‘i ~ NatScL THE IMLAY OUTLET OF GLACIAL LAKE MAUMBE, IMLAY CITY. MICHIGAN by Winifred Ann Burgis A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in the University of Michigan 1970 Masters Committee: Professor Donald F. Eschman, Chairman Associate Prcfesacr William R. Farrand ABSTRACT This study relates use of the Imlay outlet to the Huron-Erie lobe ice events which controlled formation of the three Glacial Lake Maumee stages. Pebble composition and matrix texture have differentiated two till sheets in the vicinity of Imlay City, Michigan. The older till has a high limestone pebble content and sandy matrix and is restricted in surface exposure to the interlobate area west of the Imlay channel. The younger till has a high sandstone pebble con- tent and more clayey matrix texture; it occurs at the surface east- ward from the Inlay channel at least to Yale, Michigan. The Imlay outlet thus marks a significant contrast in till petrologies. Be- cause the pebble compositions of the two tills reflect different source areas, these tills are attributed to two different ice ad- vances. A period of ice retreat is inferred between these advances. Deposition of the older limestone till in the interlobate area is correlated with Lake Maumee I which stood at 800 feet above sea level and discharged southwestward through the Fort Wayne outlet. Some lake water may have used the Imlay outlet route at this time. with subsequent ice retreat a lower, unknown outlet was opened in the "thumb" of Michigan, and lake level fell to Lake Maumee II at 760 feet; this lake discharged westward to the Glacial Grand River. when the Huronqfirie lobe readvanced, the flaumee II outlet was over- ridden, and lake level rose to Lake Maumee III at 780 feet. Sand- stone till was deposited during this readvance to and retreat from the Imlay channel during Maumee III time. Lake Maumee III drained simultaneously through the Imlay and Fort Wayne outlets. ACKNOWLEDGEMENTS This paper would not have been possible without the patient, ready, and excellent advice of Dr. Donald F. Eschman, who sug- gested this topic for research, and of Dr.‘flilliam R. Farrand; their generosity with their time in the field, laboratory, and office has been greatly appreciated. Efficient and enjoyable field assistance was rendered by Jan Miller in the summer of 1968 and subsequently by Mr. and Mrs. Richard W. Burgis, John R. Burgis, Mary Lee Muhlenkort, and Susan E. Hays. A special thanks must be said to Mr. and Mrs. Arthur Molzon, the Harold Molzon family, and the innumerable curious people who made living and working in La- peer County a pleasure. ii Introduction . . . 0 TABLE OF CONTENTS The Lake Maumee Outlet Controversy . Till: Pebble Composition . O O O Till Pebble Composition and Bedrock Patterns . Till: Gravel: Pebble Composition . Sand-Silt-Clay Ratios . 4! Morphology of the Sandstone Till Sheet . . . . . . Correlation of "Thumb" Moraines .~. . . . . . . . . . Possible Color Variations in the Limestone and Sandstone Tills 00110111810118.0000...0000000000000 Suggestions for Further o 0 0 0 0.0 a 0 o o e o ApPQnd-ix o o 0 e 0 References Cited . 0 iii Locations of the Manmee II Outlet . O O 20 51+ 36 41 ‘+9 5h 59 61 65 7o 71 73 l. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 15. 14. 15. 16. 17. LIST OF FIGURES Location map of southeastern Michigan . . . . . . . . . Topographic features of the Imlay City area . . . . . . Longitudinal profiles of channels 0 . . . . . . . . . . Map showing Lake Maumee I drainage . . . . . . . . . . . Map showing Lake Maumee II drainage . . . . . . . . . . Map showing Lake Maumee III drainage . . . . . . . . . . Till: Pebble composition . . . . . . . . . . . . . . . Sedimentary pebble composition of the limestone and Bandatone till Bheets o o o o o e o o o o o o o c o o 0 Distribution of limestone and sandstone till sheets . . Schematic stratigraphic sections of interlobate area and Dsanvills Mountain 0 o o 0 o o o o o c o o c o 0 c o o 0 Distribution of "pinkish" sediment and black shale pebbles in till . . O O O O O O O O O O O Q C O O O O O T1113 Sand-silt-clay ratios 0 0,. o o 0 o o o o o o o o Gravel: Pebble composition . . . . . . . . . . . . . . Hypothesized directions of ice movement during maximum of the sandstone-rich ice advance . O O O O O O O O O O Pebble composition of buff limestone, buff sandstone, and "pinkiflh" tillB o 0 c o o o o 0.0 o o o o o o 0 o 0 Sand-silt-clay ratios of buff sandstone and "pinkish"‘ tilla o o o o o o O O O O O C O O O O O Q Q O O O O O I Differentiation of limestone and sandstone till sheets on the basis of pebble composition and matrix texture . iv ‘ 10 17 17 18 21 22 2Q 53 35 57 42 51 62 63 66 l. 2. 3. h. 5. 7. LIST OF TABLES Summary of the Lake Maumee outlet controversy . . . . . . 1Q Pebble composition and matrix texture means and con- fidence intervals for sediments in the study area . . . . 19 Results of Student's t-tests of the significance of dif- ferences in pebble composition and matrix texture between tills of the Imlay City area . . . . . . . 26 O O U Q C Q Q Results of Student's t-tests of the significance of dif- ferences in pebble composition between tills and gravels 0f ‘the area 0 0 o 0 o o o o 0000000003]: Results of Student's t-tests of the significance of dif- ferences in pebble composition between gravels of the Imlaycityaréa.....................'+O Correlations of "thumb" moraines by Leverett and Taylor (1915’13030)000000000000000000000055 Correlations of "thumb" moraines suggested in this paper..........................57 1. LIST OF APPENDICES Student's t-test used in this paper . vi 0 INTRODUCTION Leverett and Taylor's (1915) classic monograph established the existence of three stages of Glacial Lake Maumee on the basis of topographic and morphologic evidence. This approach to study of the Cary history of the "thumb" of Michigan has, however, pro— duced controversy about the outlet or outlets used by each Lake vMaumee stage. This paper examines sedimentologically and strati— graphically topographic features near Imlay City, Michigan, to which Leverett and Taylor (1915) assigned a Lake Maumee age. The debate about Lake Maumee's outlets provides a framework in which to compare the effectiveness of the morphological and sedimenta- logical approaches in deciphering the glacial history of an area. So reverent was the reception of Leverett and Taylor's (1915) impressive monograph (Hough, 1965) that little systematic field work has been done in the "thumb" since their original horseback and buckboard surveys. In the last twenty years new evidence has prompted