 deeper w'ater from a depth of two feet. This very shallow terrace 
 in a locality favorable for its development indicates a poorly de- 
 veloped undertow. 
 
 From this point to the north end of the lake the land slopes 
 gently to the shore and the only discernible adjustment of the shore 
 has been the carving of low cliffs bj^ waves of very moderate force. 
 Some of the material which was removed from the cliffs has been 
 transported to the north and deposited in a bar which stands slightly 
 above the swampy lowland bounding the north end. This bar leaves 
 the higher ground some distance inland but borders the lake at the 
 north end. On the west side similar currents have built a strong- 
 bar at right angles to the one from the opposite shore which has 
 partially cut off an extensive swamp and caused a peculiar gourd- 
 like outline of the shore. These deposits were both formed at the 
 higher level but probably were not above water for much of their 
 extent. 
 
 The northward drift of the currents along the west shore is again 
 evident at the narrow hook which has developed in front of a low 
 exposed terrace a short distance south. In the early stages of the 
 lake waves have been energetic along this shore, and the terrace of 
 the higher level is well exi)osed, in places reaching a width of nearly 
 one-fourth mile. At the higher stage this smooth shore was broken 
 by bays north of Kokosen and in the vicinity of Three Pines. In 
 both cases complete bars were formed across the mouths of these 
 embayments, forming lagoons which are still swampy. Between 
 Three Pines and Idylwild the same process is being repeated at 
 the present level by the formation of a spit across the neck of the 
 small bay. Below this bay the moraine recedes from the lake and 
 the outwash is considerably wider. In places the waves have re- 
 moved the higher terrace and are sapping the sands of the outwash. 
 Ice-shove is effective in this locality at the present level and has 
 formed ramparts at the edge of the exposed terrace which is, there- 
 fore, poorly drained. The greater part of the eroded material is 
 carried by shore currents rather than by undertow, and the sub- 
 merged terrace is poorly developed. Most of it drifts to the south 
 
232 INLAND LAKES OF MiICHIGAN 
 
 and is dropped at Long Point wkere the currents are forced from 
 the shore. The activity at this point began during the higher level 
 and shows features similar to those found on the point north of 
 the outlet. The main part of the point is an exposed terrace of 
 this level upon which was built a strong bar from tne north, indi- 
 cating a reversal from wave to current action. The bar stands 
 about twenty feet back from the present beach and is sharply 
 truncated by the south shore. The revival of wave action, as shown 
 by the removal of the tip of the bar, was due to the inauguration of 
 a new cycle by the sinking of the water level and is still in progress 
 at periods of excessively high water such as occurred in 1913. On 
 the north -side of the point, however, the currents are still active, 
 the point having been extended approximately two hundred feet 
 offshore below the water level. As is the case at many of the points, 
 the submerged terrace is well developed on the side affected by 
 currents, here the north, and drops into deep water at four feet at 
 distances offshore which increase towards the ends of the points. 
 
 The shore from Long Point to the south end of the lake shows 
 features very similar to those just described. The long stretches 
 are generally regions of wave action which in many places has 
 removed the terrace of the former level and is working in the sands 
 of the outwash. Two strikingly similar points, B on map and 
 point to the south, break the shore line. Both are V-bars which 
 began their development at the higher stage and were interrupted 
 by the drop in level. In each case the currents from the north were 
 the more active and built strong bars which probably stood above 
 water level. Likewise, on the south side bars developed which met 
 those from the north, forming a more or less symmetrical triangular 
 embankment with enclosed lagoon, but were not completed at the 
 time of the sinking of the water level. However, at the present 
 stage the triangular form is well shown and the lagoons are filled 
 with water or are swampy. 
 
 The south shore of the lake is bounded by the so-called "beaver 
 dam" which, local tradition to the contrary, is one of the best ex- 
 amples of a bar seen by the writer. This form, shown in Plate XIV, 
 is composed entirely of sand and joins the east shore in an unbroken 
 curve. Its crest stands slightly above the level of the higher stage, 
 except for the last hundred feet at the west end, and supports a 
 fringe of trees which accentuates the linear character of this bar. 
 In the photograph, Plate XIV, a similar bar may be seen stretching 
 across the triangular lagoon isolated by the "beaver dam." This 
 bar, however, stands at a lower elevation and is poorly defined, 
 Both of these bars were formed during the higher stage and obviously 
 
Alichipiii OcnlOKical .uui 
 
 I'ulilication .".0. (icolo^ical Series 25, 
 riatc XIV. 
 
 A. "BEAVER DAM," OTSEGO LAKE. 
 
 iB. ICE-FORCMED V-BAR, LOXG LAKE. ALPEXA COUNTY. 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA 233 
 
 the outer bar developed first. For some reason the point of de- 
 parture of the currents from the shore was shifted to tlie north 
 before the outer bar reached the water level and remained in the 
 position now occupied by tlie inner bar, the "beaver dam." The 
 currents instrumental in their formation were those driven by 
 northwest winds along- the shore south of First Point. At flood 
 stages of the present level waves'have been active and have steepened 
 considerably the front slope of the bar, which normally should be 
 gentle. 
 
 Along the east shore cliffs in the outwash reveal the source of 
 the material of these bars. In places the cliffs show fresh cutting 
 which may be laid directly to the excessively high water in 1913. 
 Near First Point the cutting is less active and the terrace of the 
 higher stage appears and widens towards the point. In the early 
 stage of the lake this point was merely a blunt projection of the 
 outwash which was carved into a cliff and terrace. Then followed 
 a period of reversed activity, ascribed to the downcutting of the 
 outlet, during which currents predominated. The drift from both 
 directions left the shore at this point but that from the south was 
 the stronger. The material from this side was deposited in a strong 
 bar which splits twice before meeting the single bar on the north 
 side. Thus, on the south side of the apex there are three distinct 
 bars separated by marshy lagoons. In front of the bars a triangular 
 terrace of considerable extent developed, which was exposed by the 
 sinking of the water to the present level. Since that time, the point 
 has been extended under water for nearly three hundred feet and 
 drops at a depth of four feet. The stretch of shore on either side 
 of the point along which currents may develop is slightly longer on 
 the north side while the northwesterly winds have quite an advant- 
 age in reach over those from the southwest. The greater effective- 
 ness of the currents from the south is, therefore, somewhat unex- 
 pected. The determining factor seems to be a blunt projection Just 
 north of the point, which causes considerable deflection of the cur- 
 rents from this direction, as is shown by the distinct submerged 
 terrace. 
 
 In the broad embayment north of First Point waves are again 
 the active agent and sand cliff's are the consi^icuous shore features. 
 The terrace of the higher level has been removed except in proximity 
 to. the points. Arbutus Point is very similar to First Point aud re- 
 quires no special discussion except for slight modifications. The 
 main episodes in its development are identical but the predominat- 
 ing currents came from the north. Instead of the split bars found 
 on the south side of First Point there developed here a compound 
 
234 INLtAND LAKEiS OF MICHIGAN 
 
 hook with several spurs recurving to the shore. In front of the 
 hook on the north side is a series of parallel beaches which may 
 mean a gradual lowering of the water level. This point is not grow- 
 ing at present and some cutting may take place at flood stages. 
 However, a yerj well-defined submerged terrace has developed in 
 conformitj^ with the north side of the point. 
 
 As regards origin and general features, Otsego Lake presents 
 many interesting features. The major shore adjustments have taken 
 place when the water stood higher than at present and show many 
 variations. The sandy material of the surrounding outwash has 
 been easily eroded and each change of conditions is clearly regis- 
 tered. Although the full extent of the lake in the early stages was 
 not determined, cutting by waves prevailed generally along the 
 shores. The reversal of activity, making currents effective at the 
 numerous points, is remarkably consistent and is ascribed to the 
 downcutting of the outlet. Then followed a period of stability dur- 
 ing which the points were increased by the development of bars. 
 The drop in level to the present stage was probably accomplished 
 slowly and allowed the growth of a series of minor bars on some 
 of the points. Since the drop has caused the abandonment of the 
 outlet, it must be ascribed to a decrease in the amount of ground 
 water which is the main source of supply. The depressed level of 
 the surface of the ground water is a subject full of complications, 
 but one important factor may have been a general drying-out fol- 
 lowing the disappearance of the glacier which supplied immense 
 quantities of surface water in melting. 
 
 At the present level the adjustments are not striking and occur 
 for the most part during stages of high water. No important 
 changes are to be expected but minor adjustments may be accom- 
 plished by currents in a few localities. With wave action limited 
 to exceptional conditions of infrequent occurrence, there is little 
 probability^ of any extended growth of the points. Considering the 
 amount of current action, the submerged terrace is of very inferior 
 development. A more distinct terrace may have been formed at 
 the higher level and is now being slowly adjusted to the present con- 
 ditions. 
 
 Little sediment is being brought into the lake and filling in this 
 way is a negligible factor. A'egetation is taking hold on the muddj^^ 
 bottom but has as yet made little growth on the off-shore sands. 
 Fishing is the favorite recreation on the lake and this, with many 
 sand beaches and excellent locations for summer colonies, has made 
 Otsego one of the more popular of the interior lakes of the State. 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA 
 
 2:35 
 
 HUBBARD LAKK 
 
 Hubbard lake lies in the drainage of the Neuelon Kiver which 
 flows almost directly north from the lake to the Thunder Bay Kiver. 
 It lies just south of the northern boundary of Alcona County in 
 the central part and, thus, is not readily accessible from the rail- 
 road. Yet a lake of nearly thirteen and a half square miles in area 
 ranks as one of our larger inland lakes and is well worth a visit, 
 even though a ride of sixteen miles from Alpena is necessary. The 
 ride, however, need not be monotonous for a variety of glacial 
 formations are crossed and serve to hold the attention. 
 
 tlOKAIS/C T7LL fLA/N SAUD OtPOStT lAHC PLAiN 
 
 Fig. 1 4. Map of the glacial formations in the vicinity of Hubbard Lake, 
 bard Lalce is the white area in center of map. 
 
 Note : Hub- 
 
 The first part of the trip is across the sandj^ lake bottoms of 
 the forerunners of the Great Lakes but in the distance may be seen 
 the strong relief of the moraines to the west and south. The mo- 
 raines stand well above the lake beds and afford many opportuni- 
 ties for distance views of the neighboring region. It is impossible, 
 of course, to get a comprehensive idea of all the glacial formations 
 in a single trip across the region, but careful mapping of this lo- 
 
236 INLAND LAKES OF MTOHIGAN 
 
 cality has shown that the moraines are fragmentary rather than 
 arranged in continuons belts. The fragments often cover large 
 areas and are in a very general way aligned in a nearly north- 
 south direction. Between the moraines are till plains or sandy 
 deposits similar in topography but standing at a considerably lower 
 elevation. The distribution of the various deposits is shown in 
 Fig. 74 and from this it will be seen that Hubbard Lake is hemmed 
 in by four patches of moraine. These come to the shores in two 
 places on the east side but on the south and west stand back a 
 short distance. The intervening lowlands determine the position of 
 the larger drainage channels, including most of the inlets and the 
 outlet, at the extreme north end. '. 
 
 In a region of such complicated glacial deposits it is difficult to 
 imagine the behavior of the ice. The irregular morainic tracts 
 in themselves nearly enclose basins and show that during the melt- 
 ing the distribution of the ice was very complex and even frag- 
 mentary. Thus, it is probable that large masses of ice stood in the 
 depressions after the moraines were free from ice. In the case of 
 Hubbard Lake, it may be supposed that an exceptionally thick mass 
 of ice covered a distinct basin which was uncovered by subsequent 
 melting. The origin of the basin is, however, still a question, due 
 to the lack of knowledge as to the form of the submerged portion 
 of the basin and to the uncertainty as to the interpretation of con- 
 ditions during the waning of the glacier. 
 
 The outlet, see Fig. 75, furnishes a good starting place for a 
 physiographic study of this lake. At the time of the writer's visit 
 a dam of nearly five feet head was holding the lake above its natural 
 level. The present dam has been in for a few years only but the 
 waters for some time previously were periodically held up by log- 
 ging operations. The submergence and subsequent obliteration of 
 the former shore makes the determination of the amount of flood- 
 ing difficult, but it is quite evident that this has been sufficient 
 to cause a decided renewal of the activity of the waves and cur- 
 rents, as will be seen from the description that follows. 
 
 The outlet flows northward through a narrow channel which has 
 been cut from three to four feet below the level of the swampy 
 lowland at this end of the lake. The outstanding feature in this 
 locality is a strong bar composed of sand and gravel which separates 
 the swamp from the lake. It stands well above the present level 
 even under the flooded condition and indicates a higher level of 
 the lake, which is amply confirmed at other points on the shores. 
 This former stage must have been the original level of the lake, 
 
INTERIOR LAKES OP" THE SOUTHERN PENINSULA 
 
 2;i7 
 
 (lui'iiig Avliicli coiisiderabU' adjustment of the shore;-; took place, 
 and was interrupted by the deepening of the outlet. The bar at the 
 outlet developed from the east side and partially cut off tlie swampy 
 lowland which now lies adjacent to the stream. The material for 
 this bar was derived fi-oni the almost continuous clifl's which rise 
 from five to ten feet above the straight sliore to point E, see map, 
 
 TT Ee/V. B 7C. 
 
 r 27/V /5T£:. 
 
 Af,/e5 
 
 T-seA/. .e.se. 
 
 Fig. 75. Outline map of Hubbard Lake, Alcoua County. 
 
 Fig. 75. In one locality only, A, have the shore currents persist- 
 ently dej)osited material. The deposit is in the form of a spit which 
 swings outward with very gentle curvature for a distance and then 
 continues parallel to the shore as a sharp spit. The partially en- 
 closed lagoon is open at tlie north end showing unmistakably that 
 the efifective currents flow northward along this shore. The cause 
 
2:8 INLAND LAKES OF MIOHIGAN 
 
 is the greater reach of the waves from the south. The presence of 
 good-sized tree trunks standing in water along the bar may be 
 taken as an indication that this bar developed during the former 
 level and also that the lake has been flooded. The reason for the 
 currents leaving the shore in this locality is obscure and must 
 remain so until the configuration of the bottom of the lake is known. 
 
 Again at point B the currents have been active and the cause of 
 their departure from the shore is obviously the abrupt bend in the 
 shore line. In this case currents from both directions were involved 
 and a V-bar which enclosed a triangular swamp was built at the 
 former level. The currents from the southeast were the more ef- 
 fective and the bar is unsymmetrical, the sharper curvature occur- 
 ring on the north side. From the map it v»'ill l)e seen that the shores 
 from wdiich the material of this bar may have been derived are 
 about equal in length on each side. This balance is an apparent 
 one, for the moraine which causes Churchill Point comes to the 
 lake shores just east of the point and has furnished the greater 
 amount of material, as shown by the height of the cliffs which ex- 
 ceed twenty feet in places. In addition, the reach of the southerly 
 winds is much greater than that of those from the northwest, tiie 
 only ones which could be effective in the formation of the north 
 side of Lhis bar. 
 
 Along the shore from B to Churchill Point considerable cutting 
 has been accomplished by the waves and a distinct submerged ter- 
 race is present. The outside edge is found usually at about two 
 hundred yards out and drops to deep water from a depth of from 
 fifteen to sixteen feet. This seems rather excessive as compared 
 with the depths found on other lakes of like size, but it must be 
 remembered that this lake stands considerably above its normal 
 level. The recent renewal of wave action is here attested by the 
 fresh cliffs which frequently slump, carrying trees and sod to the 
 beach. 
 
 Around Churchill Point the moraine drops to a till plain into 
 which East Bay heads. This bay is exposed to waves of long reach 
 from the southwest which are intensified by the convergence of the 
 shores, and the shore adjustments were, consequently, exceptionally 
 well defined. The submerged terrace is wide and drops at a depth 
 of twelve feet to the channel of deeper water. Also bars were formed 
 at the higher level on each shore by currents running towards the 
 head of the bay. As in the case of the outlet, these bars cut off 
 the swampy lowland adjacent to Misery Creek, and the lagoon, 
 as well as the exposed portion of the terrace, grew up to forest 
 after the sinking of the level. Under the present flooded condition. 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA 239 
 
 llie trees si and in water most of the time aud are rapidly dying. 
 The bar on tlie west side of the bay extends to the mouth of Misery 
 Creek, which has some difficulty in keeping an open channel. But 
 wave action is the predominant agent on this side and the bar is 
 retreating landward in a veiy irregular fashion, due to the presence 
 of the drowned trees. On the east side of the bay the bar is better 
 preserved aud terminates in a graceful hook at the entrance of the 
 small emba3'ment into which Misery Creek flows. 
 
 To the south cliffs prevail as far as Sucker Creek and reach a 
 maximum height at the projection opposite Churchill Point, which 
 is caused by a morainic fragment. The broad swells and sags 
 cause an alternation of freshly carved cliffs and sandy beaches, 
 which may be definitely described as bars only in the vicinity of 
 the mouth of Sucker Creek. The shore-drift is here to the south as 
 shown by the turning of the stream in this direction. This small 
 stream has brought much material into the lake aud has deposited 
 it in a broad shoal, best described as a flooded delta. Ice-action 
 has been powerful along this shore but the only locality where 
 its eft'ects may be observed is at the attachment of the bar to the 
 cliffs north of Sucker Creek. It is likely that much more extensive 
 ramparts or boulder-lined strands were formed but have been sub- 
 merged or destroyed since the flooding. The formation of strong 
 pressure ridges on this lake in winter, one of which occurs regu- 
 larly in this vicinity, makes it almost certain that the ice-push is 
 exerted by jams rather than by expansion. Where the pressure 
 ridge occurs, the lake is two miles in width as nearly as can be 
 measured on the map, an observation of value to the physiographer 
 interested in differentiating between the two types of ice-shove. 
 
 Below Sucker Creek cliffs again predominate to the south end 
 of the lake, which is bordered by the undulating lowland through 
 which the West Branch flows. The raising of the water level has 
 flooded this lowland and caused wholesale destruction of the trees, 
 except on low knolls which stand as islands above the >swamp and 
 are sharply set off by the green foliage. During the higher level, 
 the flooding was even more pronounced than at the present time and 
 a shallow embaymeut extended inland along the course of the West 
 Branch. The development of a' bar from the east forced the West 
 Branch to the west side of the valley and cut off the embaymeut, 
 except for the channel maintained by the stream. The bar is in 
 process of destruction at present, due both to the wash of the waves 
 and the spring freshets during which the West Branch frequently 
 inundates the lowland, including the bar. The current of the stream 
 is especially strong near its channel and consequently the destruc- 
 
240 INLAND LAKES OF MICHiaAN 
 
 tion of the bar has beeu carried farthest at the east end. In fact, 
 this process in conjunction with the flooding of the lake has so far 
 obliterated the end of the bar that the only surface indication is a 
 row of dead trees standing iu water. 
 
 Along the west side of the lake the lowland runs northward in a 
 narrow strip as far as Mud Bay, where it is interrupted by a spur 
 of the moraine which usually stands from one-fourth to one-half 
 mile back from the lake shore. The lowland strip is somewhat 
 higher than at the south end and, although swampy, has no marked 
 shore features below Mud Bay. This broad bay is caused by a sag 
 at the northern extremity of the lowland strip and was formerly 
 more extensive than at present. The reduction in size was caused 
 by the development of a complete bar which now forms the shore at 
 the head of the bay. This bar is exceptionally well developed, due 
 largely to its enlargement by subsequent ice-action. 
 
 Hardwood Point is being attacked by the waves to some extent 
 and has furnished the material for the bar at Mud Bay. However, 
 to the north ground moraine again borders the shore. This stands 
 considerably higher than the lowland below Hardwood Point and 
 low cliifs are the predominating shore features. In one locality 
 only have the sags extended below the lake level and this was com- 
 pletely isolated from the main lake by a bar. As at Mud Bay, ice 
 has succeeded in intensifying the bar and has remodelled it into 
 an ice rampart which rises nearly five feet above the lake. Crow 
 Point is another interesting locality because a considerable portion 
 of the terrace of the former level is exposed here. The sandy ma- 
 terial of the ground moraine in this locality was very easily eroded, 
 and the low knob whicii extended well out into the lake at the 
 former level was planed into a terrace of greater width than usual, 
 so v^^ide that it is not entirely flooded under present conditions. 
 Ice-action is again evident in the series of parallel ridges which 
 rise above this terrace. 
 
 From the above description it should be clear that but two stages 
 stand out in the history of this lake basin. Of these, the latter is 
 now submerged by an artificial flooding of the lake and we know 
 practically nothing concerning the adjustments which took place at 
 that time. Of those described all, except possibly some of the 
 effects of ice-shove, have occurred at the higher level. However, 
 the adjustments were carried to an advanced stage of devlop- 
 ment which determined to a large extent the outline and the gen- 
 eral character of the adjustments during the later stage. It is 
 true that the shores were not completely adjusted to wave action 
 and much material was therefore being distributed by currents. 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA 241 
 
 Nevertheless, much Avas accomplished aud, in addition to the sub- 
 merged terrace, the embaymeuts were reduced by bars. In most 
 cases the bars were completely developed and in numerous instances 
 strengthened by ice-action. Among the notable changes accom- 
 plished in this way may be mentioned the embaymeuts at the outlet, 
 West Branch, Sucker Creek, and Misery Creek. 
 
 The lowering of the level was due to the downcutting of the out- 
 let which was sufficient to account for a drop of about five feet. 
 Judging from lakes whose main adjustments have occurred at 
 higher levels, we may be reasonably certain that changes of a minor 
 character only were in process previous to the flooding of the lake. 
 When a considerable depression of the level of a lake has taken 
 place, it frequently happens that the activity along a, given shore 
 may be reversed from cutting to deposition (see Otsego Lake). 
 Such reversals are likely to take place where the terrace is locally 
 wide and probabh^ occurred on this lake at Crow Point and the 
 delta of Sucker Creek. This, however, is conjectural and cannot 
 well be substantiated under the present conditions. 
 
 The most striking thing of physiographic import is the renewed 
 actirity on the shores of Hubbard Lake, due to the artificial raising 
 of the water level to within about one foot of its former position. 
 The flooding of the adjacent lagoons, previously dry as shown by the 
 forest growth, and the freshly cut cliffs are the more evident effects 
 of this rejuvenation. In places, however, unexpected eft'ects such 
 as the degradation of the bars of the higher level are found, even 
 though the water level stands lower than formerly. The sap-ping 
 of cliffs and destruction of dej)Ositional forms of a higher level is 
 due to the fact that the ponding of the lake by a dam holds the 
 water at the flood stage for a longer period than normal each year 
 and thus causes eft'ects that are greater than would be the case if 
 the lake were allowed its normal seasonal fluctuation. 
 
 Ice action exerts a powerful shove on the shores of the lake and 
 mainly by jams, although some expansion may take place in the 
 narrower parts. In general, the effects are more frequently en- 
 countered and are better developed on the west side and consist 
 almost exclusively of ramparts formed on the bars which cross 
 the mouths of identations. Boulder-lined strands are not present, 
 although conditions both as to material and topography are fav- 
 orable for their formation. Ice jams are active on all shores of a 
 lake but have a tendency to exert a greater push on the east side. 
 Likewise wave action is more intense on the east side due to the 
 prevalence of storm winds from the west, and it is possible that 
 31 
 
242 INLAND LAKES OF MilCH'IGAN 
 
 many of tlie features due to ice shove have been destroyed on this 
 side while on the opposite side they are still preserved. 
 
 During the higher stage of the lake, its area was considerably 
 reduced by the cutting off of indentations but the lagoons had not 
 been filled when the level lowered. At present they are again 
 flooded and are rapidly being filled with vegetation. As regards 
 the main lake, little filling of any kind is taking place and, since the 
 outlet is dammed, we may have little concern about the further 
 extinction of the lake. Shore adjustments are far more important 
 at the present and will continue to be so as long as the high level is 
 maintained. 
 
 LONG LAKE, ALPENA COUNTY 
 
 This lake should not be grouped with the interlobate lakes nor 
 was it ever isolated by current action from one of the former stages 
 of the Great Lakes but is appended here for want of a better place. 
 In the naming of lakes descriptive terms are very often employed 
 and those most frequently used are "round" and "long". This is 
 especially true of the inland lakes of Michigan and, since these 
 names are duplicated many times, it is necessary fo add a state- 
 ment as to their location. The lake under discussion crosses the 
 boundary between Presque Isle and Alpena counties near its east- 
 ern extremity and is divided into two almost equal parts thereby. 
 It extends seven and a half miles in a northwest-southeasterly di- 
 rection and nowhere exceeds one and a half miles in width. The 
 area is slightly greater than eight square miles, making the average 
 width in excess of a mile. The basin lies on the exposed terrace 
 of Lake Algonquin whose shore stood just west of the lake. The 
 terrace in this locality consists of a series of troughs and ridges 
 which run parallel to the glacial striations found on the bed Tock. 
 It is probable that the ridge-and-trough topography existed pre- 
 vious to glacial times and was modified, but not obliterated, by the 
 passage of the ice over it. As regards the troughs the modifica- 
 tion where the ice moved parallel to the trend of the topography 
 consisted mainly in a widening and deepening jDrocess, and the 
 effects produced" varied locally due to variations in hardness of the 
 rock, fracture systems, and other factors. Where the erosion was 
 greater the trough was deeper and this, in conjunction with the 
 damming of the trough by glacial deposits may account for the 
 Long Lake basin. Yet the irregularities of the .shoreline and in 
 the depth of the lake and the occurrence of numerous limestone 
 blocks, both on the shores and the bed of the lake as well, lead one 
 to believe that some other factors have been instrumental in the 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA 
 
 243 
 
 formation of this basin. Altliougli positive evidence is not at hand, 
 one cannot overlook the possibility of the effect of uhdergvound 
 drainage and the resulting sink holes. It is certain that a consid- 
 erable amount of the rainfall sinks below the surface in this re- 
 gion and flows to Lake Huron in subterranean channels. Sink holes 
 and disappearing streams are numerous in the area from Thunder 
 Bay to Black Lake in Cheboygan County, and the former are found 
 
 r::53N.-R.7f:. 
 
 T3aM- /Z7E 
 
 Ty2.M 
 
 Miks 
 
 Fig. 76. Outline map of Long Lake, Alpena County. 
 
 within a few rods of Long Lake. Man}' of the smaller irregularities 
 of the shore are due to the removal of blocks along the frequent 
 joint planes but it appears difficult to account for the larger em- 
 bayments unless the possibility of sinks is admitted. 
 
 The surrounding region is very thinly drift covered and in sev- 
 eral places the bed rock outcrops on or near the lake shore. These 
 outcrops, however, are not numerous and have little effect on the 
 
244 INLAND LAKES OP MICHIGAN 
 
 general outline of the lake. The numerous large boulders are large- 
 ly of local derivation, that is, are of limestone, and many have been 
 pushed on the shores from the lake bottom by ice. 
 
 The lake is reached hj a six mile drive from Alpena which brings 
 one to the southwestern shore. The south end of the lake is held 
 by a low sand plain which stands but slightly higher than the lake 
 and is swampy during high water. The sand beach which forms 
 the lower end of the lake is the best example of shore adjustment 
 to be found on this body of water. Not only is the material well 
 assorted and the curvature even but a well-defined submerged 
 terrace is present which has a maximum width of one-fourth mile 
 and drops at Ave to six feet. A low ice rampart skirts the present 
 shore but evidence of a definite current deposit was not found. The 
 rampart may possibly be a' remodelled bar but this is not certain. 
 Yet the position of the outlet at the southwestern side and its turn 
 to the east after leaving the lake indicate a drift from the east side 
 of the lake. The adjustment of this shore is due to some extent 
 to the topography and material but largely to the great reach of the 
 prevailing storm winds from the west and northwest. The outlet 
 has cut a shallow trench through the sand fiat; a matter of im- 
 portance in explaining the presence of shore features above the 
 present level found elsewhere on the lake. 
 
 About a mile inland from the southwestern side of the lake the 
 bluff of Lake Algonquin runs approximately parallel to the lake. 
 The exposed terrace of this lake has a very thin surface cover near 
 the southwestern end of Long Lake and in places the rock beds are 
 exposed. At all outcrops the rock has been planed off and is 
 marked with striations which run parallel to the lake basin. Such 
 evidence shows the part played by the glacier in the formation of 
 this basin. ■ 
 
 The shoreline north of the outlet on the west side has few irreg- 
 ularities for about one mile and shows little adjustment, A broad 
 bay at locality A, see map. Fig. 76, marks the beginning of a very 
 irregular shore .which persists nearly to the north end of the lake. 
 At A the waves and currents have accomplished little, but a small 
 indentation on the north side of the bay has been isolated by an ice 
 rampart and is now well on the way to extinction by vegetation. 
 In fact, ice has been the most ]30werful agent on this shore and 
 waves and currents have been effective only in the most favorable 
 locations. Thus, ramparts are the common shore feature and in 
 manj^ places are strong ridges of coarse material. At B currents 
 have turned from the shore and a well-defined, but small, hook 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA 245 
 
 point iiiii' u|» ilio lake lias bceu runiinl. Again at C the currents have 
 deposited detritus but in this case the spit is turned to the south- 
 east, showing; a reversal in the direction of the effective currents. 
 
 Along the south shore of the bay below point D an excellent ice 
 rampart of at least five feet in height and containing boulders of 
 several tons in Aveight was found. The rampart diminishes to- 
 wards the head of the bay and disappears at tlie low flat which con- 
 tinues to tlie northwestward more than one hundred yards. This 
 lowland was formerly a part of the lake but has been partially 
 drained and filled with vegetation. The bay is shallow and the 
 muddy bottom contains a heavy growth of reeds. In places the 
 shore and bottom are lined with a white sand which is found to be 
 composed almost entirely of shells, indicating the active formation 
 of shell marl. On the north side of the bay current action is well 
 shoAvn by the spit which has grown an appreciable distance across 
 the bay. The material averages nearly three inches in size and 
 indicates the presence of a powerful drift into the bay from this 
 side. The presence of a rampart on the spit and the coarseness of 
 the material leads one to conclude that ice has been of considerable 
 importance in the building of this form. The spit is not increasing 
 under the present conditions and its period of most rapid develop- 
 ment must antedate the cedars of more than two feet in diameter 
 growing on its crest. 
 
 On the end of point D, where one might expect current deposits, a 
 submerged terrace, composed of rocks except for a sandy S;one near 
 its outer edge, is present. This terrace, therefore, has been swept 
 free of finer material by the waves and undertow and has a 'com- 
 paratively narrow built portion. On the side of the point facing 
 the lake a narrow, but distinct, exposed terrace upon w^hich bould- 
 ers are scattered is evidence of a former level of the lake betw^een 
 two and three feet higher than that at the present time and show^s 
 that the terrace was cut, wave activity having been the predomi- 
 nant factor in the adjustment of this point in the past. The tend- 
 ency for the boulders to stand near the former shore shows, further- 
 more, that ice has been active during the higher level. 
 
 The broad embayment north of D is caused by an extensive sag 
 in the Algonquin terrace. The present shores are composed of 
 rubble and show little adjustment. For the most part the irregu- 
 larities of this shore are due directly to variations in the Algonquin 
 terrace and adjustments, if present, are found at the projections. 
 One of the most interesting was found at F. The projection itself 
 does not differ materially from those occurring so frequently along 
 
246 INLAND LAKES OF MICHIGAN 
 
 this shore, but the small point which runs to the northwestward 
 from its extremity is of considerable interest. In all but one respect 
 this minor feature resembles an unsymmetrical V-bar, see Plate 
 XIA", B. Two bars meet in a blunt point and enclose a shallow 
 depression. That on the east, or lake, side is the better developed 
 and is fronted by a submerged terrace while that on the opposite 
 side drops steeply below the water level. On the east side the 
 bar has been remodelled in part into an ice rampart. The most 
 interesting feature in connection with this point is the coarseness 
 of the material of which it is composed. The smallest size of 
 particles was not less than one inch but the maximum was as great 
 as two feet in largest diameter. Also the rubble is for the most 
 part angular in form rather than rounded, the form characteristic 
 of water-worn material. Clearly this material has not been trans- 
 ported by waves and currents unaided and yet is not the typical ice 
 rampart. It seems possible that stray blocks of ice in whose bases 
 rocks have been frozen may have been blown against the point 
 and upon partial melting dropped some of their load. The ma- 
 terial was then worked over to some extent by the pounding of the 
 ice blocks as they drifted with the waves, causing a spit -like form 
 of very coarse material. The exposure of this point to the south- 
 easterly_ winds which sweep up from the lower end of the lake ac- 
 counts for the greater development on that side. 
 
 Beyond F the irregular shore shows little adjustment until the 
 small bay designated as G on the map is reached. On the east side 
 of this bay the terrace of the former level is present and is em- 
 phasized by ice ramparts at the shorelines of both the present and 
 higher levels. On the opposite side of the bay, which has the ad- 
 vantage of a relatively smooth shore to the west, the currents, aided 
 by ice, have succeeded in building a hook of coarse piaterial similar 
 to the form found at F. That the currents along this shore are re- 
 versed with the shifting of the winds is shown by the hook which is 
 developing westward from the end of point H. The material of 
 this hook is sand and extends under water as a shoal of consider- 
 able extent. The point has been abruptly turned by the westerly 
 winds, giving the effect produced by tidal races in the ocean, as at 
 the tip of Cape Cod. In the small reentrant at I an interesting 
 series of five parallel cusps pointing towards the middle of the 
 bay was noted at the time this study was made. These low water 
 forms are similar to those found on Douglass Lake, under which 
 their manner of formation was discussed. Just west of the indenta- 
 tion, a small projection of the shore is caused by an outcrop of 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA 247 
 
 thinly bodded limostone which show.s some undercutting by the 
 waves at tlie hi<ihei' level. At present, however, the waves have not 
 been able to keep the foot of the low clitf free from talus and the 
 point is, therefore, receding at a very slow rate. 
 
 Westward to the end of the lake, the shore is relatively smooth 
 and little adjustment has taken place. One rather prominent ex- 
 ception is the point at J which breaks the continuity of the cur- 
 rents along this shore. A spit of rubble up to six inches in diameter 
 is growing in a southerly direction from the west side and may cut 
 off a small bay in time. The contrast between the" coarse material 
 of the spit and the sandy bottom of the bay in process of inclosure 
 emphasizes the importance of drifting ice as an aid to currents in 
 the formation of such features. It is interesting to note that the 
 northeasterly winds play the more prominent part in the shove 
 exerted by ice on this shore. 
 
 Beyond J there is little of interest until the shore turns to the 
 southwest to form the small haj at the end of the lake. Here tw^o 
 incipient spits are developing and the material is again coarse 
 rubble indicative of ice-shove. The probability of these spits devel- 
 oping across the mouth of the bay is remote, for vegetation is well 
 started in the bay and literally carpets the bottom. Lilies and 
 rushes have also taken hold and extinction by vegetation before the 
 spits are completed may be expected. Within the bay the notice- 
 able shore adjustment is a well-developed ice rampart w^hich stands 
 just above the high-water strand. The west end of the lake is 
 very shallow and is lined for most of the distance with a good sand 
 beach. 
 
 In general, the northeastern shore of the lake is more regular in 
 outline than the opposite shore and especially so from the west end 
 to the large bay at L. Adjustments by waves and currents are very 
 slight and are, therefore, very noticeable when present. Thus, the 
 spit which has developed from the end of the narrow point at K 
 stands out prominently and is an indication of the strength of the 
 easterly winds which have a much greater reach than those from 
 the west. But the relatively smooth s'tretch of shore to L is not 
 without interesting features, for here are found some of the best 
 examples of the effects of ice shove on the shores of the lakes of 
 our State. Ramparts are the striking features of the shore and 
 they are found in two almost continuous belts, representing the 
 former and the present levels. As a rule the rampart of the former 
 level is the stronger and stands from five to ten feet back of the 
 one formed under the present conditions. The stronger rampart 
 
24S 
 
 IXLAXD LAKES OF MICHIGAN 
 
 is a distinct ridge wliicli frequently reaches five feet or more in 
 height and contains boulders of several tons in weight. The off- 
 shore slope T^'as not swept free of boulders before the water level 
 receded and a quantity of material has been piled up since that 
 time in the present rampart. Xor is the process complete at the 
 
 «rx 
 
 Fig. 77. Boulder pushed on shore by an ice jam. Long Lake. Alpena County. This 
 boulder ^Yill travel up the beach in stages and eventually become a part of the 
 
 *=.. ... ^.....^.^^ i.^.oii^^i on snore oy an ice jam. j^ong j_.aKe. Aipena 
 boulder will travel up the beach in stages and eventuallv become 
 rampart. (.Sketch from photograph.) 
 
 ^ 
 
 ^^S^i 
 
 . J?-*" ^ 
 
 „ I . — - >'-^-\^''y»^-°^ 
 
 ,iH^*"^-* 
 
 
 
 Fig. 7S. Boulder pushed on shore by an ice jam. This boulder has nearly com- 
 pleted the journey to the strong ice rampart which stands a few feet to the right 
 and is hidden by the trees in the sketch. Long Lake, Alpena County. (Sketch from 
 photograph.) 
 
 present time for boulders are seen in the process of transportation 
 across the terrace towards the rampart. Two such boulders are 
 shown in Figs. 77 and 78 from which may be seen the trench along 
 which the rock has moved and the rubble piled in front like the 
 
Miiliii^nn I ;i'(il(ii;ic;il miu 
 I'.iold.uiral Sur\('.v 
 
 'nlilii-alion .'111. ( Icolduica I Scrii 
 I'latc XV. 
 
 MiMiiiiiMi 
 
 u....::^,iM... 
 
 -- 4.>- JtL-iwim ■jiwtiwiiwwi^wi^yy 
 
 
 
 -filial ii;.'|| 
 
 ■vj?.-- 
 
 A, u'iM'()U.\ii;i) srri', i.om; i.aki:, ali'i:.\a t'orxrv. 
 
 U. K'K KA.Ml'AK'l'. AIMM.K ISLANP. (IKCIIAKD l.AKi:. 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA 249 
 
 "boue ill the teeth" of an oeeau liner. Several of these boulders 
 were seen on this shore and the paths of all were found to be 
 practically in the same direction. This direction, moreover, was 
 not at right angles to the shore but from southwest to northeast. 
 These rocks were undoubtedly moved by the ice but the manner of 
 shove has not been observed b}^ the writer. However, the fact that 
 large ice-jams occur on this shore of the lake and that the direction 
 of the movement of the boulders over the terrace is diagonal leads 
 one to believe that ice jams have been more efifective than expan- 
 sion, although the width of the lake is well within the limit for the 
 latter. 
 
 The lake is very shallow off this shore and the bottom is com- 
 posed of large angular boulders except for a zone of sand near the 
 shore. Xo ''drop off" could be detected but the sand zone is con- 
 sidered its equivalent, its outside edge marking the outer limit of 
 the effects of the undertow. 
 
 The bay at L is shallow but, nevertheless, waves of considerable 
 power have removed the thin drift cover and exposed the lime- 
 stone on the west side. Vegetation is gradually taking hold and 
 will aid in the extinction of this arm. Another factor of importance 
 in this connection is the action of currents and ice off the point at 
 the east side of the entrance. The greater portion of the point which 
 separates this bay from the main lake was original with the lake 
 basin, but the end has developed into a hook, typical in all re- 
 spects except the size of the material, which is frequently as much 
 as six inches in diameter. Thus, the beach is evenly curved, a de- 
 pression stands to the rear, and the bottom slopes gradually off 
 shore into the lake but drops steeply ou the bay side. Currents 
 have clearly been the important factor in its formation but the 
 large size of the material leads to the conclusion that ice has been 
 instrumental to some extent. However, near the end of the hook a 
 peculiar offshoot, see Plate XV, A, has developed at right angles 
 to the general trend of the shore. This spit, which is composed of 
 large, angular material, is, furthermore, serpentine in form and 
 could hardly have been formed by currents alone. One is com- 
 pelled to account for its formation in a manner similar to that of 
 the point at F on the opposite side of the lake. Currents are effec- 
 tive on the south side of the point also and have built a blunt spit 
 of rather coarse material. Currents from both directions are con- 
 tributing to this point but are making slow progress, due to the 
 deeper water in this part of the lake. Ice push is very effective 
 on the end of this spit and has piled up a rampart fully six feet 
 in height. 
 
250 INLAND LAKES OF MICHIGAN 
 
 To the soiitlieastAvard the beach, although composed of rather 
 coarse material, is of even curvature until the bluffs of the higher 
 ground at M are reached. Considerable adjustment has taken 
 place along this shore, as is shown at the small indentation just 
 north of the bluffs which has been separated from the main lake 
 by an ice rampart. It is probable that currents were instrumental 
 in the isolation of this indentation in the early stages and that the 
 bar thus formed has been remodelled by ice. The currents in this 
 case came from the west and, after the completion of the bar, swung 
 around the apex of point M and were turned from the shore a 
 little farther on. Currents from the opposite direction also left the 
 shore in this same locality and as a consequence a blunt spit has 
 developed which is similar to the V-bars already described. The 
 material is large for such a form, the usual diameter being about 
 one-inch with a maximum of four inches. Here also the ice has been 
 effective and has formed a rampart comparable in size to that found 
 on the point below bay L. When seen by the writer the currents 
 from the southeast had the advantage and were building out on the 
 north side, but this condition may have been temporary. In gen- 
 eral, the symmetry of such a form is a safe indication of the strength 
 of the forces Avhich are active in its development. Shoals run out 
 into the lake from two hundred to three hundred feet in front of the 
 points along this shore and are sufficient to turn the currents out 
 into the lake. In most cases the shoals are of rock, but at N a 
 definite built terrace of sand which drops into deeper water at five 
 feet is present. 
 