reéevaluation of Great Lakes history, but recent research has concentrated upon glacial lake stages younger than the Maumees. The only recent field studies related to Glacial Lake Maumee deal with the morphology of the Fort Wayne outlet and Wabash valley (Fidlar, 1948; Thornbury, 1950; Wayne and Thornbury, 1951; Thorn- bury, 1958). Bergquist and MacLachlan (1951) discuss in detail the glacial features near Lake Maumee's Imlay outlet, but they rely heavily upon Leverett and Taylor‘s interpretations. The present study of till petrologies demands new interpretations of some of these features, but it substantiates Leverett and Taylor's (1915) basic concept of Lake Maumee drainage. THE LAKE MAUMEE OUTLET CONTROVERSY The highest or first stage of Glacial Lake Maumee, Maumee I, is generally thought to have stood at 800 feet above sea level and to have drained down the Fort Wayne outlet to the Wabash River. The lowest or second stage of this lake, Maumee II, is generally thought to have stood at 760 feet above sea level and to have drained westward to the Glacial Grand River. Whether a presently unknown outlet northeast of the Inlay outlet or the Inlay outlet itself carried water from this lake has been a subject for debate. Discharge from the middle or third stage, Maumee III, which stood at 780 feet above sea level, is thought to have flowed solely through the Imlay channel westward to the Glacial Grand River; or solely through the Fort Wayne outlet to the Wabash River; or simul- taneously through both the Imlay and Fort Wayne outlets. The three stages of Glacial Lake Haumee date approximately between 14,000 and 13,000 years B.P. (Farrand, 1962). Leverett and Taylor (1915) give the first and only detailed, comprehensive account of Glacial Lake Maumee written to date. A small, narrow, crescentic Lake Maumee I formed between the newly- constructed Fort Wayne moraine and the retreating Huron-Erie lobe ice front. The lake expanded northeastward with continued ice re~ treat. Lake level rose until the water found an outlet through the Fort Wayne moraine, thus initiating the "Maumee torrent" (Thornbury, 1958, p. 465) down the Fort Wayne outlet and Wabash spillway. These conclusions are based upon the presence of Maumee I beaches which converge on and enter the Fort Wayne outlet on the inner slope of the Fort Wayne moraine (Taylor, 1897; Leverett, 2 1902). Maumee I beaches are found also on both the outer and , inner slopes of the Defiance moraine. Leverett and Taylor thus believe that, although its size and shape varied with ice margin fluctuations, the lake's level and outlet remained essentially un— changed during ice retreat from the Fort Wayne moraine and during readvance to, construction of, and retreat from the Defiance moraine. Beaches on both flanks of the next younger Birmingham moraine indicate that Lake Maumee I existed during and immediately after deposition of this moraine. As the ice front retreated east- ward from the Birmingham moraine, a new, lower outlet a few miles east of Imlay City, Michigan, was opened, lake level fell to that of Maumee II, and the Fort Wayne outlet was abandoned: "... it was during the activity of this outlet that the lowest beach of Lake Maumee was made ... The location of the outlet channel at the time of the lowest beach is not definitely known. In southeastern Lapeer County ... it appears to have been overridden and destroyed." (Leverett and Taylor, 1915, p. aha) Maumee II beaches are characteristically faint and fragmentary because of their submergence and modification by the succeeding, higher Maumee III stage, about 20 feet above Maumee II. The low- est Maumee stage was contemporaneous with ice front retreat of un- known magnitude and with the subsequent readvance until the ice overrode and closed the low, unknown Maumee II outlet. The ice "... moved 17farther47'westward to a position close along the east side of the great Imlay outlet channel ... This channel shows evidence of having been crowded westward by the ice as it was building the Lflmlay_7'moraine along its east side" (Leverett and Taylor, 1915, p. 322). Dreimanis and Karrow (1965, p. 95) believe that enough time elapsed between closing of the Maumee II outlet and opening of the Maumee III outlet for a beach deposit inter- LAKE HURON SAG/NAW 0 BA)’ 0 Ubly __ r _________ .___ I I I l- I 2: I of’ SANILAC L I I10 I 210 miIes ‘\3 CQUNTY I ____J F'5"""'® CI’ ford I "'"I North—Branch e I; Brown CITY : I s MelviQ _____. I LAPEER Ifl—fl—eHW-Iole I‘ COUNTY <9 G'loodlond II Imloy ity® “ e Copoc I I l g Q A|mont __ST. CLAIR L_____fl_§__,___----L-"'“‘ ICOUNTY Q I q, 2 I ZER 5 @ \’S°B’* 0 ‘o a ‘2 Birmingham \ \ MA 2; LAKE \ ' @ Z Ann Arbor ’ ERG [so 6 e YpsIIonIi I BASE Figure 10 Location map of southeastern Michigan showing the re- lationship of the Lake Maumee and Lake Whittlesey zero isobases and the direction of maximum postglacial uplift. mediate in elevation between those lakes to have developed in southern Ontario as the water rose during this Maumee II-III transition. Because the Imlay outlet stood at a higher elevation than the obliterated Maumee II outlet, lake level rose to that of Mau- mee III. Taylor in 1897 and Leverett in 1902 tentatively iden- tified Maumee III beaches in the heads of the Fort Wayne and Im- lay outlets and therefore suggested simultaneous discharge of this lake stage through both channels. In 1915 (p. 522), however, they concluded that Maumee III "... was barely too low to overflow at Fort Wayne ...” and used only the Imlay outlet. But the present study has not confirmed the Maumee III beach features which Taylor (1897) traced into the Imlay outlet. The best criteria for deter- mining whether Maumee III used both the Fort Wayne and Imlay out- lets are the elevations of the outlet divides at the end of Mau- mee III time with respect to lake level: discharge from YBO-foot Lake Maumee III could have occurred through both outlets, for the Fort Wayne outlet head stood at 757 feet, as measured by Leverett, and the Imlay channel divide stood at 760 feet, as determined from recent topographic maps. Leverett and Taylor feel that the Imlay, Goodland, Deanville, and Yale moraines were deposited during Mau- mee III ice retreat from the eastern bank of the Imlay channel. Retreat from the Yale moraine opened a new outlet farther narth and lower than the overridden Maumee II outlet, and lake level fell to produce Glacial Lake Arkona. Therefore, according to Leverett and Taylor (1915, p. 469), Maumee I discharged through the Fort Wayne outlet and was dammed by ice which stood temporarily at