 To the southeast, the shore is broken by a number of small 
 points and the shallow bay at O. A small spit on the north side of 
 the bay indicates some current from the northwest but this is 
 necessarily feeble due to the irregularities of the shore along which 
 the currents develop. On the south, however, the currents from the 
 opposite direction are much more powerful and have built a hook 
 which, with its submerged continuation, extends two-thirds of the 
 distance across the mouth of the bay. The shore to which the hook 
 is attached has a very low slope and a broad flat is exposed during 
 periods of low water. Much of the material on this flat is coarse 
 and several boulders were noted which had halted in their journey 
 towards the shore. These have undoubtedly been pushed by ice 
 and, inasmuch as the direction of the paths, agrees with that found 
 on this shore near the upper end of the lake, ice jams are believed 
 to have been the propelling force. 
 
 The shore to point P is more regular and shows little adjustment 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA 251 
 
 save for the assortment of the beach material which is relatively 
 coarse for a lake of this size. At 1* currents, developed on 
 both sides of the projection, have left tlie shore and built a blunt 
 V-bars. This bar is, in reality, composed of two V-bars, one built 
 outside the other. The one nearer the land is the older and the en- 
 closed lagoon is now filled with vegetation, whereas the outer has 
 developed subsequently and the wet lagoon supports a growth of 
 rushes. It is probable that the older lagoon was formed at the 
 higher level but not until a strong ice rampart had been piled up 
 at the old shore. The outer lagoon is the result of the activity of 
 currents under the present conditions and is developing most rap- 
 idly during high-water. Obviously, the currents from up the 
 lake are the more powerful on account of the greater reach of the 
 winds, the longer stretch of shore, and the protection afforded by 
 point Q to the short strij^ of shore to the south. 
 
 Point Q is a peculiarly shaped projection which originally was a 
 narrow strip jutting out into the lake. Strong currents evidently 
 swing around this point from the southeast, and as a result the end 
 is a well-developed hook whose point of departure from the original 
 point may be readily detected by the ice rampart at the former 
 shore. At present the hook is growing directly towards point P 
 and may in the course of time enclose the narrow strip of water 
 between the two points. The turning of Q to the northwest is in- 
 teresting in view of the fact that the two points between this and 
 the southeastern end of the lake are both spits pointing in the 
 opposite direction. It appears, then, that the turning of Q is 
 due to the protection of the shore to the northwest by P rather than 
 to exceptionally strong currents from the opposite direction. 
 
 The first of the spits below Q was caused by a bend in the original 
 shoreline sufficiently abrupt to throw the currents from the shore. 
 The material is coarse and has been pushed into a rampart on the 
 lake side. The main shore protected by this spit is a smooth sand 
 beach but gives way to coarser material beyond the end of the spit. 
 Nearer the lower end of the lake a narrow strip of the lowland bord- 
 ers the lake and forms a fringe of swamp. The second point de- 
 veloped across a depression in the swamp during the higher level 
 and enclosed a narrow lagoon which is now grown up to marsh 
 grass and may be flooded during very high water. 
 
 The basin of Long Lake, then, may be assigned to glacial scour 
 with the possibility of solution by ground water as an added 
 factor. The irregularities of the shoreline furnished opportunities 
 for numerous adjustments of the shores by waves and currents, and 
 
252 INLAND LAKES OF MICHIGAN 
 
 the abundance of coarse material as well as the topography of the 
 shores is favorable to development of forms due to ice push. The 
 adjustments by waves and currents have seldom passed beyond the 
 early stages but those due to ice have produced some very marked 
 results. In fact, ice ramparts are by far the most pronounced of 
 all the shore features on this lake and there is hardly a locality 
 where they may not be observed. Both types of ice shove are ef- 
 fective but the writer is inclined to lay the greater stress on ice 
 jams. Furthermore, floating cakes of ice seem to have played at 
 least a subordinate role in the shajDing of some of the depositional 
 features. 
 
 The lack of adjustments by waves. and currents may be assigned 
 to several factors. Undoubtedly waves of considerable size run 
 lengthwise of the lake but lose much of their force before reaching 
 the shore on account of the uneven surface and the flatness of the 
 off shore slopes. The greater part of the work done by the waves 
 has been the assortment of the beach material, accomplished by 
 the removal of the finer particles. The paucity of current deposits 
 is due in part to the very irregular shore which prevents the forma- 
 tion of currents of sufficient continuity for large effects. In addi- 
 tion, the material upon which these forces are acting may be of 
 some importance. In general lake clays, boulder clay, and lime- 
 stone are the materials encountered. The disintegration of both 
 the lake clays and the till furnishes little sand but rather a slimy 
 mud which is carried out onto the lake bottom. This mud is very 
 calcareous and may account for the lime deposits which accumulate 
 on the pebbles near shore, although it is recognized that such de- 
 posits are formed by the action of lower plant forms. 
 
 The two stages in the histor^^ of the lake conform to the general 
 conditions found on our inland lakes. The drop in level is due to 
 the downcutting of the outlet and has occurred rather rapidly. If 
 this is continuing at the present time, the process is a slow one and 
 the probability of the lake being drained is remote. Vegetation has 
 taken hold in a few localities and some marl is being deposited 
 but it seems best to consider the lake in a youthful stage of de- 
 velopment. 
 
CHAPTER VII 
 
 INTERIOR LAKES OF THE SOUTHERN PENINSULA, CON'T 
 LAKES OF THE EASTERN INTERLOBATE AREA 
 
 ORCHARD LAKE 
 
 In a glacinted veg;ion the -largest and most diversified deposits 
 are found in tlie interlobate areas. As lias been stated heretofore^ 
 in tlie reentrants between any two lobes of ice the rock debris is 
 supplied from two ice fronts, and the accumulation is, therefore^ 
 not only much greater than that of a' single front but also the 
 forms are very irregularly distributed. In such regions the con- 
 tinuity and arrangement of the deposits are greatly disturbed, or 
 wanting entirely, and the topography consists of an intricate patch- 
 work of the various glacial forms. Undrained depressions are, 
 therefore, especially numerous but are usually small on account of 
 the lack of continuit}- of the deposits. In the interlobate area which 
 stretches southwestward from the Thumb, the surface is literally 
 dotted with lakes and from certain eminences as many as ten to 
 fifteen may be seen. Oakland County is especially favored in this 
 respect ; more than one hundred named lakes appear on the Pontiac 
 topographic map Avhicli represents about one-fourth of the area. 
 The lakes are all small, ranging from a maximum of abotit two 
 square miles down to insignificant ponds, and lie for the most part 
 in morainic basins or pits in the outwash. Two of the larger of 
 these lakes. Orchard and Cass, were selected for study on account 
 of their size, popularity, and accessibility, but the writer realizes 
 fully that the work should be extended to include groups of lakes^ 
 possibly on the basis of drainage systems. 
 
 These two lakes are part of a group which lies southwest of 
 Pontiac and drains into the Clinton River. They may be reached 
 via the Jackson Branch of the Grand Trunk R. R., the Detroit and 
 Orchard Lake intertirban cars^ or by automobile over excellent 
 roads. They both lie in rather large morainic basins and are, 
 therefore, rather simple as to origin. The shores, bounded by mo- 
 raine, are varied as to relief but uniform as to material. The stiff 
 boulder clay is not readily eroded by shore agents, and on small 
 lakes, such as Orchard and Cass, the adjustments of the shores are 
 
254 
 
 INDAND LAKES OF MICHIGAN 
 
 not on a large scale, Nevertheless, sufficient work has been done 
 both at the present level and a higher level to make the study in- 
 teresting. 
 
 ,, The nearly circular form of Orchard Lake is emphasized by Apple 
 Island which stands slightly north of the center of the lake. See 
 
 M/7es 
 
 Fig. 79. Topographic map of Cass and Orchard lakes and vicinity, Oakland County. 
 (From U. iS. G. 'S. Pontiac quadrangle.) 
 
 Fig. 79. The lake is entirely surrounded by moraine which rises 
 well above the water, dipping only occasionally to small undrained 
 swales. The shores are well drained for the most part and ad- 
 vantage has been taken of the many excellent locations for summer 
 homes. The lake is deep (reputed to be one hundred and seventy- 
 five feet) but systematic soundings have not been made, so that the 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA 255 
 
 exact (leptlis are not at hand. For our purpose it is sufficient to 
 know that the waves may develop free from interference by the 
 bottom except in the vicinity of the shores. The lake receives the 
 Avater of Pine Lake, which stands to the west, through a small 
 channel of very sluggish current as the lakes stand at about the 
 same level. Orchard in turn empties into Cass Lake, which stands 
 one foot lower, tlirough a similar channel across the narrow neck 
 of land on the north side. From Cass Lake the drainage passes 
 through Otter and Sylvan Lakes to the Clinton River which is 
 dammed a short distance below Sylvan Lake. The total fall 
 from Pine to Otter Lake is about two feet, a' fact which may be ac- 
 counted for in part by the damming of the Clinton Eiver. Yet, it 
 is doubtful if the drainage above Cass Lake is affected by the dam^ 
 and the slight dift'erence in the levels of Pine and Orchard Lakes 
 is due to the flat gradient and short course of the outlet. 
 
 The outlet of Pine Lake enters Orchard at the narrowest part of 
 the land strip v.hich separates them. This insignificant stream 
 flows through a swamp which extends to the shores of Pine Lake. 
 No indication of bars or ramparts was noted along the Orchard 
 Lake shore but a well-developed bar crosses the flat on the Pine 
 Lake side at a level slightly above the present. Thus, the two lakes 
 were originally connected, forming part of a larger and very ir- 
 regular lake which included Pine, Orchard, Cass, Otter, Sylvan and 
 possibly Elizabeth lakes. Also during this early stage Pine and 
 Orchard were separated by the development of the bar just men- 
 tioned. 
 
 South of the outlet of Pine Lake the shores of Orchard are ex- 
 posed to the powerful west winds and the waves have the full sweep 
 of the lake. The laud stands well above the lake level except "for 
 two small sags, so that the adjustment of the shore has been accom- 
 plished mainly by waves. They have formed a steep clifl; in the 
 compact till and ii well-defined submerged terrace which reaches a 
 width of one hundred to one hundred and fifty feet and drops into 
 deep water from depths of five to six feet. The recession of this 
 cliff straightened *the shoreline to some extent but currents were 
 also instrumental, as shown at the sags. In both cases the re- 
 entrants were cut off by bars which have since been remodelled into 
 ice ramparts. The first sag, A on map. Fig. 79, is a small morainic 
 depression which, since its separation from the lake, is filling rapid- 
 ly with vegetation. 
 
 Around the south side the lakeward slopes are very gentle and 
 the adjustments mucli less pronounced than on the east side. The 
 
256 INLAND LAKES OF MICHIGAN 
 
 submerged terrace swings in close to the shore and the cliffs are 
 very low. The adjustment of this shore consists mainly in the 
 transportation and distribution of material from the east side by 
 currents, forming a sand beach and built terrace. Between Orchard 
 and Upper Straits lakes a flat saddle interrupts the uniform condi- 
 tions of the south side, and the adjustments of the shores are 
 negligible. The lowest part of this saddle stands about five feet 
 above the present level of Orchard Lake and served as a very nar- 
 row divide between the two lakes during the higher level. The 
 blunt point, B on map, is a part of the saddle and shows no current 
 activity. North of this point the land rises to a rolling moraine 
 and low cliffs of variable height face the shore. The absence of 
 current deposits at B leads to the conclusion that the material from 
 these cliffs has been transported northward. This material was 
 deposited in two spits near the entrance to the embayment at the 
 northwestern part of the lake. The first spit lies directly south of 
 Cedar Island and is too small to be shown on the map. It is slight- 
 ly recurved near the tip, almost enclosing a narrow lagoon, and has 
 been converted into an ice rampart near its attachment to the 
 shore. The second spit lies west of Cedar Island and is much the 
 larger. The currents along the west shore continued beyond the 
 first spit but were forced directly out into the lake at the second. 
 Thus they dropped practically all their load and have built the 
 larger spit in this locality. It is doubtful if any material was sup- 
 plied from the west for the bay is small and the shores are mucky. 
 In fact, the bay is being filled with vegetation at a rapid rate 
 and the heaviest growth occurs on the south side. 
 
 The hills which form the west and north sides of the bay are 
 separated by a low sag which extends through to the west end of 
 Cass Lake. This channel is a' swamp from Cass Lake to the road 
 and stands less than three feet above Orchard Lake at its highest 
 point, — the road crossing. If allowance is made for some filling 
 along the highway, it seemis probable that a shallow connection 
 formerly existed between the two lakes in this locality. 
 
 North of Cedar Island a small knoll forms a point on the main 
 shore which was sufficient to turn the currents from the shore. The 
 material transported along the north shore to this point is appar- 
 ently very small in amount, since the spit has made little headway. 
 However, it is pointing directly towards Cedar Island and its con- 
 tinuation would tie the island to the mainland. A contributing 
 factor in this connection is the spit growing landward from the 
 north end of the island but much is yet to be accomplished. Cedar 
 Island is a low knoll which mav serve as an excellent index to the 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA 257 
 
 activity of the shore agents in this part of the lalce. The north 
 shore shows no adjnstnient and the waves have not prevented a 
 thriving growth of vegetation. On the south side the waves liave 
 carved the slopes into low dill's, and the material has been trans- 
 ported in both directions by reversing currents, as shown by a spit 
 at the west end as well as at the iwrth. In addition to the waves 
 and currents, ice has been effec,tive and has pushed boulders into 
 the cliffs and the landward portions of the spits into well-defined 
 ramparts. The possibility of the island attaching itself to the 
 mainland by the growth of the spits at either end suggests itself, 
 but this is improbable on account of the limited amount of material 
 in this small island. 
 
 The north shore is a succession of cliffs which drop to lake level 
 in two places only. In both cases the low areas are channels which 
 connected w'ith Cass Lake during the early stages of the higher 
 level. Bars developed from the west across both channels on the 
 Orchard Lake side and were later pushed up into ice ramparts. 
 The westerly connection was completely obstructed in this way and 
 Beebe Lake was formed midway between the two lakes. The east- 
 erly channel was never completely closed and is now occupied by 
 the outlet of the lake. In the vicinity of the .Militarj' Academy the 
 adjustments are obscured by "improvements." One might expect 
 at least 'the beginning of a spit where the land drops to the flat 
 separating Pine and Orchard lakes but, if so, it has been covered 
 by the highway. 
 
 Apple Island is a somewhat elongated cluster of morainic knobs 
 that stand well above the level, of the lake. Its location near the 
 center of the lake exposes all of its shores to agents of approxi- 
 mately equal intensity, and differences in topography are, there- 
 fore, more important in determining the adjustment, of the shores 
 than the wind direction. Waves are very active and have carved 
 steep cliffs along the entire southern and the greater part of the 
 northern shores. At the blunt projections, however, the slopes were 
 originally flatter and the effects were more j)ronounced in some re- 
 spects. Not only were the waves able to cut a wider terrace but 
 also currents were able to add to this, although no definite bars 
 were formed. In such localities the sinking of the water level ex- 
 posed relatively broad terraces which are excellently preserved and 
 furnish the clearest evidence of the former level of the lake. The 
 points on the west, north, and northeast shores are well worth in- 
 spection. The expansion of the ice during the winter is also efl'ec- 
 tive on the shores of the island and especially so on the south and 
 33 
 
258 INLAND LAKEiS OF MICHIGAN 
 
 east sides, due to tlie greater expanse of tlie lake. On the east side 
 the outer edge of the exposed terrace has been pushed into a ram- 
 part, see Plate XY, B, which obstructs the drainage and causes 
 swamp conditions in places. 
 
 CASS. LAKE 
 
 A convenient starting point for the study of Cass Lake is Dollar 
 Lake. This little pond lies in a pit in a fragment of outwash that 
 borders a part of the east shore of Cass Lake and is now a liiuck 
 basin, rapidly decreasing in size on account of the encroachment 
 of vegetation. Dollar Lake was formerly in direct connection with 
 Cass Lake but the channel became so crowded with plants that it 
 was necessary to open it by dredging. 
 
 With the exception noted, Cass Lake is surrounded by moraine, 
 composed of broad swells and sags rather than sharp knobs and 
 basins. Consequently the shores are either dry or are swampy for 
 long stretches. The south shore stands consistently Avell above the 
 lake and the shore features are so uniform that the few exceptions 
 are greatly emphasized. Leaving Dollar Lake, one's attention is at 
 once attracted to the point across the bay. The major portion of 
 the point is due to a knoll which juts well out into the lake but 
 the narrow eastward projection is clearlj^ a spit. The spit stands 
 more than two feet above the present level of the lake and is covered 
 with grass and trees. See Plate XVI, A, Near its attachment it 
 has the characteristics of an ice rampart which continues along the 
 edge of an exposed terrace surmounted by a cliff. The effects of 
 Avaves, currents and ice may be seen within a short distance and 
 indicate a higher level for this lake which corresponds exactly with 
 the former level of Orchard Lake. The disturbing feature is the 
 present activity of the waves which is removing the rampart and 
 spit, laying bare the roots of trees. See Plate XYI, B. This is un- 
 questionably a renewal of activity and is due to an elevation of the 
 water level, probabl}^ a result of the obstruction in the Clinton 
 Eiver. 
 
 South of the spit the exposed terrace soon plays out and the 
 shore is faced by a cliff which is continuous to the outlet of Or- 
 chard Lake, with the exception of a narrow gully about midway. 
 This cliff, pounded by waves driven in from the west, has furnished 
 the material for the spit above. The renewed activity of the waves 
 is clearly seen in the fresh cliffs, landslides, and undermined trees 
 and may result in further extension of the spit after it has adjusted 
 its position to the new conditions. 
 
Miclli.L;',-Ul (;cnlii,^ii';il Mill 
 r.iolo.uical Siiivi'V 
 
 I'lililic.-ilidii ;;(>. (icoliiuicMl v^orics 25, 
 I'l.ilc XVI. 
 
 ^<^. 
 
 A. SPIT, CASS LAKE. 
 
 B. DISSECTION OF ICE KA.MI'AUT. CASS EAKE. 
 
INTERIOR LAKES OF THE SOUTHERN' PENINSULA 259 
 
 Tlie outlet of Oiehard Lake and the Beebe J.ake channel interrupt 
 an abnost continuous clitt" to the west end of the lake. Wave ac- 
 tion is less intense along this part of the shore on account of the 
 restricted reach of the waves and fresh scars are not abundant 
 on the face of the cliff. Also a boulder-lined strand is still intact. 
 Farther west wave activity is still less and two neighbor] ng gullies 
 are able to maintain a small delta in the lake. Strong currents 
 are undoubtedly set up along this shore but conditions are un- 
 favorable for deposition at present. During the former stage, the 
 depressions at the Orchard Lake outlet and the Beebe Lake de- 
 pressions were crossed by currents, but the latter only was ob- 
 structed by a bar. The position of the outlet of Beebe Lake near 
 the east side of the depression shows that the bar developed from 
 the west. 
 
 Near the west arm the exposed terrace of the upper level appears 
 and Avidens towards the bay. It reaches its greatest width on the 
 south side of the bay and then diminishes. Outside the bay on the 
 west shore the cliff's rise directly from the beach. The presence of 
 the exposed terrace in the more protected places leads to the con- 
 clusion that it was formerly continuous and has been removed sub- 
 sequently to a very large extent. The cliff's extend to the neigh- 
 borhood of Gerundegut Bay, which was formerly an extended but 
 shallow arm of the lake. The dropping of the water level exposed 
 more than one-half of its bed and the shores are consequenth" low 
 and swampy at the present time. The entrance to this bay is re- 
 i-tricted to a channel which does not exceed forty fe«^t in width, 
 the remainder, as shown on the map, being blocked by very shallow 
 water grown up to rushes. The shallow water appears to be due 
 to submerged terraces which were extended from either side by cur-, 
 rent action. The greater extension of the terrace on the west shows 
 that the currents from this direction are the stronger. This may 
 be due to more favorable shore conditions combined with the greater 
 effectiveness of the southwest winds. Within the bay the adjust- 
 ments are uncertain, for the shores are lined by swamp, and filling 
 by vegetation is the dominant physiographic process. Between 
 Gerundegut and Coles bays, the shores consist of the swampy ter- 
 race for the first half of the distance but beyond rise to cliffs which 
 extend to the west side of Coles Bay. In contrast with Gerundegut 
 Bay, Coles Bay is deep and has a relatively narrow entrance. A 
 minor indentation Just west of the entrance to the bay was cut off' 
 by currents but much more important adjustments are found at the 
 bay and along the shore to the east. The small spit at the west 
 side of the entrance to Coles Bay seems rather insignificant on 
 
263 INLAND LAKES OF MICHIGAN 
 
 first sight but closer examination discloses that its continuation 
 under water nearly blocks the channel, the water above it nowhere 
 exceeding three feet in depth. Furthermore, this bar rises to the 
 higher level of the lake east of the bay and runs as far as the out- 
 let, causing it to turn to the southeast. Near the east side of Coles 
 Bay the bar stands at the jDresent shore but is fragmentary and 
 irregular in form, due to the renewed activity of the waves which 
 is causing a recession of the shore. East of the outlet the old 
 lake bed extends towards Sylvan Lake but is dotted with small 
 hills which stood as islands above the lake at its higher stage. 
 In building the shore road advantage has been taken of a well- 
 defined bar which extends southeastward from the outlet and con- 
 nects directly with one of the "island" hills whose slopes were 
 carved into a cliff and terrace during the higher stage. Farther 
 south the old lake bottom connects with the Dollar Lake depres- 
 sion without definite shore features. 
 
 A study of the shores in the vicinity of the outlet shows that 
 currents were the active agency in their development. The work 
 was practically all accomplished when the lake stood at the higher 
 level, but some deposition may now be taking place on the sub- 
 merged portion of the Coles Bay bar. The most conspicuous ad- 
 justment was the building of bars across the flat now occupied 
 by the outlet. They developed from both directions and would 
 have met, if the currents east of the outlet had run a little farther 
 to the south. As it is they dove-tail and Cass Lake was never com- 
 pletely isolated during the higher stage. 
 
 Briefly stated, the outstanding fact, in the history of the lakes 
 under discussion is the existence formerly of a large, irregular 
 body of water which not only included the basins of Orchard and 
 Cass lakes but of some of the neighboring lakes as well. During 
 this stage considerable adjustment of the shores took place and, as 
 a result, the present lake basins were at least outlined, if not iso- 
 lated. If it may be assumed that the bars stood above the water 
 level, Orchard and Pine lakes became definite basins and Cass 
 nearly so. Many of the changes, however, were interrupted by the 
 lowering of the level and little seems to have been accomplished 
 during the lower stage. A renewal of activity, however, is clearly 
 indicated at the present time, due probably to the interference 
 with the flow of the Clinton River, and adjustments are beginning 
 anew. It may not seem possible that the effects of an obstruction 
 of the drainage would be so far-reaching but it must be kept in 
 mind that an actual rise in the water level is not necessary to 
 produce an increase in the adjustment of the shores. Small lakes 
 
INTERIOR LAKES OP THE SOUTHERN PENINSULA 261 
 
 vary cousiderably in level during the season and any cause which 
 will prevent a fall during the di*y season is equivalent to a higher 
 level on the average. As has already been stated, this study was 
 not complete but it is hoped that from what has been done the 
 work may be extended by the reader to the other lakes of this 
 group. 
 
 LONG LAKE, GENESEE COUNTY 
 
 In the discussion of Orchard and Oass Lakes the confused ar- 
 rangement of the glacial formations of the interlobate area in the 
 southeastern part of the State was mentioned. Outside this area 
 on the Saginaw Bay side tlie deposits are distributed with a regu- 
 larity which, by way of contrast, is very pronounced. Morainic 
 ridges, separated by till plains and occasional outwash aprons, 
 form a series of wide, roughly parallel loops about the Saginaw 
 Lowland, extending from Genesee County westward to Ionia County 
 and thence northward to the southeastern part of Eoscommoii 
 County. In many localities the streams follow the moraines for 
 long distances but in others they flow directl}^ across the trend of 
 the ridges. Numerous small lakes occur in this region, the major- 
 ity filling depressions in the till plains or morainic basins. A group 
 of isuch lakes is located in southern Genesee County, north and west 
 of Fenton, some members of which were examined by the writer. 
 The most common type are the shallow basins in the till plain 
 which often contain great quantities of marl, utilized in the manu- 
 facture of Portland cement, for example. Mud Lake. The larger 
 'lakes show the more interesting physiographic development and 
 Long Lake, see Fig. 80, was chosen as more or less representative. 
 
 The glacial formations in this region have a general east-west 
 trend but are less regular in their distribution than farther to the 
 w^est. A high, strong moraine follows the southern border of Fen- 
 ton Township, marking a decided halt of the ice front at this lo- 
 cality. The next halt of the ice is shown by a fragmentary moraine 
 which crosses the region near the south end of the lake and was of 
 short duration. Another moraine crosses the north arm of Long 
 Lake but turns abruptly to the south just east of the lake, merging 
 with the fragmentary moraine mentioned above. From this it ap- 
 pears that the ice front held its position east of the offset but re- 
 ceded to the westward and uncovered the southern part of the Long 
 Lake basin. During the halt of the ice border which followed a 
 local outwash apron developed, fragments of which surround the 
 south arm of the lake. North of the outwash the shores are bordered 
 
262 
 
 INLAND LAKES OF MICHIGAN 
 
 by till plain, which may be interpreted as a fosse, followed by mo- 
 raine along- the northarm and till plain at the north end. 
 
 The orientation of the Long' Lake basin transverse to the trend 
 
 Outline map of Long Lake and neigliboring lakes, Genesee County. 
 
 of the moraines suggests that a depression existed prior to the 
 advance of the ice. Some scour by the advancing ice probably took 
 place, deepening the depression somewhat to form the Long Lake 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA 263 
 
 basin, iu which case Orrs I'oiut and Craues Island were deposited 
 later when the ice retreated. During the retreat of the glacier the 
 ice persisted in the basin and upon melting formed the outwash 
 about the south end of the lake. 
 
 Long Lake lies in the drainage of the Shiawassee Kiver whicli now 
 flows directly into the east end of Mud Lake through an artificial 
 channel. From Mud Lake it follows the south side of a moraine 
 westward for more than fifteen miles before crossing. On all inaps 
 which the Avriter has seen the outlet of Long Lake is placed at the 
 north end connecting with a branch of Swartz Creek. This outlet 
 undoubtedly functioned but was artificial. Discussion concerning 
 the former natural' outlet will be found below. 
 
 An excellent distance view of Long Lake may be had from the 
 high kame isouth of Fenton and from this eminence the irregular 
 outline of the lake is veiy apparent. From this view and a study 
 of the map it is seen that the lake consists of three basins set off 
 by Orrs Point, and Log Cabin Point and Cranes Island. It also 
 appears that the west shore is much more irregular in outline than 
 the east, but this is less pronounced when the adjustments of the 
 east shore are taken into consideration, as will be seen later. 
 
 At the south end of the lake the low remnants of outwash slope 
 genth^ to the lake and have been carved into a well-defined terrace 
 which stands between two and three feet above the present level. 
 Wave activity in this small arm of the lake is moderate and the 
 shores are grown over with vegetation, consequently the beach is 
 poor. Towards the outlet the land is higher and well-defined cliffs 
 were formed at the higher level. In the neighborhood of the out- 
 let a low flat which stands below the former level of the lake ex- 
 tends westward and connects with the former channel of the Shia- 
 wassee Kiver. The south arm o.f Long Lake now^ shows an interest- 
 ing anomaly in that the outlet now leaves the lake across what ap- 
 pears to be a delta which extends to and includes the small island 
 iu this part of the basin. 
 
 If this formation is correctly interpreted, it is evident that the 
 Shiawassee entered the lake at this point during the former level. 
 At this time two outlets were possible, one just north of Orrs 
 Point leading back into the present Shiawassee valley, and another 
 at the north end of the lake. Of the three channels leading from 
 the lake, the one now occupied by the outlet was the lowest and 
 naturally accommodated the outflow when the lake level dropped 
 permanently. Yet it is possible that it may have served as an inlet 
 during the former level under flood conditions of the Shiawassee, at 
 which time great quantities of material were brought into the lake 
 
264 , INLAND LAKES OF MICHIGAN 
 
 and deposited, forming the delta. On the other hand, when the 
 waters lowered during the dry season this channel may have served 
 as the outlet which had a sluggist current so that the delta was 
 left intact. 
 
 The south arm of the lake is relatively deep and a well defined 
 off-shore terrace drops into deep water from a depth of four feet. 
 This terrace has been remodelled under the present conditions to 
 an uncertain extent and inferences from it as to the strength and 
 work of the waves are of little value. As a rule the surface of the 
 terrace is covered with heavy deposits of marl which is of commer- 
 cial value on some of the nearby lakes. 
 
 North of the outlet the shores are low and in places the terrace 
 of the upper level is exposed. Similar conditions prevail on the 
 south side of Orrs Point, a m'orainic ridge which runs nearly to the 
 east shore, forming a narrows of less than one hundred feet in 
 width. During the former level considerable material was carried 
 outward along the south side of the point and deposited in a spit 
 which threatened to close the channel. Under present conditions 
 little work is being done on this shore and the spit is not growing 
 appreciably. A like condition exists on the north side but the spit, 
 although possibly of greater extent than that on the south side, 
 had accomplished relatively less in the filling of the channel. The 
 north shore spit is attached to a narrow terrace of the upper level 
 which stood thirty inches above the water level at the time of 
 measurement. Ice shove is here evident in the line of boulders 
 along the shore and a partly demolished rampart at the outer edge 
 of the terrace. Westward into Black Bass Bay cliffs rise from the 
 present shore, showing that the waves not only have removed the 
 upper level terrace but are still furnishing material some of which 
 is being added to the spit at the end of the point. Thus, there is 
 still a possibility of the isolation of the south arm. 
 
 Black Bass Bay stands on a till plain and is very shallow with the 
 exception of a small hole near the west end. This probably was the 
 location of an isolated ice block. Within the bay a slight swell- 
 near the north shore forms Duck Island. The low relief, as indi- 
 cated by the shallowness of the bay, extends westward from the lake 
 to the old Tchannel of the Shiawassee. This flat stands slightly be- 
 low the level of the former stage and, if not an outlet, was a broad 
 arm which connected this basin with the lake basins which lie to 
 the west. 
 
 The west shore north of Black Bass Bay is lined with a gravel 
 beach above which stands an almost continuous terrace of the 
 upper level. Currents as well as waves are active along this shore 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA 265 
 
 and those from the south leave the shore at the end of Piue Point. 
 Southeasterly winds are the only ones which cause northward drift- 
 'ing currents on this sliore and, since these winds are not of long 
 duration nor great intensity, the amount of material transported 
 is relatively small. Nevertheless, a well developed spit was formed 
 at the point during the upper level and is continuing its growth 
 under the present conditions. Ice action is also effective near 
 the attachment of the spit and has formed ramparts at both the 
 upper and present levels. 
 
 Within Cranes Cove wave action is slight and the exposed ter- 
 race is well preserved. Near Crane's cottage ice has formed a ram- 
 .part at the present shore which obstructs the drainage of the ex- 
 posed terrace locall}^, forming a lagoon-like swale. At the head of 
 the cove the terrace widens to a swamp which extends westward to 
 the road and was a shallow arm of the lake during the upper level. 
 
 The shallow water between Cranes Island, or Cases Island, and the 
 west shore establishes a close relationship between the two. The shal- 
 low water, however, is not due to current action, and any direct con- 
 nection of the island with the mainland w^ill be accomplished by 
 filling by vegetation which has already established itself. The 
 island is a low knoll of nearly twenty-five acres in extent. The great-' 
 est activity along its shores at present occurs on the south side 
 and is due to waves. They have removed the terrace of the former 
 level and are cutting back cliffs of from twelve to fifteen feet in 
 height. On all other shores the exposed terrace is well preserved. 
 The preservation of this terrace on the north side seems singular in 
 view of the exposure of this shore to strong winds of considerable 
 reach. It is due to the flatness of the off-shore slope, which has 
 been intensified by the sinking of the water level and greatly re- 
 duces the effectiveuss of the incoming waves. The adjustment of 
 the shores during the higher level was accomplished mainly' by the 
 waves, as shown by the exposed terrace on all sides but the south. 
 However, westward moving currents developed on the north shore 
 and succeeded in building a small spit at the northwest corner of 
 the island. 
 
 North of Cranes Island on the main shore, the terrace of the 
 higher level makes a slight notch in the gentle slopes and is con- 
 tinuous to the swampy depression which may have functioned 
 as the outlet of the lake. This channel is partially closed by a 
 strong ice rampart, the best example of this feature found on the 
 lake. It is possible that this channel served as an outlet during 
 the higher stage but, if so, was secondary to the Shiawassee River. 
 
266 INLAND LAKES OF MICHIGAN 
 
 In case both outlets functioned, the northern channel- was 
 abandoned bj^ the deepening of the channel of the Shiawassee Kiver. 
 
 Moraine borders the north shore and the slopes have been cut 
 into almost continuous bluffs ranging from fifteen to twenty feet 
 in height. The northeastern shore is low and the exposed terrace 
 appears along the shores of the bay at the northeast end of the 
 lake. High bluffs again come to the shore at Lookout Point but 
 drop to a low sag near the end of the point. This sag stands slight- 
 ly below the former level and the tip of the point was, therefore, a 
 small island during that time. South of the point the bluffs recede 
 and below Bayport a small crescent-shaped lagoon was cut off by 
 current action during the higher level. 
 
 Between Bayport and Log Cabin Point the shores are faced for 
 most of the distance by bold cliffs. A well defined ''drop off" follows, 
 this shore and extends outward as much as two hundred yards in 
 places, dropping at a depth of six feet. The depth of the "drop off", 
 it will be noted, is g-reater than on the west side of the lake, and 
 this is due to the greater power and size of the waves developed 
 by winds from the westerly quadrant. Weves are active under the 
 present conditions on this shore and breakwaters are necessary in 
 places to prevent the recession of the cliffs. Sags in the cliffs are 
 infrequent but where present show both current and ice action at 
 the bars that developed during the former level. 
 
 Log Cabin Point is an extension of the moraine into the lake but 
 originally had a gentle slope. Consequently it now shows a wide 
 terrace. The original slopes were much flatter on the south side of 
 the point and the submerged portion of the terrace is, therefore, 
 much wider on this side. A small embayment on the south side of 
 the point was completely cut off by a bar at the higher level and this 
 has isince been remodelled into an ice rampart. Bluffs line the shore 
 to the vicinity of the narroAvs with the exception of another indenta- 
 tion south of Lakeside which was cut off by the usual combination 
 of bar and ice rampart at the higher level. This indeniration runs 
 back of the bluff's and appears again at the lake shore north of the- 
 narrows where it is similarly cut off. 
 
 The two small points on the east shore opposite Orrs Point ar& 
 due to current and ice action during the former stage and were 
 built of material brought in from the cliffs to the north. In this 
 vicinity the shore consists of an alternation of cliffs and swamps,^ 
 the latter being as a rule cut off. An excellent rampart-bar iso- 
 lates a small swamp just south of Coopers Landing. In the same 
 way a nearby swamp, which extends a considerable distance to the 
 east, was cut off, definitely separating a small pond from the Loug 
 
INTERIOR LAKES OF THE SOUTHERN PENINSULA 267 
 
 Lake basin. The point below the narrows consists of several small 
 knolls surrounded by swamp on the east. This swamp was undoubt- 
 edly open water at the former stage but has since been filled to 
 some extent by vegetation. 
 
 The east shore to the south end of the lake is lined by low cliffs 
 and vegetation grows to the shore. The exposed terrace is not well 
 developed and in many places is wanting entirely. Also the sub- 
 merged terrace is but moderately developed and drops usually 
 within one hundred feet of the shore. Vegetation is much more 
 abundant along this shore than in other parts of the lake. In gen- 
 eral, wave action is moderate in this part of the basin, due prob- 
 ably to the restricted reach of the waves. 
 
 There appears to be sufficient evidence to show that Long Lake 
 has stood for a considerable period of time at a level about thirty 
 inches higher than at present, an almost universal occurrence on the 
 inland lakes of Michigan. When first formed the basin was con- 
 nected with those of Squaw and Mud Lakes and also included a 
 number of small and large indentations which were either isolated 
 by current action or drained by the sinking of the water level. That 
 the former stage was of long duration is shown by the fact that 
 practically all of the shore adjustments took place at that time. 
 Thus, strong cliffs and an almost continuous terrace were cut by 
 the Avaves. Currents succeeded in throwing bars across many of 
 the indentations and added to the terrace formed by the waves. In 
 addition, spits were started from some of the points, which show 
 little growth under the present conditions. The greatest change 
 has been the (reversal of the Shiawassee drainage with the cessation 
 of delta building. In general, the adjustments were carried to 
 such an advanced stage that the present outline of Long Lake was 
 determined at that time. At present, the waves are doing most of 
 the work on the shores and in places are cutting back the cliffs, the 
 exposed terrace having been removed. The change of greatest 
 moment that one may foresee is the possibility of the growth of 
 the spit at the north end of Orrs Point which will eventually iso- 
 late the south arm. Vegetation has taken a firm hold on the sub- 
 merged terrace, following a more or less abundant accumulation of 
 marl. Some of the shallower parts of the basin are in process of 
 filling in this way, but a large part of the lake is too deep for such 
 method of extinction. The absence of entering streams precludes 
 filling by sedimentation. 
 
CHAPTEK \ni 
 LAKES OF THE NORTHEKN PE:S'IXSULA. 
 
 BREVORT LAKE 
 
 Brevort Lake is one of the largest lakes of the Northern Penin- 
 sula and lies approximately twelve miles northwest of St. Iguace 
 and within one mile of Lake Michigan. Allenville on the Duluth, 
 South Shore and Atlantic R. R. is the nearest stop and a short 
 drive brings one to the east end of the lake. 
 
 The lake is five and one-half miles long and nearly twa and one- 
 half miles in greatest width. See Fig. 81. The south shore runs 
 nearlj^ parallel to that of Lake Michigan, that is, northwest- 
 southeast, but the north shore takes a nearly east-west direction so 
 that the lake tapers towards the w^est in general outline. It covers 
 an area of 6.7 square miles and its average width is about one 
 and one-fourth miles. 
 
 The lake lies in a shallow basin surrounded by low ground for 
 the most part. Near the southeastern end two low knolls rise 
 above the sandy plain. These knolls are of hard rock thinly covered 
 with glacial material. They are surrounded by a sand plain which 
 formed a portion of the bed of Lake Nipissing and now borders 
 the east end of the lake. On the north side a flat swamp extends 
 from the lake shore northward to the Carp River. This swamp 
 closely resembles outwash in that its surface is flat and the material 
 is sand, but stands somewhat lower in elevation than the outwash 
 plain at the west end of the lake. The south shore from Luepnitz 
 Bay to the west end is bounded by a belt of sand dunes which rise 
 nearly one hundred feet above the lake and are the highest land 
 in- the vicinity. These dunes form the inner border of a narrow 
 sand strip that separates this lake from Lake Michigan. Beyond 
 the dunes is a series of bars, parallel to each other and the Lake 
 Michigan shore, which are separated by shallow troughs grown 
 np to swamp grass. Near Lake Michigan another belt of dunes 
 is found. From the elevation of the bars it is clear that Brevort 
 Lake was a part of Lake Nipissing in its early stages but was cut 
 off by a bar which stood near the location of the inner belt of 
 dunes. This stage was of relatively long duration and the constant 
 
270 
 
 INLAND LAKES OF MICHIGAN 
 
 addition of sand to the sliore by curreuts furnished the paaterial 
 for the dunes. Then followed a recession of the shore accompanied 
 by rapid encroachment of vegetation on the exposed bars and 
 
 Fig. 81. Outline man of Brevort Lake, Macl^inac County. 
 
 dunes, which fixed their position. The final belt of dunes prob- 
 ably represents another halt in the recession of Lake Nipissing 
 but may belong to the present stage. Just what may have been 
 
LAKES OF THE NORTHERN PENINSULA 271 
 
 the cause of the depression in wliich Brevort Lake lies is not clear 
 at present but it is certain that the lake has been isolated by cur- 
 rent action and tlius may be classed as a lagoon. 
 
 Tlie relatively smooth east-west shore on the south side of the 
 east end is caused by one of the low rock knolls covered with a 
 veneer of drift. Along the shore the waves have removed the finer 
 drift and exposed blocks of limestone which have been forced into 
 the low cliff by ice. Some expansion of the ice takes place in 
 Avinter but ice jams are the more powerful, according to the state- 
 ments of the inhabitants of the region. Along this shore as far 
 Avest as the point, A on map, which is caused by the southward 
 turn of the shore into Luepnitz Bay, is a wide' submerged terrace 
 which drops into deep water at six feet. The west edge of this 
 terrace is in line with the extension of the shore of Luepnitz Bay. 
 On the west side of the point a distinct boulder-lined strand stands 
 above the sand beach of the present level. This strand probably 
 represents a natural level of the lake, which is low at the present 
 time, due to the blowing out of a log jam at the outlet. The blow- 
 ing out of the dam accomplished more than its purpose and deep- 
 ened the outlet. 
 
 Along the east shore of Luepnitz Bay two small swamps drop be- 
 low the general level. No current action was noted at the swamp 
 nearer point A but the more westerly swamp is completely cut off 
 by two parallel bars. The bar farther inland is the stronger and 
 indicates a stage during which the lake stood nearly six feet above 
 the present level. This stage will be referred to as the Higher 
 Level. The second bar was formed durin^ the Upper Level which 
 preceded the deepening of the outlet. West of this swamp a smooth 
 sand beach runs to the southwest shore of the lake. The shore, of 
 Lrcpnilz Lay is exposed to strong westerly and northerly winds 
 which have caused a well-defined submerged terrace extending 
 approximately one hundred yards off shore. Also much material 
 is carried eastward along this shore and deposited on the terrace 
 oft" point A, forming a broad, submerged hook. 
 
 The entire south sliore is lined with sand dunes whose back 
 slopes rise steeply from tlie lake. At present, the dunes are fixed 
 but formerly they encroaclied on the lake from the west, causing 
 a projection of the shore at each dune. Such projections were 
 more exposed to wave action than the intervening shore and Avere 
 cut back rapidly on account of the easily eroded sand. In this way 
 a relatively smootli shoreline developed, but below the Avater level 
 a triangular terrace extended oft' shore from each dune. The reces- 
 sion of the cliffs and the formation of the triangular terraces prob- 
 
272 INLAND LAKES OF MICHIGAN 
 
 ably took place duriug the Higher Level of the lake. With the 
 recession of the water to the Upper Level, conditions were reversed 
 and currents began to deposit along the edges of the triangular ter- 
 races. Thus, at the present time these terraces are deeper in the 
 center and resemble V-bars. From the vicinity of the outlet to the 
 west end of the lake, westward moving currents are the more power- 
 ful and the triangular terraces have been remodelled in form as well 
 as surface. They, therefore, are extended in a northwesterly di- 
 rection. The best example of this is point B which extends north- 
 west two hundred j'^ards beneath the water. 
 
 The outlet bends abruptly to the southeast soon after leaving 
 the lake and flows In a direction parallel to the southwest shore 
 of Brevort Lake for nearly three miles before turning towards 
 Lake Michigan. It then doubles back on its course for more than 
 a mile and finally enters Lake Michigan on a reverse curve, that is, 
 swings back to the southeast. The turning of the upper part of its 
 course is due in part to the growth in a southeasterly direction of 
 the original bar which separated Brevort from Lake Michigan 
 and also to the general trend of the topography where not covered 
 by dunes. The outlet was not followed by the writer to Lake 
 Michigan but the final turn to the southeast is x)robably due to the 
 drift along the shore in this direction. 
 
 The sags between the dunes vary in elevation above Brevort 
 Lake and in many places are low enough to cause swamp condi- 
 tions along the shore. At such places the exposed terraces of the 
 higher levels are wider than usual and have no distinct line of 
 separation, giving the effect of a single terrace of exceptional 
 width. No definite bars cross the mouths of these indentations 
 which were relatively broad, but occasionally an ice rampart is 
 found, for example, west of point B. These ramparts may be re- 
 modeled bars but this is uncertain. 
 
 A well-defined submerged terrace fringes this shore and drops 
 consistently at six feet, except off the projecting dunes. As has 
 been intimated, the outline of its outer edge is much more irregular 
 than the present lake shore. Fox Point is a blunt sand spit which 
 developed mainly at the Upper Level and nearly closed the channel 
 into the west arm of the lake. At present the growth is very slow 
 but, nevertheless, may complete the spit. The west arm is shallow 
 and very muddy, and there is a possibility that the rapidly en- 
 croaching vegetation may fill the basin before it is isolated by the 
 spit at Fox Point. 
 
 The entire north shore of the lake is bordered by a broad ex- 
 posed terrace which terminates at a low cliff approximately one 
 
LAKES OP THE NORTHERN PENINSULA 273 
 
 mile north of the lake. This terrace correlates with the great bar 
 along the south shore which isolated Brevort Lake and was formed 
 during the late Nipissing stage of the Great Lakes. For the greater 
 part of its extent the terrace drops to the shore of Brevort Lake 
 in a low clitif at the foot of which rests a line of boulders, indica- 
 tive of ice action. The points are caused by excessive accumula- 
 tions of boulders which, obviously, retard the action of the waves. 
 On the other hand, minor indentations were present along this 
 shore but were virtually all cut off by bars during the Higher 
 Level. In front of the bars the beach is of clear sand and has been 
 pushed into a rampart in many places. Just east of Black Point 
 a small terrace, standing below the elevation of the bars, is evi- 
 dence of the Upper Level. The evidence of the latter, although 
 sufficient to establish this stage, is meager and the obvious con- 
 clusion is that it was of short duration. 
 
 A -submerged terrace of variable width and slope follows, this 
 shore. In general, the depth at the drop off increases from two 
 feet near the west end to six feet at the east end. This variation 
 cannot be entirely accounted for but may be ascribed in part to the 
 increased effectiveness of the storm winds from the west as the reach 
 becomes. greater. It was noted that the lesser depths at the drop- 
 off were in many cases due to a low ridge near the edge 
 of the terrace and may be accounted for to some extent by 
 ice action. Powerful ice jams occur on the north shore and reach 
 heights of ten to twelve feet at times. The interesting fact is that 
 they not only drive on the shore but are also piled on the edge of 
 the terrace where the low ridges are located. The effects are more 
 pronounced in the shallower water, that is, nearer the west ei'd, 
 and the stronger ridge is formed at the edge of the terrace in this 
 part of the lake. It seems reasonable, then, that the large varia- 
 tion in depth at the drop off is due in part to ice action. The 
 effect of these jams on the shore is best seen west of Black Point 
 where an excellent rampart has been formed. 
 
 In the vicinity of Cut Eiver Point currents have produced some 
 decided effects. The point itself is a V-bar which is better de- 
 veloped on the west side, indicating stronger current action from 
 this direction. It stands Avell above the present lake and was 
 formed during the Higher Level. But the most striking effect is 
 the formation of the Cut Eiver reef which is a submerged bar ex- 
 tending far out into the lake in accordance with the curvature of 
 the west side of the point. It has been somewhat modified on the 
 east side, but the abrupt drop into deep w^ater indicates that the 
 activity of waves and currents is slight on this side. Cut River 
 35 
 
274 INLtAND LAKES OP MICHIGAN 
 
 Bay has a sand beach of even curvature back of which is a low ice 
 rampart. The possibility of a bar, remodelled by ice action, natur- 
 ally suggests itself but is dififlcult of proof. The river is forced 
 to the east before entering the lake, showing the prevalence of 
 currents from the opposite direction. A large quantity of silt has 
 been deposited by the stream and has been distributed on the east 
 side of the hRj, forming a broad, submerged terrace. 
 
 At C currents of about equal intensity left the shore from both 
 directions and built a symmetrical V-bar of greater dimensions 
 than that noted at Cut Eiver Point. This also was formed at the- 
 Higher Level but the enclosed lagoon is still a good-sized pond. 
 In addition to the exposed V-bar, a submerged bar similar to the 
 Cut River Reef extends well out into the lake. The balance of the 
 currents on either side of this point is due to the protection afforded 
 the shore north of the point by the Cut River reef. Thus, only 
 local currents are set up on the north side and by winds of short 
 reach, but on the south side the currents are formed along a greater 
 stretch of shore and by winds of longer reach, although of less 
 power. However, part of the material derived from the shore be- 
 tween C and D was carried southward and deposited at T> in a; 
 spit. The greater part of this deposit is submerged and, together 
 with the submerged terrace on the opposite side of the lake, leaves 
 only a narrow channel of deeper water leading into the east bay. 
 The north and east shores of this bay are low and the only shore 
 feature of importance is a well-delined ice rampart. The effect 
 of vegetation as an aid in the formation of ramparts where the 
 material is sand is here well shown, for the rampart is not present 
 across the clearings. During the higher levels a distinct submerged 
 terrace was formed which now supports a heavy growth of rushes. 
 The edge of the terrace is the off-shore limit of rushes and may 
 thus be readily traced during the growing season. 
 
 Brevort Lake was originally an arm of Lake Michigan but was 
 completely shut off from the main body of water by a bar of large 
 dimensions which has since been heaped into dunes. During the 
 Nipissing stage the broad, exposed terrace bordering the north 
 side of the lake was formed. Later the water subsided and halted 
 for a relatively long time at a level six feet above the present stage. 
 At this level the greatest adjustment of the shores took place but 
 the permanent effects were accomplished by currents. Undoubtedly 
 waves were active and aided materially in the formation of a sub- 
 merged terrace, but this has been subsequently remodelled. Numer- 
 ous spits began their development at this time and in some cases 
 became bars, but no very striking changes in outline resulted. 
 
LAKES OF THE NORTHERN PENINSULA 275 
 
 Tlie strongest action took place on the north side, as shown by the 
 cut off indentations and the ^^-ba^s at Cut llivei' Point and locality 
 C It is possible that the reef at Cut Kivei- Point and the projec- 
 tions to the east were at least started during this stage. 
 
 A second subsidence halted at an intermediate level whose fea- 
 tures are poorly preserved and w^as probably of short duration. 
 Nevertheless, many of the forms of the preceding stage were re- 
 modelled and in some cases added to. The spits were either planed 
 down to the water level or greatly extended. The terrace of the 
 Higher Level was generally cut down to conform to the lower water 
 level, and, on the south shore, parts of the remodelled terrace have 
 become areas of deposition. The present stage was initiated by 
 human interference and minor shore adjustments are progressing 
 slowly. The work to be done consists in remodelling the forms of 
 the previous levels by waves and the extension of the previous de- 
 posits by currents. The most prominent change will result from a 
 continuation of Fox Point across the entrance to the west arm. 
 
 Ice jams have been and still are very active on the lake with, 
 marked results where conditions are at all favorable. Some head- 
 way has been made towards the extinction of the lake but this is 
 of local importance. Vegetation is well established at the ends of 
 the lake and considerable filling has been accomplished. This is 
 especially true of the west arm where the process is aided by the 
 silt of North Creek. Another locality where sedimentation is im- 
 portant is at the mouth of Cut Kiver, but, in general, it may be 
 .stated that the process of extinction is in an incipient stage. 
 
 THE MANISTIQUE LAKES 
 
 The moraine which extends across southern Luce and western 
 Mackinac counties has already been mentioned in the discussion of 
 the Lowlands of the Northern Peninsula in Chapter I. Within 
 this belt the local relief is low, and the topography consists of ir- 
 regularly placed knobs of gentle slope and shallow basins w^hich are 
 larger than is usual in moraines of strong relief. Subsequent to its 
 formation, the moraine was covered by Lake Algonquin and, inas- 
 much as the depth was moderate in this locality, received a veneer 
 of sand over its surface. Numerous lakes stand in the basins, and 
 attention is called in particular to the Manistique lakes which are 
 situated near the western borders of Luce and Mackinac counties. 
 These three lakes are named North Monistique, Monistique, and 
 South Monistique on the Land Survey maps but are now called 
 Kound, Manistique, and Whitetish respectively, the o in Monistique 
 having been changed to a. 
 
276 
 
 INLAND LAKES OP MICHIGAN 
 
 These lakes lie in the peculiar drainage system of the Manistique 
 River, a brief description of which follows. The main stream has 
 
 T4SA/. /?/2^V 
 
 T44M /3./ZIV 
 
 T 43 Af /S /2tV 
 
 d M.'/^S 
 
 Fig. 82. Outline map of tlie Manistique Lakts, Luce and Mackinac Counties. 
 
 a general southwesterly direction and, in its upper course, seems 
 to follow the border between the moraine on the east and the 
 
LAKES OF THE NORTHERN PENINSULA 277 
 
 swampy plain on the west which rises very gradually in elevation 
 to the northwestward and extends to within a few miles of Lake 
 Superior in the vicinity of Munising'. The greater part of the water 
 of this drainage system flows southeastward in numerous parallel 
 tributaries which join the trunk stream at approximately right 
 angles. On the other hand, the tributaries from the east are not 
 only smaller but less numerous and, moreover, have the hap-hazard 
 pattern typical of the drainage of a morainic country. Round 
 and Whitefish lakes drain into Manistique Lake and thence to 
 tlie Manistique River, forming the largest tributary entering the 
 main stream from the east. 
 
 The two larger lakes, Manistique and Whitefish, were examined by 
 the writer. They may be reached by conveyance either from Gould 
 on the Soo Line or from McMillan on the Duluth, South Shore and 
 Atlantic R. R. A railroad spur runs from iSeney to Curtis which is 
 located between the two lakes, but the train schedule was uncer- 
 tain at the time when this study wa,s made. 
 
 MANISTIQUE LAKE 
 
 Manistique is the larger of the two lakes, its area being 15.8 
 square miles as compared with 6.4 square miles for Whitefish. It 
 probably floods se^^eral contiguous morainic basins and ha;s, there- 
 fore, a very irregular shore-line. A few knobs rise above the sur- 
 face of the lake, forming small islands. The lake is very shallow, 
 the reported maximum depth being twenty-five feet. Random 
 soundings were made in the course of this work, but no depths 
 greater than twelve feet, were obtained. But, assuming the larger 
 value to be correct, it is to be expected that the larger waves 
 agitate the water to the bottom of the lake and their development, 
 therefore, is greatly impeded. That this is actually the case is 
 shown by the excessive turbidity of the water after storms. In 
 general, then, adjustments of the shores are on a small scale, even 
 though the topography is favorable to the development of forms by 
 both waves and currents. 
 
 Along the entire north shore the most consistent adjustment is 
 the submerged terrace which, however, is not sharply defined. It 
 is merely a zone of sand whose outer edge is marked by a change in 
 material rather than by a steeper slope. In many places the posi- 
 tion of the outer edge may be approximated from the reeds of cir- 
 cular cross-section which seldom grow beyond the terrace and often 
 form a fringe at the edge. On account of the gentle otf-shore slope, 
 a small amount of material has sufficed to form this terrace, and 
 
278 INLAND LAKES OF MICHIGAN 
 
 the effects of waves and currents at the shore are not pronounced. 
 At the higher land clay bluffs line the shore, as at A and along the 
 south shore, see map, Fig. 82, but along the flatter slopes the activ- 
 ity of the waves and currents seems to have been limited to the 
 formation of beaches, the material of which is well assorted. 
 
 On the west shore in the vicinity of the outlet, the off-shore slope 
 is so flat that it is doubtful if even the largest waves strike the 
 beach with any considerable force. The key to the activity in this 
 locality is the long, narrow point, B on map, at the end of which 
 currents from either direction must run directly out into the lake. 
 A small bar has been built outward from the end but is submerged 
 for most of its extent. Thus, some current action is indicated, but 
 wave action is not sufficient to pound the material above the water 
 level. The greater width of the submerged terrace on the north 
 side of the point is evidence that the currents on this side (possibly 
 return currents from the northwest shore) are the better developed. 
 This is due not only to the greater reach of the waves affecting the 
 north side of the point but also to the smoother shoreline which al- 
 lows stronger currents to develop, other conditions being equal. 
 
 The blunt projection C stands well above lake level, rising like 
 an island above the lake on the east and the swamp on the west. 
 This elevation is composed of till which furnishes coarse beach ma- 
 terial, consisting of gravel and boulders. The presence of a faint 
 terrace ten to twelve feet above the present level on the lakeward 
 slopes of this point is an indication of a higher stage of the lake. 
 This in itself is not sufficient to establish this stage but, inasmuch 
 as another fragmentary terrace at the same elevation was found 
 at the opposite end of the lake, it seems reasonably certain that 
 the lake formerly stood at this level. This point, therefore, must 
 have been a low island at that time. South of this former island 
 the vegetation of the swampy lowland is encroaching on the lake 
 and there is no definite beach. 
 
 The land bordering the south shore of the lake stands higher in 
 general than that along the shores already described, and cliffs of 
 moderate height are more frequent. Thus, from the low shore 
 south of C to F an almost continuous bluflf faces the lake at varying 
 distances from the present shore. The bluff does not follow point D 
 but extends directly across to the head of Cooks Bay where it 
 rises twenty feet above the lake. This point is a low swamp and 
 is apparently the terrace of a level of the lake which was inter- 
 mediate between the highest and present stages. From the head 
 of Cooks Bay the bluffs make a large loop eastward and return to 
 the shore about one-half mile east of point E. A small spit on 
 
LAKES OF THE NORTHERN PENINSULA 279 
 
 the west side of this point is the only deiiuite current deposit noted, 
 on this shore of the lake. The adjustments have been slight, and 
 the fact that the spit is composed of cobbles, ranging in size from 
 tliree to six inches in diameter, suggests ice as an aid at least in 
 the formation of this feature. 
 
 Beyond E the topography is undulating and causes a series of 
 broad headlands and wide embayments. Invariably the beaches at 
 the headlands are of coarse material and in the embayments they are 
 of sand. At the apex of the broad outward curve between E and F 
 the ice has pushed up a distinct ridge of boulders off shore. This 
 was probably formed by ice jams, since the lake is somewhat la'rge 
 for expansion. At F a bluff and beach of the intermediate level, 
 mentioned in connection with point D, are readily detected. The 
 alternation of headland and embayment continues along the east 
 shore and offers little of additional interest except at Gr. Here a 
 patch of the terrace of the highest level was found. 
 
 In conclusion, it seems unnecessary to add to the statement con- 
 cerning the shore adjustments made in the iutroductoiy part of this 
 description. Yet from the study it seems probable that the lake 
 has stood at two higher stages in the past, one at ten to twelve feet 
 above the present level and another at three to four feet. Obviously, 
 the highest level must have covered a muck greater area than the 
 present lake and possibly included Round and Whiteflsh. As re- 
 gards extinction, vegetation appears to be the most important 
 factor. A heavy growth of rushes forms an almost complete 
 fringe about the lake, but the accumulation of their dead parts is 
 not important as yet. The entering streams apparently bring in 
 little sediment for, if such were the case, deltas would be formed 
 at their mouths. 
 
 WIIITEFISH LAKE 
 
 .Whitefish Lake lies within a half mile of Manistique and extends 
 four and one-half miles to the southwest. The east shore winds 
 about in broad curves, but the opposite side is very irregular, 
 consisting of long points and narrow intervening bays. The long- 
 est of the points almost crosses the lake near the middle, point B 
 in Fig. 82. The lake is deeper than Manistique and, although less 
 than one-half the size (6.4 sq. mi.), shows considerably greater ad- 
 justments of the shores. Yet, as compared with those found on 
 the shores of many of the lakes of the Southern Peninsula', the ad- 
 justments are on a small scale. 
 
 At the narrows between this lake and Manistique, the low shore 
 is lined by a sand beach which soon gives way to the westward to 
 
280 INLAND LAKES OF MICHIGAN 
 
 bluffs which rise to a maximum height of tJiirty feet. Along the 
 west side the shores are poorly adjusted and little of interest was 
 found north of locality A. The sand spit at this point has been 
 built b}^ currents moving eastward along the south side w^ith but 
 slight additions from the north. The total amount of material de- 
 posited here is relatively small, a fact confirmed by the lack of ad- 
 justment of the shore to the west. In fact, the shore of the entire 
 bay to the south has suffered little change and is being encroached 
 upon by vegetation. The lack of adjustment must be due to the 
 shallow water, which interferes with tlie development of the waves, 
 and to the short reach of the strongest winds. Point B is the index 
 to the activity of the shore agents in this part of the lake and 
 the essential feature is a small sand bar which extends north- 
 eastward from the end of the point. Unquestionably here, as at A, 
 the currents along the south side of the point have accomplished 
 most but apparently are not iDOwerful. No great amount of cut- 
 ting has taken place but much of the finer material has been re- 
 moved, leaving a beach of coarse material interspersed with num- 
 erous boulders. 
 
 No adjustments worthy of description were noted between point 
 B and the south end of the lake, although the irregular shore offers 
 many opportunities for adjustment. Minor indentations at the 
 heads of the first two bays south of the point would surely have 
 been cut oft' if currents had been active along these shores. Ob- 
 viously, with no currents developed in the bays, spits would not be 
 formed at the headlands. 
 
 At the south end, the beach becomes sandy and has a more even 
 curvature, which is suggestive of better adjustment. Weak cur- 
 rents are driven eastward along this shore and ^wing out into the 
 lake at C on the east side. A small spit is growing here but has not 
 reached the islands which stand off shore. Inasmuch as the spit is 
 located south of the main bend in the shore line, currents from the 
 north have not contributed to its formation. Aside from this spit 
 the shore features on the east side of the lake are of little interest. 
 The alternation of sand and boulder beaches at the embayments and 
 headlands respectively persists to the outlet. As mentioned pre- 
 viously, the boulder clay is covered with a veneer of sand. At the 
 embayments the sand is not removed but merely adjusted along the 
 beach. At the headlands wave action Is sufficient to remove the 
 sand and enough of the finer material of the till to concentrate 
 the boulders on the beach. 
 
 In general, wave action is the predominant force in the adjust- 
 ment of the shores of this lake. In the more exposed locations 
 
LAKES OF THE NORTHERN PENINSULA 281 
 
 rather i)roininoiit dills have been cut. The quarried material, liow- 
 ever, was very hirgely distributed on tlie otf-shore slopes, forming a 
 zone of sand which corresponds to a distinct submerged terrace. 
 Bushes have established themselves on the terrace quite generally 
 but are frequently limited to a thin fringe at the outer edge, giving 
 a ready means of determining its location. No evidence of the 
 higher levels noted on Manistique Lake was found, although the 
 highest level of Manistique must surely have flooded this lake. 
 
 INDIAN LAKK 
 
 Indian Lake lies in the drainage of the Manistique river, forming 
 a catch basin for the Indian River, whose head waters lie more 
 than twenty-five miles to the northwest. It empties into the Manis- 
 tique about two miles above the mouth through a short outlet 
 which is navigable for small craft in its upper part. Thus the lake 
 may be reached by boat from the outskirts of the city of Manistique. 
 
 Indian Lake is five and three-fourths miles in length and has an 
 area of tliirteen square miles. Thus, the average width is two and 
 one-fourth miles but the maximum width, measured from the outlet 
 to the head of Big Spring Bay, see Fig. 83, reaches nearly double 
 this figure. From one vicAvpoiut this lake may be considered a la- 
 goon, for it was isolated from the Lake Michigan basin by. a short 
 sand bar which bridged the rather narrow connecting channel. On 
 the other hand, the lake basin lies directly transverse to a pre- 
 glacial ridge of rock which is thinly covered with glacial material. 
 Limestone outcrops in Manistique, at Millers Point on the east 
 side of the lake, and on the west side between Silver Creek and 
 the south end. From this it seems probable that the Nipissing bar, 
 which cut off this lake, extended between rock buttresses. 
 
 The rocks in this region are stratified and dip towards the center 
 of the Southern Peninsula, that is, slope down to Lake Michigan. 
 The trend of the layers at the surface is, therefore, approximately 
 parallel to the Lake Michigan shore. Previous to glacial times, 
 harder rock layers came to the surface in this vicinity and formed a 
 somewhat elevated ridge, or cuesta (see Chapter I), Avhile on either 
 side broad troughs were formed by the more rapid erosion of the 
 weaker layers. In the vicinity of Indian Lake a stream flowing in 
 a southeastward direction crossed the cuesta through a narrow 
 gap. It so happened that the movement of the glacier was also 
 towards the southeast, that is, along the course of the stream which 
 flowed through the gap, and the valley may have been enlarged in 
 this way. The enlargement of the valley by the glacier was more 
 pronounced in the softer rocks north of the gap than at the gap, 
 
282 
 
 INLAND LAKES OF MICHIGAN 
 
 and aT)road basin with a relatively narrow outlet to tlie south was 
 formed. After the melting of the glacier, the basin was partially 
 filled by the sand plain which lies north of the lake and the southern 
 end was closed by a bar. 
 
 In the journey up the outlet which flows over bare rock in 
 places, it is interesting to note that the channel has been deepened 
 
 7r42/\/.-/3/€ lA/. 
 
 Fig. S3. Outline map of Indian Lake near Manistique, Schoolcraft County. 
 
 approximately three feet. This arouses expectations of a higher 
 level of the lake which is not apparent along the shore from the 
 outlet to Millers Point. This stretch of shore is faced by a low cliff 
 of limestone rubble. The rock is a thinly bedded, highly fractured 
 limestone and has been greatly disturbed by the push of ice on the 
 face of the cliff. Around Millers Point, however, the cliff recedes 
 and between it and the shore stands a well-defined terrace which 
 
Miclii.:;an ( Icolosifiil and 
 J'.iolojiii-al Survo.v 
 
 rulilication ;jO. floc.hmical Series 25, 
 I'late XVII. 
 
 A. ICE KA.Ml'Ai; I', INDIAN LAKE, 
 
 B. VEGETATION, LAGOON AND BAR, NEAR INDIAN LAKE. SCHOOLCRAFT COUNTY. 
 
LAKES OF THE NORTHERN PENINSULA 283 
 
 corresi>oii(ls in elevation with tlie outlet l)efore it was deepened. 
 At the point of depaitnre of the clilTs from the present shore a 
 strong rampart of live to six feet in height has been pushed up by 
 ice. Northward the rampart splits into two distinct ridges, one of 
 which folloAvs the present shore. The other swings back from 
 the shore and extends to Orrs Point as a ridge on the terrace. Near 
 tlie northern end the material of this ridge becomes clear sand 
 and the front slope flattens, or in other words, the rampart merges 
 into a bar. 
 
 Back of the cliffs of the higher level mentioned above, the land 
 slopes so gently towards the lake that it appears flat. A few higher 
 spots rise above this flat and in each case the foot of the normally 
 gentle slope has been carved into a low cliif . This is interpreted 
 as a terrace and shore of a still higher level of the lake, which stood 
 higher than the Nipissing stage of the Great Lakes, but below the 
 Algonquin, Tlius, evidence is at hand to establish two higher stages 
 of the lake, the Ujiper and the Intermediate. The activity during the 
 Upper stage was largely that of waves in this vicinity and resulted 
 in the formation of low cliffs and a well-defined terrace. During the 
 early stages of the Intermediate Level, wave action likewise pre- 
 dominated and similar forms developed. Later, however, conditions 
 changed, due possibly to a slight lowering of the level, and cur- 
 rents, aided by ice, built a rampart-bar along the terrace, which 
 served as a breakwater for the cliffs to the rear. At the present 
 level, waves and currents are engaged in removing the finer ma- 
 terial from the beach, leaving the coarser to be pushed into a ram- 
 part by the ice. 
 
 At Orrs Point rock again outcrops and the shore is lined with 
 angular fragments of limestone. North of this point this coarse 
 material is piled into one of the best ramparts found on our lakes. 
 See Plate XVII, A. The rampart rises fully six feet above the 
 lake and is a sharp ridge with steep front and rear slopes. 
 
 At the point of deflection of the shore to the northwest, A on map, 
 the cliffs turn to the northeast and the high land drops to a wet, 
 grass-covered plain which rises barely above the present lake level. 
 This swamp borders- the entire north end of the lake as far as the 
 south shore of Big Spring Bay and extends for miles to the north, 
 rising with imperceptible slope. Obviously, the lake covered a con- 
 siderable part of this lowland during the higher levels but no at- 
 tempt was made to determine its limits. Sand ridges, which may 
 be either spits or off-shore bars, were noted crossing the lowland, 
 but the only one traced was the spit which extends from the cliffs 
 
284 INLAND LAKES OF MICHIGAN 
 
 at A to Smith Creek. The present shore is mucky and has no 
 definite beach. 
 
 Indian River enters the lake just north of Big Spring Bay and, 
 in fact, has been instrumental in forming the north shore of this 
 haj. Great quantities of silt have been brought to the lake by this 
 stream and deposited in a typical delta across which runs a num- 
 ber of distributaries. The shore line has been built out in this 
 way and the delta is encroaching on Big Spring Bay from the north. 
 In addition to the exposed part of the delta, a broad submerged 
 platform extends fully a mile off shore. Little sediment is brought 
 in at low water stage under present conditions but there is a 
 possibility of large additions to the submerged portion of the delta 
 at least during times of flood. 
 
 Big Spring Bay takes its name from a large spring nearby. This 
 spring has no effect on the lake shore but as the ^'wonder point" 
 of the vicinity deserves mention. It is a large pool about 200 feet 
 by 100 feet in surface extent and nearly fifty feet deep. The water 
 issues with considerable force through a fracture in the bottom and 
 is so transparent that the cloudlike effects of the sand fountain at 
 the bottom may be clearly seen. 
 
 The west side of the lake from Big Spring Bay to the south end 
 is bordered by high ground, and the prominent shore features are 
 the cliff and terrace of the Intermediate Level. The limestone out- 
 crop, mentioned in the introductory remarks on this lake, occurs 
 at the small point about one mile north of the south end of the lake. 
 
 Carpenters Bay is lined with an adjusted sand beach which has 
 been thrown up into a storm beach in places. A lagoon of about 
 fifty feet in width stands to the rear and in turn gives way to the 
 beach of the Intermediate Level. These beaches run to the dunes on 
 the east side of the bay. The dunes are composed of the sands of 
 the Upper Level bar which cut off the Indian Lake basin and are 
 situated farther inland than the original bar. From this locality 
 south to Lake Michigan, the material is all beach sand and is dis- 
 tributed in a succession of topographic forms similar to those found 
 on the Mpissing bar at Brevort Lake. Thus the gentle slope be- 
 tween the dunes at the east side of Carpenters Bay and Lake Mich- 
 igan consists of a series of nearly parallel bars and narrow lagoons. 
 The swamp condition of the lagoons is interrupted by the bars, and 
 the two forms may be readily differentiated by the vegetation which 
 consists of lanes of swamp grass between lines of evergreen trees, 
 the latter marking the bars and the former the lagoon. The con- 
 trast between the two types of vegetation may be appreciated from 
 Plate XVII, B. The number of these bars was not determined but, 
 
LAKES OF THE NORTHERN PENINSULA 285 
 
 judging from the i)late, must be large. It is probable that those 
 nearer Lake ]\Iiehigan run ai)proximately parallel to the Michigan 
 shore, but in the vieinity of Indian Lake there is a tendency for 
 them to swing to the shore south of C from both directions, that 
 is, to come together and form a ^'. Those from the northeast are 
 the stronger and are truncated in some instances at the present 
 shore, forming slight projections of the shore line and increasing 
 the height of the clitt's. From the vicinity of C to the outlet, the 
 bars follow the direction of the jjresent shore. The growth of these 
 bars in a northeasterly direction has forced the outlet to the north 
 side of the depression through which it flows and accounts for the 
 southward bend of the stream after leaving the lake. Obviously, 
 these bais were fom.ed after the isolation of the Indian Lake 
 basin. The tendency of the exposed bars along this part of the 
 lake to come to the shore south of C is puzzling. The shore is 
 exposed to winds having a westerly component and, on the average, 
 the stronger currents flowed to the northeast. The material car- 
 ried by the currents must have been derived, from the south, inas- 
 much as the outlet Avould have prevented acquisitions from the op- 
 posite direction. Some source of supply of material near point C 
 seems essential and the large otf-shore shoal in this locality is at 
 least suggestive. 
 
 Summing up, it may be stated that two higher levels may be 
 recognized from a study of the shores of Indian lake. Also the 
 outlet is dammed at present and the water is somewhat higher 
 than formerly. The flooded condition, however, is not serious 
 and has resulted in but a slight increase in the activity of the shore 
 agents. During the highest level, the lake was much larger than 
 at present, the principal.extension being to the north and northeast. 
 During this stage, the lake was separated from the main lake, and 
 its northern border very probably adjusted by the development 
 of bars across the sand plains. The latter was not determined 
 by the writer, but the presence of sand bars in proximity to the 
 present shore substantiates the inference. The subsidence to the 
 Intermediate stage caused a large reduction in size and brought 
 the shores near their present position, except at the north end. 
 However, in this locality important adjustments were made by 
 current action, so that there was little change in the outline of 
 the lake after the final subsidence. The activity of ice on the shores 
 of the Intermediate and present levels at least is clearly shown 
 and the ramparts thus formed are among the best found by the 
 writer. This lake, especially on the northeast shore, is well beyond 
 the maximum limit of size for ice expansion, and ice jams, there- 
 
-286 INLAND LAKES OF MICHIGAN 
 
 fore, have exerted the shove. It is evident that sedimentation by 
 streams from the north has greatly reduced the extent of this lake 
 in the past but is of less importance at the present time. The lake 
 is of sufficient size to act as an efficient settling basin, so that 
 there is little probability of rapid deepening of the outlet, ignoring 
 the presence of the dams. A slight revival of activity has resulted 
 from the obstruction of the outlet and modification due to this may 
 be anticipated in the future. 
 
 HURON MOUNTAIN LAKES 
 
 In general, the surface of Michigan stands at a relatively low 
 elevation. In the Northern Peninsula the western part, or High- 
 lands, rises considerably above the eastern Lowlands and includes 
 the highest elevations in the State. Two areas only, the Porcu- 
 pines- and Hurons, rise above 1800' feet and have been dignified by 
 the term mountains. The appropriateness of this appellation de- 
 pends largely on the viewpoint. Speaking more particularly of the 
 Huron Mountains, the summits certainly do not tower ajjove the 
 land to 'the south, but from the north they rise sharply a thousand 
 feet and more above the level of Lake Superior. Such elevations 
 may seem insignificant to one accustomed. to the lofty, snow-capped 
 peaks of our western mountains but to the plains-dweller or navi- 
 gator might appear formidable. The writer is unaware of the 
 history of the naming of the Huron Mountains but suggests that, 
 perhaps to the explorer whose paths are guided largely by the water 
 ways, the name mountains may not have seemed inappropriate, 
 much less ridiculous. 
 
 The so-called Huron Mountains consist, of a number of hard rock 
 knobs, bare or sparsely covered with vegetation, which form the 
 western terminus of a narrow belt of ancient cr3^stalline rocks ex- 
 tending from the vicinity of Marquette to the Huron River. This 
 belt outcrops along the shore of Lake Superior for about ten miles 
 above Marquette and then extends slightly more to the west than 
 the lake shore, leaving a narrow coastal strip which is underlain 
 by the brown Lake Superior sandstone. The mountains rise above 
 the coastal strip, which is considered a part of the Lowlands of 
 the eastern half of the Peninsula, and were never reduced to the 
 level of the ancient peneplain which was formed in this region. The 
 last great episode in the geological historj^ of this region was its 
 erosion by glaciers. The entire surface of the land was covered by 
 the ice and the most important work in the Huron Mountain area 
 was degradational. By this the soil was removed and the 
 rock surfaces were smoothed and rounded. Thus, the elevations 
 
LAKES OF THE NORTHERN PENINSULA 287 
 
 were fashioned iuto kuobs and the valleys into broad depressions 
 with gently nndulating floors. Deposition by the ice occurred 
 largely south of Ihe mountains and was sligiit within the mountains 
 and on the lowland belt. On tlie coastal belt thin deposits of till 
 and sands of Lakes Algonquin and Nipissing are present", but in the 
 mountains the veneer of disintegrated material, where present, 
 seems barely sufficient to support the heavy forest growth. 
 
 The region has suffered from forest fires and but little 
 virgin forest still remains. The country is as much a wilderness as 
 may be found in the State and furnishes ideal conditions for the 
 get-back-to-nature recreatiouists who desire a complete change. 
 Quite naturallj^, this locality has been selected b}^ the Huron Moun- 
 tain Shooting and Fishing Club for its private grounds. The fascin- 
 ation of this region need not be analyzed but it 'may be ventured 
 that not the least important factor is the numerous small lakes. 
 Ten of these are named on the map issued by the club and none 
 exceeds two square miles in area. They may be divided into two 
 groups: Those lying in basins within the mountains and those on 
 the piedmont, or on the lowland belt. Mountain and Ives lakes 
 are described here as illustrations of the former type, and Conway, 
 Pine, and Kush of the latter group. 
 
 Mountain and Ives lakes, see Fig. 84, stand more than a hundred 
 feet above those lying on the lowland, and their outlets cascade 
 steeply to the lower level, exposing the underlying rock foT part of 
 their courses. They appear to stand in rock basins which were 
 gouged out by the ice and later partially covered with drift or 
 lake deposits of Algonquin age. The elongated form of Mountain 
 Lake is due to the fact that it rests in a valley which crosses the 
 range and extends into the thick glacial deposits to the south. The 
 character of this valley will be better understood when the geology 
 of the broad embayment between Huron River and Piwe River 
 points is known. Cliffs of brown sandstone face the Lake Superior 
 shore along these points and for some distance within the bay. They 
 then recede from the shore and converge towards Pine Lake and the 
 mouth of the valley in which Mountain Lake lies. Along the 
 Superior shore the cliffs give wa}' to a great sand deposit which 
 forms the head of the bay. This sand formation holds back Pine 
 Lake and is, in reality, a great sand bar, the north slope of which 
 consists of no less than twenty-five small bars, conforming in di- 
 rection with the Lake Superior shore. From this it seems clear 
 that this embayment formerly extended back to the Mountain Lake 
 valley and was cut off by current action during a higher stage of 
 Lake Superior, probably the Nipissing. Moreover, the elongated 
 
288 
 
 INLAND LAKES OP MICHIGAN 
 
 form and alignment of Howe, Rush and I'ine lakes approximately 
 parallel to the border of the crystalline rocks is suggestive of a 
 cross-channel at or near the contact of the crystalline rocks and 
 
 -J 
 
 I I 
 
 A 
 
 ^ 
 
 .vU^= 
 
 the sandstone. This is further strengthened by the fact that the 
 sandstone north of Howe and Rush lakes rises well above the 
 level of these lakes. It seems probable, then, tihat in pre-glacial 
 times a stream flowed northward to the Lake Superior basin, cross- 
 

 
 S5 
 
LAKES OF THE NORTHERN PENINSULA . 289 
 
 iug the crystaline rocks of the Huron Moimtaiu range and the 
 brown .sandsilone. The latter was much more easily eroded and not 
 only a wider but deeper valley was cut, causing the steep slope be- 
 tween Mountain and IMne lakes. Also tributaries developed near 
 the contact of the different formations and entered the main stream 
 from both sides in the vicinity of the western expansion of Pine 
 Lake. 
 
 ■ Conway Lake lies in a separate depression which was Ukewise 
 cut off from Superior during Nipissing time by a bar which de- 
 veloped between Pine River and Conway Points. 
 
 riNE LAKE 
 
 The Club House and Cabins of the Huron Mountain Club are 
 located on the Superior shore in the vicinity of Pine River. A 
 short walk from the Club House across the sand beaches brings one 
 to Pine Lake near the outlet. The presence of a dam prepares one 
 for the fringe of dead timber and large quantity of driftwood. 
 AYest of the outlet the waves are moderately effective and a beach 
 of even contour extends to ithe west end of the lake. Near the 
 outlet the beach material is rubble and is taken as an indication 
 of the sorting action of tlie waves. Farther west, the waves wash 
 the clear sands of the retaining bar and have formed a beach of 
 even curvature, faced by a sand cliff of varying height up to ten 
 feet. The variation in height of the clift' is due to the fact that 
 the shore does not run parallel to rthe bar. There is also a well 
 defined submerged terrace in this vicinity. The projection at the 
 vrest end of the lake is caused by a knoll of till. This material is 
 in itself more resistant than the unconsolidaited sand but, in ad- 
 dition, contains numerous boulders which accumulate on the beach, 
 due to the selective action of the waves and possibly to ice push, 
 and effectually hold up wave action. South of this projection 
 the trough which connects Pine Lake with Rush is encountered. 
 Near the inlet a low flat, "scarcely above the lake level, extends 
 for a short distance northeastward. The flat surface, black soil, 
 swamp shrubs, and dead tamaracks are adequate evidence that this 
 was formerly an arm of the lake but has been entirely filled by vege- 
 tation. See Plate XVIII. 
 
 South of the inlet itill borders the shore and little adjustment 
 has taken place. The betach is uneven and of coarse material, and 
 no submerged terrace is present. To the southeast tlie slopes are 
 flatter and the material is sandy. Currents have carried consid- 
 erable material to the southeast and deposited it in a spit which 
 runs out on the lowland surrounding the inlet from Mountain Lake. 
 37 
 
290 INLAND LAKES OF MICHIGAN 
 
 On this bar are two ridges which are probably storm beaches. The 
 projection at the mouth of the inlet appears to be a delta but is 
 badly flooded on account of the dam. Continued growth of this 
 delta would surely close the entrance to the middle arm of 'the lake, 
 Second Pine. This, however, is not probable because the waters of 
 Mountain Lake carry little sediment and the course of the outlet 
 is too short for much to be acquired after leaving the lake. The 
 present delta probably consists of iHie material eroded from the 
 channel of the outlet in the early stages of the existence of thisi 
 stream. 
 