the Defiance and Birmingham 1 moraines; the final location of the retreatal ice dam of Maumee II is unknown, but the lake drained northward and westward through a channel near Imlay City, Michigan, which was subsequently oblit- erated by ice advance to the eastern bank of the Imlay channel; and, contemporaneously with deposition of the Imlay through Yale moraines, Maumee III discharged through the Imlay channel, with only small-volume, intermittent drainage possible through the Fort Wayne outlet. Bergquist and MacLachlan (1951) suggest brief use of the Im- lay outlet during post-Birmingham Maumee I time simultaneous with the main Maumee I discharge through the Fort Wayne outlet. "... retreat of the ice front Z_from the Birmingham moraine ultimately opened an outlet at the present site of Imlay City, Michigan, which was several feet lower than the Wabash outlet at Fort Wayne. At first the channel past Imlay City was neither low enough nor wide enough to carry the full discharge of the expanded lake and for a while both the Imlay outlet and the Wabash outlet operated simul- taneously. This initial use of the Imlay outlet was of but short duration, for the ice front continued to retreat un- til it reached a position somewhere northeast of Imlay City and uncovered an outlet about fifteen feet lower than the Imlay outlet. The use of this low outlet marks the second or lowest Maumee stage." (Bergquist and MacLachlan, 1951, p- 5) Thus, ice retreat during late Maumee I time uncovered a natural low in the position of the present Imlay channel. Running water first began to modify this depression as ice-border discharge and drainage from Lake Maumee I. Because modification began in the infant Imlay channel much later than initial cutting of the Fort Wayne outlet, the Imlay channel was less well defined topograph- ically than the Fort Wayne outlet in late Maumee I time. Berg- quist and MacLachlan therefore feel that the increase in discharge capacity produced by opening the infant Imlay channel was insuf- ficient to lower the level of Lake Maumee I. 30mm 'BNNVHO AV'IW! amveow 31w smvaow a'rn/wvao M 8 d U . . . (I amvsow owvwoooe 5'; m U U) U] amvsow En'uAsuo . a 5 f, HNWUOW AV'IWI _--_.-________ KHZ) Rom" \g (‘P/‘B VBHV E. " v‘ p33 BLVQOWHBLNI W )1 (J , ' / minus a.” mwwws , “0* n 'IHNNVHO ; a? v Maaso 'nm Q a“ 0 Q Q 'oo_sn_rje_1s 9- '00 ovwmxié‘ ' \ 'BNNVHQ "I3NNVHOl 3'] mveo szmvam _. mo UMOJQ AV-IWI NIVlNflO / / zfiwwvao .I I I dlo' 0'0 2% r_ wanuvm seaAsuvsl .fq g'g/ _. new? Rigor: 3J3 / \\ -‘ ooomsems .\_ ./ __.-__..___._ _-~___1 3>l\7'l ‘B'IVA ‘3111mm 'HONVUB HiUON ‘Alla AV'MI ‘oeoaeno ‘OVdVO ‘amsusns ‘Alla NMose 'Avmoose ‘aw’! semeve 'momv =03|OfilS SBWQNVBOVHO BlnNlw-Znl. ~03 333dV1 '00 V'lOOSfiJ. VBHV XIII) A‘V—HAH 3H1. jO SEEMS/3:1 OIHdVHSOdOi '3 3809B Bergquist and MacLachlan's proposal of brief Maumee I dis- charge through the infant Imlay channel is conceptually logical given lake level and the present Imlay channel divide elevation corrected for uplift. According to Leverett and Taylor's (1915) interpretation of Lake Maumee III drainage, however, the Imlay channel is also the product of main Maumee III discharge. The original elevation of the Imlay channel depression and the amount of possible downcutting during Maumee I time relative to possible downcutting during Maumes III time are unknown. I looked for the gravel terraces that Taylor (1897) claims to have found in the Imlay channel to see if they might provide evidence of two periods, Maumee I and III, of discharge through the outlet, but no such ter- races were identified in the field. Therefore, the exact Imlay channel divide elevation at the end of Maumee I time, the eleva- tion which was to determine whether Lake Maumee II water could flow through the Imlay channel, remains unknown. I The presence of high kamic masses on both sides of the Imlay channel along the western flank of Deanville Mountain indicates that the original Imlay channel low was not, contrary to Bergquist and MacLachlan's (1951) suggestion, opened by the simple retreat of active ice during Maumee I time. Ice blocks were probably trapped in a pre-existing depression in the present position of the Imlay channel during pre-Maumee I or Maumee I time as the kames were built. These abandoned ice blocks may have stood too high in Maumee I time to have permitted brief northward lake discharge. Use of the Imlay channel by the various stages of Lake Maumee was dependent not only upon the original elevation of the topographic low there but also upon the time of abandonment, size, and rate of melting of these ice blocks and upon the timing and magnitude of any downcutting in the channel. 0f the controlling factors only the present Imlay channel divide elevation is known, and therefore Maumee I discharge through an infant Imlay channel must remain an unproven possibility. Bergquist and MacLachlan (1951) feel that the ice which closed the Maumee II outlet readvanced to and deposited the Mount Clemens moraine, a correlative of the Deanville moraine (Leverett and Taylor, 1915, p. 50), at the beginning of Maumee III time. Leverett and Taylor (1915), however, conceive of a more extensive ice readvance from the final Maumee II retreatal position and iden— tify the Imlay and Goodland moraines as the oldest Maumee III moraines. Bergquist and MacLachlan imply that the Imlay and Good- land moraines were deposited as the ice front retreated from their postulated infant Imlay channel during Maumee I time. Petrology of the till in the Imlay through Yale moraines reported in this paper supports Leverett and Taylor's age assignments, however. In 1958 Hough agrees that the Maumee II outlet has been over- ridden by a subsequent ice advance and that its location is unknown. But by 1967 he identifies the Imlay channel as the outlet of Lake Maumee II (J. L. Hough, personal communication). This conclusion is based upon Leverett and Taylor's statement that there are three flat divides in the Imlay channel, none of which stands higher than 802 feet (1915, p. 276-277). Uplifted beaches of the three Maumee stages rise at the rate of one foot per mile north of their zero isobase and have not experienced differential uplift (Farrand, 1962). Confusion about the exact location of the Maumee zero iso- base exists in the literature. In their text Leverett and Taylor UL IMLAY CHANNEL ' DhHDE 85C>- - O\\ A../\ 830'-- “my, .,\ \ s|0'- LAKE MAUMEE sorrow A_ \ ‘W .."'