 Within Second Pine the till soon gives way to the smooth, side 
 slopes of a prominent rock knob upon which the waves have made 
 no impression. The abrupt projection beyond the rock shore is 
 caused by a till knoll. The shore has the usual breakwater of bould- 
 ers and is but slightly affected by the waves. Nearer the entrance 
 to Third Pine abundant snags form a still more effective break- 
 water and the shore shows still less adjustment. Third Pine is a 
 narrow embayment ui^ou which small waves only are possible and, 
 consequently, little adjustment is to be expected. In addition, the 
 hard crystalline rock is exposed on most of the south shore and 
 shows no effects of wave action.' Along the north shores of Third 
 and Second Pine the material is sandy drift, and the beach is of 
 sand or fine rubble. Nevertheless, there has been little adjustment, 
 otherwise bars would have been thrown across the restricted en- 
 trance ta these basins. 
 
 As regards shore adjustments, then, this lake has little to offer. 
 This is due largely to its small size, but other factors are hard 
 rock exposures and, at present, the large quantity of driftwood. 
 The extinct embayment at the northwestern end, however, is a 
 perfect example of complete extinction by vegetation. 
 
 CONWAY LAKE 
 
 Conway Lake lies about a mile northeast of the upper end of 
 Pine Lake and is easily reached by trail. This lake is also a lagoon 
 of Lake Nipissing, which was not drained by the sinking of the 
 water level. The narrow retaining bar may readily be recognized. 
 This lake, although of good size, has no outlet and is probably 
 drained by seepage through the bar to Lake Superior or across the 
 swamp which extends from the northeast end of the lake to the 
 Salmon Trout River. The lake is evidently very shallow because 
 small waves disturb the lake to the bottom, making the water very 
 turbid, and tfliis is to be expected from the swampy lowland upon 
 which the lake stands. The only evidence of shore adjustment by 
 
LAKES OF THE NORTHERN PENINSULA 291 
 
 waves and currents found was at the soutliwest end where a small 
 indentation has been completely isolated by a sand bar. The in- 
 teresting thing here is that the lagoon is dry, although its bed is 
 apparently below the level of tihe lake. The absence of effective 
 waves and currents, liowever, has proved favorable to the preserva- 
 tion of the ramparts formed by the expansion of the ice, even 
 though the sandy material is not especially conducive to their 
 formation. An almost continuous rampart surrounds the lake 
 but never exceeds two feet in height. In places it is so small that 
 even a slight cutting by the waves would destroy it. Vegetation in 
 the form of rushes and lily pads has taken hold and one may predict 
 a relatively rapid extinction of the lake from this cause. 
 
 RUSH LAKE 
 
 The adjustment of the shores of Rush Lake seems all out of 
 proportion to the size of the lake, after a visit to Pine and Conway. 
 The lake is irregular in shape and not greatly different in size from 
 the foregoing. It, nevertheless, exhibits shore features which are 
 indicative of intense activity in spite of the fact that the eroded ma- 
 terial is to a large extent the Lake Superior sandstone. 
 
 In the vicinity of the depression through which the outlet flow;^ 
 the beach is of sand and of even curvature. To the south the beach 
 cun^es around a blunt projection, the cause of which is not ap- 
 parent at the shore. A short distance back, however, stands a 
 well-defined cliff' cut in brown sandstone, which was formed at a 
 level approximately four feet higher than the present. Inasmuch 
 as Rush Lake stands more than sixty feet above Lake Superior, 
 this must have been a transitional stage between Algonquin and 
 Kipissing. The projection, therefore, was more prominent during 
 the early stage of this higher level and was cut back by wave action. 
 .Below this point along the south shore is a beach, smooth except 
 for one small point, caused by an accumulation of boulders. The 
 beach material increases in size from sand to rubble and where 
 coarse has been pushed up into a low ice rampart which extends 
 practically uninterruptedl}^ beyond the boat house. Fringing the 
 entire south shore of this bay is a well-defined submerged terrace 
 of varying width up to fifty feet, which drops into deep water at 
 about four feet. This terrace is evidence of strong wave action 
 along this shore which, furthermore, may be corroborated on the 
 slope to the south. Proceeding back from the shore, one encounters 
 first the cliff" of the four-foot level which in places is undercut in the 
 brown sandstone. Still higher, thirty feet above the lake, may be 
 found fragments of a shore line formed during another of the transi- 
 
292 INLAND LAKES OF MICHIGAN 
 
 tional stages between Lakes Algonquin and Nipissing. An especial- 
 ly good example of an undercut cliff may be seen at this level above 
 tlie south, side of the narrows of the bay. Above the thirty-foot 
 level there appears to be a broad terrace with cliff which may be 
 the Algonquin shore. This, however, is not distinct and the inter- 
 pretation is uncertain. 
 
 The trough in which this bay lies rises gently to the west and 
 extends through to the south shore of the lake. During the four- 
 foot level the shore stood about one hundred paces to the west 
 and at the thirty-foot stage completely flooded the narrow channel, 
 making an island of the present peninsula which forms the north 
 shore of the bay. The peninsula is composed of brown sandstone 
 which has been carved into an almost continuous cliff at the four- 
 foot level along the south side, being undercut in places as much 
 as fifteen feet. The amount of recession of the shore during this 
 period is shown by the width of the exposed terrace which reaches 
 a maximum of twenty feet. For a body of water so small as this 
 lake must have been, the wave activity appears excessive, especially 
 so since its effects are negligible at the present level. Since there 
 was little difference in the size of the lake during the four-foot level 
 and at present, the excessive cutting of the former stage must be 
 attributed largely to a longer period of time during which the waves 
 were acting. 
 
 Wave cutting is also the predominant factor in the development 
 of the north shore of the peninsula. The greatest effects are found 
 at the four-foot level and in the brown sandstone, as on the south 
 side. An eastward drifting current along the south shore of the 
 main part of this lake is formed during northwest ^^blows" but is' 
 dissipated in the bay near the end of the peninsula. Thus, no de- 
 posit is found at the tip and none will be formed until the bay 
 is closed. A small sand deposit on the west side of the entrance to 
 the bay is taking the form of a spit but its growth is apparently 
 very slow. Farther west the activity of the waves is especially 
 noticeable, inasmuch as the shore of the thirty-foot level was entire- 
 ly removed by the recession of the cliff of the four-foot stage, which 
 is exceptionally high. 
 
 Near the west end of the lake the heavily timbered slopes rise 
 steeply to Mount Huron, in places making a 40° angle with the 
 horizontal, and show little erosion by the shore agents at the higher 
 levels, although hard rock was not encountered. The shore of the 
 thirty-foot level, is not distinguishable and that of the four-foot is 
 relatively faint. The lack of strong shore features along this part 
 of the lake is due, to some extent, to the short reach of the effective 
 
LAKES OP THE NORTHERN PENINSULA 293 
 
 winds, but the possibility of their partial obliteration by the slump- 
 ing of material down the steep slopes is suggested by the peculiar 
 submerged projections along this shore, which may be material 
 brought down in landslides. 
 
 At the west end of the lake a sand beach skirts the trough which 
 continues westward towards Howe Lake. This trough rises 
 above the thirty-foot leA^el, so that the Rush and Howe lake basins 
 were not connected during that stage. Along the north shore near 
 the west end, the material is a sandy till but soon gives way to 
 the brown sandstone. The four-foot level is marked by a continuous 
 cliflf as far as point B on map, but the thirty-foot shore is frag- 
 mentary, appearing infrequently as a notch in the cliffs. Along 
 this stretch of the shore the submerged terrace is narrow, varying 
 from five to twenty feet in width, but east of point B it widens to 
 more than one hundred feet, the depth at the outer edge remaining 
 at four to five feet. North of this shore stands a flat topped ridge 
 of sandstone thinly covered with glacial material, beyond which is 
 a dry trough extending parallel to that in which Rush Lake lies. 
 The ridge continues to the boat house at the northeast corner of the 
 lake, rising well above the lake level. Along this naturally smooth 
 share there was little opportunity for current deposits so that the 
 features consist almost solely of cliffs, notched at the four and 
 thirty-foot levels. However, some activity by currents from the 
 west is shown by the gradation of the material on the present 
 beach, the size diminishing toward the east end w^here the beach 
 is of fine sand. The coarser material has been pushed into local 
 ramparts by ice, but the action is evidently of but moderate in- 
 tensity. 
 
 Concluding, it may be emphasized that wave cutting has been the 
 most important factor in the adjustment of the shores of Rush lake 
 at the various levels. Furthermore, it may be stated that the re- 
 sults accomplished by the waves on this small lake, which probably 
 never exceeded one-half mile in average width and two and one- 
 half miles in length, were exceptionally great and surprisingly 
 so after a study of the nearby lakes of like size. 
 
 IVES AND MOUNTAIN LAKES 
 
 Of the two lakes within the Huron Mountains, Ives is by far the 
 simpler. This nearly circular basin was covered by the waters of 
 Lake Algonquin, but this fact is disclosed by the elevation rather 
 than by distinctive shore features. At this time its shores were open 
 to the buffeting of the waves of the main lake and, if these shore 
 features are indistinct, the adjustment at the present level must be 
 
294 INLAND LAKES OF MICHIGAN 
 
 slight indeed. As a matter of fact, where the hard rock is exposed 
 the waves have made no impression. The best defined shore feature 
 on the lake occurs on the northeast shore south of the outlet where 
 the sandy material has been washed into a shallow submerged ter- 
 race of approximately thirty feet in width. 
 
 Mountain Lake, on the other hand, is long and irregular in out- 
 line, and presents numerous opportunities for adjustments by all 
 of the shore agents. The shore materials are the hard rocks of the 
 Huron Mountain mass and glacial material, including both till 
 and sand. Obviously, the adjustments are more pronounced in 
 the glacial material. 
 
 In the vicinity of the outlet the material is sandy, and an excel- 
 lent beach and submerged terrace are prsent. The latter is more 
 than one hundred feet in width and drops into deeper water from a 
 depth of thirty inches. At the north end the sand has been cut into 
 low cliffs but towards the outlet these drop to a sand beach upon 
 which stands a low ice rampart. The outlet cuts through these 
 forms and, within one-quarter mile of the lake, begins to cascade 
 over the granite rock. 
 
 South of the outlet the material changes to till and the beach 
 contains many boulders. The shore as far as Mt. Homer is irregular 
 due to a succession of minor projections and small bays. The 
 shores consist uniformly of boulder strands at the projections and 
 sand beaches in the bays. This would indicate wave action on 
 the points and a gradual filling process by currents in the bays. 
 In addition to the effects of waves and currents may be seen some 
 excellent ramparts, formed by the expansion of the ice during the 
 winter. In fact, the rampart is almost continuous but is much 
 better developed at the projections because of the coarser material. 
 At such locations these ridges contain boulders up to three feet in 
 diameter and rise steeply to heights of several feet, the largest 
 standing seven feet above the lake. 
 
 At the foot of Mt. Homer the crystalline rock comes to the shore 
 and very little adjustment is noticeable. Where the rock slopes 
 gently to the lake, the surfaces appear as smooth as when uncovered 
 by the glacier, but on the steeper slopes a slight roughening of the 
 rock at the water level was noted. This roughening is due to the 
 breaking off of small angular blocks of the rock along fissures. It 
 seems probable that frost action has been more effective than the 
 waves in this process, inasmuch as the water deepens rapidly at 
 such locations and the fragments sink below the reach of the waves, 
 thus depriving them of the tools which are necessary for abrasion. 
 
 South of the narrows the east shore is more irregular than to 
 
LAKES OF THE NORTHERN PENINSULA 295 
 
 the north ami, iu addition to the projections of till, has two rock 
 promontories otf which lie small rock islands. The glacial ma- 
 terial is till and the beaches along such stretches of shore are 
 lined with boulders. On the rock promontories the rock exposed 
 on the north side is roughened as describes! above but that on the 
 southern exposure shows little change. The greater effectiveness, 
 of the northerl}^ winds is clearly shown here and this may be at-> 
 tributed to their long reach as well as their high, velocities. The 
 small islands off the rock points are likewise of solid rock but 
 are not outliers which have become detached from the points by 
 wave action. They are formed by ice scour and their smooth convex 
 surfaces appear so like the backs of sheep that such forms are 
 known as roches mofttomioes (sheep backs). 
 
 At the south end a forked bay extends to the southeastward. The 
 depression which causes the north fork leads to the small but in- 
 teresting Canyon Lake which was discussed in Chapter I under 
 lake basins due to faulting. The Cliff Eiver enters through a de- 
 pression at the southwestern end of the lake, and this sluggish 
 stream is depositing its sediment in a delta which, as yet, has 
 Caused no great projection of the shore. This is not necessarily 
 an indication that the delta is small for the valley of the lower 
 course of the stream is a low swamp and possibl}^ was an arm of the 
 lake which has been filled by the deposits of the stream. A similar 
 form lies at the mouth of a small stream entering the lake one-fourth, 
 mile north. In both cases the front slope of the deltas drops steeply 
 into deep water from depths of not more than fifteen inches, indi- 
 cating active growth of the delta and moderate wave activity. 
 
 The west shore is xevj similar to that on the east side but, iu 
 general, the slopes are flatter. Thus, the shores of the bays are in- 
 clined to be swampy and are fringed with alders. The beaches are 
 bouldery except in the bay which lies south of Lumber Camp Bay 
 and is connected with it by a swampy trough. The trough is sand- 
 filled and the beach of the bay in the south arm is due to the work- 
 ing over of this material rather than to current action. Northward 
 the shores are cut in till as far as Lumber Camp Bay and the beach, 
 contains many boulders. Within this ba}^, the beach, material 
 gradually reduces in size to sand at the head. This is due largely 
 to a variation in the glacial material but may result in part from 
 transportation of the finer material southward along the beach by 
 waves and currents. 
 
 Northward along the west side, conditions are similar to those 
 on the opposite side but the outline of the shore is more regular. 
 At the foot of Mt. Ida the rock outcrops on the shore but does not 
 
295 INIiAND LAKES OF MICHIGAN 
 
 form a projection. The rock shore gives way to a boulder beach 
 and this in turn is interrupted by the only current deposit on the 
 lake which can be recognized with certainty. Currents from both 
 directions, but particularly from the south, have turned from the 
 shore and built an embankment nearly two hundred feet out into 
 the lake. The exposed part is similar to a V-bar but lacks the cen- 
 tral depression. The submerged portion drops steeply into deep 
 water from a depth of thirty inches. 
 
 Within the bay at the northern end the boulder beach presents 
 little of interest. This bay is a part of the same depression in 
 which Portage Lake lies. At the north end of the lake the hard 
 rock of Mt. Huron comes to the shore and shows a very slight 
 amount of abrasion. This gives way to the sand cliff described at 
 the beginning of this discussion. 
 
 In general, it may be stated that the lake is too small to show 
 the effects of intense wave action. The presence of a submerged 
 terrace which drops at about thirty inches is an indication of 
 waves which may reach a length of eight feet and a height of less 
 than one foot. Such waves have had no elfect on this hard rock 
 in the many years they have been active. The glacial deposits, 
 however, show wear consistent with the size of the lake. The 
 greater effectiveness of the winds having northei^ly and westerly 
 components is a natural consequence of the form of the lake and 
 the prevalence of storm winds from these directions. This is seen 
 in the tendency towards stronger wave action on the east shore 
 and current action on the west. The effects are also greater In the 
 northern arm than in the southern, due to its greater size and 
 regularity in outline. The lake is in a youthful stage and a dis- 
 cussion of the possibilities of extinction seems futile. 
 
 LAKE MICHIGAMME 
 
 On the extreme western border of Marquette County lies the 
 "Big Lake," Michigamme. With an area of seven square miles, it 
 seems insignificant compared to the Great Lakes, which were well 
 known to the Indians who frequented this country, but locally, its 
 importance probably justified its Indian name. For miles about it 
 is the largest inland lake, but its claim to our interest is not based 
 on size alone. Unique in form, picturesque in location, studded 
 with numerous islands and broken by bays, this lake possesses a 
 charm equalled by few, if any, of the lakes of the State, and in 
 addition offers abundant material for physiographic study. 
 
 In shape, it resembles a large Y, spreading its arms to the west 
 and southwest a distance of about six miles from its eastern end, 
 
LAKES OF THE NORTHERN PENINSULA 
 
 297 
 
 see Fig. 85. The first view of tlie lake is usually obtained from 
 tlie railroad whose tracks follow the north shore. The part seen is 
 the island-studded north arm set among the bald hills of the Mar- 
 quette Eange, which project above the tree tops. 
 
 Although the glacier covered this region and the rounded and 
 smooth rock knobs present striking evidence of its action, it is 
 necessary to go to the rock formations and structures in order to 
 understand the origin of this peculiar basin. Th« ridge-and-valley 
 topography of this section is due to the varying resistance of the 
 
 r48M-/5J/ M 
 
 T47 M- /?.J/ W 
 
 i fines 
 
 Fig. 85. Outline map of Lake Michigamme, Marquette County. 
 
 rock formations of the Marquette Range which have a general 
 east-west trend. In this region the range is a great trough or 
 syncline which is deformed by a minor fold on the north side 
 just east of the lake, as shown in Fig. 86. 
 
 .Granite /yib/k S/amo 
 
 nuorfz/h: 5/otc 
 
 Fig. 86. Geological section across the Marquette Iron Range, showing relation of 
 Lake Michigamme to the formations. (After Van Hise and Bayley) 
 
298 INLAND LAKES OF MICHIGAN 
 
 The upturned edges ruu aloug the north side of the lake in a 
 narrow belt, but the range spreads out to the southwest, and from 
 this expansion two small troughs extend to the south and south- 
 east, the latter reaching to Republic. The different rock forma- 
 tions vary greatly in their resistance to erosion and of these the 
 Michigamme, which underlies the greater part of the lake, is the 
 least resistant. However, this formation is variable in hardness 
 and is very resistant where changed by metamorphic processes. 
 This is illustrated on the islands and along the southwest side of 
 the lake, where the rocks stand in bold hills. These rocks were 
 greatly eroded by running water previous to the invasion of the 
 ice, and a general system of east-west valleys and ridges was 
 formed. This trend is consistent along the north side of the lake, 
 but on the opposite side swings towards Republic in a broad curve 
 to the south and southeast. The valleys followed the least resistant 
 rocks and, in particular, the Michigamme formation which in itself 
 appears to be softer in proximity to adjacent formations. Thus, in 
 addition to the east-west valley in which the northern part of the 
 lake lies, a branch extended towards Republic and is now occupied 
 by the south arm and the outlet of the lake. 
 
 Such probably was the topography in its main aspects when the 
 ice advanced from the northeast. The presence of islands indicates 
 that the scouring action of the ice was able only to modify the exist- - 
 ing surface by rounding off the hills and deepening the basin. The 
 passage of the ice was across the main trend of the ridges and, 
 while its general movement was independent of the topography, 
 that of the ice border must have been greatly influenced by the 
 relief features over which it passed. Thus, it is probable that 
 the ice poured through the gaps in the ridges on the north side of 
 the lake and spread laterally into the valley, deepening it locally 
 into a basin. As the glacier was disappearing a remnant of ice 
 filled this basin and deposited great quantities of sand at the east 
 and west ends and along the outlet. In this way the Michigamme 
 basin, due largely to pre-glacial conditions, was modifie4 and iso- 
 lated by glacial action. 
 
 In a lake as irregular as Michigamme frequent adjustments to 
 wave and current action are to be expected, and the observer is 
 not disappointed in this case. Hard rock outcrops in places on 
 the shores but the greater part of the material within reach of the 
 shore agents is of glacial origin. The drift is sandy as a rule and 
 is, therefore, easily attacked. Along the sides it is thin or absent 
 but reaches considerable thickness at the ends, especially the west. 
 The adjustments may best be described in the order of a traverse 
 
LAKES OF THE NORTHERN PENINSULA 299 
 
 around the lake, begiuuing at the west end, a trip easily made by 
 hoat but offering difficulties on foot. 
 
 At the west end, the Three Lakes Eiver enters tlie coiislricted arm 
 of the lake from a' narrow sand-plain which extends to the nortli- 
 west beyond Nestoria. About a mile to the west the land rises to 
 elevations of a hundred feet above the lake and has a rolling surface 
 composed of heavy sand interspersed with large boulders. A well 
 record in this locality gave twenty feet of heavy sand underlain by 
 ninety feet of quicksand which, therefore, extends below tlie level 
 of the lake. This sand plain, although not typical outwash, is 
 broken by several pits in which small lakes and swamps lie. Bass 
 Lake in T. 48 N., R 31 W. is an excellent example and is rapidly 
 being filled with vegetation which completely encircles the lake and 
 is in the floating bog stage. Just before entering the lake the Three 
 Lakes River has deposited the sands brought from its upper stretch- 
 es and nearl}'^ filled a former bay of the lake through wliich it now 
 flows in a broad serpentine course. In addition to the sedimenta- 
 tion, vegetation is now rapidly completing the filling of the bay. 
 The bay was caused by a projection of till from the north shore 
 along which the road runs to a bridge across the narrows. The 
 lakeward side of the projection has been straightened by shore 
 action under the influence of easterly winds, which here have a 
 long sweep, and a short but complete bar cuts off a narrow swamp 
 now grown up to vegetation. Along the north side as far as the 
 town of Michigamme, the strand is of large boulders swept clear of 
 smaller material. ,tWave action is slight at present, but soundings 
 indicate a terrace of about one hundred feet in width which drops 
 into deeper water from a depth of seventeen feet. A rock bottom 
 was encountered out to depths of eight feet, but beyond this firm 
 gravel was encountered to the south side of the lake. A terrace 
 extending to a depth of seventeen feet indicates an intensity of wave 
 action far too great for a lake of this size. The writer is, therefore, 
 inclined to believe that there is present here only a cut-terrace, 
 represented by the rock-paved bottom, and that the gravels are due 
 to a strong return current at this end of the lake rather than to 
 local undertow. 
 
 The boulder beach continues around the peninsula east of the 
 town of Michigamme where it is interrupted by a railroad fill. 
 North of the railroad tracks in this vicinity precipitous rock cliffs 
 foiTu the south side of a ridge running parallel to the lake shore. 
 The base is generally of uniform elevation several feet above the 
 present lavel of the lake but is obscured by a talus of large rocks. 
 Below the cliffs is a flat upon which the South Shore tracks are 
 
300 INLAND LAKES OF MICHIGAN 
 
 built. The formation bears a striking resemblance to a wave-cut 
 cliff and terrace but the writer was unable to find further evidence 
 of a former level of the lake at this height. 
 
 Beyond the cliffs the tracks leave the shore and turn northeast 
 acro/ss a bay that is half filled with floating bog. Between this bay 
 and the indentation, designated as A on the map, the shore is of 
 sandy drift which has been cut back recently into low cliffs. The 
 recent cutting is due to a renfewal of activity caused by the arti- 
 ficially raised vv^ater level. The material of this shore has been car- 
 ried in both directions and deposited in- bars at the turnings of the 
 shore. The bar at the eastern extremity of this beach is submerged 
 but its presence is indicated by the breaking of waves over it during 
 storms, see Plate XIX, A. On the west side, however, the bar 
 stands above water but is being cut away rather than added to at 
 the present time, and has been bisected as shown in sketch, Fig. 87. 
 
 Fig. 87. Sketch showing plan of the spits on the north shore of Lake Michigamme. 
 
 The eastern side of the entrance shows much less action, indicating 
 less powerful waves from the east winds of restricted reach. The 
 north shore of the arm to the east, B on map, is likewise composed 
 of sandy drift and is cut into fresh cliffs ten to fifteen feet high. In 
 this case the material is transported to the east end of the bay 
 where, augmented by the sediment of an entering stream, it is de- 
 posited in a sand bar which cuts off the swamp to the rear. The 
 undertow has been active in carrying the sand out at the head of 
 this bay and an excellent submerged terrace has developed, the 
 surface of which was covered with well developed ripple marks 
 when seen by the writer. The point forming the south side of the 
 bay is caused by a ridge of hard rock veneered with drift and 
 shows little wave action on the bay side. The main shore of the 
 lake along this point shows active cutting, the beach varying from 
 fine sand to boulders, depending on the character of the drift. 
 This material is broken up by the waves and transported eastward 
 by the prevailing westerly winds. Otherwise we should expect some 
 deposition at the tip of the point. Along this shore, especially well 
 shown at point C, was found a wave-cut notch in solid rock about two 
 feet above the present level. The cliffs vary from two to ten feet in 
 height and are bordered by a terrace ten to fifteen feet wide at its 
 foot. At the present level the strand is lined with boulders accumu- 
 
MicliiS'an (Joolosicnl iiiui 
 'I'ioloj^ii-al Surv<'.v 
 
 I'niilication •".(». ( Ji'dlnnical Scries 2."», 
 I'latr XIX. 
 
 Mi 
 
 A. WAVES r.ltlOAKIXC OVKK SrmrKUCKI) r.AK. LAKK MICIIKJAMME. 
 
 v.. STORM I'.EACII. EAST EXI) OF LAKE M KTIIGAM.ME. 
 
LAKES OF THE NORTHERN PENINSULA 
 
 301 
 
 lated from the drift which covers the liard roclc on its lower slopes. 
 This upper level, as indicated by the notch, does not correspond in 
 elevation with the base of the high, cliffs farther to the west and 
 probably marks the higher water level of the lake. 
 
 About a mile from the east end of the lake the Bijiki River brings 
 down, for the most part in the spring, great quantities of sand 
 which is carried out into the lake in a great bar extending almost 
 to the south shore. This bar has a pronounced effect on the waves, 
 in that they are suddenly shortened in length and increased in 
 height, making a rather treacherous bit of water during storms. 
 The river is able to keep its channel open on the average but during 
 a strong west wind 'at the time the writer visited this locality a 
 sand bar was being formed across the channel. East of the Bijiki 
 River rock hills again stand near the shore. An interesting bar 
 was noted just to the east of a rock projection where the railroad 
 embankment begins. The bar which resembles a cuspate foreland, 
 extends onward in line with the shore to the "west for some distance 
 and then turns abruptly back to the shore, as shown in the ac- 
 companying sketch. Fig. 88. The enclosed lagoon is compound. 
 
 i'WiJ/lst r/. 
 
 
 
 c_„- 
 
 Fig. 88. Cuspate foreland on the north shore of Lake Michigamme. (Sketch 
 
 from photograph) 
 
 the different parts being separated by sand bars, and shows a de- 
 velopment in steps similar to the formation of a hook. This is 
 further emphasized by the heavy growth of vegetation in the older 
 part to the west. During the writer's visit a storm beach was 
 being piled up on the front of the bar and enclosed a narrow lagoon. 
 This probably was not permanent but illustrates the importance 
 of storms in the building of such features. One exceptionally 
 heavy storm may do more work than a long period of moderate 
 
302 INLAND LAKES OF MICHIGAN 
 
 winds. As may be inferred from rts shape, tlie bar lias been built 
 mainly under the influence of westerly winds but it is probable 
 that easterly and southerly winds have played some part in form- 
 ing the portion which extends from the point to the shore on the 
 east side. 
 
 Beyond the bar a railroad embankment of stone borders the lake 
 nearly to the eastern end. At this end of the lake the beach receives 
 the full force of the waves thrown up by the westerly winds, which 
 regain their form after crossing the sand bar of the Bijiki River, 
 and is a perfect curve except where 'littered with drift wood. The 
 slope is flat and a broad beach of fine sand is exposed. The force 
 with which the waves strike this shore is indicated by the presence 
 of two storm beaches which are too low to appear in the photo- 
 graph, Plate XIX, B, but are denoted by the lines of drift wood 
 resting in the shallow trenches to the landward of each beach. The 
 upper or landward storm beach was formed during higher water, 
 probably in the spring, while that next the shore was in process of 
 formation when the picture was taken. The broad submerged 
 terrace which ex'tends ofl: this shore indicates an exceptionally 
 strong undertow. 
 
 Along the southeast shore, bouldery drift is abundant 'and the 
 shore is lined with large rocks which hold up wave action. This 
 continues to the point north of the dam, D on maj), where deposi- 
 tion has increased 'the length of the point. In addition to the ex- 
 posed part, this point extends for some distance under water, as 
 shown by the growth of rushes'. At present, however, it is being 
 cut away rather than added to. In the bays on the west side of the 
 south arm little wave action takes place, but the promontories are 
 being attacked. At point E the material is sandy and the cutting 
 has been rapid. However, the currents lose their force in crossing 
 the mouth of the bay and a spit runs south from this point, con- 
 tinuing under water nearly to the opposite shore. North of this 
 point the headlands are due to hard rock which outcrops on the 
 shore in places. The cutting by the waves in such places is a 
 matter of a few inches only at the present level, due in part to the 
 resistance of the rocks but chiefly to the fact that they extend 
 steeply below the surface and the fragments quarried by the waves 
 drop into deep water and do not serve as tools. In many places a 
 higher water level which agrees wdth that found on the north shore 
 was noted. 
 
 Along the south shore of the west arm hard rock comes to the 
 shores locally but much of the beach is of boulders. Opposite the 
 town of Michigamme and lying close to the shore is Sundstrom's 
 
LAKES OF THE NORTHERN PENINSULA 303 
 
 Islrtiul, F ou map, a small island of till which is slowly beiug tied to 
 the mainlaml at the southwestern corner. The bar runs to the 
 southeast and is submerged at the present time but may have been 
 dry during low water before the dam was put in. From this point 
 to the west end of the lake the shore is composed of large, in some 
 cases huge, boulders. 
 
 The islands of the lake form one of its most picturesque features. 
 With the exception of some of the smaller otf-shore islands, they 
 are composed of the more resistant rock. The larger ones in the 
 main lake are surrounded by low rock cliffs or a boulder strand 
 but show little otf-shore deposition. They have been rounded off 
 by the glacier and, in the south arm, are elongated in the direction 
 of the ice movement. The latter may be considered as roches mou- 
 tonnees (sheep backs) partly submerged. 
 
 PORTAGE AND TORCH LAKES 
 
 The Portage and Torch Lakes under discussion are situated in 
 Houghton County and together form a very irregular shaped basin 
 which is most interesting with regard to its origin. In a general 
 way, it may be considered as consisting of two trouglis which 
 intersect 'at an angle of about 50^. See Fig. 89. The broader, or 
 main trough, runs approximately parallel to the west shore of 
 Keweenaw Bay and also to the trend of the rocks which form the 
 Keweenaw Peninsula. The narrower trough winds across the 
 peninsula in a general northwest-southeast course but has been 
 closed at the north end by current activity during an earlier stage 
 of Lake Superior. Dredging operations readily converted this 
 trough into a ship canal which affords this important copper region 
 direct shipping facilities both to the east and west. The expansion 
 at the intersection of the trough is occupied by the main body of 
 Portage Lake. Closer examination, however, shows that the main 
 trough is cousistant only from Dollar Bay on Portage Lake north- 
 eastward, and is occupied by Torch Lake and the Trap Eock River; 
 also that the continuation of this trough across Torch Lake is 
 followed by the Pilgrim Eiver and is a much less conspicuous topo- 
 graphic feature. The depression in which- the main body of Portage- 
 Lake lies runs slightly west of south, joining the Torch Lake trough 
 at an angle of about 35°. It is followed by the Sturgeon River 
 which in its lower course meanders across a valley flat formed by 
 the deposition of its heavy load of silt. Briefly stated, then, the de- 
 pressions in which these lakes lie consist of three troughs: One 
 running parallel to the Copper Range, a second which crosses the 
 
304 
 
 INLAND LAKES OF MICHIGAN 
 
 range, and a third extending almost due south from the intersection 
 of these two. 
 
 In order to make clear the manner of formation of these troughs, 
 the more important episodes in the physiographic history of this 
 
 op£R fwr^^^vj.^ 
 
 Fig. S9. Outline map of Portage and Torch Lakes and vicinity, Houghton County. 
 (rFrom U. S. Lake Survey cliart.) 
 
 region will be briefly sketched. The first great event of interest 
 in this connection was the complex folding of the ancient rocks 
 of the Lake Superior region. The axes of the folds are parallel to 
 Lake Superior, and, as a result of this, the rock layers dip beneath 
 the lake and extend in an approximately east-west direction. A 
 
LAKES OF THE NORTHERN PENINSULA 305 
 
 notable exception to the latter is the copper-bearing ^(>ck;4 of the 
 Keweenaw I'eninsula Avliicli have a northeast-sonthwest trend. 
 After the folding, these rocks were subjected to prolonged erosion 
 during which the surface was peneplaned. However, considerable 
 relief existed on the peneplaned surface, due to the differences in 
 the resistance of the rocks; also the complexities of the folding 
 caused varied topographic features among which the ridge-and- 
 valley type was prominent. The period of erosion was followed 
 by a sinking of the land below the level of the sea, during wliich 
 sediments of great thickness were deposited. These filled the valleys 
 and covered the ridges in the vicinity of Lake Superior so that, 
 when an uplift occurred, the streams flowed directly to the Lake 
 Superior depression over the surface of these nearly flat-lying rocks. 
 Then followed the removal of the greater part of the sediments by 
 erosion, exposing the former topographic features. The ridges were 
 encountered first but the trunk stream's were able to maintain their 
 courses across them. The tributaries, however, developed along 
 the former valleys in many cases'. In this way a system of trans- 
 verse drainage was imposed upon the former topography. Finally, 
 the region was invested by glaciers which modified its surface both 
 by abrasion and deposition. 
 
 Applying this sequence of events to the region under considera- 
 tion, the resistent rocks of tlie Copper Range were folded and, at 
 the close of the first period of erosion, formed a prominent ridge. 
 This ridge was buried later by sediments which also filled the de- 
 pression between it and the Huron Mountains. When the land 
 was again lifted above the sea, the drainage flowed in a northwest- 
 erly direction from the Huron Mountains across the buried Copper 
 Range. During the period of erosion which foUow^ed, the range 
 was uncovered but the trunk streams succeeded in maintaining 
 their courses across it for a time, forming numerous gaps. In 
 addition, tributaries developed along the southeastern edge of the 
 ridge. Later, streams developed from the northeastward in the 
 less resistant sediments of the Keweenaw Bay depression and di- 
 verted the head waters of the trunk streams, forcing them to 
 abandon their courses across the range. Within the Keweenaw 
 Peninsula proper but one of the gaps has been worn down to the 
 level of Lake Superior and this is occupied by Portage Lake. The 
 reason for the persistence of this particular gap is uncertain but it 
 seems probable that it follows a fault plane along which erosion 
 was more easily accomplished. 
 
 If the existence in former times of a trunk stream flowing north- 
 eastward through this gap is conceded, the troughs occupied on the 
 39 
 
305 INLAND LAKES OF MICHIGAN 
 
 one hand by Torch Lake and the Trap Rock River, and on the other 
 by the Pilgrim River may be easily accounted for as tributary 
 valleys on the upper side of the ridge. In addition, it is known 
 that the rocks on the east side of the entire range have droi:)ped 
 along a great fault and it is believed that this has had some in- 
 fluence in the location of the courses of these tributary streams. 
 
 The meager information at hand does not warrant a definite 
 conclusion as to the origin of the Sturgeon River depression. The 
 occurrence of hard rock outcrops between it and Keweenaw Bay 
 seems to be an argument for a pre-glacial valley. If siuch is the 
 case, it must have been a tributary to the trunk stream w^hich 
 crossed the peninsula through the Portage Lake gap, and its size 
 seems disproportionately large. Enlargement of this valley by 
 glacial scour is not probable for the movement of the ice which af- 
 fected the east side of the peninsula was almost directly across this 
 valley. On the other hand, the lobe of ice which ran into the Ke- 
 weenaw Bay lowland deposited moraines in a festoon which con- 
 forms more or less closely to the outline of the bay, and the possi- 
 bility that the valley stands between morainic ridges cannot be 
 dismissed without further study. 
 
 As in the case of a number of the inland lakes of Michigan, the 
 Portage and Torch Lake basins were flooded by the predecessors 
 of Lake Superior — Lakes Algonquin and Nipissing. The beaches 
 of Lake Algonquin stand more than four hundred feet above Lake 
 Superior and take one far afield. During Algonquin time the 
 greater part of the peninsula was submerged and only the tops of 
 the higher hills stood above the water, forming a chain of islands 
 which w^ere aligned in conformity to the trend of the peninsula and, 
 for the most, elongated in the same direction. The Nipissing shore, 
 however, stands but thirty-four feet above the present lakes, and 
 may be easily recognized a short distance back from the shore for 
 most of its extent. It is, however, usually beyond the reach of the 
 waves of tlie present level and is well preserved. The greatest de- 
 parture from the present outline of the lakes during this stage 
 occurred along the course of the Sturgeon River, whose lower 
 course was flooded for a distance of twelve miles. The present 
 connections with the main lake existed but were necessarily some- 
 what greater in width. The Upper Entrance (north) was closed by 
 a sand bar in Nipissing time, but the Portage Entry remained open. 
 The closure of the Upper Entrance was due not only to its lesser 
 width and more favorable shore conditions, but also to its position 
 on the west shore of the peninsula which is exposed to the full 
 force of waves driven by the strongest winds. 
 
LAKES OF THE NORTHERN PENINSULA 307 
 
 The narrow nortli arm oT Portage Lake, see Fig. 81), nowliere 
 reaches a half mile in width and, furtlierinore, is protected from 
 strong winds by the high flanking hills, so that the shore features 
 are not noteworthy. The most pronounced features are the delta- 
 like accumnlations of sand. In some cases these occur at the 
 mouths of streams and appear to be natural, but others are the 
 stamp sands from the copper mills. The accumulation of the latter 
 was rapidly obstructing tlie channel and it became necessary to 
 establish harbor limits for tliis part of the lake. At Pilgrim Point 
 the deposit is a delta of the Pilgrim River which lias shifted its 
 course on the Hat at least once. The submerged part of thiis delta 
 shows clearly a northerly drifting current along the shore south 
 of the point. The broad submerged terrace which extends eastward 
 from the point may be, in part, the deposits of such currents pre- 
 vious to the present extension of the delta of the Pilgrim, but the 
 deposition of a large quantity of stamp sand from the Isle Royal 
 mill makes this interpretation uncertain. 
 
 The records of the weather bureau station at Houghton show that 
 the prevailing storm winds in this locality are from the west, 
 northwest and north. Thus, the entire west shores of both l*ort- 
 age and Torch Lakes lie in protected positions and 'should, there- 
 fore, show less pronounced effects of the shore agencies. This is 
 well shown along the shore which extends almost due isouth from 
 Pilgrim Point to Pike Bay at the south end of the lake. Not only 
 the features of the Nipissiug shore but those of the present level as 
 well are meagerly developed. Cutting has been the predominant 
 process bnt, as a rule, has been (slight in amount, as shown by the 
 low cliffs and narrow submerged terrace. A number of minor 
 projections, which need not be specifically mentioned, occur along 
 this shore and are due to one of several causes, among which may 
 be mentioned deltas of streams entering Lake Nipissing, similar 
 forms, at the present level, old docks, and vegetation. The latter 
 protects the shores and, where removed, allows a recession of the 
 shore line by wave cutting, see Fig. 40, Mullet Lake. Obviously, 
 the projections occur where the vegetation is intact. The abundant 
 growth of rushes and, in places, lily pads is an indication that this 
 shore is nearing complete adjustment to the present conditions. 
 
 The extension of the lake far to the southward during Nipissing 
 time has already been mentioned and also a hint as to the activity 
 of the Sturgeon River been given. A better appreciation of the 
 work of Sturgeon River may be gained from a study of its lower 
 course. At its mouth stands a large delta which causes the irregu- 
 lar projection of the south shore of the lake. The stream does not 
 
308 INLAND LAKES OF MICHIGAN 
 
 split into distributaries but has shifted its course a number of 
 times in the past, causing the ragged growth of the delta. An 
 abandoned channel at the northwestern point indicates the position 
 of the stream when the growth of the delta threatened the complete 
 isolation of Pike Bay. Other abandoned channels exist southeast 
 of the j)resent mouth of the Sturgeon. Deposition at the mouth of 
 the Sturgeon in its present position will in time tend to obstruct 
 the outlet which has a feeble current. The small channel leading 
 into Pike Bay is artificial, having been cut for logging purposes. 
 In the vicinity of this cut a sudden rise in the surface of the delta 
 occurs on both sides of the Sturgeon River. This rise, although 
 slight, is interesting because it indicates two stages in the growth 
 of the delta. The higher, or up-streami, part developed during the 
 Mpissing .stage and is evidence of the great quantity of silt car- 
 ried by the Sturgeon during that time. Some idea of the amount 
 of material deposited may be gained from the submerged portion 
 of this delta alone, which fills a valley twelve miles long and more 
 than two miles in Avidth to a depth of at least twenty feet and 
 possibly double this figure, a volume estimated at twenty million 
 cubic yards. In addition, the exposed part of the delta, the extent 
 of which is undetermined, must be considered. The lower stage 
 of the delta has grown since the drop to the present level and is 
 relatively insignificant compared with the Nipissiug delta. Includ- 
 ing both the exposed and submerged parts in the estimate, it prob- 
 ably has an extent less than one-tenth the submerged part of that of 
 the Nipissing stage. 
 
 The channel through which the Portage River flows is of con- 
 siderable width and carries on its 'side slopes the well-defined cliflf 
 and terrace of the Nipissing level. The undisturbed frontal slope 
 of the terrace drops to a low swamp across which the stream winds 
 in broad curves which closely resemble meanders. At the Portage 
 Entry the river has been turned to the right before entering the 
 lake by a sand bar which grew from the bluffs to the east and has- 
 nearly closed the entrance. 
 
 Along the east side of Portage Lake the slopes are uniform and 
 gentle. Consequently, the shores are unbroken by large indenta- 
 tions but still have minor irregularities which indicate a lack of 
 perfect adjustment to waves and currents. Wave cutting pre- 
 dominates, as on the west side, and low cliffs line the shore for 
 most of its extent. The activity on this shore is somewhat greater 
 than on the west side but the effects are hardly more noticeable. At 
 present the accumulation of drift wood affords considerable pro- 
 tection to the shore. 
 
LAKES OF THE NORTHERN PENINSULA 309 
 
 The irregularities of the shore-line are greater in the narrow 
 Torch Bay than on the shores of the main lake and consist of pro- 
 jections rather than of enibayments. A number are due to artificial 
 structures but the numerous natural points in this protected arm 
 lead one to suggest that perhaps more emphasis should be placed 
 on the adjustment of the shores of the main lake. 
 
 Within Torch Lake the east shore is exceptionally smooth for 
 about one mile, a fact which may be attributed to current action. 
 The blunt projection northeast of McCallum Creek is clearly a hook 
 but formed at a level higher than the present. From this locality 
 to the north end of the lake, the shore features are similar to those 
 on the main shore of Portage Lake. lAround the depression at the 
 north end of the lake the contour of the shore has been modified 
 somewhat by the silt of the Trap Rock iRiver, but no definite delta 
 has been formed. As at the south end of Portage Lake, Lake Ni- 
 pissing flooded the lower part of the depression through which the 
 Trap Rock River now flows but extended hardly more than two 
 miles from the present shore. 
 
 The Nipissing terrace is well defined along the west shore of 
 Torch Lake and upon it are located rail and wagon roads, as well 
 as numerous stamp mills and smelters. To the sands of the former, 
 e. g., the Calumet, Hecla, Osceola and Quincy stamp isands, are due 
 the large projections of the shore line. So prominent are the pro- 
 jections that the natural shore agents on this protected shore are 
 of little effect. But on the opposite shore in the narrow part of 
 the lake below Ureux Point, the waves and currents are much more 
 efl'ective. Cliffs line the shore from the point to within a half mile 
 of the southwestern end of the lake, and for a considerable part of 
 the distance are cut in red sandstone. The material quarried from 
 these cliffs has been transported southwestward and deposited in 
 a spit at the end of the lake. The Torch Lake depression continues 
 through to Dollar Bay on Portage Lake but nowhere rises above 
 the Nipissing level. It therefore formed a second connection be- 
 tween the two lakes at that time. 
 
 Before leaving the discussion of these lakes, the shore of Portage 
 Lake from Dollar Bay to Grossepoint demands consideration. The 
 Mpissing shore features, although distinct, are on a small scale, 
 and the activity is not great at the present level. Thus, the small 
 projection opposite Pilgrim Point, a delta of Nipissing time, has 
 neither been reduced nor added to. Yet off Grossepoint there 
 exists a submerged hook, the Middle Ground, which in extent far 
 surpasses any similar feature to be found on the lake. Clearly 
 this liook is much too large to have been formed bv wave and cur- 
 
310 INLAND LAKES OF MICHIGAN 
 
 rent action on the short stretch of shore northward to Dollar Bay. 
 This discrepancy may be readily accounted for by considering that 
 the hook is an incipient form built during the Mpissing stage, 
 when the currents had full sweep of the shore northward to Ureux 
 Point on Torch Lake. 
 
 GOGEBIC LAKE 
 
 For the origin of the euphonious name of Gogebic Lake, we must 
 go back to the Indians. Its derivation, however, is uncertain, some 
 authorities suggesting that it is a contracting of agojehic, meaning 
 "rocky" or "rocky shore," and others that it comes from gogebing, 
 "^'dividing lake." The former seems the more appropriate on ac- 
 count of the rock outcrops in the vicinity of the lake, notably the 
 Alligator Head which occupies a commanding position about the 
 southwestern shore of the lake. 
 
 Gogebic Lake, see map. Fig. 90, has a total length of fourteen 
 miles, if the eastward extension at the north end is included, and 
 covers an area of somewhat more than twenty square miles. It is 
 remarkably consistent in width and is relatively free from promi- 
 nent projections and deep embayments. In fact, where widest 
 it does not exceed two and one-half miles and nowhere narrows to 
 less than three-fourths of a mile except at the ends. It extends 
 in a general direction which is somewhat east of south and departs 
 from this only at each end. The direction of the south end is al- 
 most due north-south while at the north end an abrupt bend to 
 the east occurs. These changes in direction of the lake will be better 
 understood after a discussion of the origin of this basin. 
 
 Obviously, much of the surface of a region over which continental 
 glaciers have passed is covered with a variable thickness of drift 
 which obscures the underlying rocks and increases the difficulty of 
 interpreting the pregiacial conditions. If, however, the relief of 
 the land over which the ice passed was great, as was the case in 
 this region, the depressions were quite consistently covered by the 
 glacial deposits, while the uplands received a thin veneer of drift 
 or were left bare. Nevertheless, the deposits are usually not of 
 sufficient thickness to conceal the former topography, and the gen- 
 eral features of the pre-glacial landscape may be deciphered. 
 
 The prominent topographic features of this region are two rock 
 ranges which stood well above the surrounding lowlands in pre- 
 glacial times and still form the commanding elevations. The more 
 northerly is the Copper Range which forms the backbone of the 
 Keweenaw Peninsula and extends southwestward into Wisconsin, 
 following the trend of the Lake Superior shore. Near the western 
 
LAKES OF THE NORTHERN PENINSULA 
 
 311 
 
 Mi/es 
 
 Fig. 90. Outline map of Lake Gogebic, Ontonagon and Gogebic Counties. 
 
312 
 
 INLAND LAKES OF MICHIGAN 
 
 boundary of the State it comes in contact on the south with the 
 Gogebic .Kange which is the eastward extension of the Penokee 
 Range of Wisconsin. This range has a nearly east-west trend in 
 Michigan but ends abruptly at the west shore of the south end of 
 Gogebic Lake. These two ranges thus form a westward-pointing V, 
 the southern limb of which is relatively short. See Fig. 91. The 
 
 LIMITS or 
 
 omtonagon 
 
 drain/^ge: 
 
 Mg. 91. Map of western part of the Northern Peninsula shiowing the general dis- 
 tribution of the Copper and Gogebic Ranges and also the basin of the Ontonagon 
 River. 
 
 pre-giacial drainage developed in a manner described for the re- 
 gion to the northeast (see Portage and Torch Lakes), and the 
 streams flowed across the ranges into the Lake Superior basin. 
 Thus, numerous gaps were formed, some of which are still occupied 
 by streams. Many, however, have been abandoned and are now 
 wind gaps. 
 
 In order to attain our immediate purpose, that is, the orig'in of 
 the Gogebic Lake basin, it is necessary to consider the Copper 
 
Micliiiian Geologicjil and 
 Hiolo.srii'al Survey 
 
 J'lililic.-itioii ".O, Ueolosical SiM'ieS 25, 
 Plate XX. 
 
 A. i.ooKixt; xoirni ox i.akk (lOiiKiuc. 
 
 B. JIOUNTAIN LAKE, HURON MOUNTAINS. 
 
LAKES OF THE NORTHERN PENINSULA 313 
 
 Range only, inasmuch as the basin apparently does not cross the 
 Gogebic Range. See Fig. 91. From the map it will be seen that 
 Gogebic Lake occupies a narro^y depression which skirts the east 
 end of the Gogebic Range, extends northward to the Copper Range, 
 and then turns abruptly to the east along the edge of the Copper 
 Range. The Ontonagon River which drains the lake likewise fol- 
 lows the south side of the range and extends fully twenty miles to 
 the east before breaking through. Directly north of Gogebic Lake 
 is a low wind gap through the range, see Plate XX, A, in which are 
 located the headwaters of the Iron River. From the gap this river 
 flows almost due north to Lake Superior. Furthermore, Gogebic 
 Lake is relatively shallow, and it therefore seem,s reasonablj^ certain 
 that the two depressions — the Iron River valley and the basin of 
 Gogebic Lake — were formerly continuous. If such were the case, 
 the cause of the abandonment of the gap ofifers a problem. 
 
 The solution of this problem is difficult on account of the glacial 
 deposits, in particular the moraiuic material in the bottom of the 
 gap, and will remain uncertain until further work has been done. 
 In fact, the problem of the drainage of the entire Copper Range 
 is one that the writer desires to study further. As it is, two ways 
 by which the gap may have been abandoned suggest themselves. 
 The simpler way is to account for this by glacial action. The ice 
 passed over this region in an almost southerly direction, as may 
 be seen from the striations on Pilot Rock at the south end of Gogebic 
 Lake. It seems certain that the ice must have passed directly up 
 the pre-giacial valley whose course is now marked by Iron River 
 and Gogebic Lake. Some abrasion was, of course, accomplished 
 but, inasmuch as Gogebic Lake is relatively shallow, it is felt that 
 the basin was not greatly enlarged in this way. Soundings taken 
 off Six Mile Bay show a gradual increase in depth to tifteen feet in 
 more than one-third the distance across the lake. These soundings, 
 although very incomplete, seem to indicate a flooded channel rather 
 than a definite basin. 
 
 During the recession of the ice, the border halted along the 
 heights of the Copper Range. At this time the waters to the south 
 were impounded and formed a large lake, Ontonagon, which spread 
 eastward into Houghton County but discharged through an outlet 
 to the west. Further recession of the ice, the details of which 
 we may omit, uncovered the gaps and that of the Ontonagon 
 River proved to be the lowest in the range between the Fire Steel 
 and Presque Isle rivers. This gap then served to drain the ponded 
 waters south of the Copper Range and still continues to drain 
 not only tlie area occui^ied by this lake but also a large area to 
 
314 INLAND LAKES OF MICHIGAN 
 
 the south which extends to the Wisconsin line south of Gogebic 
 Lake. During the halt of the ice, however, morainic material was 
 deposited along the range but in greater amount in the gap 
 north of Gogebic Lake than in that occupied by the Ontonagon 
 River. Thus, the jjortion of the stream flowing in the Gogebic 
 Lake-Iron River valley south of the range was diverted east- 
 ward to the Ontonagon. The lowest course open to this water lay_ 
 just south of the Copper Range, due j)robably to preglacial tribu- 
 taries which followed the south side of the range and flowed into 
 the Ontonagon on the one hand and to the Gogebic drainage on the 
 other. Obviously, the divide between these two streams originally 
 stood higher than the gap of the Gogebic drainage. Thus, when 
 the water was forced to flow across this divide by the plugging of 
 the gap north of Gogebic Lake, the lower portion of the diverted 
 part of the stream was flooded, forming Gogebic Lake. 
 
 Another conception of the origin of the basin of Gogebic Lake is 
 based on the action of pre-glacial streams. As stated previously, 
 i: iS' believed that this region was once completely covered with 
 sediments. After an elevation of the land which lifted the region 
 above the level of the sea, stream courses develojoed on the surface 
 of the sediments, flowing northwestward into the Lake s^uperior 
 basin. This drainage system was maintained for the most part as 
 the streams cut through the sediments and encountered the buried 
 i-ocks below. In this way there was imposed on the folded rocks 
 below a system of drainage, the trunk streams of which ran across, 
 or transverse, to the rock ridges. Less is known of the tributaries 
 inasmuch as tlie valleys are masked with drift, but, drawing an 
 analogy from well known regions where this type of drainage pre- 
 vails, it may be stated with some confidence that many of them 
 developed along the upper sides of the ridges. The headward ex- 
 tension of such tributaries of adjacent streams brought about con- 
 flicts for territory which eventually resulted in the formation of 
 secondary divides. The trunk streams, however, varied in their 
 ability to deepen their channels, mainly on account of differences 
 in volume, and the larger streams cut the deeper gaps. This, in 
 turn, gave an advantage to the tributaries of such streams which 
 were able to lengthen their courses at the expense of the weaker 
 streams on the opposite sides of the divides. In this way the divides 
 shifted from the larger streams towards the smaller and at 
 the same time became lower in elevation. In some cases this 
 process continued until the upper courses of the weaker streams 
 were captured by the tributaries of the stronger and the water 
 gaps abandoned. 
 
LAKES OF THE NORTHERN PENINSULA 315 
 
 According to this conception, the Ontonagon River, the largest 
 stream in Michigan which crosses the Copi)er Kange at the present 
 time, was of like magnitude in pre-glacial times. Tributaries de- 
 veloped along the south side of the Copper Range, but that worlcing 
 eastward was much the smaller on account of a conflict with the 
 drainage of the Keweenaw Bay depression. On the west side, 
 moreover, the tributary had a decided advantage over the tribu- 
 taries of the adjacent streams and steadily worked headward until 
 it captured the upper course of the stream flowing in the Gogebic 
 Lake-Iron River valley. See Fig. 92. Between this depression 
 
 Fig. 92. Map to show the change in drainage which has taken place within the 
 present basin of the Ontonagon River. Map to the left illustrates probable drainage 
 in former times, and map to the i-ight shows the series of stream captures and 
 present drainage system. 
 
 and the Ontonagon gap other cases of capture by the same stream 
 are, of course, possible. Following the capture, continued down- 
 cutting by the Ontonagon resulted in a deepening of the upper part 
 of the Grogebic Lake-Iron River valley. 
 
 The flooding of the valley to form Gogebic Lake is still to be 
 accounted for, and consideration of this phase of the problem 
 brings to our attention the abrupt eastward turn of the north end 
 of the lake. This arm is consistent in size with the remainder of 
 the lake and may be interpreted as a portion of the pre-glacial 
 valley which continued to the eastward, thus corroborating the 
 idea of stream capture. If this be the case, the flooding must be 
 due to glacial deposition in the course of the stream below Gogebic 
 Lake, although some deepening by glacial scour is possible. 
 
 With this discussion of the two ways in which Gogebic Lake may 
 have originated, we leave the subject for future work to settle and 
 
316 INLAND LAKES OF MICHIGAN 
 
 pass on to a description of tlie shores of tlie lake. The study of 
 the lake, aside from the origin of the basin, was less interesting 
 than was to he expected, for two reasons. The absence of prom- 
 inent headlands and deep embayments gives little opportunity for 
 the distribution of Avave-cut material in forms that produce strik- 
 ing results. In addition, the lake has been dammed for power pur- 
 poses and at the time of the writer's visit stood nearly three feet 
 above its normal level, obscuring to a large extent the natural shore 
 features. 
 
 The popular Gogebic Resort at the south end of the lake may be 
 reached by a five mile drive from Grogebic Station on the Chicago 
 & ^Northwestern R. R. At the opposite end, however, the Duluth, 
 South Shore & Atlantic skirts the north shore, stopping both at 
 Bergland and Lake Gogebic. The latter route is perhaps the most 
 convenient for a. physiographic study because the flooded condition 
 of the lake and the reason therefor are at once apparent. 
 
 In general, the slopes rise more steeply from the lake on the east 
 side than on the west. The relief of the lake, however, is consid- 
 erably more on the west side than on the opposite side, and this is 
 due to the existence of the heights of the Gogebic Range at the 
 south end and of a low swamp which leads westward from the 
 northwestern shore. The swamp borders the lake from Eight Mile 
 Bay to the north end, a' stretch of more than four miles. 
 
 In Bergland Bay and, in fact, along most of the north shore, the 
 flooding of this lake is very apparent. iWhere the shores are high, 
 wave action is particularly efl'ective, but along the low shores and 
 for an undetermined distance inland, the forest trees stood in 
 water when seen hj the writer. The normal shore features were 
 so obscured that little could be made out except at the sharp point 
 on the south side of Bergland Bay. This appeared as a submerged 
 sand bar upon which stood dying trees and is interpreted as a sand 
 spit in process of disintegration. It is claimed by the interests who 
 maintain the dam that the summer of 1913 was the only time since 
 the dam was built (1906) that the water remained consistently 
 high. This, however, does not mean that the effects of the revival 
 of the shore agents will not be felt, for the water stands abnormally 
 high in the spring which is a time of frequent and powerful storms. 
 The onaterial 'along the east shore as exposed by the waves is a 
 sandy till, and the topography of the land may be described as 
 gently rolling. Thus, there are minor points and bays but not of 
 sufficient prominence, as a rule, to turn currents from the shore. 
 Waves, then, are the prevailing agent of erosion and cliffs of vari- 
 able height the prominent physiographic feature. At the slight 
 
LAKES OP THE NORTHERN PENINSULA 317 
 
 projections, boulders are concentrated on tlie boacli but in the eui- 
 bayments the material Is smaller in size. The indications of cur- 
 rent action are slight indeed and are, thus, the more noticeable. 
 Those found were opposite Eight Mile Baj' a short distance north 
 of the township line and consisted of two short stretches of gravel 
 beach and a small indentation which was cut off by a bar. The 
 material of the bar was coarser towards the north end and shows 
 that the effective currents are driven southward along the shore. 
 
 The abrupt turn of the shore into iSix Mile Bay appears favorable 
 for current deposition, but none was found. East of the stream 
 which enters this bay, a well defined spit separates the low swamp 
 along the stream course from the lake, but joins the cliffs a short 
 distance south. iSouth of the bay a boulder ridge below the cliff 
 indicates strong ice shove which is not generally evident along 
 this shore. Once discovered, however, the numerous boulders lined 
 on the beach become significant. Cliffs prevail to the south end of 
 the lake but are interrupted at the mouths of the streams. Evi- 
 dence of ice push is found in places, but the most noticeable feature 
 along this stretch of the shore is the spit on the south side of the 
 mouth of Trout Brook. As was the case at Six Mile Bay, northward 
 moving currents left the shore and built a small spit which has 
 turned the mouth of the stream in the same direction. 
 
 A reversal in direction of the effective currents takes place be- 
 tween Trout Brook and the south end of the lake, for the shore east 
 of Ice House Bay is lined by a smooth sand beach which continues 
 into'the bay as a spit. Under normal conditions this spit would 
 in time close the entrance to the bay but at present is being re- 
 moved. The promontory between Ice House Bay and the Slate 
 Kiver is a hard rock knob upon which the glacier recorded the di- 
 rection of its movement by striations. These show the movement 
 to have been parallel to this part of the basin. 
 
 The high banks on the east side are wooded to the beach and 
 show but moderate activity of the waves as comj^ared with the- 
 opposite shore. Just soutli of Alligator Point a most unexpected 
 V-bar was found. This, of course, is a deposit built by currents 
 from both directions, but no reason for the currents leaving the 
 shore is ventured at present. Above Alligator l*oint a' large rock 
 outcrop of suggestive shape peers through the woo Is and has been 
 descriptively dubbed the "Alligator Head." 
 
 North of Alligator Point the shore swings to the northwest and 
 is somewhat more irregular. Nevertheless, wave action predomin- 
 ates and but one current deposit was noted, a V-bar between Pabst 
 and Bingham bays. Along tlie low shore from Eight Mile Bay 
 
■318 INLAND LAKES OF MICHIGAN 
 
 nortliward, ice ramparts are the striking feature. The singular 
 number might almost be used in referring to them, for the ridge is 
 nearly continuous and reaches a height of fully five feet in some 
 places. 
 
 In conclusion it may be stated that the most striking fact brought 
 out by the study of the shores of Gogebic Lake is the revival of the 
 activity ot the shore agents. This is of very recent date and the 
 most, pronounced effects are those produced by waves. The cliffs 
 are universally freshened unless protected by vegetation, and, shore 
 currents are; ineffectual. Examples of active deposition under the 
 present conditions are few indeed, and in some cases the forms 
 built under former conditions are being removed. In general, 
 greater activity is displaj-ed on the east shore than on the west and 
 this, is to be expected on account of its position on the lee of the 
 prevailing storm winds. 
 
 Ice shove is apparently very strong but shore conditions are 
 usually unfavorable for the formation of ramparts. On the north- 
 western shore they are very well developed but elsewliere the lining 
 of boulders on the beach is the usual occurrrnce. The way in which 
 the ice works on this lake is not known to the writer but, judging 
 from the width of the lake, both jams and expansion may be effec- 
 tive. 
 
 CHICAGON LAKE 
 
 This small lake is situated a short distance north of the Wis- 
 consin-Michigan boundary line in Iron County. It lies for. the 
 most part in a northeast corner of T. 42 N., R. 34 W., but small por- 
 tio.ns spread out into the adjoining townships, from which it flows 
 northeast to the Paint River. It is most readily reached from Iron 
 River over the excellent Iron Kiver-Crystal Falls road, which 
 crosses the outlet about a mile from the lake. 
 
 Its shape is elongated, as would be expected, for it lies in a valley 
 between the drumlinoidal hills which are characteristic of this 
 region. As may be seen from the map. Fig. 93, the valley is blocked 
 by morainic knobs across which the Crystal Falls road runs and 
 which separate this basin from the one to the north occupied by 
 Trout Lake. The lake, then, may be assigned to the class formed by 
 morainic dams to which type Fortune Lakes, situated about two 
 miles to the west, also belong. 
 
 The shores along the sides are generally high and dry and afford 
 many excellent locations for cottages which have as yet not been 
 taken advantage of to any great extent. Both ends of the lake are 
 swampy and this is especially true of the north end. The lake at 
 
LAKES OF THE NORTHTCRN PENINSULA 319 
 
 4 Aif'/cs 
 
 Fig. 93. Topographic map of Chicagon Lake and vicinity, Iron County. 
 
320 INLAND LAKES OF MICHIGAN 
 
 this end is shallow and supports a growth of rushes. Shrubs grow 
 to the water's edge and hold up any slight wave action that may 
 occur in this restricted part of the lake. The encroachment of 
 vegetation is marked in this locality and has reached the advanced 
 stage of quaking bogs which appear as flat, grass-covereil areas. 
 
 The east shore is comparatively straight with the exception of a 
 projection near the north end and is flanked by smooth, rounded 
 slopes which are steep in places. On a lake of this size wave action 
 is not intense and shore forms are not to be expected on a large 
 scale. Along the northeast shore as far as Park's farm the shores 
 are lined with a pavement of boulders which extends a short dis- 
 tance above the present water level. This gives way in places to a 
 small but definite boulder ridge, elevated about three feet above 
 the lake and a smaller amount above the narrow strip to the land- 
 ward side. The regularity of the boulders on the shore has given 
 rise to stories to the effect that Indians had paved these shores in 
 times past but this may readily be accounted for on physiographic 
 grounds. The material of the banks is a sandy till and the wave 
 action weak, so that only the sand is carried away. The boulders, 
 thus concentrated, are then shoved to the shore by ice expansion 
 which is sufficient to form definite ridges locally. The elevation 
 of the boulders indicates a higher level for the lake and this is 
 well supported by the frequent occurrence of an elevated shoreline 
 about three feet above the present level. 
 
 This higher level marks -the original position of the shores but 
 may also mark a temporary high level. Forty years ago an ex- 
 tensive beaver dam is known to have been maintained on the outlet 
 of the lake and has since been destroyed. The water may have 
 stood at the upper level during the existence of the dam but it 
 hardly seems probable that this would have been maintained long 
 enough to allow the cutting off of indentations, as is the case just 
 below Park's farm where a bar completely closed an opening at the 
 higher level. The lowering in level seems to be better accounted 
 for by the cutting down of the outlet. 
 
 At Point A, see map, an interesting change has taken place. 
 This point is incorrectly shown on the map for the tip of the point 
 is in reality a rather elongated knob with a' low, flat tract between 
 it and the main shore. On the slopes of the knob and also the main 
 shore, the strand of the upper level ma}^ be seen above the flat. 
 This indicates clearly that the knob once existed as an island, 
 separated from the mainland by shallow water. From this point 
 to the south end, the shores are high and are nipped by the beach 
 of the upper level. The extreme south end of the lake is bordered 
 
LAKES OF THE NORTHERN PENINSULA 321 
 
 by a good sand beach whicli gives way to an ice rampart along point 
 B, where the laud is higher. The west shore of the lake presents little 
 of interest until the point C is reached^ at which wave action seems 
 to be most effective. On the very tip of the point the beginnings 
 of a spit of coarse cobbles iDoiutiug to the south was noted. The 
 effective winds, therefore, come from the north and northeast. Be- 
 yond this point the shores merge into the swamp at the north end of 
 the lake. 
 
 From the description, it is clear that wave action is slight on this 
 lake. It seems to be limited to the transportation of sand, except 
 in one case, point C. The effects as to cliff formation are almost 
 negligible because of the presence of numerous boulders which 
 rapidly become concentrated on the beach and hold up the wave 
 action. The absence of a cut-and-built terrace is in keeping with 
 the slight amount of wave action and the lack of adjustment by 
 current deposition. Ice action is relatively prominent and is of 
 the expansion type, for the probability of any extensive ice jams 
 on so small a lake is slight. 
 41 
 
CHAPTER IX 
 
 LAKES OF THE KALAMAZOO MORAINIC AND OUTWASH 
 
 SYSTEM. 
 
 As stated in tlie Introduction, tlie work, the results of which are 
 given in this chapter, was undertaken in order to include a much 
 wider representation of the lakes in the various parts of the State 
 but has been confined to the Southern Peninsula. It is possible 
 that the distribution of these lakes would have been emphasized 
 more had a county amit been used as a basis for their grouping. 
 Such a unit, however, is not only artificial but difficult to adhere 
 to and it was decided to group them according to their relation to 
 the various positions of the border of the glacier during its re- 
 treat. FolloM ing this system the lakes may be referred to the three 
 morainic systems shown in Fig. 3 with a fourth class of lakes bor- 
 dering the Michigan-Huron shores. 
 
 The order of discussion is chronological in that the groups are 
 taken up in the order in w^hich they were uncovered by the ice but no 
 attempt has been made to arrange the individual lakes in each 
 group in the succession of their appearance. This grouping is 
 used largely as a matter of convenience and emphasis should 
 not be placed on the sequence for which no claim of accuracy is 
 made. 
 
 GROUP I 
 
 in this group are included the lakes which lie in that portion 
 of the State not covered by the ice at the time of formation of 
 the Kalamazoo morainic system, see 2, Fig. 3, and also those 
 within the moraine itself. They occur in Lenawee, Washtenaw, 
 Jackson, Hillsdale, Branch, Cass, Kalamazoo, Calhoun, and 
 Barry Counties, those of St. Joseph County having already been 
 described. During this stage the reentrant between the Erie 
 and Saginaw lobes of the glacier was located in northwestern Wash- 
 tenaw County. The strong relief of the surface at the angle of 
 the V formed by the junction of the two moraines presents a strong 
 contrast to the greatly pitted outwash plain that spreads out like a 
 fan to the southwest into Jackson County. Numerous lakes are 
 found both in the morainic basins and in the pits in the outwa'sh. 
 
324 INLAND LAKES OP MICHIGAN 
 
 The greater number of morainic lakes is found within Washtenaw 
 County and of these Sugar Loaf was selected as typical. 
 
 SUGAR LOAF LAKE 
 
 Sugar Loaf Lake is located somewhat more than five miles 
 north of west of Chelsea and can be reached by conveyance from 
 that city. The basin is larger and more irregular than the outline 
 of the lake suggests and consists of several connected basins. A 
 more flooded condition formerly existed and there is a probability 
 that a much greater territory was covered by this lake at that time. 
 The former shore line is well shown just east of the inlet at the 
 north side. At this higher stage some of the smaller indentations 
 were cut off by bars, a good example of which may be found a 
 short distance east of the inlet mentioned. The material for the 
 bars wa^-: carved from the bold cliffs which line this shore of the 
 lake, but was limited in amount for bars were not thrown across 
 the larger indentations and the exposed terrace is narrow. At the 
 present level the slight development of the shores is consistent 
 with the small expanse pf the lake which is not conducive to strong 
 wave action and vegetation is getting started a't the shore. This 
 lake is not so popular as some in the vicinity but cottages were 
 in the course of construction at the time of the writer's visit. 
 
 CAVANAUGH LAKE 
 
 This small lake lies four miles west of Chelsea and a mile and 
 one-half north of the Sylvan Road stop on the interurban line. It 
 is one of the most popular lakes in this section of the State. The 
 many excellent locations for cottages have been used, especially 
 on the south and east sides, and a number of excellent summer 
 homes have been built. The lake lies in one of the numerous basins 
 formed in this locality. Many are morainic basins but that of 
 Cavanaugh is a fosse. Several of the lakes in this vicinity are 
 separated by low sags which served as connections during a former 
 flooded condition of drainage. Thus Doyle Lake to the east was 
 once connected with Cavanaugh but a bar developed across the 
 opening and may have separated the lakes during the later stages 
 of the high water. This bar which swings around the southeast 
 part of the lake may serve as a starting point for a study of the 
 shores. Other features of the higher level are an exposed terrace 
 at the foot of a cliff which rises locally to considerable height. 
 
LAKES OF THE KALAMAZOO SYSTEM 325 
 
 CROOKED LAKE 
 
 Crooked Lake is located a short distance northwest of Cavanaugh 
 and, although longer, is much narrower and more irregular in out- 
 line. The east shore of the lake for most of its extent is faced by 
 steep cliffs consisting of the sand of the outwash plain west of 
 Cavanaugh Lake. On the west, however, the land rises with more 
 gentle slope to the moraine on the west. Such a trough-like de- 
 pression between an outwash plain and a moraine has been termed 
 a fosse and it seems logical, therefore, to call a lake situated in 
 such a depression a fosse lake. The lake, however, is more true to 
 type along the northern part than near the south end. The east 
 side of the lake is used to some extent for resort j^urposes and af- 
 fords good locations. This lake has also stood at a higher level but 
 the shore features are not as pronounced as on Cavanaugh. 
 
 In the extreme southeastern part of the outwash, and also of 
 Jackson County, are located two interesting lakes, Wamplers and 
 Vineyard. 
 
 WAMPLERS LAKE 
 
 Wamplers Lake lies on the line of Jackson and Lenawee counties. 
 It is not on a railroad but near the Chicago Pike and is reached by 
 automobile over good roads. This pleasing body of water extends 
 nearly two miles in an east-west direction but is less than one mile 
 in width. The lake is much visited during the summer months and 
 has a well appointed hotel in addition to numerous cottages. The 
 greater number of buildings are located on the northeast shore and 
 in particular on a split bar of a former level which stands several 
 feet above the present level. This bar developed westward from the 
 high cliff at the east end of the lake, splitting into two separate 
 bars in the vicinity of The Farm. Soon after leaving the cliff it 
 crosses a swale which formerly served as a connection with a 
 small lake basin to the east. An exposed terrace may also be made 
 out in places but this is rapidly being worn away by a revival of 
 wave cutting due to damming of the outlet. A complete study of 
 the drainage and shore conditions of this lake should prove interest- 
 ing. The basin lies at the junction of the moraine to the south 
 and outwash on the north but has the characteristics of a pit rather 
 than those of a fosse. 
 
 VINEYARD LAKE 
 
 Vineyard Lake lies a mile or more west and north of Wamplers 
 Lake and may be reached most conveniently from Brooklyn about 
 
326 INLAND LAKES OP MICHIGAN 
 
 three miles distant on the Ypsilauti, Hillsdale Branch of the New 
 York Central Lines. The lake occupies a large, irregular pit in 
 the outwash which has a length of more than two miles and a 
 width in excess of one half mile. The principal irregularities are 
 two long bays which extend to the north and to the northwestward. 
 The lake is not as attractive as many in the vicinity because of 
 the shallow water near the shore which is covered with a heavy 
 growth of rushes and weeds. This condition is due to the partial 
 exposure of a well-developed terrace, which was formed at a higher 
 level. Steep and prominent cliffs face a large part of the shores, 
 indicating strong wave action at the higher level, but most of the 
 quarried material seems to have been carried out by the undertow 
 to form the terrace, inasmuch as no bars were seen in the course 
 of the rapid study made of this lake. 
 
 CLARK LAKE 
 
 Clark Lake, situated ten miles south of Jackson on the Cincinna'ti 
 Northern K. R. which maintains a station stop at the west end of 
 the lake, is another popular summer resort in southeastern Jackson 
 County. This narrow basin — the lake measures less than one- 
 half mile in width but more than two miles in length — appears to 
 be part of an east-west drainage line Which extends eastward 
 from the lake to the Raisin River north of Brooklyn. In the imme- 
 diate vicinity of the lake the channel is flanked on both sides by 
 patches of moraine which rise above the till plain but to the east 
 runs through outwash. The writer's information as to the depth 
 of the lake is unsatisfactory but indicates relatively shallow water. 
 It is evident, however, that the basin is not true to any one type. 
 Inasmuch as it clearly is not a pit and also is not characteristic 
 of morainic basins, a process of elimination leads to its classifica- 
 tion tentatively as a sag in a till plain. 
 
 The long stretches of high ground along the sides due to the 
 presence of moraine furnish excellent locations for summer cottages 
 and the lake is well populated during the summer months. 
 
 The east-west orientation of the lake allows the stronger storm 
 winds full sweep of the lake and consequently considerable adjust- 
 ment of the shores has taken place. This adjustment occurred 
 during a level between two and three feet above the present and 
 the shore features stand a short distance back from the water. 
 These features include forms resulting from the action of waves, 
 currents and ice, the latter two frequently combining in the forma- 
 tion of a' single form. In particular, the rampart-bar at Pleasant 
 View may be mentioned. iWell-developed bars encircle the ends of 
 
LAKES OF THE KALAMAZOO SYSTEM 327 
 
 the lake, that at the east end proving useful in damming the outlet. 
 Among the more interesting features are the cuspate foreland and 
 completely enclosed lagoon in the vicinity of Eagle Point. 
 
 DEVILS AND ROUND LAKES 
 
 The moraine which flanks the outwash just discussed on the east 
 and south is of exceptionally high relief and includes the well 
 known Irish Hills which lie south of Wamplers Lake. From one 
 locality in these hills as many as seven lakes may be seen. These 
 lakes are for the most part too small to be used extensively as 
 summer resorts and were not studied. Many of the basins, how- 
 ever, are separated by very low land and, furthermore, show some 
 alignment, suggesting the problem of former drainage conditions, 
 the solution of which could not be undertaken at this time. 
 
 Among the larger lakes of this district are Devils and Round. 
 These lakes are very popular with summer visitors as maj^ be sur- 
 mised from the extensive resort at Manitou Beach at the south- 
 western end of Devils Lake. Round Lake is less than a mile in 
 diameter and is well named. Devils Lake is considerably larger 
 than its neighbor, having a maximum length of more than two 
 miles, but is much more irregular in shape, the outstanding feature 
 being a narrow arm which extends nearly a mile and one-half to 
 the northward. The lakes may be reached via the Cincinnati North- 
 ern R. R.but the greater number of visitors come in automobiles. 
 
 As regards origin these basins may be classed as morainic. That 
 of Devils Lake appears large for basins of this type. Soundings 
 may show, however, that the lake floods several basins. This was 
 the case formerly with this lake and Round. The merging of the 
 two lakes occurred when the water level stood about four feet 
 higher than at present and the important adjustments of the shores 
 took place at this stage. The most consistent feature is the par- 
 tially exposed terrace which stands at the foot of the numerous 
 cliffs. Currents also were active and deposited their loads at the 
 indentations. A striking example of this may be seen at the narrow 
 neck of land which now separates the two lakes. Currents in both 
 lakes aided in building a strong bar which extended nearly across 
 the opening and, with the lowering of the water level, divided the 
 continuous sheet of water into the present Devils and Round Lakes. 
 This locality serves as an excellent starting point for a physio- 
 graphic study of these lakes. 
 
328 INLAND LAKES OF MICHIGAN 
 
 BAWBEESE^ COLDWATER AND MARBLE LAKES 
 
 The lakes visited in Hillsdale and Branch counties are discussed 
 as a group because of the similarity of the origin of the basins. 
 They all lie in channels through which the waters of the melting 
 ice escaped to the south and southwestward. Within these channels 
 tongues of ice of uneven thickness become stagnant and were cov- 
 ered by the deposits of the streams, forming narrow strips of out- 
 wash of even surface below the general elevation of the surround- 
 ing country. The subsequent melting of the ice caused a settling. 
 or pitting of the surface which took the form of channels at a still 
 lower elevation through which the present drainage flows. Due to 
 uneven thickness of the ice tongues these lower channels are un- 
 graded, the deeper parts being tilled by the present lakes. 
 
 One of these channels runs in a southeasterly direction through 
 Hillsdale and forms the setting for a string of five lakes the largest 
 of which is Bawbeese on the outskirts of the town. This lake, 
 named after an Indian chief, is ver^^ popular locally, the south 
 shore being lined with cottages. For a lake of this size, one and 
 one-half miles in length by less than one-half mile in width, the 
 adjustments of the shore are strong. These adjustments all took 
 place at a level five feet above the present level and may be ac- 
 counted for by the sandy character of the material of the outwash 
 which is easily worked. Cliils are neither frequent nor prominent 
 but, nevertheless, strong bars were built across indentations along 
 the sides, the most prominent standing a short distance back from 
 the present shore along the south side and that connecting Wolff 
 Point on the north shore to the mainland. The outlet of the lake 
 is now dammed and the rejuvenation may best be observed in the 
 rapid cutting at Wolff Point. 
 
 About three miles west of Hillsdale a similar channel almost par- 
 allels that just discussed. Within this channel are numerous 
 small lakes whose shores show little development but whose basins 
 were formed in the same manner. A third channel runs nearly 
 north-south about one mile east of the line between Hillsdale and 
 Branch counties. Likewise in this channel are numerous lakes, the 
 largest of which are Hamlin and Long, which are too small to show 
 decided adjustments of the shores. The fourth and last of these 
 channels lies in southeastern Branch County and is much wider 
 and less defined as a channel than those just mentioned. The pits 
 are correspondingly larger and less elongated in form but are ar- 
 ranged along the course of the channel. Our interest lies in the 
 larger lakes at either end of the string, Marble and Goldwater, 
 
LAKES OF THE KALAMAZOO SYSTEM 329 
 
 the intermediate ones, Long, Mud, Bartholemew and Middle being 
 too small for special consideration. 
 
 COLDWATER LAKE 
 
 Coldwater Lake is the southern member and also the largest of 
 the chain, having a length of three and a maximum width of one 
 and one-half miles. The lake is triangular in shape and contains 
 a large island in the southeastern part. A road skirts the east 
 shore for most of its extent and forms an approach for the large 
 number of cottages which have been built on this part of the lake. 
 It is the only lake of the chain whose extensive marl deposits have 
 not been tapped for the manufacture of cement and is therefore 
 the most popular, although an eight mile drive from Coldwater is 
 necessary to reach the lake. This lake probably stood at a higher 
 level formerly but, if so, the drop in level has been not more than 
 two feet, an amount so small that the present high water stage in 
 the spring might well be confused with a definite higher level. 
 The shore features are relatively simple. Waves have been \evj 
 active in forming the steep cliffs. Most of the material carved from 
 the cliffs seems to have been distributed by the undertow rather 
 than by long-shore currents, inasmuch as a well defined submerged 
 terrace lines most of the shore and little evidence of the closing of 
 indentations by bars was found. 
 
 MARBLE LAKE 
 
 Marble Lake extends three miles south from the edge of the town 
 of Quincy on the Hillsdale-Adrian branch of the Michigan Central 
 R. R. and is therefore easily accessible by rail. This lake, although 
 of practically the same length as Coldwater Lake covers less area 
 on account of its nearly uniform width of slightly more than one- 
 half mile. It is essentially rectangular in shape except for the 
 large embayments at either end. The lake has been of consider- 
 able economic importance in that its extensive marl deposits have 
 been utilized in the manufacture of Portland cement at the Quincy 
 mill. These deposits are now virtually exhausted in this lake but 
 are being worked in the lakes to the south. The removal of the 
 marl with steam shovels has caused an abrupt drop into deep water, 
 a condition which makes the lake dangerous from the viewpoint 
 of the resorter and has worked to the detriment of the lake. The 
 workable deposits of marl appear to occur off the west shore mainly 
 and consequently the east side of the lake has not suffered from 
 artificial destruction of the shore. In addition, the east shore 
 
330 INLAND LAKES OP MICHIGAI^ 
 
 from Cedar Point southward is favored with one of the best loca- 
 tions for summer cottages the writer has seen, consisting of a flat 
 well above the water and covered with a grove of beautiful oaks. 
 This location has, of course, been "discovered" and numerous cot- 
 tages line the shore. Here also is found the key to the shore ad- 
 justments of the lake. A well defined submerged terrace extends 
 off-shore and slopes outward so gently that a zone of considerable 
 width is exposed during low water. Great quantities of small 
 shells which have contributed to the formation of the marl may be 
 found on the beach, thrown there by the waves. An almost 
 continuous cliff lines the shore but breaks at an elevation of five 
 feet above the present level. This cliff with the occurrence of com- 
 plete bars at the same elevation crossing the occasional indenta- 
 tions is sufficient to establish a former higher level of the lake. The 
 action of currents is best shown at Cedar Point which extended 
 out into the lake as a spit at the higher level. 
 