.o.e.o....m 790'— 770-1 Almont IMLAY lmlcy City Divide CITY Divide IMLAY CHANNEL ---—-~-— WEAVER DRAIN CHANNEL _ _ _ _ ___ 00000000000000 ~—---—- NORTH BRANCH TRANSVERSE CHANNEL Figure 3. Longitudinal profiles of channels discussed in this paper. Profiles drawn from topographic Vertical exaggeration approximately 335 X. Profiles not corrected for effect of postglacial uplift. Imlay and Weaver Drain channels flowed northwestward in study area. Elm Creek and North Branch Transverse channels flowed westward to the Imlay channel. Gradient of Mill Creek channel slopes eastward. Almont and Imlay City "divides" in Imlay channel located by Leverett and Taylor (1915). maps issued in 1963. BRANCH 11 (1915) refer to Maumee beaches as undeformed as far north as h to 5 miles north of Birmingham, Michigan, the location of the Lake 'Whittlesey zero isobase (Figure 1). In Plate XX, however, they show Maumee beaches beginning to rise farther south; the zero iso- base used in this paper is taken from Plate XX and passes through a point 2 miles south of Plymouth, Michigan (Figure l).1 Leverett and Taylor thought that the channel divide near Imlay City, 52 miles north of the Maumee zero isobase, stood at approximately 750 feet before uplift and was low enough for Lake Maumee II water at 760 feet to have covered it. The divide farther north at Dean- ville Mountain was uplifted 60 feet, standing at 7%2 feet before uplift according to Leverett and Taylor's elevations; in Houghis 1967 (personal communication) interpretation the Deanville Moun- tain divide thus lies farther downstream on the gradient of Maumee II drainage than the Imlay City divide. Topographic maps issued in 1963, however, show that this northern divide has a present elevation of just over 820 feet, not 802 feet, but that the Imlay City divide does in fact stand between 800 and 810 feet (Figure 3). Therefore the northern divide controlled northwesterly flow through the Imlay outlet. The lmlay City divide, which stood at 750 feet before uplift, and the Almont divide, which stood at ap- proximately 755 feet before uplift, are simply small irregularities on the lake bottom; the southern portion of the Imlay outlet is not a true river channel but rather a long, narrow arm of Lake Maumee 1In the following discussion of Maumee II drainage, possible evolutionary changes in the configuration of the Imlay channel are ignored. The channel during Maumee II time is assumed to have been morphologically the same as at the end of Maumee III time. Although the assumption is probably invalid, it simplifies the discussion. 12 restricted by the Imlay moraine, which was built into it in Mau- mee III time. The northern divide elevation before 60 feet of up- lift was about 760 feet according to recent maps. This divide stood approximately at the level of Lake Maumee II, not at 7h2 feet as calculated from Leverett and Taylor's elevations. There- fore, Lake Maumee II could have discharged through the Imlay chan- nel only during periods of high water. If the effects of probable evolutionary changes on the outlet are taken into account, even such small-volume, intermittent flow from Maumee II through the Imlay channel may have been impossible. Another outlet in the "thumb" must have carried most, if not all of Maumee II's discharge. Hough's (1967, personal communication) erroneous conclusion that the Imlay outlet was low enough to carry large—volume Maumee II drainage may have resulted from choice of a Maumee zero isobase farther south than the one used in this paper or from failure to check Leverett and Taylor's (1915) northern divide elevation against new topographic maps. Struck by the fact that the Fort Wayne outlet head stands about 20 feet below Lake Maumee 111 1evel,1 Bough (1958) resur- reots the idea of simultaneous discharge of this lake through the Fort Wayne and Imlay outlets. This idea had been proposed by Tay- lor in 1897 and by Leverett in 1902, but they discarded it in 1915. "... the present elevation of the bed at the Fort Wayne outlet is 757 feet A.T. ... quite low enough to carry an appreciable part of the discharge from a lake standing at 780 feet A.T. It is suggested here that 1The elevation of 757 feet A.T. cited in the literature for the Fort Wayne outlet head was determined by Leverett in 1902 (p. 712). As in the case of the northern Imlay channel divide, his altitudes are not always accurate. ' 13 the Fort Wayne outlet was cut to this low level during the third or Middle Maumee stage, and that during this process most if not all of the discharge was transferred from the Imlay to the Fort Wayne outlet." (Bough, 1958, P0 In 1963, however, Hough strongly advocates the exclusive dis- charge of Lake Maumee III down the Fort Wayne outlet: "This stage must have discharged down the Wabash River because the divide at the head of the drainage at Fort Wayne, Indiana, is 757 feet above sea level. It is not known whether this divide was cut to this elevation during the waning Stages 0? h18h68t,171_7 Lake Maumee, "or during Middle [111'] Lake Haumee, but the latter pos- sibility seems more probable ... [During Maumee III time there was;7 discharge westward down the Grand River Val- ley in Michigan, but it is probable that this discharge was not from Middle Lake Maumee but was melt water from the local ice front." (p. 92) Hough gives no reason why Lake Maumee III could not have flowed over the Imlay channel divides which he believes were as low as or lower than the Fort Wayne outlet head during Maumee III time. Hough's growing conviction that Lake Maumee III drained exclusively through the Fort Wayne outlet may have caused him to reinterpret Naumee II history erroneously. No major lakes older and as high as or higher than the Maumees are known in the "thumb," and the well-developed Imlay outlet is too high to have drained lakes younger and lower than the Maumees. If not used by Maumee III, the Imlay channel, according to Hough's divide elevations, can be explained only as the route of brief Maumee I or of main Maumee II discharge. Houghfs (1958; 1963) hypothesis that the Fort Wayne outlet divide was downcut during Maumee III time suggests that examina- tion of Fort Wayne outlet and Wabash sluiceway terraces may deter- mine whether two Maumee stages, I and III, used this discharge route. Two terraces are generally recognized in that drainage 14 Table l . SUMMARY OF THE LAKE MAUMEE OUTLET CONTROVERSY ARTICLE MAUMEE I V _ MAUHEE II MAUMEE III Taylor Fort Wayne Stage not re~ Imlay outlet (1897) outlet cognized (Fort Wayne outlet ?) Leverett Fort Wayne Stage not re- Imlay and Fort (1902) outlet cognized Wayne outlets Leverett Fort Wayne Outlet near Imlay outlet & Taylor outlet Imlay City (1915) Malott Fort Wayne Outlet lower Not Fort Wayne (1922) outlet than Fort outlet Wayne outlet Bay ‘Fort Wayne Outlet north New outlet at (1936; 1938) outlet of Imlay City Imlay City Bergquist & Fort Wayne & Outlet north- Imlay outlet MacLachlan Imlay (brief east of Imlay (1951) use) outlets City Hough Fort Wayne "Thumb" ice- Imlay and Port (1958) outlet border dis- Wayne outlets charge Hough Fort Wayne "Thumb" ice- Fort Wayne (1963) outlet border dis- outlet charge (Imlay outlet ?) Hough Fort Wayne Imlay outlet Fort Wayne (1967, per~ outlet outlet sonal commun- ication) Dreimanis & Fort Wayne West across West across Karrow outlet Michigan Michigan (1965) are Wayne & Fort Wayne Edge of Sagi- Fort Wayne Zumberge outIet naw lobe outlet (1965) 15 system: the upper terrace represents the surface of a Tazewell- Early Cary valley train, and the lower is the floodplain-erosional surface left by the Maumee I torrent (Malott and Shrock, 1929). The lower terrace can be firmly associated with Lake Maumee dis- charge because this surface is preserved almost unaltered in the Fort Wayne outlet, and it joins the Maumee lake plain accordantly (Fidlar, 1948, p. 103). Only one terrace represents Maumee dis- charge, whether from one or two lake stages. Wayne and Zumberge (1965) believe that Lakes Maumee I and III discharged exclusively through the Fort Wayne outlet and state that the "... erosion surface along the Wabash Valley planed by the overflow waters of the two high phases of Lake Maumee stands 16 to 22 feet above the modern floodplain and is called the Maumee terrace" (p. 77). Be- cause Lakes Maumee I and III differed only 20 feet in surface elevation, it is perhaps possible that even if two distinct epi- sodes of Maumee discharge did occur through this outlet, their features would be virtually indistinguishable. The single Fort Wayne outlet-Wabash sluiceway terrace of Maumee age does not settle the controversy about Lake Maumee III discharge. Interpretation of this terrace's significance apparently depends upon a precon- ception of Maumee drainage patterns, not upon field evidence. Mau- mee II and much of Maumee III discharge flowed through the Glacial Grand River channel, but no separate Maumee terrace is found there; rather, one terrace represents discharge from Lakes Maumee, Arkona, and Whittlesey (Leverett and Taylor, 1915, p. 361). On the other hand, the Wabash sluiceway does display a terrace that conclusive- ly proves at least one episode of Maumee discharge. Evaluation of the literature about the Lake Maumee outlet 16 controversy leads to the following conclusions: 1) 2*) 3) 4) Lake Maumee I, after cutting its main outlet through the Fort Wayne moraine, existed during deposition of the Defiance and Birmingham moraines (Figure 4). If ice blocks trapped in a pre-existing depression at the present site of the Imlay channel northern divide stood low enough during late Maumee I time, this lake could have briefly discharged through an infant Imlay channel as well as at Fort Wayne. If the Imlay out- let did function then, its addition to the volume of Maumee I discharge was not sufficient to lower lake lave 1 0 Ice retreat opened a new, lower outlet northeast of Imlay City, Michigan, and lake level fell to the Maumee II stage (Figure 5). The Maumee I outlet or outlets were abandoned. The Imlay channel northern divide stood too high at the end of Maumee III time to have permitted continuous large-volume Maumee II discharge through this outlet; the divide may have stood even higher during Maumee II time, thus elim- inating even intermittent, small-volume flow from the lake. The location of the Maumee II outlet is unknown, and that channel has probably been largely obliterated by a later ice advance. Ice readvance closed the Maumee II outlet, and lake level rose until Lake Maumee III water flowed north- westward through the Imlay channel. Fort Wayne out- let terraces do not give evidence of two separate Maumee discharges; however, the similarity of the Fort Wayne outlet head and Imlay channel northern divide elevations at the end of Maumee III time in- dicates that simultaneous discharge through both outlets must have occurred during Maumee III time (Figure 6). Glacial Lake Arkona formed when ice retreat from the Yale moraine uncovered an outlet even lower than the buried Maumee II outlet. The Imlay and Fort Wayne outlets were permanently abandoned. Cary time ended with the drop in lake level from Maumee III; Lake Arkona existed during the Cary/Port Huron interval (Bergquist and MacLachlan, 1951, p. 2Q). These conclusions, based upon morphological evidence, will now be examined in the light of new sedimentological and stratigraphic evidence. __g—__-I_-O__ \ LAKE / QHICAGQ' \ / * — ~ — -—- |\~a/ I LAKE MAUMEE:[ L__—- Figure 4. Map showing Lake Maumee I draining southwest- ward through the Fort Wayne outlet in post-Defiance time. Dotted line indicates position of the newly deposited Defiance moraine. Glacial Grand River flowed from edge of the Saginaw lobe into Lake Chicago, which discharged to the west and south via the Chicago outlet. Arrows indicate direction of ice movement. (Modified from Rus- sell and Leverett, 1908, p. 15) VA?‘ \ LAKE \ CHICAGO ) Figure 5. Map showing discharge of Lake Maumee II through an unknown outlet northeast of Imlay City, Michigan, west- ward to the Glacial Grand River and Lake Chicago during a period of ice retreat. Lake Chicago drained to the west and south via the Chicago outlet. Arrows show direction of ice movement. (Modified from Russell and Leverett, 1908, p. 15) 18 MN’ \ LAKE I) CHICAGO \ I I I / MAUMEE 3111‘ Figure 6. Map showing simultaneous discharge of Lake Mau— mee III through-the Imlay outlet westward to the Glacial Grand River and Lake Chicago and southwestward through the Fort Wayne outlet during a period of ice readvance. Lake Chicago drained to the west and south via the Chicago out- let. Arrows indicate direction of ice movement. (Modified from Russell and Leverett, 1908, p. 15) 61 % CRYSTALLINE % CLAY °/<> SAND SEDIMENT %SS;+Sh %SILT LIMESTONE TILL 5L9 1' 2.6 l6.6 ‘-"- 2.3 3I.4 1'1 2.I 44.8 i 4.7 34.3 i 2.0 20.8 1': 3.0 SANDSTONE TILL 32.0 i 2.: 5L9 : 2.8 |e.| : 1.4 28.8 :1: L8 40.3i0e 30.9: 2.0 DEANVILLE ‘ MOUNTAIN 25.6 i 5.2 5L3 i 6.3 23.I i 3.7 40.0 i 4.4 33.8 1'- 2.l 26.2 7-"- 3.6 TILL SANDSTONE- DEANVILLE 30.5 i‘ 2.0 5L? 2!‘. 2.6 I7.8 i2 I.5 3L3 i L9 38.91‘: I.0 29.8 1'- I.7 MTN. TILL LIMESTONE + + + ______ GRAVEL 52.4- 3.0 I7.2 - 3.5 30.5 - 3.6 SANDSTONE GRAVEL 40.02 6.I 38.3 2!: 7.9 2L? 1*: 4.4 --—-— ___..