 GOGUAC LAKE 
 
 Goguac Lake is located just outside the southwestern limits of 
 Battle Creek. The street car system of the city has extended its 
 service to the lake which has become a' recreational center. On 
 this account and also the fact that a considerable part of the shore 
 is utilized for a public amusement resort, Willard Park and a golf 
 course and Country Club, the number of private cottages are not 
 as numerous as might be expected. 
 
 This small lake lies in a pit which in outline is roughly like an 
 hour glass and is nearly twenty-five feet below the general level of 
 the outwash. The shore is faced by steep cliffs of sandy material 
 broken occasionally by minor indentations. The adjustments of 
 the shores, therefore, were largely confined to the attack of waves 
 and the accompanying undertow which formed bold cliffs and a 
 well defined submerged terrace. In places a part of the terrace is 
 exposed at the foot of the cliff, indicating a former level between two 
 and three feet above the present. This is further shown by the ex- 
 posed bars and spits at the indentations and at the cuspate foreland 
 at Willard Park. In addition to the shore features of the lake, an 
 interesting example of the extinction of a lake by vegetation may be- 
 seen on the (Willard farm just east of the road leading to Willard 
 Park. 
 
 GULL LAKE 
 
 Gull Lake is one of the best in the southern part of Michigan. It 
 is nearly five miles in length and more than one mile in width for 
 
LAKES OF THE KALAMAZOO SYSTEM 331 
 
 most of its extent. It is located about equidistant from Kalamazoo 
 and Battle Creek but some three miles north of the main road be- 
 tween these cities. The popular means of transportation to the 
 lake is by automobile but interurban service is maintained by the 
 Michigan K. E, Co., connecting Grand Eapids as well as the 
 cities mentioned. However, it must not be inferred that the appeal 
 of this lake is limited to the locality for many of its visitors come 
 from far without the State. Cottages are numerous, especially 
 along the east shore, but, in addition, there are a number of im- 
 posing estates. The popularity of the lake, however, is not due 
 entirely to its fortunate location and large size. Another reason is 
 the excellence of the shores, due in large part to the physiographic 
 developments that have taken place. 
 
 This basin is sunk below the surface of the extensive outwash 
 plain which developed in the wide angle between the Michigan and 
 Saginaw lobes when the ice border stood a few miles from the lake. 
 The lake is large for a typical pit but may be included in this class 
 until more detailed studies are made. 
 
 One of the first observations to be* made at Yorkville which is 
 situated on the outlet is that the lake has been dammed, causing 
 what appears to be a serious flooding of the lake. The flooding 
 of the outlet is clearly in excess of eight feet and it is, therefore, 
 surprising to find the shores of the lake uniformly dry. The ex- 
 planation is soon found in unmistakable evidence of an abandoned 
 level more than six feet above the present lake at which stage the 
 major adjustments of the shores took place. 
 
 The lake furnishes a wealth of material for physiographic study 
 which could not be attempted at the time this work was done and, 
 therefore a brief description of conditions neaT Midland Park is 
 given as a key to other adjustments. At this locality the broad flat 
 upon which the cottages are built is the exposed terrace of the 
 higher level. It ends abruptly at the foot of a steep cliff, twenty 
 or more feet in height, which marks the former shore line. On the 
 terrace may be found two distinct sand bars which swing north- 
 ward towards Brytyit Point. The bar nearer the lake stands at 
 the lower elevation, indicating a halt in the lowering of the lake 
 level. 
 
 At present the adjustments are not important. Ice action is 
 perhaps the most effective but prominent ramparts are not found 
 on account of the sandy character of the material on the shore. 
 
332 INLAND LAKES OP MICHIGAN 
 
 CROOKED LAKE^ BARRY COUNTY 
 
 Of the numerous small lakes which lie in the outwash that de- 
 veloped in the reentrant angle between the Michigan and Saginaw 
 lobes of the glacier only Crooked and Wall lakes were visited by 
 the writer. Crooked Lake is located in southwestern Barry County 
 and is reached most conveniently by automobile, although the Chi- 
 cago, Kalamazoo and Saginaw R. R. passes the resort on the north- 
 eastern shore. 
 
 The lake lies near the apex of the elbow shaped outwash plain and 
 occupies either a single pit of very irregular outline or a' number 
 of connected pits which have a striking northeast-southwest orien- 
 tation. It has a length of more than four miles but consists of 
 narrow parallel channels and bays separated by islands and penin- 
 sulas so that its area is relatively small. Lower Crooked is, in fact, 
 an oblong channel which surrounds a large island and is not only 
 narrow but shallow. Little adjustment of the shores has taken 
 place and they are consequently muddy. The chief interest in 
 this part of the lake lies in the heavy growth of vegetation which 
 will eventually convert Lower Crooked into a marsh. 
 
 At the northeastern end, Upper Crooked, the lake expands to 
 more than a mile in dimensions and some adjustment of the shores 
 may be seen at a level nearly five feet above the present beach. One 
 of the best localities to observe this higher level is at Stony Point 
 which is a spur of the nearby moraine covered with a thin veneer 
 of outwash material. The end of the point is a steep wave-cut cliff 
 from which the material to form the spits on either side was carved. 
 This locality indicates the adjustments that may be expected on 
 other parts of the shore of Upper Crooked. In addition to the de- 
 velopment of the shores considerable filling by vegetation has been 
 accomplished and is especially evident on the numerous shoals or 
 "blind islands". 
 
 WALL LAKE . 
 
 Wall Lake is located two miles- northeast 04 Crooked in the very 
 apex of the outwash plain which developed in the reentrant between 
 the Michigan and Saginaw lobes. The north shore is bounded for 
 the most of its extent by moraine but from the opposite side out- 
 wash stretches southward for miles. The basin was caused by the 
 melting of buried ice and is, therefore, classed as a pit. 
 
 This small lake has a length of less than one and one-half miles 
 and a width of somewhat more than one-half mile but is interesting 
 nevertheless. In contrast to Crooked Lake its shores are broken by a 
 
LAKES OF THE KALAMAZOO SYSTEM 333 
 
 single long peninsula and it is deep, so that the waves have an un- 
 obstructed sAveep. Many adjustments of the shores, therefore, may 
 be observed and those on the north shore are selected as typical. 
 
 The most striking adjustments in this locality are the steep 
 cliffs at the headlands, the perfect bar which rises fully five feet 
 above the present water level and is responsible for the name of the 
 lake, and the decided off-shore terrace. In addition, strong ice ram- 
 parts were noted along the west side. This almost perfect adjust- 
 ment means excellent shore conditions for resort purposes and the 
 lake is deservedh' popular during the summer months. 
 
 AUSTIN GROUP 
 AUSTIN^ LONG;, WEST (pIKE) AND GOURD NECK LAKES 
 
 This group of lakes lies some seven miles directly south of Kala- 
 mazoo on the Grand Kapids and Indiana R. R., trains of which stop 
 at Austin Lake. The lakes are discussed as a group because they 
 all occupy pits in the outwash plain first mentioned in the account 
 of Gull Lake. Furthermore they all drain southward through the 
 Portage River into the St. Joseph and were once connected, with the 
 possible exception of Gourd Neck, during a former swollen condi- 
 tion of the drainage which appears to have been general in the State. 
 
 Austin Lake is the central member of the group and has a length 
 of two and one-half miles- and a maximum width of slightly more 
 than one mile. It is very shallow a fact which evidently has effec- 
 tively hindered wave action for there are few shore adjustments to 
 be found. The lake has been artificially lowered and as a result a 
 broad sand flat is exposed along the shoije. The flat is covered 
 with reeds which with other water plants are encroaching on the 
 shallow lake. The lake is popular as a fishing ground but the cot- 
 tages are built on the ueigliboring lakes which have more attractive 
 shores. 
 
 West, or Pike Lake, lies a few rods directly west of Austin and 
 is nearly circular in form, the largest diameter being somewhat 
 more than one mile. The sudden darkening in color of the water 
 off shore gives the impression of a deep lake and a decided sub- 
 merged terrace. Such is not the case, the effect being produced 
 by the change from sand to mud which marks, nevertheless, the 
 limit of effective wave action. The shallowness of the water has 
 not interfered with the wave action as on Austin and decided ad- 
 justments of the shore are to be found. An excellent and conven- 
 ient starting point for the study of these features is the resort at 
 
334 INLAND LAKES OP MICHIGAN 
 
 the east end where the land between this lake and Austin rises but 
 slightly above lake level. The most decided feature is the strong 
 sand bar at an elevation of about four feet above the lake which 
 crosses the flat between the lakes and shows that the lakes were 
 connected during a higher stage and that later West Lake became 
 an independent basin. This part of the lake is the most popular 
 and the bar, although obscured somewhat by cottages, is, neverthe- 
 less, easily recognized. Similar adjustments are to be expected 
 on other parts of the shore but probably are not as well developed 
 inasmuch as the east shore is exposed to the strongest winds. 
 
 LONG LAKE 
 
 Long Lake is situated a short distance to the northeast of Austin 
 into which it drains across a flat so low that it must have been 
 flooded at the earlier stage indicated on West Lake. In this case 
 no bar crossed the flat and the separation of the lakes must have 
 been caused by the lowering of the water level. Elsewhere along 
 the shore no evidence of current action was found but sharp cliffs 
 indicate work by waves. This, however, was not great inasmuch 
 as both bars and submerged terrace are lacking. The lack of these 
 features has not proved a deterrent influence on its use as a sum- 
 mer resort as shown by the resorts on the south and southwest 
 shores, 
 
 DIAMOND LAKE 
 
 The last lake in Group I to be considered is Diamond Lake. It 
 is located just east of the limits of Cassopolis which is the junction 
 point of the Grand Trunk and Air Line Division of the Michigan 
 Central so that no diflflculty need be encountered in reaching the 
 lake. The lake is attractive and of good size for this section of 
 the State, its dimensions being two and one-half miles in length 
 and more than a mile in width. Many cottages and costly summer 
 homes have been built along the shores near Cassopolis and on the 
 island. 
 
 From the physiographic viewpoint the lake is of little interest 
 except as to the type of basin which is one of the numerous pits 
 which break the surface of the outwash plain. The level of the 
 lake is high due to the obstruction of the outlet by a dam at Browns- 
 ville and, although the shores are not badly flooded, the waves are 
 gradually wearing back the banks and undermining trees. 
 
CHArXER X 
 
 LAKES OF THE VALPARu\ISO-CHARLOTTE MOEAINIC AND 
 0UTAVA8H SYSTEM 
 
 In Group II are included the lakes which lie in the area between 
 the Kalamazoo morainic system, 2, Fig. 3, and the Valparaiso- 
 Charlotte morainic system of Southern Michigan, 3, Fig, 3. It will 
 be noted from the figure that both the eastern and western inter- 
 lobate areas became greatly accentuated as the ice shrank back ta 
 the position indicated by the Valparaiso-Charlotte moraine during 
 which stage they existed as narrow valleys with ice walls. Tha 
 western interlobate was much more pronounced, extending from 
 the vicinity of Grand Rapids northward beyond Cadillac, a distance 
 of nearly one hundred miles, as compared with less than half that 
 distance for its eastern counterpart. Both, however, are made up 
 of a patch-work of deposits which is characteristic of such locations. 
 Between the two interlobates the formations are much more regular 
 and have an east-west trend. 
 
 WALLED LAKE^ OAKLAND COUNTY 
 
 This lake is roughly triangular in shape and has a length of one 
 and one-half miles and a maximum width of somewhat more than 
 one mile. Its shores are regular in outline with the exception of 
 one long point on the northwest shore, thus the sweep of the waves 
 is unobstructed. The lake lies for the most part in a morainic 
 depression but at the south end overflows on the till plain which 
 spreads southward from the lake. A road circles the lake and fol- 
 lows the shore on all but a part of the west side so that the shores' 
 are readily accessible. As to train service, the Jackson Branch of 
 the Grand Trunk runs within one-half mile of the north end and 
 the Detroit-Saginaw Line of the Pere Marquette has a stop at 
 Wixom, three miles west of the lake. 
 
 For physiographic study the west shore is the best, although' 
 sharp cliffs and an exposed terrace are present on the east side. 
 The sharp point on the west side is a spit which may be followed ' 
 without difficulty southward for nearly a half mile. Soon after 
 leaving the point one notes that the bar takes on the characteristics 
 of an ice rampart or wall which is so distinctive as to give the lake 
 
336 INLAND LAKES OF MICHIGAN 
 
 its name. The absence of bars at other points on the shore where 
 they might be expected leads to the conclusion that ice push has 
 been the dominant force active on the shores. The lake is very 
 popular as a summer resort and nearly all the available frontage on 
 the west, east and southeast has been built upon. 
 
 WHITMORE LAKE 
 
 Whitiiiore is another of the popular lakes of tnis group. It is 
 located ten miles north of Ann Arbor on the boundary between 
 Washtenaw and Livingston counties. Trains of the Ann Arbor 
 R. R. stop at the south end, but with the improvement of the roads 
 the automobile is the more popular means of reaching the lake. 
 
 The lake lies in a region of complicated glacial deposits and 
 many interesting features of this type as well as shore features 
 may be seen. The high hills at the north end are kames. On the 
 ^east a strip of outwash borders the lake but gives way to ground 
 moraine which extends to the south end of the lake. The south 
 side consists of a flat outwash plain, on the west of which lies a 
 narrow stretch of ground moraine followed by a moraine of low 
 relief along the greater part of the west side. The lake has neither 
 outlet nor inlets and for most of its extent the shore is bounded by 
 banks which are of moderate height but steep for unconsolidated 
 material. The basin appears to be a pit formed by a rather large 
 block of ice of irregular thickness buried by outwash which de- 
 veloped from the southward. The outwash is a broad channel 
 which carried the water escaping from the ice when the front of 
 the glacier stood ,a short distance northwest of Ann Arbor. 
 
 As to the shore features, the chief interest lies in a higher level of 
 the lake during which the water discharged through an outlet at 
 the southeastern part. After leaving the lake the former outlet 
 turns abruptly to the north and extends to the Huron valley. The 
 development of the shores at the higher level was exceptional for 
 a lake of this size and resulted not only from the activity of waves 
 and currents but of ice as well. The best locality for study is along 
 the northwestern shore where excellent examples of cliffs, spits, 
 bars and ice ramparts may be found. At present the encroachment 
 of vegetation is beginning in parts of the basin and is progressing 
 rapidly on the spit-like form extending out from the south shore 
 of the lake. 
 
LAKES OF THE VALPARAISO-CHARLOTTE SYSTEM 337 
 
 HURON RIVER GROUP 
 PORTAGE^ BASE LINE, STRAWBERRY, ZUKEY, ETC. 
 
 The lakes included in this groui) are located in the valley of the 
 Huron River southwest of Lakeland, situated at the junction of 
 the Ann Arbor R E. and the Jackson Branch of the Grand Trunk 
 R. E. They all occupy parts of an elongated pit that extends in a 
 northeast-southwest direction from Lakeland to Portage Lake. 
 This pit was formed by the burial of an ice mass of very irregular 
 outline and thickness in a former drainage channel through which 
 the water from the glacier escaped in a northwesterly direction 
 beyond Pinckney and thence to the southwest towards Jackson. The 
 subsequent melting of the ice left a depression below the level of 
 the outwash, which conformed in general outline to the ice mass 
 and also contained a number of deeper basins which contain the 
 present lakes. 
 
 The several lakes of the group are not discussed independently 
 because the adjustment of the present shores is iusignificant and 
 also because a higher level was found at an elevation sufficient to 
 have merged all into one large lake having numerous bays, penin- 
 sulas and islands. The principal indication of this stage is the 
 gently sloping flat which extends from the lake shores back to the 
 sharp cliffs and is interpreted as an exposed terrace. In addition, 
 bars which swing out from steep cliffs were found. Thus, the study 
 of the former shore will give the best results. Some of the most 
 interesting localities are the sand point on the east side of Portage 
 Lake north of the cottages, the north shore of Base Lake near 
 the outlet, and the great bar which swings southwestward from the 
 cliff on the southeast side of Bass Lake. 
 
 These lakes are used extensively as summer resorts, the most 
 popular being the pairs, Portage-Base and Zukey-Strawberry, lo- 
 cated at the ends of the chain. 
 
 DUCK LAKE 
 
 Duck Lake is located in northeast Calhoun County almost equi- 
 distant from Albion, Charlotte and Eaton Rapids, from which 
 cities it draws many summer visitors. The nearest railroad stop is 
 Springport on the Hillsdale-Lansing Branch of the New York Cen- 
 tral, but good roads have made the automobile the principal means 
 of reaching the lake. This small lake of one and one-half miles in 
 length and somewhat more than one-half mile in width lies for the 
 
 43 
 
338 INLAND LAKES OP MICHIGAN 
 
 ■ most of its extent on a till plain, the land rising to tlie moraine 
 at the sonth end. The basin is, in general, shallow and has, further- 
 more, numerous shoals. Thus, it may be considered as one of the 
 larger sags in the till plain modified by minor sags and swells. 
 
 The material of the shores is compact till except for the sandy 
 morainic material on the south end; and consequently the attack 
 of the waves has been slight. The adjustments have taken place 
 at a level four feet above the present stage and are found at either 
 end and the east side. At the south end clean sand beaches indicate 
 an assorting of the material by the waves; and a fairly well de- 
 fined submerged terrace is the result of the accompanying undertow. 
 
 At Charlotte Resort a strong bar crosses the mouth of a former 
 bay and at the north end there is the possibility of a similar form, 
 obscured by the road. Waves and currents have had little force on 
 the west shore as shown by the presence of marl beds. The -ram- 
 parts on this shore, however, show a powerful shove by ice. A heavy 
 growth of rushes on the offshore terrace indicates the beginning, at 
 least, of the extinction process by vegetation. 
 
 CHIPPEWA LAKE 
 
 A number of small lakes are situated in the large morainic tract 
 of north-central Mecosta County, and Chippewa Lake, which Jics 
 twelve miles east and north of Big Rapids, was selected as repre- 
 sentative of these lakes. There are no rail connections but a good 
 road from Big Rapids insures a comfortable trip by automobile 
 or other conveyance. 
 
 Except for a narrow strip of outwash at the north end, the lake 
 lies in a very sandy moraine of strong relief. The knobs and basins 
 of the surrounding land have counterparts in the shoals and deep 
 holes of the lake bottom so that the lake is typical of those which 
 flood several adjacent morainic basins. 
 
 The slopes to the lake are a series of cliffs and flats, thus furnish- 
 ing numerous possibilities for adjustments which were readily car- 
 ried on in the sandy material but at a level nearly five feet above 
 the present lake. In addition to the steep, wave-cut cliffs, the work 
 of currents at the entrances to bays is noticeable. Many of the 
 smaller indentations of the original lake were completely cut off 
 by bars which have been remodeled into ice ramparts in some 
 cases. However,' at the larger bays spits are found. An interesting 
 example of this was noted along the southwest end of the lake 
 where the road follows the spits which developed from each side 
 of the large bay, now nearly dry. Along the outwash at the north 
 
LAKES OF THE VALPARAISO-CHARLOTTE SYSTEM 339 
 
 eud the teiTcace of the higher level is well exposed and is from two 
 hundred to three hundred feet wide. Another of the many interest- 
 ing- features is the land-tied island of the former level on the east 
 shore. This lake has exceptionally interesting physiographic fea- 
 tures whicli, together with the sinking of the water level, account 
 for the present rather regular outline of the lake. It is becoming 
 a' very popular summer resort, the more favored locations being the 
 village of Chippewa Lake, the north shore, and the land-tied island 
 with connecting bar. 
 
 HESS LAKE 
 
 Hess Lake lies about two miles southwest of Newaygo, located on 
 the Grand Kapids-Petoskey Branch of the Pere Marquette R. R. 
 This lake has a length of nearly one and one-half miles and a 
 rather consistent width of about three-fourths of a mile. The 
 shores are regular in outline on all but the south side, which is 
 broken by a number of bays and promontories. The lake is shallow 
 — it probably does not exceed thirty-five feet in depth — and is sur- 
 rounded by outwash except at the southwestern shore. It is, 
 therefore, classed as a pit. 
 
 From the study of this lake it is apparent that the northerly 
 and easterly winds have not been effective in the adjustment of the 
 shores. The irregular south shore consists of headlands, which 
 show few effects of wave action, and mucky bays, sparsely grown 
 up to lily pads. The power of the westerly winds, however, is well 
 shown on the north shore. As usual the adjustments were found 
 between three or four feet above the present level and bars first ap- 
 pear near the middle of the north shore. The bars increase in 
 strength towards the east end where former bays have been com- 
 pletely cut off. Ice action is also powerful and ramparts of sand, 
 bound by the roots of trees, were noticed. The north shore of the 
 lake is a popular summer resort and, due to the favorable shore con- 
 ditions and its accessibility, is almost completely lined with cot- 
 tages. The south shore is reached by a roundabout route over 
 sand roads, and but one location, a camping ground, was found. 
 
 REED LAKE 
 
 The location of Reed Lake within the city limits of East Grand 
 Rapids accounts for the great popularity of this small lake of but 
 slightly more than a mile in length and less than a half mile in 
 width. It lies in a morainic basin that is irregular both as to 
 depth and outline. The shore is generally well drained except 
 
340 INLAND LAKES OF MICHIGAN 
 
 about the muddy bay at the east end. As a' recreational center 
 the lake is interesting but as an illustration of physiographic de- 
 velopments little can be said. The shore agents working in hard 
 till accomplish little on a lake of this size and the adjustments of 
 the shore are few and simple. Waves have succeeded in steepening 
 the banks and forming a moderate ofE-shore terrace, a part of 
 which is now exposed due to a drop of at least three feet in the 
 water level. Bars are found only at the smaller indentations and 
 stand back of the present shore. One example of a bar completely 
 enclosing a small bay was found in the vicinity of Pierce's Landing. 
 A heavy growth of vegetation both in the lake and on the shore 
 indicates the initiation of the extinction process. 
 
 MINER LAKE , 
 
 Miner Lake is a small lake one and one-half miles in length and 
 less than a mile in width but is, nevertheless, one of the larger 
 lakes of Allegan County. It is located about three miles northeast 
 of Allegan and lies in an irregular depression of a till plain near 
 its junction with the moraine. As is usually the case for small 
 lakes whose banks are till, the adjustments of the shore are not 
 far advanced. Inasmuch as there has been a lowering of the level 
 of the lake, the adjustments are now beyond the reach of the waves 
 and have virtually ceased to develop. 
 
 Furthermore, vegetation has taken hold and extinction appears 
 to be the next step. Aside from wave cut cliffs and a partially 
 exposed offshore terrace little of interest was noted. The lake is 
 not popular as a summer resort but is frequented mainly by fisher- 
 men. 
 
 PAW PAW LAKE 
 
 Paw Paw Lake lies within two miles of the northern border of 
 Berrien County, within a mile of the towns of Coloma and Water- 
 vliet and six miles west of Hartford. These towns are all located 
 on the Chicago-Grand Kapids Branch of the Pere Marquette. The 
 lake may .also be reached directly by interurban from Benton 
 Harbor. In addition, the main roads are excellent. 
 
 The total length of the lake is two and one-half miles but its 
 width is but slightly more than a half mile, the elongation being 
 nearly east-west. The chief irregularities are bays at the southwest 
 and northeast ends, the latter being set off from the main lake by 
 two distinct peninsulas. The peninsula on the east side and the 
 east shore of the lake near the outlet are the favored locations for 
 
LAKES OF THE VALPARAISO-CHARLOTTE SYSTEM 341 
 
 sumnner homes which, line the shore even though the shore condi- 
 tions are not uniformly good. 
 
 The lake lies on a till plain but is not tj'pical of such lakes in that 
 it is reported to be one hundred feet deep. The banks slope gently 
 to the sliore on the north side but on the south rise more abruptly 
 from the lake, due possibly to a general rise in the till plain towards 
 the moraine whose border stands about one mile to the south. 
 
 The lack of the adjustments of the shores is very noticeable con- 
 sidering the size of the lake and may be accounted for in part by 
 the fineness and compactness of the till of which the shores are 
 composed. The features found w'ere all at a higher level and con- 
 sisted mainly of sharp cliffs and an exposed terrace fronting the 
 elevations. One small spit was noticed at the west end but none 
 were seen at the other favorable localities, such as the peninsula 
 at the east end which was an island at the higher level. 
 
 In recent times wave action has been renewed and is actively 
 cutting back the shores wherever the vegetation has been removed. 
 Such a condition suggests an obstruction in the outlet, in this case 
 a dam four miles down the outlet. 
 
 This lake is a good example of the service of a lake as a reservoir 
 for during the spring floods the Paw Paw Kiver backs into the lake, 
 causing exceptionally high water and the flooding of some of the 
 low ground upon which cottages unwisely have been built. 
 
 CORA LAKE 
 
 That part of the surface of the outwash plain extending south- 
 westward through Paw Paw in Van Buren County and drained 
 by Dowagiac Creek which was not covered by glacial Lake Dowa'- 
 giac is dotted with numerous small pits. A number of these pits 
 hold water and are very tj^pical of this kind of basin, having in 
 many cases no outlets and insignificant inlets. 
 
 A group of lakes of this type is located some six miles southwest 
 of Paw Paw and of these Cora, Three Mile, Little and Big Reynolds 
 and Eagle lakes were visited. Conditions -are so similar in these 
 lakes that the brief description of Lake Cora will serve as a key 
 to the many interesting examples of shore adjustments that may be 
 found on all of them. 
 
 This lake may be reached via the Kalamazoo, Lake Shore and 
 Chicago R. R. which stops at the summer resort on the north side, 
 although the automobile is proving a more popular means of reach- 
 ing Cora Lake on account of the excellent roads. The resort, con- 
 sisting of a commodious hotel and a number of cottages, is an in- 
 
342 INLAND LAKES OF MICHIGAN 
 
 dication of the popularit}^ this small lake has eujoyed. The basin 
 of the lake consists of a main pit surrounded by numerous smaller 
 connecting pits, making a rather irregular depression. Yet the 
 shores are of exceptionally even contour, the greatest break being 
 Paradise Point, and the disparity may be accounted for by the 
 shore adjustments that have taken place. This lake which has no 
 outlet has varied in level within recent times and now stands low. 
 It has stood between three or four feet higher and at this level 
 the adjustment of the shore occurred. The change in color off- 
 shore indicates a well defined terrace, the upper part of which is 
 now exposed as a sand flat. At the higher ground steep and prom- 
 inent cliffs rise from the terrace, and at the indentations completed 
 bars are found in virtually all cases. Paradise Point is of sufficient 
 interest to warrant special mention for short bars have tied a 
 former island to the mainland. As stated earlier, similar condi- 
 tions hold for the other lakes of this group and their study should 
 prove interesting. 
 
 SISTER LAKES 
 
 Another group of pit lakes in the outwash drained by Dowagiac 
 Creek, including the' Sister Lakes (Round and Crooked), Dewey 
 and Magician lakes, was visited by the writer. They are located 
 five to six miles northwest of Dowagiac on the northern boundary of 
 Cass County and have no direct railroad connections. These lakes, 
 which are classed as pits, stand near the edge of the moraine to 
 the west and are sunk well below the surface of the thin veneer of 
 outwash, so that till is frequently exposed on the shores. Further- 
 more, the pits are so numerous that the outwash in places has a hill 
 and dej)ression topography very similar to the morainic knobs and 
 basins. Since the moraine is very sandy, these two types of 
 topography are best differentiated from the hill tops which rise to 
 approximately the same elevation within the outwash but are of 
 variable height in the moraine. 
 
 Considerable variation is found as to shore features on the 
 different lakes of the group. Crooked, Round and Dewey are similar 
 in development but Magician presents a decided contrast; 
 
 The latter is vei^^ irregular in form, having many bays and 
 points and some islands. Strong wave action is possible only on 
 the headlands and islands, and currents have little chance to de- 
 velop on the irregular shores. Consequently the adjustments are 
 limited to wave cutting in favorable localities and consist of sharp 
 cliffs and an off-shore terrace. The outer edge of the terrace now 
 
LAKES OF THE VALPARAISO-CHARLOTTE SYSTEM 343 
 
 appears as a flat at the foot of the cliffs, due to a lowering of the 
 water level. 
 
 The lake is very attractive from the scenic viewpoint and numer- 
 ous cottages line the shore. 
 
 The remaining lakes of the group, Crooked, Round and Dewey, 
 are verj- similar, except in form, and present a contrast to 
 Magician in that the adjustment of the shores by both waves and 
 currents is very decided. They have all stood at a level four to 
 five feet higher than at present and are, therefore, fringed by a' 
 broad sand flat. Improved roads lead to almost all parts of the 
 lakes and they are, consequently, developing as summer resorts. 
 
 Among the more interesting features are the very noticeable cliffs 
 fronting the high ground. Considerable recession of the cliffs has 
 taken place, and a sharp off-shore terrace is of general occurrence. 
 In addition, strong currents built bars across the mouths of in- 
 dentations and spits at some of the points. Thus, on Crooked Lake 
 a* completely enclosed lagoon was found at the east end of the lake. 
 Also the first point on the south side has been increased in length 
 by a spit, and at the west end a bar may be made out, although 
 very much obscured by the road, which indicates a probable con- 
 nection between Round and Crooked lakes. 
 
 On Dewey Lake the most interesting developments have occurred 
 at the east end which leads into a marsh nearly as large as the lake. 
 This marsh was formerly a shallow arm of the lake connected by 
 a narrow but deep channel. Spits started at the higher level on each, 
 side of the channel and succeeded in closing about one-half of the 
 gap. The marsh is a good example of almost complete extinction 
 by vegetation which obtained an early start in this shallow part of 
 the lake. 
 
 INDIAN LAKE 
 
 Indian Lake is located in the northwestern part of Cass County 
 just off the western boundary line. It lies nearly six miles west of 
 Dowagiac and the roads leading to it are improved. From the 
 geological standpoint it is located on an outwash plain within a 
 short distance of a moraine which skirts the west side of the lake. 
 It is, thus, a pit but a large one for the lake is nearly two miles in 
 length by almost a mile in width and is regular in outline. The 
 regularity of the shore is due in large part to the adjustments of the 
 shore which are numerous and interesting. They were accom- 
 plished at a level some three feet higher than the present and on 
 the sides of the lake show a close correspondence to the strength 
 and frequency of the winds. 
 
344 INLAND LAKES OF MICHIGAN 
 
 On the east side the waves have formed bold cliffs and sand 
 beaches, and currents have thrown complete bars across the em- 
 bayments. On the opposite side, however, the cliffs are less steep, 
 the beaches muddy and the bars with a single exception conspicu- 
 ous by their absence. This is a spit which developed southward 
 across the entrance to a large huckleberry swamp and probably 
 (lid not rise above the former level for much of its extent. This 
 contrast, which is emphasized by the encroachment of vegetation 
 at the more protected places on the west side, shows clearly the 
 greater effectiveness of the westerly winds. 
 
 At the ends of the lake the topography of the shores is very dissim- 
 ilar, and the more pronounced effects have been produced at the 
 north end under the drive of southerly winds, although the 
 spit noted on the west side indicated stronger northerly winds. 
 Thus, the land at the north end is low with the exception of the 
 island-like hills at Highland from which complete bars swing in 
 either direction around the head of the lake and cut off a broad 
 swamp. At the south end, however, the land is higher and cliffs 
 predominate. 
 
 With these general considerations the specific features may be 
 left to those interested to work out. As a final consideration atten- 
 tion may be called to the geographic relationship between the 
 physiographic development of the shores and the settlement of all 
 but the west shore of this very popular lake. . 
 
CHAPTER XI 
 
 LAKES OF THE PORT HURON MORAINIC AND OUTWASH 
 
 SYSTEM 
 
 The lakes included in this group are located within the Port 
 Huron moraiuic system, 4, Fig. 3, and in the area between this 
 system and the Valparaiso-Charlotte moraine, 3, Fig. 3. In general, 
 the glacial deposits of this area are irregular in distribution with 
 the exception of two areas : The lake plains of the southeastern part 
 of the State and the regailar series of formations which mark the re- 
 tirement of the Sagijiaw lobe. In the latter of these areas the lakes 
 are few in number and relatively unimportant and in the Erie 
 plains they are almost entirely absent. The one lake of the Erie 
 lowland (Ottawa) seen by the writer is little more than a mud hole, 
 a large part of which dries up during the summer, and is interesting 
 only as an example of a lake located in a sink. There are also no 
 lakes of importance in this section of the eastern interlobate area, 
 which extends well into the ''Thumb" region. Therefore, the lakes 
 in this group are located in the north central and western parts of 
 the Southern Peninsula but do not include the lagoons along the 
 west coast which are discussed in the following chapter. 
 
 CRYSTAL LAKE^ MONTCALM COUNTY 
 
 Crj-stal Lake lies seven miles south of west of Stanton in eastern 
 Montcalm County and may be reached by a drive of about ten 
 miles over roads that were found in excellent condition. It is one 
 of the few popular lakes which are located in the regular series of 
 deposits of the Saginaw lobe. 
 
 The glacial formations have a north-south trend in this locality 
 and the lake lies in an irregular moraiuic depression that originally 
 consisted of several basins. This depression is situated on the 
 edge of a sandy moraine which gives way to a till plain just east 
 of the lake. The easily eroded material, the irregular outline of 
 the lake, and the succession of hills and swales at the shore made 
 favorable conditions for the numerous adjustments found on all 
 the shores except the north, in which locality the lakeward slopes 
 are uniformly gentle. At the uplands sharp cliffs show cutting 
 by the waves, at the swales bars, usually complete, indicate the 
 
346 INLAND LAKES OF MICHIGAN 
 
 activity of currents, and on the gentle slopes ramparts signify shove 
 by the ice. 
 
 In addition, the shore adjustments enable one to decipher an 
 interesting series of events in the physiographic history of this 
 lake. The most favorable locality for this study is along the south- 
 west shore where a former bay has been cut ofE by ,a series of bars 
 and rampart-bars. The highest is a large isand bar which stands 
 fully six feet above the present level of the lake. Between this 
 bar and the lake are two distinct rampart bars at intermediate 
 elevations which mark halts in the lowering of the lake to its 
 present level. This locality requires careful study but, once 
 solved, will isimplify further study of the shores. 
 
 The most significant adjustment was the dosing of the mouths 
 of embayments by bars which, with the lowering of the water levely 
 reduced the original area of the lake by at least one-half. In this 
 connection the spit which developed along the northeast shore on 
 the flat between Crystal and Mud lakes may be mentioned. 
 
 COLD WATER LAKE^ ISABELLA COUNTY 
 
 Coldwater Lake is ten miles northwest of Mt. Pleasant in west 
 central Isabella County; but a drive of fourteen miles from this 
 city is necessary to reach the lake. The lake rests in a. pit in a 
 narrow strip of outwash, which fronts a moraine to the eastward 
 and was formed by border drainage, that is drainage running par- 
 allel to the ice front. The pit is located so close to the moraine 
 that the southeastern shore is composed of the till of this forma- 
 tion. The relief along this shore, therefore, is great and presents 
 a decided contrast to the surface of the outwash which rises barely 
 more than ten feet above the level of the lake. 
 
 The adjustments are slight indeed along the shores bounded by 
 outwash although many lakes similar in size — the length is ap- 
 proximately one mile and the width one-half mile — ^have decided 
 shore features. The cliffs on the east shore show some wave action, 
 but the southea'stem shore, where the uneven morainic topography 
 furnish'ed more favorable conditions for adjustments, is the most 
 interesting. In particular, may be mentioned the large amphi- 
 theatre which opens into the lake. Two well developed spits^ 
 which stand at elevations of three and five feet above' the present 
 level, swing out into the opening in parallel courses from the 
 south shore and show that this indentation would have been cut 
 off eventually, if their development had not been stopped by a' 
 lowering of the water level. 
 
LAKES OF THE PORT HURON MORAINIC SYSTEM 347 
 
 It is evident from tlie spits just mentioued tliat the lake has 
 stood at two higher levels but the presence -of an old dam on the 
 outlet a few rods from the lake makes uncertain the relationship 
 between at least one of the spits and a natural level of the lake. 
 The level of the lake when the dam was operative and the length 
 of time that this level was maintained was not learned, and it is 
 possible, then, merely to make the suggestion that the lower and 
 weaker spit was formed at this artificial stage. 
 
 Aside from the shore features at the southeast end, the lake 
 furnishes an example of the partial filling by marl, formed in part 
 of the minute shells that are abundant on the beach, 
 
 MISSAUKEE LAKE 
 
 Missaukee Lake is the largest of a group of ten or more lakes 
 which are located in west central Missaukee County. It may be 
 reached via the Lake City Branch of the Grand Rapids and Indiana 
 R. R. which stops at Lake City on the east shore of the lake. 
 
 All of these lakes lie in pits in an outwash plain and of these 
 the Missaukee Lake depression is by far the largest; it is nearly 
 circular with a diameter of somewhat less than two and one-half 
 miles. The lakes either have no outlets or drain eventually into 
 Missaukee, the southeastern member of this group. Inasmuch as 
 the drainage of the region in general is to the southeast, Missaukee 
 Lake occupies the key position and it has no natural outlet, the 
 artificial channel operating only at infrequent periods of excep- 
 tionally high water. The Missaukee Lake depression, although it 
 contains the deep holes characteristic of pits, is nevertheless, very 
 shallow. The writer's information is that the general depth of 
 water is approximately ffiteen feet, and, inasmuch as the surface 
 of the lake stands about ten feet below the surface of the outwash, 
 the total depth is twenty-five feet. The shallowness of the water 
 must hinder the development of the larger waves but, nevertheless 
 adjustments of the shore have taken place, due probably to the 
 ease with which the outwash material is worked. 
 
 The adjustments are found above the present surface of the lake 
 and indicate the usual higber level in past time. Along the south 
 and east shores the depression has regular wall's, and an almost 
 unbroken, wave-steepened clitf faces the lake. At the foot of the 
 cliffs there is now exposed a broad sand flat wTiich continues be- 
 neath the water to a decided "drop-off" wherever the water is 
 deep, as at the east end. The effects of current action are best 
 seen along the very irregular north shore and are too numerous for 
 specific mention in this brief report. Before leaving Lake City a 
 
348 INLAND LAKES OP MICHIGAN 
 
 well developed bar may be seen, and along the north shore examples 
 of exposed spits, bars enclosing lagoons, and land-tied islands are 
 numerous. The north shore bars are distinct but stand at a lower 
 elevation than that at Lake City which is interpreted as an indica- 
 tion that these forms were in process of formation when the lake 
 level subsided. 
 
 The extinction of former embayments by draining and vegetal 
 accumulation is another interesting phase of a physiographic study 
 of this lake, the details of which should prove to be well worth 
 while. 
 
 The lake has not kept pace with many others as a summer resort 
 but has qualifications which are superior to some of the more popu- 
 lar, including good fishing and favorable shore conditions on the 
 south side. 
 
 MANISTEE LAKE 
 
 Manistee Lake is in northeastern Kalkaska County and may be 
 reached from the town of Kalkaska on the Grand Kapids and 
 Indiana and the Pere Marquette railroads. A drive of ten milesi, 
 is necessary but the writer found the road good with the excep- 
 tion of the last mile. 
 
 This lake is located on the southwestern extension of the inter- 
 lobate area between the Michigan and Huron lobes which joined 
 at approximately right angles in the vicinity of Gaylord. The 
 formations involved are a moraine, the Port Huron, and a rather 
 broad strip of outwash lying to the southeast. The lake lies on the 
 outwash at the junction of these two formations. It is the largest 
 lake in the region, having a length of two miles and a width that 
 averages nearly a mile. The basin is regular in contour but varies 
 most on the west side, due possibly to the influence of the moraine 
 whose border runs not far from this shore. The basin forms a 
 part of a broad drainage channel which is followed by the Manistee 
 River to the southwest. Either the outwash is excessively pitted 
 near the lake or the basin is not a pit, for the slopes quite generally 
 rise gently from the lake. Whatever the origin of the basin may be,, 
 a small amount of material was furnished by wave action but 
 this was distributed to such advantage that pronounced changes 
 were effected. Thus, at a higher level the waves cut a broad ter- 
 race on the east shore and the material was worked southward 
 into a strong sand bar which crosses the lowland at the south end 
 as far as the outlet. On the opposite side of the outlet the material 
 derived from the short stretch of low shore was carried in both 
 directions and deposited in small spits near the outlet and on the 
 
LAKES OF THE PORT HURON MORAINIC SYSTEM 349 
 
 south side €f tlie entrance to an enibayment to the northwest. 
 Other adjustments were found on the west shore including well 
 defined ice ramparts where conditions are favorable for their for- 
 mation. The lalce is not popular as a summer resort so far, as 
 there are no cottages but it is visited by fishermen and. campers 
 who can get boats at the south end. 
 
 BIG STAR LAKE 
 
 Big star Lake is located in the southwestern part of Lake County 
 and has an area of about two square miles. Its length is more 
 than two and one-half miles but its outline is so irregular that the 
 width varies from less than one-fourth mile to more than one mile^ 
 It is not so popular as similar lakes nearer Lake Michigan but, 
 nevertheless, is an attractive lake which should develop as a sum- 
 mer resort in the future. The nearest railroad stop is Baldwin on 
 two lines of the Pere Marquette system, eight miles distant by 
 road. The writer found the road from Baldwin to the lake in good 
 condition and advises this route for those unfamiliar with the wind- 
 ing sand roads of the jack pine and grub oak plains. 
 
 The nature of the Star Lake depression is very easily recognized 
 for one may almost fall from the edge of a most monotonous sand 
 plain to the lake shore, so sharp are the bluffs which surround the 
 lake. It is clearly a pit but is large and much more irregular 
 than the present crooked outline of the lake. This disparity be- 
 tween the outline of the pit and the lake is due to the adjustments 
 of the shores at a higher level, the greatest changes having been, 
 caused by the deposition of bars across the mouths of bays. Such 
 adjustments are best defined on the south side and are well illus- 
 trated along the soutliwest shore where a bar completely closes a 
 narrow lagoon. Similar forms, but not complete at the larger, 
 embayments, may be found on the south shore. The source of the 
 material of these bars is the numerous sand cliffs which rise nearly 
 twenty feet above the lake. Apparently a small amount of re- 
 cession of the cliffs was sufficient to supply the material inasmuch 
 as the off-shore terrace, although well defined, is narrow, and the 
 strip exposed by the recession of the water has been removed from 
 the foot of the cliffs.. 
 
 The north shore consists of broad bays and headlands, and cur- 
 rents have developed along the shores of the bays rather than across 
 the entrances, eventually dissipating in the lake off the ends of 
 the points. Thus the changes in the outline on this shore are less 
 striking than on the south side. 
 
350 INLAND LAKES OP MICHIGAN 
 
 FREMONT LAKE 
 
 Fremont Lake is situated on the southern edge of Fremont in 
 southwestern Newaygo County, the city limits extending to include 
 a park on the north shore of the lake. The town is located on the 
 Muskegon-Big Kapids Line of the Pere Marquette R. R., and the 
 lake is not an unreasonably long walk from the station. 
 
 The basin lies at the border between outwash and till plain and 
 is rather difficult to classify as to type. The till plain borders the 
 north and east sides of the lake, the remainder of the shores con- 
 sisting of the sands of the outwash. Also the eastern part of the 
 basin is shallow, a characteristic of lakes in the sags of till plains, 
 but the western part is deep. Thus, it appears that we have here a 
 pit which is open to the east and that the water is sufficient in 
 amount to fill not only this pit but also flood an adjoining depres- 
 sion of the till i)lain to the east. 
 
 The basin is very regular in outline and consequently very few 
 adjustments by shore currents are to be found. The two bars 
 noted stand at an elevation which denotes a former level of the 
 lake about three feel; above the present surface. One, a very in- 
 distinct spit, runs northwestward on the low flat at the southwest- 
 ern part of the lake, and the second, also a spit but much better 
 developed, extends from the bluff on the west side of the outlet to 
 the stream channel. A much greater adjustment resulted from 
 the action of waves and undertow and consists of numerous cliffs 
 but more especially a decided submerged terrace. The terrace is 
 uniformly present but varies in width, being greater along the 
 shores bordered by outwash. The outer part of this terrace is 
 exposed by the sinking of the water level, forming a sand flat 
 which has a width of more than one hundred feet in places on the 
 south shore. Ice shove is also effective as shown by the decided 
 ramparts in the vicinity of the outlet. In some places as many as 
 four ramparts were found, giving the impression of an ice-shove 
 terrace although it is probable that such is not the case. 
 
 Another interesting feature is the delta which is the site of the 
 park on the north shore. This delta consists of the silt deposited 
 by the stream which enters at this point and was built at the 
 higher level. The stream now carries a dift'erent type of load, 
 consisting of the refuse of the canning factory and tanner}^ which 
 makes a very unpleasant condition of the shore. 
 
LAKES OF THE PORT HURON MORAINIC SYSTEM 251 
 
 r.IG BLUE LAKE 
 
 Big Blue Lake is in the northeast corner of Muskegon County 
 and may be best reached from Whiteliall on the Pere Marquette R. 
 R. One may go in from Twin Lakes or Holton, but, in any case, 
 the roads are not good for automobile traffic at the present time. 
 They are sand roads Avhicli otTer difficulties even to those who are 
 familiar with their peculiarities. A favorite description of such 
 roads is parallel snake tracks through second growth woods, which 
 means constant attention to the steering of the machine in order 
 to avoid getting out of the tracks and striking trees or stumps. 
 Also the sand becomes loose in dry weather and the traction is 
 very heavy making it almost impossible for heavy machines to 
 "plow through.''' Frequently, the familiar north-south, east-west 
 system of roads is sadly lacking, and numerous branch roads offer 
 oj^portunities to stray, none too i^leasant an experience for most 
 people in such thinly settled areas. The writer suggests the road 
 from Whitehall. 
 
 It is obvious from the foregoing statements that Big Blue Lake, 
 although within ten miles of a railroad, is not readily accessible 
 at the present time. Yet it is a very attractive lake and is frequent- 
 ed by those who desire a complete change from the formalities of 
 city life. The visitors come almost entirely from the vicinity of, 
 Chicago and have built a number of cottages which are grouped 
 at the east end and at two localities on the north shore. 
 
 The lake measures nearly two miles in length by about three 
 quarters of a mile in greatest width and lies in a deep pit in an 
 outwash plain. In addition to the main depression, small pits are 
 so numerous that the lake shores have a rolling topography, and 
 the varied shore conditions made possible numerous adjustments 
 which have been readily accomplished in the loose sand. These 
 changes occurred at a level live feet above the present and are so 
 numerous that mention of each feature would entail a description 
 of almost the entire shore of the lake. The bluflfs have been steep- 
 ened by wave action and drop to a narrow exposed terrace which 
 frequently continues under water to a' "drop-off". At the embay- 
 ments the bars, which rise well above the present level, are so well 
 developed that it is possible to follow the shore around the lake 
 dry shod except at the small brook which drains the lake. In 
 addition to the work of waves and currents evidence of strong ice 
 push may be found. This is the only noticeable adjustment which 
 is taking place under present conditions. 
 
352 INLAND LAKES OF MICHiaAN 
 
 TWIN LAKES 
 
 Under this heading are four small lakes which are located about 
 ten miles northeast of Muskegon on the Muskegon-Big Rapids Line 
 of the Pere Marquette R. R. They are now known as East, Middle, 
 North and West lakes. All of these lakes lie in shallow pits in 
 the same outwash plain as Big Blue Lake, six miles to the north. 
 None of the lakes has an area of one-half square mile but, inas- 
 much as they are connected, the group may be considered as one. 
 
 The water level now stands about ten feet only below the out- 
 wash surface. The lakes have no surface outlet and consequently 
 the water level varies over periods of years, the highest level as 
 recorded by the shore features having been three feet above the 
 present or about seven feet below the plain. Thus, at that time 
 the waves were able to accomplish much without the removal of 
 an excessive amount of material and as a result there is present 
 an off-shore terrace that appears out of proportion for a lake of 
 this siz^e. A part of the material quarried by the waves was dis- 
 tributed along the shores also and formed a number of distinct 
 bars. One of the best localities to observe these is on the north 
 shore at the connection between East and Middle lakes. The bars 
 on Middle Lake are stronger than those on East, one in particular 
 nearly crossing the lowland connecting the two lakes. Also in the 
 same vicinity an island Avas tied to the mainland of Middle Lake. 
 Two spits extend out from the south end of this island and show 
 clearly the difference in activity of the currents in the two lakes at 
 this point. The larger spit developed in Middle Lake, due to the 
 greater power of the westerly winds. Another interesting example 
 of the development of a spit occurs at the east end of North Lake 
 where a broad arm of this lake was nearly isolated and has since be- 
 come a marsh. Other similar features may be found on the shores of 
 these lakes, and their discovery may be left to those interested. 
 
 The lakes are connected with Muskegon by an improved road in 
 addition to the railroad and are popular summer resorts. A camp 
 ing ground and boat livery are located at the east end of East Lake 
 but the majority of the summer inhabitants own cottages many of 
 which are much more pretentious than the usual summer resort 
 cottage. 
 
 WOLF LAKE 
 
 Wolf Lake is six miles east of the city limits of Muskegon and 
 within one-fourth mile of the main road between Muskegon and 
 Grand Rapids. Auto bus service to the lake is maintained in sum- 
 
LAKES OF THE PORT HURON MORAINIC 353 
 
 mer so that no difficnilly need be exi>ei'ieueed in reaclilng it. The 
 ride, liowever, is most nionotonons for one travels over a sand 
 plain so tiat that even a slight variation in the surfa)ee claims the 
 attention, and tlie variations are few indeed. Once there, the lake 
 conies into view abruptly and is recognized at once as a pit for 
 the sand plain near the lake is outwash. and steep bluffs rise con- 
 sistently from the water to a height of about twenty-five feet on 
 iiU sides. 
 
 The regular outline of the lake as shown on the map is mis- 
 leading as far as the original basin is concerned for there are small 
 embayments on all but the south side. This change in outline is 
 due to the natural development of shores which occurred when the 
 lake stood nearly five feet higher than at present. The lowering 
 of' the water exposed the former shore and, thus, facilitated the 
 study of its features. Inasmuch as the lake has no outlet, the drop 
 in level is an indication of a' climatic change, the significance of 
 which is not as yet clear. 
 
 On a lake the size of Wolf, — less than one square mile in area and 
 with a greatest diameter of less than a mile, — the waves and cur-: 
 rent>s are comparatively feeble, but the high bluffs of loose sand 
 were easily eroded, thus furnishing abundant material without 
 seriously reducing the power of waves or currents. Consequently, 
 the bluft's, although steepened, have not receded to any great ex- 
 tent and there is no decided ott'-shore terrace. The w^ork of the 
 shore currents, however, is much more noticeable in that a rela- 
 tively small amount of material was deposited in such a way as to; 
 cause very decided changes in the outline of the lake. The refer-' 
 ence here is to the bars which developed across the necks of em- 
 bayments. T-hese bars do not completely close the openings in all 
 cases but are consistent in that their development shows a direct 
 relationship to the prevailing winds. Thus, along the east shore 
 which is exposed to the strongest winds one may follow the shore 
 irrespective of cliflf or swale because the bars are complete, but on 
 the north and west shores only spits are found. 
 
 The lake has insignificant inlets and is fed mainly by springs, 
 which accounts for the exceptionally clear water. This in combina- 
 tion with clean beaches, excellent locations for cottages along the 
 bluffs, and good fishing has made this lake a favorite with the 
 residents of the vicinity-. 
 
CHAPTER XII 
 
 BOEDER LAKES AND LAKES OF DIVERSE ORIGIN OUT- 
 SIDE THE PORT HURON MORAINIC SYSTEM 
 
 The lakes described iu this chapter coustitiite our fourth group 
 and include those which lie outside the limits of the Port Huron 
 Morainic system, 4, Fig. 3, as well as the border lakes which are 
 within the area of the third group. The border lakes are so 
 numerous that the fourth group may be conveniently subdivided 
 into border lakes and those of diverse origin. 
 
 BORDER LAKES 
 
 The border lakes are all former coastal embayments which have 
 been cut off by the development of great sand bars across the. 
 openings and are, therefore, lagoons. The bars developed during, 
 higher stages of Lake Michigan — Algonquin or Nipissing — and 
 were left well above the water level when the waters of these lakes 
 subsided. Then there followed a period of eastward movement of the 
 finely assorted sand of the bars caused by westerly winds and de- 
 position of the sand in great rows of dunes which are unsurpassed, 
 at least as regards size, so far as is known. Later the dunes be- 
 came covered with vegetation which so etfectuallj^ stopped their 
 movement, always to the east, that they have remained fixed in 
 position to the present time, except for occasional blow outs. 
 
 A blow out is merely the renewal of the movement of a fixed dune 
 but apparently originates in a limited area and works up the front 
 elope in a narrow zone to tiie very crest. This results in a great 
 trough on the front slope, the transformation of the crest into a 
 saddle, and a fresh deposit of sand on the back slope which is very 
 steep. Often the blow outs encroach on the lagoon, causing a 
 projection of the shore, and in this way contribute to the filling of 
 the basin. It is recognized that the dunes are but indirectly re- 
 lated to the study of lakes, but they are most attractive and furnisli 
 such excellent locations for summer homes that this brief sketch 
 of their history seems warranted. 
 
 Even though the manner in which these embayments were iso- 
 lated be known, there still remains the problem of the type of em- 
 bayment, of which there are several. Some are easily recognized 
 
356 INLAND LAKES OF MICHIGAN 
 
 but others, for example Pine, Walloon, Torchlight, etc., which are 
 described in an earlier chapter, present difficulties. Among the 
 types of basins easily recognized are the numerous drowned mouths 
 of streams. The causes of the drowning, or partial submergence, of 
 the mouths of the streams entering the southern part of Lake 
 Michigan is due to the uplift of the land in the northeastern part 
 of North America following the retirement of the glacier. It is 
 not necessary to go into the details of this complicated subject to 
 realize that, if the lower part of Lake Michigan were not affected 
 by the uplift while the northern part was being elevated, the water 
 would pile up in the southern part of the lake and, thus, rise with 
 reference to the land. Such was the case and in the tributaries 
 of Lake Michigan as far north as the Betsie River at Frankfort the 
 water backed into the mouths of the stream valleys. 
 
 The outline of such lakes is very irregular and, in a typical case, 
 consists first of a main channel which may, or may not, be winding. 
 Farther inland the main channel ramifies and each ramification 
 may in turn divide so that the pattern resembles that of a de- 
 ciduous tree. In all cases in Michigan the lakes brauK^h to the 
 east and, inasmuch as the strongest winds are from the west, the 
 force of the waves is so largely dissipated in the diverging chan- 
 nels that the effects are insignificant. Conditions were unfavorable 
 for adjustments by shore currents and the occurrence of features 
 due to these agents are very exceptional. ,Waves and undertow 
 were effective in forming cliffs and terraces which are found with, 
 monotonous regularity at levels above the present, in particulaT 
 the level of Lake Nipissing which stood about fifteen feet above 
 the present lake along this part of the shore. One other comlmon 
 characteristic of these lakes is the tendency for the heads of bays 
 which have entering streams to be silted up, forming a deltadike 
 flat which has pushed a singularly even front into the lake. 
 
 This degression, which possibly over-emphasizes the characteris- 
 tics of the droAvned-river lagoon, was made purposely in order to 
 avoid the dull repetition which became very apparent when the 
 attempt was made to discuss individuallj^ each lake of this type. 
 Nine of these lakes were visited by the writer, as follows: Kalama- 
 zoo near Saugatuck, Black near Holland, Spring near Grand Haven, 
 Muskegon, White near Whitehall, Pentwater, Pere Marquette near 
 Ludiugton, Manistee, and Betsie at Frankfort. A number of these 
 are well known ports which need no discussion here except that 
 they are thereby made more accessible. Their use as summer re- 
 sorts, however, is of interest. The exposed terrace of the higher 
 level serves as an excellent site for cottages and their location just 
 
BORDER LAKES AND LAKHS OF DIVERSE ORIGIN 357 
 
 east of Lake ^licliiiian at-.suics at least a teiii]ieiiiij>- ol the suiiuuer 
 Iieat. It is not suri>iisin<;-, therefore, to fiiitl some of tlieiii almost 
 lined \^•ith snnunci' liomes many of which are costly estates. Of 
 these mentioned JJlack, S])ring- and AVhite are the most popular. 
 
 The lagoons of diverse origin, obviously, cannot be discussed as a 
 group as were the drowned streams nor can the probability of shore 
 adjustments be postulated in advance. Consequently they are dis- 
 cussed individually, althougli this method may give undue prom- 
 inence to some of these lakes. 
 
 BASS LAKE 
 
 Bass Lake is located about five miles north of Pentwater and 
 t^vice this distance south of Ludiugtou, both on branches of the 
 Pere Marquette R. R. Ludington is also accessible by boat, but 
 in either case a drive is necessary to reach Bass Lake, which is 
 situated near the excellent West Michigan Pike. 
 
 The most distinctive feature of this lake is that it j)arallels the 
 Michigan shore instead of extending inland. The west shore con- 
 sists of an exposed sand bar which is not wholly obscured by dunes 
 and on the opposite side the land slopes gently upward to a sharp 
 rise at the same elevation as the bar, this elevation being that of 
 Lake Nipissing. Furthermore, the lake gradually increases in 
 depth towards the west side and at most hardly exceeds twenty- 
 five feet. The basin, therefore, is the deeper part of a shallow em- 
 bayment which existed during Lake Nipissing and is masked by 
 sand so that the surface indications are of little aid in the deter- 
 mination of its origin. It appears to be part of a narrow crescentic 
 lagoon which was isolated during Nipissing time, the position of 
 the bar having been determined by the then prominent headland 
 a few miles north. This type of lake is rare in Michigan, the only 
 other example seen by the writer being Devil's Lake near Alpena, 
 an unattractive lagoon rapidly filling with vegetation. 
 
 The shore features of Bass Lake are very similar to the lagoons 
 already discussed. The Nipissing shores are best developed and 
 stand about fifteen feet above the lake. They consist of a broad 
 terrace and cliff on the east side, and of a sand bar and flat on 
 the west. Below the Nipissing terrace a lower exposed terra'ce is 
 evident on the east shore but no deposits by shore currents were 
 found. Aside from the terraces the most interesting physiographic 
 development is the large delta formed by a stream entering the 
 east iside of the lake. This delta is triangular in shape but is 
 exceptional in that it extends into the lake as an apex rather 
 
358 INLAND LAKES OF MICHIGAN 
 
 than one of the sides of the triangle, due to the fact that the stream 
 did not form distributaries. 
 
 The lake is developing as a summer resort and cottages have 
 been built on both sides. The west shore in the dune area is by far 
 the better location and is also the more popular. 
 
 rORTAGE LAKE^ MANISTEE COUNTY 
 
 Portage Lake is located eight miles north of Manistee and may be 
 reached by the boats of the Michigan Transit Co., the Manistee and 
 Northeastern K. E. from Manistee, and by automobile over ex- 
 cellent roads. The lake is three and one-fourth miles in length. 
 Two broad points, one each on the north and south shores, break 
 the otherwise regular shores but are offset in position so that 
 thej^ give the effect of a sinuous channel of about one mile in width. 
 This is misleading, as may be determined from the very excellent 
 Lake Survey Chart No. 777 which shows that the lake consists of 
 two basins separated by a narrow submerged ridge over which the 
 water is but sixteen feet in depth. The greatest depth of the 
 basins is sixty feet in each case but the eastern basin is much the 
 larger in extent. The lake is separated from Lake Michigan by 
 a narrov/ row of high dunes which is continuous except for two 
 gaps, one across which the present channel has been dredged and 
 another one mile north which was the natural outlet of the lake. 
 
 The basin resembles an elongated amphitheatre, the walls of 
 which are moraine of high relief, the floor is till plain and the stage 
 the dunes. Such a distribution of glacial formations is evidence 
 that the depression was in existence at the time of the last advance 
 of the glacier and that a small lobe flowed into the depression, the 
 front of the ice halting for a time sufficient for the formation of the 
 moraine. The bar which closed the open end was formed during 
 Nipissing time and the dunes were formed subsequent to the sub- 
 sidence of that lake. 
 
 The shore features of Portage Lake are very simple and require 
 little more than brief mention. The specific features are all found 
 at the Nipissing level and consist almost exclusively of a cliff, 
 below which a broad terrace slopes to the lake. The only example 
 of a bar seen on other than the west shore was at Onekama where 
 a small stream is turned westward by it. Another physiographic 
 form is the broad flat at the east end of the lake which is interpreted 
 as a delta built during Nipissing time. The northern half of the 
 submerged ridge which crosses the lake off North Point, as shown 
 on the Lake Survey Chart, is at least, suggestive. The material is 
 
BORDER LAKES AND LAKES OF DIVERSE ORIGIN 359 
 
 sauil, aud the form and location are cliaractei-istic of the beginnings 
 of a shore current deposit. 
 
 Tliis lake shares tlie i>opularit3' of the lagoons of the Lake Mich- 
 igan coast but has not developed as a summer resort to the same 
 extent as some of those farther south. Numerous cottages are 
 scattered along the shores but the favorite location is on the dunes 
 at the west end. Here is a large colony which includes a well 
 api)ointed hotel, a recreation pavilion and a large number of pri- 
 vate cottages. A water supply system is maintained so that a very 
 comfortable vacation may be spent on this lake. 
 
 PLATTE LAKES 
 
 The I'latte Lakes, Big and Little, are situated in western Benzie 
 County on a triangular-shaped flat bounded on the east and south 
 by moraine and on the west by Lake Michigan. The nearest rail- 
 road stop is Honor on the Manistee and Northeastern, but railroad 
 connections from the south are not convenient. A drive of less 
 than three miles brings one to the narrow strip of land which sep- 
 arates the two lakes. Another route is to drive from Frankfort 
 Avhich is the northern terminus of the Ann Arbor E. R. and also 
 a stop for the boats of the Michigan Transit Co. The latter route 
 involves a ten mile drive the last half of which was found to be 
 over heav}^ sand road. The more comfortable trip for automobiles 
 is through Honor. 
 
 The writer attempts no explanation of the origin of the lake 
 basins in this flat because the area was covered by one of the prede- 
 cessors of Lake Michigan and is, therefore, masked by sand. AVhat- 
 ever the type of basin, it is known that this area w^as once flooded, 
 with the exception of a group of high morainic hills on the north 
 shore of Big I*latte, and furthermore was connected with Crystal 
 Lake through the Round Lake depression which is followed by the 
 road. The narrow connection between the two lakes was closed by 
 a bar along the Crystal Lake shore and the main dex)ression at 
 least partially separated from the main lake in a similar manner. 
 
 The shore of Lake Algonquin in this vicinity washed 'steep cliffs 
 that stand well back of the present lake shores and furnish the 
 chief source of interest in the study of the Platte Lakes, for the 
 adjustment of the present shores is negligible. A complete study 
 of this former lake involves an. area of more than twenty square 
 miles and a' similar number of miles of shoreline. Time was not 
 available for such a study but a hint of the possibilities was gleaned 
 from the bars noted near the outlet of Big Platte. Near the lake 
 the bars follow the outline of the shore and, therefore, are evi- 
 
360 INLAND LAKES OF MICHIGAN 
 
 dence that this lake was isolated from Lake Algonquin and that 
 this level was maintained for some time. A study of these lakes 
 might well begin in this locality and should by all means include 
 the area along the outlet to Lake Michigan. Another locality 
 that should prove interesting is that to the northwest of Little 
 Platte. 
 
 The Platte Lakes are extensively fished but in other respects 
 are not so well adapted to general summer resort purposes as 
 many of our lakes. There are a number of excellent locations for 
 cottages oil Big Platte but a large part of the shore is low and, 
 not desirable as a building site. 
 
 GLEN LAKE 
 
 Glen Lake and surroundings form one of the most attractive 
 bits of scenery of its kind the writer has had the pleasure of seeing. 
 We may pass over the details of its location in western Leelenau 
 County and the routes to the lake very briefly. The trip by the 
 Michigan Transit Company's boat is pleasant but the sailings are 
 infrequent. The trip in from Empire, seven miles distant, must be 
 tedious on account of poor railroad connections. The trip from 
 Traverse City by automobile is long, although the roads are not bad. 
 The inference that the lake is somewhat difficult to reach is correct^ 
 but the effort is well worth while. 
 
 The Glen Lake depression may be likened to a great oval race 
 track in the center of which is a large island, the present Glen 
 Lake filling tha entire eastern and the southwestern part of the 
 oval course. The stands which surround all but the north side 
 of the depression show an evolution, the originals consisting of a 
 loop of high hills on the east and south. Later developments 
 erected a west stand, the sands of the great Sleeping Bear. The 
 best seats are located in the south and east stands for from here 
 one may see below the cobalt blue water of the circular main 
 basin of the lake with its fringe of maize, and ofi' to the w^est is 
 the narrow arm of the lake which bows to the Sleeping Bear. 
 
 The depression was very probably in existence previous to the 
 final invasion by the glacier which slightly overran the depression 
 and persisted in its position. During this time earthy material 
 was constantly beng brought forward and dropped as the ice 
 melted. This material was piled higher and higher in hugh hum- 
 mocks so that when the ice finally disappeared a high moraine or 
 wall was left, which constitutes the south and east limits of the, 
 basin. But this explanation does not account for the high land, 
 (island) in the central part of the basin and, therefore, is qualified. 
 
BORDER LAKES AND LAKES OF DIVERSE ORIGIN 3G1 
 
 This high laud was once certainly an i.sland for the Avaters of Lake 
 Algonquin entered the" basin through cliannels on the north and 
 on tLe west sices isolating it. For a time the waves and cur- 
 rents of Algonquin coursed the track but eventually the western 
 entrance* was closed by a sand bar. Then followed a subsidence 
 of the lake to the level called Nipissing and the west stand (the 
 Sleeping- Bear Point) was erected, the material being the sand of 
 tlio bar and the builder the wind. The great Sleeping Bear is noW' 
 a blow out, or better a moving dune. The colors of the lake are 
 signiticant. The blue is due to a modification of the sun's rays as 
 they pass through the water and the depth of color varies with the 
 thickness, therefore the lake is deep. Similarly, the narrow but 
 sharply defined fringe is the yellow of the sands seen through a 
 thin layer of shallow water, signifying a submerged tei'race. A 
 final step vras necessary for the complete isolation of the lake, 
 namely the closing of the nortli entrance during Nipissing time by 
 a series of bars which still retain the water at an elevation of 
 seventeen feet above Lake Michigan. 
 
 Our interest in tlie lagoons lies primarily in deciphering their 
 geologic history but Glen Lake also furnishes many examples of 
 shore features so generally lacking on lakes of this type farther, 
 south. These features, formed to a laTge extent when one or both, 
 of the entrances were open, are naturally on a large scale. Thus, 
 the cliflfs are high and steep, the bars complete and large, and the 
 otf-shore terrace wide and distinct. A final episode is due to the 
 interference of man and is unfortunate. Reference is made to the 
 damming of the outlet which has not seriously flooded the lake 
 but has renewed the activity of the waves. The consequent re- 
 cession of the cliffs is serious and Avill result in the destruction of 
 portions of the newly graded boulevard, for eiffective preventive 
 measures are too costly to be feasible. 
 
 This lake is beautiful, its physiography is most interesting, and 
 the shore conditions are excellent, but, nevertheless, cottages are 
 not numerous. It appears to lag on account of its isolation but 
 must de'\elop rapidly as a summer resort once its qualifications 
 iire appreciated. 
 
 GRAND LAKE^ TRESQUE ISLE COUNTY 
 
 A number of lakes that would naturally be discussed here. Elk, 
 Torchlight, I'iiie, etc., have already been described in de/tail, so 
 we must pass on to the Huron shore to complete this part of the 
 group. Grand Lake is situated in western Presque Isle County 
 Avithin three miles of Lake Huron and a somewhat greater distance 
 
362 INLAND LAKES OP MICHIGAN 
 
 directly north of its neighbor Long Lake, which is described in 
 detail in an earlier chapter. A long ride is necessary to reach the 
 lake from Alpena, the most convenient railroad stop, but the road 
 is excellent for most of the drive. Eeference is made to Long 
 Lake because of the great similarity of the two lakes in form, size, 
 orientation, shore conditions and type of basin. With slight varia- 
 tions a single general description could be made to serve for both. 
 Grand Lake varies in two respects : It has numerous islands and is 
 a lagoon or at least an arm of the predecessor of Lake Huron. 
 
 The basin probably does not exceed thirty feet in depth and on 
 the average is considered shallow. Its total length is more than 
 eight miles and the? maximum width not greater than two. A con- 
 clusion as to the origin of this elongated depression is a matteir 
 of considerable difficulty, inasmuch as a number of factors which 
 cannot be accurately determined must be considered. 
 
 The lake is located in a region in which the bed rock is thinly 
 covered with glacial deposits and in places outcrops at the sur- 
 face, A logical inference, therefore, is that the basin is not due 
 primarily to glacial action insofar as deposition is involved. How- 
 ever, the course of the lake is parallel to the movement of the glacier 
 and the possibility of glacial scour suggests itself at once. In fol- 
 lowing this suggetstion, the distribution of the underlying rocks 
 is an important aid. They are all tilted sediments and, if not cov- 
 ered, would appear at the surface in belts which correspond quite 
 closely with the orientation of the lake and the movement of the 
 ice as well. The geological map of the region shows us that the 
 lake stands on a narrow belt of shale on e'ither side of which is 
 limestone and also that the lake shores along the sides coincide 
 fairly well with the boundaries of these formations. Inasmuch as 
 shale is much more easily eroded than limestone, it is logical to 
 conclude that a trough was eroded in the shale. The presence of 
 the numerous islands, however, is evidence that this was accom- 
 plished to a greater extent by running water than by ice for ice 
 would have swept the basin clean. The final isolation of the basin 
 was accomplished by shore currents of Lake Mpissing which closed 
 the ends of the trough. This conclusion, however, is too definitely 
 drawn. The glacial cover is more complete on the east side than 
 on the west, and the position of the boundaries of geological forma- 
 tions, where not actually exposed, are relative at best. Therefore, 
 the hianner of formation of this basin as given should be considered 
 as a basis for future work which will furnish the facts necessary to 
 the solution of the problem. 
 
BORDER LAKES AND LAKES OF DIVERSE ORIGIN 363 
 
 The study of thoi shore of tiiis Likoi was a <risappoiiitmeut. The 
 effects of ice shove were noted in several localities but aside from 
 this little was found. The beaches are almost unifomily of coarse 
 material and show little assortment, the one exception being the 
 sand beach near Birch Hill. Such a condition is, however, not sur- 
 prising- if certain features of the lake are kept in mind. The shal- 
 low water and numerous islands greatly hinder the normal de- 
 velopment of the waves, and the moderate waves that are formed 
 are fnrtlier reduced in crossing the off-shore shoals. Finally when 
 they strike the beach the}' encounter a compact till, which is 
 heavily laden with large rock slabs, and can accomplish little. 
 Also the currents are of a like order with the additional deterrent 
 factor of an irregular shoreline which, obviously, does not furnish 
 any extended stretch of shore along which the currents may develop. 
 The most conspicuous feature about the lake is the sheer rock cliff 
 which follows the west side. The cliff stands above and back from 
 the lake and is probably the shore of Lake Xipissing the waves of 
 which were capable of a i^owerful attack. 
 
 The lake is especially attractive on account of the numerous 
 bays, headlands and islands and is extensively visited during the 
 sumnier months. Numerous cottages occur in groups at the more 
 favorable locations on the east side and future development is to be 
 looked for. 
 
 VAN ETTEN LAKE 
 
 Van Etten Lake is located within a mile of Lake Huron about 
 two mile'S north of the village of Oscoda on the Detroit and Mack- 
 inaw R. R. The lake has a length of nearly four miles and a width 
 of less than one. This narrow basin extends from Lake Huron in a 
 northwesterly direction to an extensive swamp and apparently 
 continues as the valley of the Pine River. It drops below the sur- 
 face of a sand flat in bold cliff's and is exceptionall.y regular in out- 
 line. The sand flat is a long strip which lies between the Algonquin 
 and Xipissing beaches and is, thus, the off-shore terrace of Lake 
 Algonquin. Upon this eastward sloping terrace Xipissing bars 
 developed hemming in several shallow lakes such as Tawas, which 
 is little more than a marsh, and Cedar. Van Etten Lake is also lo- 
 cated between the two shores, the Algonquin passing just west of the 
 upper end of the lake and a'X^ijDissing bar serving as a retaining 
 wall at the lower end. The basin is much deeper than those of the 
 other lakes on this plain and is clearly of different origin. Just 
 what the origin may be is difficult to determine and one can do 
 little more than conjecture until detailed studies are made. As a 
 
364 INLAND LAKES OF MICHIGAN 
 
 working liypothesis the writer suggests that, when the configuration 
 of the basin is known, it may be apparent that the basin is a broad 
 channel cut through the sand flat by the Pine Kiver after the re- 
 cession of the lake to the Nipissing level. 
 
 The shore features of Van Etten lake are relatively simple. 
 Cliffs predominate and step down to the shore showing a former 
 level about four feet above that at present, probably the Nipissing. 
 Although tlie west sliore is less regular than the east there are no 
 decided emba^anents other than the valleys of entering streams, 
 and bars are not found. Nevertheless, currents were active and 
 left the shore at the points, forming poorly developed cuspate fore- 
 lands, all at the higher level. The most decided physiographic 
 feature, however, is the partially isolated basin at the upper end 
 which is an excellent example of filling by vegetation. 
 
 The excellent East Michigan Pike parallels the west shore a 
 few rods back from the lake and possibly accounts for the growing 
 popularity of this side of the lake. This shore is rapidly being 
 built up but the east side is apparently being deserted, although 
 skirted bv a aood road. 
 
 LAKES OF DIVERSE ORIGIN 
 
 The lakes described in this chapter, Group IV, which are not 
 lagoons are located with one exception between the limits of Lake 
 Algonquin and the Port Huron Moraine, 4, Fig. 3. The lakes vis- 
 ited occur in two groups with the exception of Carp Lake near 
 Mackinaw City and Bear Lalce on the west side of the State. One 
 group is located along the boundary between Iosco and Ogemaw 
 counties and the other a few miles southwest of Traverse City. 
 
 The lakes of the first group lie on a strip of moraine of very 
 irregular shape and of great relief. The sharp slopes and sandy 
 material of the moraine are not conducive to good roads, and a 
 trip through this region by automobile is not an unalloyed pleasure. 
 Ellake at the west end of Long Lake is a station stop on the Rose 
 City Branch of the Detroit and Mackinaw E. R. but a visit to Sage 
 Lake necessitates a drive. 
 
 LONG LAKE, IOSCO COUNTY 
 
 Long Lake appears more like a channel than a lake for its width 
 barely exceeds one-fourth mile but its length is more than two and 
 one-half miles, if its sinuous course is measured. The eastern part 
 of the basin is a very irregular morainic depression which con- 
 
BORDER LAKES AND LAKES OF DIVERSE ORIGIN 305 
 
 tiiuies wostwai'd onto ;> till i)laiii. JNEauy of the iiTegulai-ities of 
 tlie basin do uot: appear on the map because their depths are not 
 snlTicient to bring tlieni below the present level of the lake. Tn the 
 past, however, a higher level was maintained which, although it 
 stood only four feet higher than that at the present time, flooded 
 several large embayments on the north side. From the study of 
 the shores it is apparent tliat these indentations were abandoned 
 because of the sinking of the water level rather than by their isola- 
 tion by bars. A number of small bays were cut off by bars and 
 some of the points lengthened by spits but, in general, the adjust- 
 ments were still in an early stage of development when interrupted. 
 
 Among the specific localities where adjustments may be seen 
 may be mentioned the sharp point on the southeastern shore. The 
 greater j)art of this point is not a' current deposit, but the distal 
 end is clearly a spit which continues as a submerged bar across 
 the narrow channel between it and the north, shore. Another in- 
 teresting locality is at Kokosen Eesort. Currents were especially 
 active on the west side of the point and, in addition to a straight- 
 ening of this shore, have built a small hook which curves to the 
 southeast from the ti[) of the point. East of the point several 
 small indentations are completely isolated by bars and, at the 
 large ba^^ that runs north to a small circular lake, a long spit 
 extends eastward part Avay across the depression. These forms 
 are all at the higher level and may be readily observed. 
 
 The lake has a large group of excellent buildings at Kokosen 
 Resort which apparently has declined in popularity in recent 
 years. 
 
 SAGE LAKE^ OGEMAW COUNTY 
 
 Sage Lake is not easily reached and a long drive is necessary 
 whatever route is taken. The roads were not in good condition 
 at the time of the writer's visit to the lake and the route taken was 
 so roundabout that no directions as to roads are suggested. 
 
 The lake is more than two and one-half miles in length and has 
 a maximum width of one mile. Perhaps the homely comparison of a 
 meal bag tied near each end will serve to describe the shape of the 
 lake. The main part of thi» lake lies in a depression consisting 
 apparently of a number of morainic sags so placed as to form a 
 large basin Avhicli completelj' trenches a strip of very sandy moraine, 
 and the narrow ends are the extension of the basin onto the till 
 plains which flank the moraine. The basin was not studied in 
 sufficient detail, however, to venture a conclusion as to the cause 
 of this peculiar grouping of the sags. 
 
366 INLAND LAKES OP MICHIGAN 
 
 The moraine is one of very strong relief in this locality and the 
 road which skirts the northeast shore is a succession of hills and 
 swales. The swales are large and the shore, therefore, consists 
 of an alternation of broad bays and prominent headlands. The 
 southwest shore, howev-er, is much more regular and is fronted 
 by a cliff that is almost continuous. The adjustments of the shores 
 of this lake all stand at an elevation between three or four feet 
 above the present level and indicate an even earlier stage of tle- 
 velopment than those found on the shores of its neighbor. Long 
 Lake. The headlands and high banks have been carved into steep 
 cliffs but the off-shore"terrace, where present, is poorly developed. 
 Conditions are favorable along the northeast shore for deposition 
 by currents but little had been accomplished when the sinking of 
 the water level interrupted the work. No case of a complete bar 
 was found but several spits were noted on the northeast shore. 
 In all cases the spits are attached to the east sides of the head- 
 lands, indicating prevailing currents from the w^est. The best 
 example is located just west of the Lake View House and is a hook 
 about one hundred feet in length. 
 
 Sage Lake has not been developed as a summer resort but a be- 
 ginning has been made. The writer is inclined to believe that this 
 is due very largely to the difficulty encountered in reaching it. 
 When the roads are improved this lake will be better known and 
 should attract an enthusiastic summer colony. 
 
 CARP LAKE 
 
 Carp Lake is located on the boundary of Emmet and Cheboygan 
 counties about seven miles south of Mackinaw City. The Dixie 
 Highway and the Grand Rapids and Indiana R. R. both pass the 
 west end of the lake. The lake is more than three miles in length 
 and one and one-half miles in greatest width but is shallow. The 
 basin lies within the limits of Lake Algonquin which flooded the 
 region with the exception of a small patch of moraine which borders 
 the northwestern shore of Carp Lake. The original characteristics 
 of the basin are, thus, concealed by the sand which was distributed 
 over this area by the waves and currents of Lake Algonquin. How- 
 ever, the presence of a moraine suggests the possibilities that the 
 depression may be a large morainic basin which was partially filled 
 by sand or a sag in a till plain with a small amount of filling. 
 
 The shores of the lake are rather uniform, and, therefore, no great 
 changes in outline have resulted from the adjustment of the shores. 
 Yet interesting shore features are to be found and among them a' 
 
BORDER LAKES AND LAKES OP DIVERSE ORIGIN 3G7 
 
 series of bars and lainpai'ts aloug tlie shore at Carp Lake Village. 
 These bars are spaeeil on the gentle slope which extends from the 
 lake sliore back to a low cliff and mark former levels of the lake. 
 Two well defined sand bars skirt this shore but in places as many as 
 fonr rampart-bars were seen. Inasmuch as the writer could not 
 determine the elevation of these bars in this work, the relation 
 of the lake levels represented hj them to Lake Algonquin cannot 
 be given. However, the very strong bar of cobbles which cuts off 
 the marsh along the northeast shore must be an Algonquin beach 
 because currents of sufficient power to move rock fragments of 
 cobble size do not develop on lakes of the dimensions of Carp Lake. 
 Another Algonquin bar swings part way around the east end of 
 the lake and is located about two hundred and fifty paces back 
 from the shore. The action on the present shores is limited 
 very largely to ice shove and the main interest in the study of the 
 lake lies in its relation to Lake Algonquin, a more extended study 
 than could be made for tliis work. 
 
 The north shore of the lake is well adapted to summer resort 
 purposes and is very easily reached from the main highway. As a 
 result several groups of cottages have been built and future de- 
 velopment may be expected. 
 
 GRAND TRAVERSE LAKES , 
 
 DUCK^ GREEN^ LONG AND SILVER 
 
 Another group consisting of more than a dozen lakes, large and 
 small, is located a few miles southwest of Traverse City and of 
 these lakes Long, Silver, Duck and Green were visited. These 
 lakes all occupy pits in an outwash plain and those seen by the 
 writer have a north-south trend. Some of the lakes show similar- 
 ities in other respects, which are well brought out from a study of 
 Green and Duck lakes. 
 
 GREEN AND DUCK LAKES . 
 
 These two lakes, which are separated by a strip of outwash of 
 hardly more than one-fourth mile in width, might well have been 
 named Twin Lakes. In addition to a like orientation and origin, 
 their area and form are very similar. Furthermore each has a 
 projection on the east side, that on Duck being the more prominent. 
 Also the adjustments of the shores are alike in character, in both 
 cases being in a veiy early stage. In fact, very little adjustment 
 
368 INLAND LAKES OF MICHIGAN 
 
 has taken place, although the lakes are approximately three miles 
 in length and one mile in width, and they are not especially inter- 
 esting from this viewpoint. The most prominent features are 
 the steep cliffs, but apparently little recession has taken place. 
 Currents have accomplished almost nothing except to remove the 
 material from the cliffs, and depositional forms are rare. At the 
 end of the peninsula on Green Lake, for instance, one might expect 
 a large deposit but finds instead a relatively small- submerged bar. 
 Likewise on Duck Lake but one or tw^o bars were found and they 
 crossed very minor indentations. 
 