____ DEANVILLE ‘ MOUNTAIN 57.3: 2.5 14.0: 2.2 28.3: 2.0 -———— -———- GRAVEL Table 2. Pebble composition and matrix texture means and confidence intervals (0.05 level of significance) for sediments in the study area. Three samples of non-surface limestone till excluded from calculations. TILL: PEBBLE COMPOSITION Pebble composition, determined by counts of 100 stones 1 to 3 inches in maximum dimension from unleached till, sharply differ~ entiates the till west of the Imlay channel from the till east of it. Pebbles from both the till and gravel of the Imlay City area fall into four main groups: sandstone, black shale, limestone, and crystalline rocks. The crystalline group consists primarily of acidic and basic intrusive igneous rocks, quartzite, and basic and acidic volcanics, all derived from the Canadian Shield. The limestone originates from the Early to Middle Paleozoic sediments which encircle the Michigan Basin and underlie the present Lake Huron basin. The sandstone and black shale from the Mississippian Marshall sandstone and Goldwater shale, respectively, are the local bedrock of the study area. To facilitate data treatment and to emphasize different source areas for sedimentary till pebbles these indices have been defined: 1) L3 = limestone + dolomite + chert till pebbles. Although from various formations, these lithologies are related geographically and in age. They originate from a source area located between the distant Canadian Shield source and the immediate study area. Because most carbonate pebbles effervesced within a few seconds of H01 application, few true dolomite pebbles were present. 2) SS - sandstone + siltstone + conglomerate + iron clay- stdne concretion till pebbles. Most of these lithologies are derived from the Marshall sandstone, but some come from sandstone and siltstone lenses in the Goldwater shale. Sandstone of various grain sizes is the dominant lithology in this group, often constituting 80% or more Of it. 3) 88 + Sh a 88 + black shale till pebbles. These lithologies together represent local bedrock. Because it is soft and easily abraded, many pebble counts con- tain no black shale, so that frequently 881: 881 + Sh. 20 "/0 CRYSTALLINE ‘20 20 \ TILL= X Eos1 of Imlay Chonne! ~ Decnville Mountain PEBBLE COMPOSITION lnterlobcte Arec Not Surfcce Till 40 40 [Z 80 0/6 I I I I I 1 I I I 0/0 FIGURE 7. 22 TILL= PEBBLE; COMPOSITIQN IOO - 03 .J o\‘’ ' l 20 4O 60 ~80 IOO %$S;+Sh o Interlobate Area -)(- Goodiand Moraine ' Deanville Mountain El Deanville Moraine X irnlay Moraine A Yale Moraine + Otisville Moraine U Nat Surface Till Figure 8. Limestone and sandstone till sheets differentiated on the basis of sedimentary pebble composition. Sample points iden- tified by topographic feature. 2'5 Thus the ratio LSi/SSi + Sh is a measure of the abundance of more distantly derived lithologies relative to the abundance of local bedrock in the sedimentary pebble fraction.cf a till in the Imlay City area. Pebble compositions of 154 till samples have been plotted + Sh, and percent crystalline rocks with percent LS percent SS 1’ i as coordinates (Figure 7). Two distinct point clusters representing different geographic areas are defined graphically: till from the interlcbate area west of the,lmlay channel has a mean pebble com- positicn of 51.91i|2.6% L31, 16.6‘: 2.5% $5 + Sh, and 31.4 i 2.1% i crystalline rocks, while till from the area east of the Imlay chan- nel has a mean pebble composition of 32.0 i 2.1% L81, 51.9 i 2.8% 881+ Sh, and 16.1 i,l.4% crystalline rocks. Student's t-tests of the difference between the means for the two areas show that this difference is significant at the 0.05 level for each of the three pebble components (Table 5). Thus different tills lie on either side of the Imlay channel (Figure 9). The interlobate till, char- acterized by a large percentage of pebbles derived from the Lake Huron basin, is here informally named the limestone till, and the till east of the Imlay channel, characterized by a large percentage of local bedrock, is named the sandstone till. I contend in this paper that these tills represent two ice advances separated by a period of ice retreat: the limestone till is a deposit of the older ice advance; readvancing later, the ice picked up large amounts of the sandstone and black shale bedrock local to the study area, the limestone and crystalline pebble content of the ice was diluted, and the sandstone till was deposited. The occurrence of limestone till 5 to 15 feet below the western 24 FIGURE 9. __._T9§<=_<_"-1_¢_9-_.;E>____-__7 DISTRIBUTION 0F LIMESTONE LAPEER CO. SANDSTONE TILL SHEETS I_I>< YoIe Goodlond Church I I I l r' Im , , g I?‘ o Llmestone TIII Sample m '33 In . r- X Sandstone TIII Sample I? X 8| ----- -- lmIoy Channel Divide 0 I9 0 . 2 3 mu” Almont I I Xasswes. C0- ST. CLAIR CO. 25 edge of the Imlay moraine demonstrates the extension of the older, stratigraphically lower limestone till sheet east of the Imlay chan- nel. Limestone till also occurs in the channel of Mill Creek near Yale, Michigan. The creek has a sharply incised valley and has presumably cut through the younger sandstone till of the neighbor— ing ridges of the Deanville moraine to encounter the underlying limestone till. Till pebble compositions from Deanville Mountain are extremely variable: LS varies from 2 to 56%, SS 1 i crystalline content from 8 to #4%; calculated variances are + Sh from 14 to 80%, and largest for SS + Sh and smallest for L81. Three of the five peb- i ble counts containing 5% or less limestone are from the eastern edge of Deanville Mountain. Unusually shallow depths to the Mar- shall sandstone somewhere just east of the mountain may explain the very high sandstone content of these samples and some of the variance of SS + Sh in Deanville Mountain till. i Deanville Mountain till has a mean pebble composition of 25.6 i 5.2% L81, 51.3 i 6.3% as + Sh, and. 25.1 i 3.7% crystalline i rocks. Student's t-tests at the 0.05 level show that Deanville Mountain till is significantly different from the limestone till in all pebble-component means, and it is also significantly dif- ferent from the sandstone till in mean limestone and mean crystal- line rock content (Table 3). Deanville Mountain till does not, however, differ significantly from the sandstone till in sandstone content (Table 5). The Deanville Mountain and sandstone tills are identified as deposits of the same ice advance on the basis of their similar, high sandstone content. Student's t-tests of the difference between the limestone till and combined sandstone- I 26 Table 5. Results of Student's t-tests+ of the significance of dif~ ferences in pebble composition and matrix texture between tills of the Imlay City area. *Limestone Till vs. Sandstone Till Parameter F-Test for Equal t-Test of Significant Variances Difference % LSi Equal Significant %»SSi+ Sh Not Equal Significant % Crystalline Equal Significant % Sand I Not'Equal Significant % Silt Equal Significant % Clay Equal Significant *Limestone Till vs. Deanville Mountain Till MM- i-mziaizszzfiw % LSi ’ Not Equal Significant % 881+ Sh Not Equal Significant % Crystalline Not Equal Significant % Sand Equal Not Significant %»Silt Equal Not Significant % Clay ' Equal Significant + See Appendix for tests performed at 0:05 level of significance. * Three samples of non-surface limestone till excluded from calculations. 