 The problem in connection with the study of these lakes is to 
 account for the lack of adjustment of the shores. We may imme- 
 diately eliminate the size and time elements in this consideration, 
 for nearby lakes which are smaller and have been existent for a 
 similar length of time show a much greater development of the 
 shores, but farther than that the problem is as yet unsolved. 
 
 Long stretches of the shores are suitable for resort purposes 
 and the fishing is considered good, so that the lakes are popular. 
 Green Lake, however, appears to be the choice for those who build 
 summer cottages but for the itinerant recreationist with camping 
 outfit the State Park between the two lakes furnishes an excellent 
 location on either. This park of eight}^ acres is almost unique for 
 it is one of the few remaining tracts of Michigan's once famous 
 pine forests. A trip to this region is well worth while and is easily 
 made from Interlocken, the junction point of the Manistee and 
 Northeastern and Pere Marquette R. R. 
 
 LONG LAKE, GRAND TRAVERSE COUNTY 
 
 Long Lake w^hich is located some six miles southwest of Traverse 
 City is one of the most attractive lakes in the vicinity. It is more 
 than four miles long and nearly two miles in greatest width but 
 is so irregular that its size is not appreciated on first sight. The 
 view^ is also obstructed by several islands which add to the picture. 
 One must drive to the lake but will surely encounter difficulties with 
 sand roads to the south if the attempt to circle the lake is made by 
 motor. Such a trip, however, will emphasize the sandy character of 
 the outwash in which this lake lies for the basin is a pit, but a most 
 irregular one. This irregularity in addition to the islands, which 
 interfere Avith the full development of waves, may give rise to ex- 
 pectations of poorly adjusted shores; however adjustments have 
 taken place. 
 
 At the north end one is in doubt as to whether the swamp was 
 
BORDER LAKES AND LAKES OP DIVERSE ORIGIN 369 
 
 cut off from the main lake or uot, for the bar, if such it be, is 
 faint, but evidence may be found of a former level of the lake 
 about three feet above the present level, at which stage the swamp 
 was surel}' connected. Also the off-shore terrace is well developed 
 here, and in i)laces tlie part exposed by the sinking of the water is 
 still intact. To the southward the bars increase in development 
 and completely close the narrow oi)enings of the embayments. One 
 of these closed embayments on the shore extends some distance in- 
 land and is occupied by Mickey Lake. 
 
 It is not our purpose to cite all of the individual examples of 
 adjustments. As a working basis, however, it may be assumed 
 that the smaller embajmieuts are cut off. The larger bays are open 
 to the lake as would be expected from the exceptionally tortuous 
 shoreline along which continuous currents could not develop, 
 even though no islands obstructed! the reach of the waves. The 
 headlands have furnished the material for the deposits and have 
 retreated somewhat under the attack of the waves. But the prog- 
 ress of the adjustment of the shores was interrupted at an early 
 stage by subsidence of the water and little has been done since 
 that time. 
 
 As a summer resort Long Lake has lagged in comparison to many 
 others that are less attractive. This is due in part to its location 
 near the beautiful Grand Traverse Ba}- but also to some extent to 
 its inaccessability. Koads are being improved and people are 
 using lakes more and more for recreation, so that this lake will 
 surely share in summer resort development. 
 
 SILVER LAKE, GRAND TRAVERSE COUNTY 
 
 Another lake of the group under consideration is Silver, lo- 
 cated about two miles east of Long Lake. It also lies in a pit in the 
 same outwash plain but very near the margin of the moraine, so 
 near in fact that the narrow^ basin continues northward from the 
 north end of the lake as a stream valley in the moraine. The lake 
 averages less than one-half mile in width but is more than five 
 times this figure in length. It actually appears muck narrower 
 on account of the numerous islands which obstruct the view across 
 the lake. Furthermore, the lake is ver^- deep and the black water 
 comes close to the shore. This makes the lake verv' dangerous 
 and is detrimental to its development as a summer resort. 
 
 It is, however. nu>st interesting as a physiographic study. It has 
 no outlet and varies greatly in level over a period of years. At 
 present it stands nearly five feet below the highest water mark 
 47 
 
370 , INLAND LAKES OF MICHIGAN ' 
 
 which has determined the elevation of the features formed by the 
 adjustment of the shores. This higher level may be reached during 
 exceptionally wet periods, but the higher levels which are almost 
 universally found on other lakes make it more probable that this 
 is a permanently abandoned higher level. 
 
 It is difficult to select a starting point for the discussion of the 
 numerous shore features of Silver Lake unless one describes them 
 all in detail. Waves have been very active but there has been 
 little cliff recession. Of the material derived from the cliffs some 
 has undoubtedly been carried out by the undertow but it was not 
 sufficient in quantity to build a wide off-shore terrace, which would 
 require an excessive amount of material in this deep lake. The 
 material which was worked along the shore by currents, however, 
 although poissibly not greater in amount, has produced more 
 noticeable effects. All of the indentations except two are small 
 and have been cut off by bars. And of the two larger embayments 
 one only, the large bay on the southeast side, is open. The other 
 is at the north end and is completely bridged by a' strong bar which 
 separates the small Mud Lake from the main basin. In this con- 
 nection the islands are interesting. The original islands were 
 nine in number most of which are distributed off the east shore 
 opposite Silver Lake Kesort, but this number is now reduced by 
 at least three by the development of bars which either tied 
 islands to each other or the mainland. The details of these bars 
 were not worked out and should prove an interesting part of the 
 study of the numei'ous physiographic features on the shore of 
 Silver Lake which oft'ers such a decided contrast to the nearby 
 Duck and Green lakes. 
 
 BEAR LAKE 
 
 Bear Lake is located in northwestern Manistee County about 
 midway between Manistee and Beulah, eighteen to twenty miles 
 distant. Such distances seem rather long but the road in each 
 case is the West Michigan Pike which was found to be in excellent 
 condition. An automobile ride of this distance is preferable to 
 the trip of seven miles from Norwalk which has a one-train-a-day 
 service on the Manistee and iS^ortheastern R. R. 
 
 The lake is an open expanse of water of very regular outline and 
 has dimensions of two and one-fourth by one and one-fourth miles, 
 the longer axis having an east-west direction. The western part 
 of the basin hardly exceeds fifteen feet in depth but depths of fifty 
 to sixty feet are reported for the eastern part. The glacial forma- 
 tions may be readily made out as one approaches the lake on the 
 
BORDER LAKES AND LAKES OF DIVERSE ORIGIN 371 
 
 Pike frain the south. At first tlio route is through a rugged inoraiue 
 from the crest of which a broad out wash phiiu may be seen below 
 in the distance, extending northward to another moraine some six 
 miles awar. Beyond the crest the moraine slopes sharply to the 
 north and the entire lake appears in view, extending from the foot 
 of the slope onto the outwash. The basin is limited on the east side 
 by upland but stretches eastward as a low, heavily-wooded swamp 
 to Bear Creek, several miles away. Briefly stated the lake basin is 
 part of a depression in the outwash at its junction with the moraine. 
 This depression has some of the characteristics of a fosse but is in 
 part a pit, as shown by the deep basin which forms the eastern part 
 of the lake. 
 
 The shore features of Bear Lake are very simple because the 
 shore conditions are singularh* uniform. The shore is lined by 
 wave-cut cliffs except at the east end. Where the depression con- 
 tinues eastward conditions were favorable for deposition by cur- 
 rents, and. in spite of the presence of a' road around this shore, a 
 well defined bar ma^' be traced across the flat to the higher ground 
 on the north side. Also a similar form developed eastward in 
 front of the swamp at the northeastern part of the lake but was 
 never completed. These forms stand well above the present level 
 and mark a higher stage of the lake. At present the shore agents 
 are rejuvenated and the recession of the cliffs is very evident. The 
 reason for this revival of activity' is high water but the cause of the 
 high water is uncertain. The natural conditions of the outlet 
 which flows from the east end have been altered by the building of 
 a road, and, in particular, the position of the outlet has been shifted 
 south of its normal position. This may account for the filling of 
 the artificial outlet which functions only at high water and has, 
 no definite channel for tsome distance east of the bar. 
 
 The lake has not yet developed as a summer resort, although it is 
 well suited for such use. This seems to be due to the fact that it 
 was formerly isolated, for, since the construction of the West Mich- 
 igan Pike and the thereby more convenient access to the lake, it is 
 becoming popular. 
 
INDEX 
 
INDEX 
 
 A 
 
 Algonquin Lake: Page 
 
 described 11,12 
 
 cliffs, shores, terraces, etc., of 11,12 
 
 Black Lake 100, 102 
 
 Burt Lake 80 
 
 Carp Lake 366,367 
 
 Cheboygan River Basin 71 
 
 Crooked Lake 72 , 73 , 77 
 
 Crystal Lake 169 , 172 , 175 
 
 Douglass 108,109,113 
 
 Elk Lake 159,163,164,165,166 
 
 Glen Lake 361 
 
 Grand Traverse Region 122 
 
 Houghton County 306 
 
 Ives Lake 293 
 
 Long Lake, Alpena County 242 , 244 
 
 Mullet Lake 90,95,96,97 
 
 Pine Lake 135,139,140,141 
 
 Platte Lakes 306 
 
 Rush Lake 292 
 
 Torchlight Lake 145,146,147,149,150,151,153,154,155,156,158 
 
 Van Etten Lake 363 
 
 AValloon Lake 129,133 
 
 Altitude of Northern Peninsula 15 
 
 Southern Peninsula 18 
 
 Arch, sea, development of 52 
 
 Atmosphere, the work done by 3 
 
 Austin Lake, Kalamazoo County, description and discussion of 333-334 
 
 B 
 
 Bar, development of by currents of present and former lakes 53 . 54 
 
 Bass Lake 357 
 
 Bear Lake 371 
 
 Big Blue 351 
 
 Big Star 349 
 
 Black, Cheboygan County 102,103,104,105 
 
 Brevort 271 , 272 , 273 
 
 Burt 79,81,84,85 
 
 Cadillac 204 
 
 Carp 367 
 
 Cass 260 
 
 Cavanaugh 324 
 
 Chippewa 338 
 
 Clear 187,189,190,191 
 
 Conway 290 
 
 Corey 183,184,186 
 
 Crystal, Benzie County 169,171,172-4,175 
 
 Crystal, Montcalm County 345-6 
 
 Devils and Round 327 
 
 Douglass 109,110,111,115 
 
 Duck 338 
 
 • Elk 162,163,164 
 
 Fremont 350 
 
 Glen .361 
 
 Gogebic 316,317 
 
 Gull 331 
 
 Gun 197 
 
 Hesse 339 
 
 Higgins. . .-. 217-8,220,221,222 
 
 Houghton y 211,212,213,214 
 
 Hubbard 236 , 238-9 , 240 
 
 Khnger 179,180,181 
 
 Long, Genesee 266 
 
 Long, Grand Traverse County 369 
 
 Long, Iosco County 365 
 
 Manistee 348 
 
 Manistique 278-9 
 
 Michiganune •. 300 , 301 
 
 Missaukee 348 
 
 Mitchell 205,206,207 
 
376 INDEX 
 
 Mullet 89,97 
 
 Orchard 255,257 
 
 Otsego ;^^ 230,231,232,233 
 
 Pine, Charlevoix County v 135,137,138,141,142 
 
 Pine, Huron Mts 287 , 289 
 
 Platte 359 
 
 Portage, Crawford County 225,226 
 
 Portage, Houghton County ' 308 
 
 Portage, Manistee County 35S 
 
 Silver 370 
 
 Sister Lake 343 
 
 Torchlight 148 , 153 , 155 , 157 
 
 Twin Lakes 352 
 
 Van Etten 363 
 
 Wall 333 
 
 Walloon 126.128,131 
 
 Wamplers 325 
 
 Whitefish , 280 
 
 Wolf 353 
 
 Bar, submerged development of 53 
 
 Barrier beaches, development of 50 
 
 Bass Lake, Washtenaw County, discussion of 337 
 
 Basins of Lakes, origin of 22-35 
 
 Bass Lake, Baraga County, extinction of 299 
 
 Bass Lake, Mason County, discussion of 357-8 
 
 Bawbeese Lake, Hillsdale County, discussion of 328 
 
 Beach barrier, development of 50 
 
 Beach defined, formation of. See also individual lakes 48 
 
 Beach, shingle, formation of 48 
 
 storm, formation of 48 
 
 Bear Lake, Manistee County 370-371 
 
 Beaver dam, a bar 232 
 
 Beebe Lake, Oakland County, formation of 257 
 
 Betsie Lake, Benzie County, discussion of 356 
 
 Big Blue Lake, Muskegon County, discu.ssion of 351 
 
 Big Clam (Lake Mitchell), Wexford County, description and discussion of 203 
 
 Big Spring, description of 284 
 
 Big Star I-ake, Lake County, discussion of 349 
 
 Black Lake, Cheboygan County, description and discussion of 69,99-106,356 
 
 Bog, floating on Gun Lake 199 
 
 Border Lakes, discussion of 355,371 
 
 Boulder clay, defined 8 
 
 in ground moraine 9 
 
 Boulder wall, of Crystal Lake 171 
 
 Breakers, formation of 42 
 
 Brevort Lake, Mackinaw County, description and discussion of 269-275 
 
 Burt Lake, Cheboygan County, description and discussion of 78-87 
 
 C 
 
 Cadillac Lake, Wexford County, description and discussion of 203 
 
 Canyon Lake, Marquette County, description and discussion of 26,295 
 
 Carp Creek, delta of 82 
 
 Carp Lake, Emmet and Cheboygan counties, description and discussion of 366,367 
 
 Cass Lake, Oakland County, description and discussion of 253 ,258-261 
 
 Cavanaugh Lake, Washtenaw County, description and discussion of 324 
 
 Cave, limestone, formation of 6 
 
 sea, development of by waves 51 
 
 Charlevoix, location of 119, 136 
 
 Cheboygan moraine, described 69 
 
 Cheboygan River Basin, lakes, topography of 69-117 
 
 Chemical precipitation, lake extinction by. See Marl 64,66 
 
 Chicagon Lake, Iron County, description and discussion of 318-321 
 
 Chippewa Lake, Mecosta County, description and discussion of 338-339 
 
 Cincinnati Point on Crooked Lake 73 
 
 Clark Lake, Jackson County, discussion of 326 
 
 Classification of lake basins 3 , 22-35 
 
 Clay, boulder, defined 8 
 
 Clear Lake, St. Joseph County, description and discussion of 187-190 
 
 Cliffs, undercut formation of 49,51 
 
 Coldwater Lake, Branch County, discussion of 328,329 
 
 Isabella County, discussion of 346-347 
 
 Comber, defined 43 
 
 Continents, area of 3 
 
 Conway, location of on Crooked Lake 73 
 
 Conway Lake, Marquette County, discussion of 287,290-291 
 
 Copper Range, location of 311 
 
 Cora Lake, Van Buren County, discussion of 341-342 
 
 Corey Lalce, St. Joseph County, description and discussion of 182-187 
 
 Crooked Lake, Barry County, discussion of 332 
 
 Emmet County, description and discussion of 69,72-78 
 
 Washtenaw County, discussion of 32 , 325 
 
 Crooked River, described 76 , 78 
 
 Crystal Lake, Benzie County, description and discussion of 119, 167-175 
 
 Montcalm County, discussion of ' 345 
 
 Cuestas, defined 17 
 
 of Northern Peninsula, described 17 
 
INDEX 377 
 
 Page 
 
 Current action, effect of on shores. See also bars, hooks, lakes, spits R^-^d 
 
 Currents, characteristics, definition, work of 43-46 , 5,3-o.t 
 
 C"usp. description and explanation of 111-112 
 
 on Douglass Lake Ill 
 
 Houghton Lake 213 
 
 Lake Mitchell . . • 207 
 
 Loiif? Lake, Alpena County 246 
 
 Ciispate foreland defined .5.5 
 
 on Lake Michiganime 301 
 
 on Van Etten Lake 364 
 
 Cut-terrace defined 48 
 
 Cycle of shore development 66-67 
 
 Cyclonic storms, character of 60 
 
 D 
 
 Delta, in Bass Lake 3.57 
 
 Mlack Lake 10-1 1 
 
 Burt Lake 80,81 
 
 drowned valley lakes 356 
 
 Fremont Lake 350 
 
 Indian Lake, of Indian River 284 
 
 Mountain Lake '. 295 
 
 Pine Lake, Marquette County 290 
 
 Portage Lake, Houghton County 307,308 
 
 of Sturgeon Kiver 65 
 
 Deposition in Brevort Lake 274 
 
 Devils Lake, Lenawee County, discussion of 327 
 
 Dewey Lake, Cass County, discussion of -. . 342 
 
 Diamond Lake. Cass County, discussion of 334 
 
 Diastrophism, lake basins caused by 26-27 
 
 Divides, formation of 4 
 
 Douglass Lake, Cheboygan County, description and discussion of 106-117 
 
 Dowagiac, glacial lake, referred to 196 
 
 Drift, glacial, defined 8 
 
 Drowned valleys cause of, lakes in 356 
 
 Drumlins, described 9 
 
 in Iron County 16 
 
 in Pine Lake region 134,140,142 
 
 near Elk Lake 161 , 165 
 
 near Torchlight Lake 146 , 1.55 
 
 Duck Lake, Calhoun County, discussion of 337-3.38 
 
 Grand Traverse County, discussion of 367 
 
 Dunes, defined, de.scribed 4 
 
 of Northern Lowlands 18 
 
 Dunes, near Black Lake 102 
 
 Brevort Lake 269 
 
 Crystal Lake, Benzie County 170, 171 , 172 
 
 Douglass Lake 144 
 
 Elk Lake 165 
 
 Glen Lake 361 
 
 Indian Lake 284 
 
 Lake Michigan 355 
 
 Little Traverse Bay 71 
 
 Pine, Charlevoi?: County 142 
 
 Portage, Manistee County ." 3.58 
 
 Torchlight 143 
 
 E 
 
 East Lake. Muskegon County, discussion of 3.52 
 
 Eastern Interlabate Area, Lakes of 25.3-267 
 
 Elk Lake, Antrim and Grand Traverse counties, description and discussion of ... . 119, 159-166 
 
 Erie Lowland, location of 18 
 
 Esker, defined, described 9,10 
 
 Extinction of lakes, discussion of 61,66 
 
 See also discussion of individual lakes. 
 
 by chemical precipitation 64 
 
 by climatic changes 64 
 
 draining ■ 64 
 
 sedimentation 62 
 
 vegetation 02-64 ,91 
 
 F 
 
 Fosse, defined, lakes of 31 , 32 
 
 Cavanaugh Lake, a 324 
 
 Crooked Lake, Washtenaw County 32 , 325 
 
 Fremont Lake, discussion of 350 
 
 G 
 
 Garden Peninsula. Algonquin shore of 11 
 
 Genesee County, Lakes of 261 
 
 Geological Time Scale 15 
 
378 INDEX 
 
 Page 
 
 Geology and geological history of Lake Superior district 303-306 
 
 Marquette Range 296-7 
 
 vicinity of Grand Lake, Alpena County 362 
 
 Huron Mountain area 286-289 
 
 Indian Lake 281 
 
 Glaciers, continental work of 6-13 
 
 in Highland Province 16 
 
 in Michigan 7-13 
 
 Glen Lake, Leelanau County, description and discussion of 360-361 
 
 Gogebic Lake, Gogebic and Ontonagon counties, description and discussion of 310-318 
 
 Gogebic Range, location of 312 
 
 Goguac Lake, Calhoun County, discussion of 330 
 
 Gourd Neck Lake, Kalamazoo County, discussion of 333-334 
 
 Grand Lake, Presque Isle County 361-363 
 
 Grand Traverse Region, Lakes of 119-175 
 
 Great Lakes, predecessors of 11 
 
 Great Salt Lake, referred to 65 
 
 Green Lake, Grand Traverse County, discussion of 367 
 
 Ground moraine, defined 9 
 
 Ground water, work of 6 
 
 lake basins caused by 25 
 
 Gull Lake, Kalamazoo County, discussion of 330-331 
 
 Gun Lake, Barry County, description and discussion of 195 
 
 Gun River, vegetation in 196 
 
 H 
 
 Hay Lake, Emmet County, referred to 76 
 
 Hess Lake, Newaygo County, discussion of 339 
 
 Higgins Lake, Roscommon County, description and discussion of 215-224 
 
 Highland Province altitude, boundaries, topography of 15-16 
 
 Hooks, defined 54 
 
 described on Black Lake 105 
 
 Burt Lake 82 , 84 
 
 Corey Lake 186 
 
 Crooked Lake, Emmet County 76 
 
 Elk Lake 164 
 
 Long Lake, Alpena County 244,246,249,250,251 
 
 Otsego Lake 234 
 
 Torch Lake 309 
 
 Torchlight Lake ■ 152 
 
 Walloon Lake 122-3 , 128 , 129 
 
 Houghton Lake, Roscommon County, description and discussion of 30,209,215 
 
 Hubbard Lake, Alcona County, description and discussion of 235-242 
 
 Huron-Erie Moraine, described 10, 11 
 
 Huron Lobe, characteristics of 209 
 
 Huron Mountain Lakes, description and discussion of 286-296 
 
 Huron Mountain Club, location of 289 
 
 Huron Mountains, elevation and description of 286 
 
 geology and geological history of 286-289 
 
 I 
 
 Ice, effect of on shore 56-61 ,83 
 
 Ice-jams discussion of 59-61 
 
 ramparts produced by 60 
 
 on Black Lake 104 
 
 Brevort Lake 271 ,273 
 
 Burt Lake 82,83-84,91 
 
 Crvstal Lake 170, 171 
 
 Elk Lake 161 
 
 Higgins Lake 219 , 220 
 
 Houghton Lake 213 
 
 Indian Lake 286 
 
 Klinger Lake 182 
 
 Lake Mitchell 206 
 
 Long Lake, Alpena County 248,249,250 
 
 Manistique Lake 279 
 
 Pine Lake 139 
 
 Torchlight Lake 153 
 
 Walloon Lake 123,129,131 
 
 See also ramparts and 161, 170 
 
 Ice-push terrace, formation of 59 
 
 Ice ramparts, formation of 56-59 
 
 on Brevort Lake 272 ,273 
 
 Burt Lake 82 
 
 Cadillac Lake 204 
 
 Carp Lake 367 
 
 Chicagon Lake 321 
 
 Chippewa Lake 338 
 
 Corey Lake ^ _'i^8S 
 
 Crooked Lake, Emmet County 73 , 74 . 75 , 76 
 
 Crystal Lake, Benzie County 169,170,175 
 
 Crystal Lake, Montcalm County 345-6 
 
 Douglass Lake 111,113 
 
 Elk Lake 161 , 165 
 
INDEX 379 
 
 Page 
 
 Fremont Lake 3o0 
 
 Gogebic Lake 318 
 
 Gun Lake 107,200 
 
 Hesse Lake 339 
 
 Higgins Lake 210,223 
 
 Hubbard Lake 239,240,241 
 
 Indian Lake, Schoolcraft County 283 
 
 Klinger Lake 180, 181 
 
 Long Lake, Alpena County 244, 24'., 246, 247, 249, 250, 251 
 
 Long Lake, Genesee County 265,266 
 
 Mitchell Lake 205,206,207 
 
 Mountain Lake 294 
 
 IMullet Lake 91,92,96,08 
 
 Orchard Lake 2.56,257 
 
 Otsego Lake 230,231 
 
 Pine Lake, Charlevoix County 139,140,141,142 
 
 Portage Lake, Crawford County 225,226 
 
 Torchlight Lake 151 , 153, 156 
 
 Wall Lake 333 
 
 Availed Lake 335 
 
 Walloon Lake 122-3,125,127,128,129,131,132 
 
 Ice shore, how exerted, effect of 56-61 
 
 . on Burt Lake 84 
 
 Grand Lake, Alpena County 363 
 
 Gun Lake, force of 198,200 
 
 Long Lake, Alpena County 248,250 
 
 Long Lake, Genesee County 264 
 
 Orchard Lake 257 
 
 See also Ice-ramparts, Ice-jams. 
 
 Indian Lake, Cass County, discussion of 343 
 
 Indian Lake, Schoolcraft County, description and discussion of 281-286 
 
 Indian River, outlet of Burt Lake described 79 
 
 "Inland Route" defined 69 
 
 Irish Hills, lakes seen from 327 
 
 Iron County, glacial deposits of 16 
 
 Iroquois. Lake, referred to 11 
 
 Islands, land-tied development of 54 
 
 in Chippewa Lake 339 
 
 Corey Lake 185 
 
 Gun Lake 200 
 
 Lake Michigamme 303 
 
 Middle Lake 352 
 
 Missaukee Lake 348 
 
 Silver Lake 370 
 
 Ives Lake, Marquette County, discussion of 287,203,296 
 
 K 
 
 Kaiser Lake, St. Joseph County, referred to ' 183 
 
 Kalamazoo, Allegan County, a border lake, location of 356 
 
 Kalamazoo, morainic and outwash system, lakes of 323-334 
 
 Kames, defined, location of 8,9 
 
 Klinger Lake, St. Joseph County, description and discussion of 178-182 
 
 L 
 
 Lagoon, formation of 34 , 50 
 
 Brevort Lake, a 270 
 
 Conway Lake, a 290 
 
 Crystal Lake, a 172 
 
 Indian Lake, a 281 
 
 Lake basins, origin, classification of 3 ,22-35 
 
 classes of in Michigan 23 
 
 Lake shores, development of (See also various lakes) 37-67 
 
 Lake Superior sandstone, location of 286 
 
 Lakes, extinction of 61-66 
 
 by change of climate 64-65 
 
 chemical precipitation 64 , 66 
 
 drainage 64 , 65 
 
 marl 62 
 
 vegetation 62-64 , 66 
 
 of Alpena County 242-252 
 
 Cheboygan River Basin 61-117 
 
 Grand Traverse Region 1 19-175 
 
 St. Joseph County 177 
 
 of Western Interlobate Area 177-242 
 
 Lake types, discussion of ♦ 23-50 
 
 drowned valley 27 , 356 
 
 glacial 28-35 
 
 fosse 31 
 
 glacial scour 28 
 
 inter morainal 30 
 
 lagoon 34,50 
 
 morainal 28 
 
 morainal dam 31 
 
 pit 31 
 
380 INDEX 
 
 Page 
 
 oxbow i 23 
 
 rift valley 26 
 
 sink 25 
 
 Land-tied islands, development of 54 , 74 , 75 
 
 in Chippewa Lake 339 
 
 Corey Lake 185 
 
 Gun Lake 200 
 
 Lake Michigamme 303 
 
 Middle Lake 352 
 
 Missaukee , 348 
 
 Silver Lake ,. 370 
 
 Level of lakes ; lowering of, effect of x 64 
 
 raising of, effect , . x 76 
 
 Leverett, Frank, referred to 35 
 
 Limestone, cliff of, on Black Lake 103 
 
 near Indian Lake 281 , 284 
 
 near Long Lake, Alpena County 243 ,247 
 
 Little Clam, or Cadillac Lake, Wexford County, described. 203 
 
 Little Traverse Bay, Nipissing bar of 71 
 
 Long Lake, Alpena County, description and discussion of 242-2.52 
 
 Long Lake, Genesee County, description and discussion of 261-267 
 
 Long Lake, Grand Traverse County, description and discussion of 367-368 
 
 Long Lake, Iosco County, discussion of 364 
 
 Long Lake, Kalamazoo County, discussion of 333,334 
 
 Long Lake, St. Joseph County, discussion of 190-195 
 
 Lowland Province; extent, altitude, topography, etc 16-18 
 
 M 
 
 Magician Lake, Cass County, discussion of 342 
 
 Manistee Lake, Kalkaska County, discussion of 348-349,356 
 
 Manistique Lake, Luce and Mackinaw counties, description and discussion of 275,277-279 
 
 Manistique River, drainage system 18,276-277 
 
 Marble Lake, Branch County, discussion of 328,329 
 
 ]Marl, a source of lake filling, precipitation of 62-66 
 
 beds of near A lanson 77 
 
 near Higgins Lake 217 
 
 in Burt Lake 79 
 
 in Coldwater Lake, Branch County 329 
 
 in Corey Lake 184 
 
 in Crooked Lake, Emmet County 77 
 
 in Klinger Lake 181 
 
 in Long Lake, Alpena County 245 
 
 in Long Lake, Genesee County 264 
 
 in Marble Lake 329 
 
 m Mud Lake, Genesee Coimty 261 
 
 in Mullet Lake 92 , 98 
 
 in Pine Lake, Charlevoix County 137 
 
 in Sturgeon River Channel 62 
 
 in Walloon Lake 125 , 128 
 
 Marquette, elevation of Algonquin Shores, near 11 
 
 Marquette Range, geology of 296-7 
 
 Michigamme Lake, Marquette County, description and discussion of 296-303 
 
 Michigan, Lake, characteristics of 10, 11 ,209 
 
 Aliddle Lake, Muskegon County, discussion of 352 
 
 Miner Lake, Allegan County, discussion of 340 
 
 Missaukee Lake, Missaukee County, discussion of 347 
 
 Mitchell, Lake, Wexford County, description and discussion of 203,205-208 
 
 Monadnochs, defined 6 
 
 Morainic system, map of 10 
 
 Moraines, descriptions and discussions of 8,9,10,11,12,16,209,210,227,236,261 
 
 Moraine, ground, defined 9 
 
 Mountain Lake, Marquette County, description and discussion of 287,293-296 
 
 Muck, areas of 197,199,258,284 
 
 Muck areas of Cass Lake 258 
 
 Gun Lake 197 
 
 Indian Lake 284 
 
 Mud Lake 199 
 
 Mullet Lake, Cheboygan County, description and discussion of 87-99 
 
 Muskegon Lake, Muskegon County, discussion of 356 
 
 N 
 
 Natural bridges, formation of , 6 
 
 Niagara Falls, referred to 65 
 
 Nipissing Lake, described 12 
 
 shores, terraces cliffs of, near Bass Lake 367 
 
 Black 100 
 
 Brevort 269 
 
 Burt 78,79,80,81,82,83,84,85 
 
 Cheboygan River Basin 71 
 
 Crooked Lake, Cheboygan Co 72.73,75 
 
 Elk Lake 159,161,162,163,164,165,166 
 
 Glen Lake 361 
 
 Houghton County . 306 
 
INDEX 381 
 
 Page 
 
 Indian Lake, Sclioolcraft Co 281 
 
 Indian River. 79 
 
 Lake Michigan 356 
 
 Mullet Lake 90,91,92,93,95,96,97 
 
 Portage Lake, Manistee Co 358 
 
 R iish Lake ,<i 292 
 
 Torch Lake 309 
 
 Torchlight Lake 145, 146, 149, L50, L51 , 153, 155, 156, 157, 1.58 
 
 Van Etten Lake 363 
 
 North Lake, Muskegon County, discussion of 362 
 
 Northern Peninsula, location, area, topography of 13-16 
 
 Lakes of 269-321 
 
 O 
 
 Oakland County, lakes of 253 
 
 Ontonagon, Lake, area of former 313 
 
 Ontonagon River, history of 313-315 
 
 Orchard Lake, Oakland County, description and discussion of 253-258 
 
 Osceola County, elevations in 18 
 
 Oscillation, waves of, defined and described 38 
 
 Otsego Lake, Otsego County, description and discussion of 227-234 
 
 Otter Lake, Oakland County, discussion of 255 
 
 Ottawa Lake, Monroe County, a mud hole 345 
 
 Ottawa Lake, Monroe County, a sink • 26 
 
 Outwash plain, defined 9 
 
 Oxbow lakes, defined and described 23-25 
 
 Palaeozoic rocks in Iron River District 16 
 
 I^owland Province 17 
 
 Southern Peninsula 18 
 
 Paw Paw Lake, Berrien County, discussion of 340-41 
 
 Peat bog, development of, in Long Lake, St. Joseph County 191 
 
 Peneplain, defined 5 
 
 Pentwater Lake, Oceana County, discussion of 356 
 
 Pere Marquette Lake, Mason County, discussion of 356 
 
 Phase, of wave, defined 38 
 
 Physiography of Michigan 13-21 
 
 Pike Lake, Kalamazoo County, discussion of 333 
 
 Pilot Rock, striations on 303 
 
 Pine Lake, Charlevoi.x County, description and discussion of 119, 134-143 
 
 Pine Lake, Huron Mountains, description and discussion of 287,289-290 
 
 Pine Lake, Marquette County, description and discussion of 289-290 
 
 Pine Lake, Oakland County, discussion of 255 
 
 Platte Lakes, Benzie County, discussion of 359-360 
 
 Porcupine Mountains, elevation of 15,286 
 
 Port Huron morainic and outwash systems, lakes of 345-353 
 
 Portage Lake, Crawford County, discussion of 224 
 
 Portage Lake, Houghton County, description and discussion of 303-310 
 
 Portage Lake, Manistee County, discussion of 358-359 
 
 Portage Lake, Marquette County, discussion of 337 
 
 Post-Nipissing shores, terraces of on Burt Lake 81 ,85 
 
 on Crooked Lake 74 , 76 , 77 
 
 Elk Lake 166 
 
 Mullet I-ake 90 , 91 92 , 93 , 95 , 97 
 
 Pine Lake 135,137.138,140,141,142 
 
 Torchlight Lake 150, 158 
 
 Preface vii-xiii 
 
 Pre-Paleozoic rocks of Northern Peninsula 15, 16 
 
 R 
 
 Ramparts, ice, see Ice-Ramparts 56-61 
 
 Reach, of wind, defined 37 
 
 Reed Lake, Kent County, discussed 339 
 
 Rice, Wild, in Houghton Lake 210 
 
 Rift-valley lakes, described 26 
 
 River systems, formation of 4 
 
 Roches montonnes, defined and described 295 
 
 Round Lake, Benzie County 163, 171- 
 
 Round Lake, Kalkaska County, references to 136 , 157 
 
 Round Lake, Lenawee County, di-scussion of .327 
 
 Running water, work of 4-6 
 
 Rush Lake, Huron Mountain, description and discu.ssion of 287,291-293 
 
 Sage Lake, Ogemaw County, discussion of 365 
 
 St. Joseph County, Lakes of 177-195 
 
 Sandstone, Lake Superior on shores of Rush Lake 286, 291 ,293 
 
 Schall, M. Eugene., cited 193 
 
 Sedimentation, lake extinction by 64,65 
 
 Shale, outcrop of on Pine Lake 127 
 
 on Walloon Lake 138 
 
 Sheridan Hill, a Paleozoic formation 16 
 
382 INDEX 
 
 Page 
 
 Shingle beach, development of 48 
 
 Shore, currents. See Studies of lakea, spits, hooks and 53-55 
 
 .Shore, development cycle of 50 , 66-67 
 
 recession of 94 
 
 Shore, of Algonquin and Nipissing Lakes. See also Algonquin, Nipissing 11 ,71 
 
 Silver Lake, Grand Traverse County, discussion of 367,369 
 
 Sink holes, formation' of, lakes in 6 
 
 NtiMl near Long Lake, Alpena County 21 ,25,242-243 
 
 Southern Peninsula of Michigan, altitude, topography, etc 13-21 
 
 Spring Lake, Ottawa County, discussion of 356 
 
 Spits, development of by currents 54 
 
 on Black Lake 101 , 105 
 
 Brevort Lake 272 ,274 
 
 Burt Lake 80 
 
 Cass Lake 258,260 
 
 Chicagon 321 
 
 Chippewa 338 
 
 Clear 187 ,189 , 190 , 192 
 
 Coldwater, Isabella County 346 
 
 Corey 183,184,186 
 
 Crooked 343 
 
 Crooked , Emmet County 74 , 75 
 
 Crystal, Benzie County 169, 171 
 
 Dewey 343 
 
 Douglass 109,110,113,114 
 
 Elk 161 
 
 Fremont 350 
 
 Gogebic 317 
 
 Gun 199,200 
 
 Higgins 218,219,220 
 
 Hubbard 237 
 
 Indian, Cass County 344 
 
 Indian, Schoolcraft County 283 
 
 Klinger 179 , 180 
 
 Long, Alpena County 245,247,249,250,251 
 
 Long, Genesee County 264,265 
 
 Long, Iosco County 365 
 
 Manistee 348 
 
 Manistique 278 
 
 Middle 352 
 
 Michigamme 302 
 
 Missaukee 348 
 
 Mullet 89,91,96 
 
 Orchard 256 
 
 Otsego 231 
 
 Paw Paw 341 
 
 Pine, Charlevoix County 139 , 140 
 
 Pine, Marquette County 289 
 
 Portage, Crawford Coimty 225 
 
 Round 163 
 
 Rush 292 
 
 Sage 366 
 
 Torchlight 148 , 154 , 155 , 157 
 
 Walloon 123,124,125,126,127,128,129,130,131 
 
 Whitefish 280 
 
 Wolf 353 
 
 Stacks, development of 52 
 
 State Forest, lake terrace near 219 
 
 State Park, pine forests of 368 
 
 Storm Beaches, development of 48 
 
 on Black Lake 101 
 
 Elk Lake 164 
 
 Indian 284 
 
 Michigamme 301 ,302 
 
 Pine. 136,137 
 
 Torchlight 151 , 156 
 
 Strawberry Lake, discussion of 327 
 
 Sturgeon River, delta of 65,79 
 
 diversion of 77 
 
 marl in channel of 62 
 
 Sugar Loaf Lake, discussion of 324 
 
 Sunken Lake, a sink 26 
 
 Swamps of Northern Peninsula 17 
 
 Sylvan Lake, Oakland County, discussion of 255 
 
 Terrace, cut-and-built, description of 49 
 
 ice-push, formation of 59 
 
 submerged. See various lakes. 
 
 " Thousand Acre Swamp " referred to 206 , 208 
 
 Till, definition of 8 
 
 Till plain, defined, topography of 9 
 
 Tombola, defined. See also Islands, land tied 54 
 
 Topinabee, location of, discussed ., • 90 
 
INDEX 3S3 
 
 Page 
 
 Topography of Cheboygan River Basin 99, 108 
 
 "Indian Route" region. . . : 69,88 
 
 Genesee County 261 
 
 HighlaiKi Province 15 
 
 Houghton County 303-310 
 
 Lowland Province 17 
 
 AValloon Lake region 121-122 
 
 Torchhght Lake, Antrim County, description and discussion of 119, 143-1.')9 
 
 Translation waves of, defined characteristics of 42 
 
 Transverse drainage in Lake Superior region 30.5 
 
 Twin Lakes, Muskegon County, discussion of 3.52 
 
 U 
 
 l^ndertow, defined 44 
 
 work of 48-49 
 
 University of Michigan Camp on Douglass Lake 106 
 
 V 
 
 Valley flat, defined 5 
 
 Valley flats of Sturgeon River described 303 
 
 Valparaiso-Charlotte morainic and outwash system, lakes of 335-344 
 
 Van Etten Lake, Iosco County, discussion of 363 
 
 V-bar, development of 55 , 56 
 
 on Brevort Lake 273 ,274 
 
 Clear Lake 188 
 
 Gogebic Lake 317 
 
 Hubbard Lake 238 
 
 Little Corey Lake 184 
 
 Long, Alpena County 61 ,246,251 
 
 Long, St. .Joseph County 192 
 
 Mountain Lake 296 
 
 Otsego Lake 232 
 
 Torchlight Lake 1,50, 1.52, 154, 155,156 
 
 Vegetation, extinction of lakes bv 02-64 
 
 See also various lakes and. ..77, 80, 91 , 125, 129,163,196, 199,204,207,215,320,343 
 Vineyard Lake, Jackson County, discussion of 32.5. 
 
 W 
 
 Wall Lake, Barry County, discussion of 332 
 
 Walled Lake. Oakland County, discussion of 33.5-336 
 
 Walloon Lake, Charlevoix County, description and discussion of 119, 120-134 
 
 Wamplers Lake, Jackson and Lenawee counties, discussion of 325 
 
 West (Pike) Lake, Kalamazoo County 333-334 
 
 West Lake, Muskegon County 352 
 
 Watersmeet, headwaters of Wisconsin, Ontonagon and Menominee River near 16 
 
 Wave, characteristic and development of 37-43 
 
 of oscillation, defined 38 
 
 work of. See also various lakes 46,47-53,55-56,80,82 
 
 Wave base, defined 40 
 
 Weathering, agencies, of, definition of 3 
 
 AVhite Lake. Muskegon County, discussion of 356 
 
 White cap, defined 40 
 
 Whitefish Lake, Mackinaw County, discussion of 279-281 
 
 AVhitmore Lake, Washtenaw County, di.scussion of 336 
 
 Wind as a weathering agent. (See also wave, current and discussions of lakes.) ....'. 4 
 Wind and ice, activity, of. . .82.8.3-4,91.104,124,129,131,153,161,170,171,206,213,248,250 
 
 Wisconsin glacier, moraines of, recession of 11 
 
 Wolf Lake, Muskegon County, di.scussion of 352-353 
 
 Z 
 
 Zukey Lake, Livingston County, discussion of 337