27 Table 3. (Continued) Sandstone TiTl vs. Dsanville Mountain Till Parametar FbTest for Equal t-Test of Significant Variances Difference % Lsi Equal Significant % sci. Sh Equal Not Significant % Crystalline Not Equal‘ Significant % Sand Equal Significant % Silt . 5 Equal Significant % Clay Equal Significant *Limestonc Till vs. Sandstone-Deanville Mountain Till MW . “'T“3.3£.i1522“°m % LSi Not Equal Significant % SSi+.Sh 1 Not Equal Significant % Crystalline Equal Significant % Sand Equal Significant % Silt Equal Significant % Clay Equal Significant i- Three samples of non-surface limestone till excluded from calculations. 28 Deanville Mountain till support this conclusion by showing that all pebble-component means of these two sediments differ signifi- cantly at the 0.05 level (Table 5). The slight but significant difference in mean limestone and crystalline content between the sandstone and Deanville Mountain tills is probably the result of incorporation of older Deanville Mountain sediments of different pebble composition by advancing sandstone-rich ice. Some till samples from Deanville Mountain have pebble compositions‘characteristic of the limestone till. They were taken from the basal few feet of till directly overly- ing limestone-rich gravel which has essentially the same pebble composition as the limestone till; these samples may be interpreted as a mixture of this gravel with sandstone till. The stratigraphy of Deanville Mountain requires revision of Leverett and Taylor's (1915) morphological interpretation of this feature and its relationship to the interlobate area west of the Imlay channel.’ Leverett and Taylor believe that the Defiance and Birmingham moraines of Maumee I age can be followed northward into the interlobate area almost continuously to and beyond Imlay City. However, the contradictions evident in using Leverett and Taylor's (1915, p. 30) correlation of "thumb" moraines (Tables 6 and 7) and a lack of convincing field evidence indicate that these moraines cannot be traced easily, if at all, into the interlobate area in Lapeer County. The bulk of the sediment constituting the topographically sharp interlobate features near Imlay City is send and gravel; this indicates the kamic nature of the interlobate area. The major deposits of Maumee I age in the "thumb" were left by stagnant ice, not by the active ice responsible for deposition 29 of the Defiance and Birmingham moraines to the south. In many interlobate area gravel pits great thicknesses of sand and lime- stone-rich gravel are capped by a layer of limestone till. The sharp contact between these sediments suggests that the till may not be of ablation origin but rather the deposit of an ice read- vance over the interlobate kames; if such a readvance occurred, it may have correlated with readvance to the Defiance or Birming- ham moraine farther south. The dominant morphology of the inter- lobate area is kamic with superposed till, not morainic topped with occasional kames as Leverett and Taylor claimed. Leverett and Taylor's (1915) interpretation of Deanville Mountain seems to rest upon the assumption that all features along the eastern bank of the Imlay channel have the same origin and are morphologically continuous. They describe the Imlay and Goodland moraines near Imlay City as faint but distinctly separate as far north as the Mill Creek channel. Between Mill Creek and what Leverett and Taylor call the "Deanville kames" these moraines cannot be distinguished, but from the "Deanville kames" northwest to North Branch, Mflchigan, the moraines are thought to be once again separate ridges up to a mile apart. Leverett and Taylor maintain that the cluster of high "Deanville kames" rest on top of the combined Imlay and Goodland moraines where these are indis- tinguishable and that the kames are composed of gravel with till at their base. Leverett and Taylor divide the "Deanville kames" into two groups about a mile apart, each elongated parallel to the northwesterly trend of the combined Imlay and Goodland moraines north of Mill Creek. "If the kames were inset a little from the front of the ice, as seems probable, their relations and the in- 50 terval between them seems to accord well with the two moraines referred to" (Leverett and Taylor, 1915, p. 272). In other words, although the Imlay and Coodland moraines cannot be distinguished from each other near the "Deanville kames," the two lines of kames represent two ice advances: the western kames were deposited by the earliest Maumee III Imlay advance and the eastern kames by re- advance to the younger Goodland moraine. Leverett and Taylor be- lieve that the "Deanville kames" are superposed on slightly older or contemporaneous moraines. Deanville Mountain stratigraphy should therefore consist of till overlain by sand and gravel. However, just as in the interlobate area, the stratigraphy characteristic of Deanville Mountain is sand and gravel overlain by till. Furthermore, Leverett and Taylor's (1915) morphological description of the Imlay and Goodland moraines needs revision. These moraines are distinct topographic features as far north as Goodland Church where they merge, but north of the church they dov not exist. Leverett and Taylor identify these moraines in and north of Deanville Mountain because they believe that these ridges should continue to line the eastern bank of the Imlay channel. They offer no evidence for their belief that the Goodland moraine was deposited by readvancing ice rather than during a retreatal stillstand. The thick, extensive Deanville Mountain gravel de- posits prove that that feature is a kamic rather than a morainic ‘E1388 0 The limestone till does not differ significantly in mean pebble composition from the interlobate limestone gravel (Table a); moreover, this till does not differ significantly from Dean- ville Mountain gravel in mean sandstone content (Table Q); and 31 TablQ ii‘. Results of Student's t-tssts+ of the significance of dif- ferences in pebble composition between tills and gravels of the Im- lay City area. *Limestone Till vs. Interlobate Limestone Gravel Pa ate FeTbst for Equal t-Test of Significant ram r Variances Difference % LS1 Equal Not Significant % 881+ Sh Equal Not Significant % Crystalline Equal Not Significant ‘Limestone Till vs. Deanville Mountain Gravel Deanvillc Mountain Till vs. Deanville Mountain Gravel I F-Test for Equal t-Test of Significant Q Parameter Variances Difference % LS1 Equal Significant % 881+ Sh Equal Not Significant % Crystalline Equal Significant F-Test for Equal t-Test of Significant Param'ter Variances Difference % LS1 Not Equal Significant % 881+ Sh Not Equal Significant % Crystalline Not Equal Significant + See Appendix for tests performed at 0.05 level of significance. ‘l- Throe samples of non-surface limestone till excluded from calculations. 52 the interlobate limestone gravel does not differ significantly from Deanville Mountain gravel in mean sandstone and mean crystal- line rock content 62 °/o CRYSTALLINE 2C) 4C) 6C) BC) I | I I I °/o $3; + Sh 6C) LOCALITY A ' Buff Sandstone Till 0 "Pinkish" Till LOCALITY s X Buff Limestone Till $14 "Pinkish" Till Figure 15. stone, and "pinkish" tills. Pebble composition of buff limestone, buff sand- °/o CLAY 2() 2C) 40 ' 40 6() 6C) 0 // o m o X 8C) 8C) QGEQABHD I I I I I I I I I QGfSHQT 20 4O 60 80 LQCALITY A X Buff Sandstone Till "Pinkish" Till LOCALITY B ' Buff Sandstone Till o "Pinkish" TiH Figure 16. Sand-silt-clay ratios of buff sandstone and "pink- iBh" 121-118 . 61+ central North Dakota has been explained by this mechanism (Kelly and Baker, 1966). No systematic study of the pebble-free, silty and clayey sed- iment in the "pinkish" vertical and horizontal streaks has been made. These streaks may, however, consist of older "pinkish" lacustrine sediments which, like the "pinkish" till or tills to which they may be related, were also incorporated by late Cary ice. CONCLUSIONS Sedimentological analysis of till in the vicinity of Imlay City, Michigan, has led to the differentiation of two till sheets on the basis of pebble composition and matrix texture (Figure 17). The older of these tills, its relative age established by know- ledge of the general direction of ice retreat in the Huron-Erie and Saginaw lobes, is characterized by higher limestone pebble content and a sandier matrix than is the younger sandstone-rich, more clayey till. Surface exposure of the older limestone till is restricted to the interlobate area west of the Imlay channel, while the younger sandstone till sheet occurs eastward from the eastern bank of the/Imlay channel at least to the vicinity of Yale, Michigan. This contrast in till petrology along a clearly defined topographic boundary, the Imlay channel, indicates that two distinct drift sheets do occur in the study area. The high percentage of Huron basin carbonate pebbles and relatively low percentage of local pebbles in the limestone till suggest that the older limestone-rich ice picked up the bulk of its sediment load some distance east and northeast of the study area; this ice incorporated relatively little of the local "thumb" sandstone and shale bedrock as it moved westward and southwest- ward across these lithologies. The younger sandstone till, how- ever, is characterized by a high percentage of local bedrock peb- bles and a low percentage of Huron basin carbonate pebbles: the sandstone-rich ice readvanced southward and derived the bulk of its load from local sandstone and shale sources, thus diluting the more distantly derived limestone and crystalline pebble components 65 66 DlFFERENTIATlON OF TILL SHEETS 713W 6!)“ 5X)“ 4!)- LS; /SSi+Sh 3!)‘ 2K)" %SAND in MATRIX (eZmm) x East of lmloy Channel 0 lnlerlobo’re Areo ' Deonvillev Mountain [I Not Surface Till Figure 17. Differentiation of limestone (upper cluster) and sandstone (lower cluster) till sheets on the basis of pebble composition and matrix texture of 117 samples. Till type at each sample point in Figure 9 was determined from this scatter plot. 67 to produce the sandstone till. The petrologic difference between the limestone and sandstone tills reflects the dominance first of a more distant sediment source and later of a local sediment source. The high sandstone pebble content of the Deanville Mountain and sandstone tills is derived from the same local bedrock source area. Deanville Mountain till is a coarse facies of the sandstone till sheet and was produced by overriding and incorporation of older limestone-rich kamic sands and gravels by the sandstone-rich ice. The similar pebble compositions of the interlobate limestone gravel, limestone till, and Deanville Mountain gravel indicate that these sediments are deposits of the same limestone-rich ice source. The layer of limestone till which overlies the interlobate limestone gravel may be ablation till or basal till. Deanville Mountain is a northeastern extension of the interlobate area. It was built before the advance of the sandstone-rich ice which flow- ed over this massive kamic obstacle to the eastern bank of the Imlay channel. v The petrolcgic contrast between the limestone and sandstone tills records ice advance, retreat, and readvance in the Imlay City area during Lake Maumee time. This sequence coincides with the general pattern of ice movement which Leverett and Taylor (1915) deduced from morphology. The following history of Glacial Lake Maumee is based upon sedimentological, stratigraphic, and morphological data: 1) Lake Maumee I formed upon ice retreat from the newly- built Fort Wayne moraine and discharged through the Fort Wayne outlet. During Maumee I time the Defiance and Birmingham moraines were constructed south of Imlay City, and in the Imlay City-North Branch area limestone-rich kamic gravels and limestone till were deposited in the interlobate area which included 68 Deanville Mountain. If the topographic low now occupied by the Imlay channel stood below lake level in Maumee I time, it may have served as an infant Imlay channel. Because elevations in the vicinity of the present Imlay channel divide dur- ing Maumee I time are unknown, Maumee I discharge through an infant Imlay channel is an unproven pos; sibility. 2) Retreat of the limestone-rich ice opened a still unidentified outlet northeast of Imlay City; this outlet was lower than the Fort Wayne outlet and any possible infant Imlay channel. Lake level fell to Maumee II, and the outlet or outlets of Lake Maumee I were abandoned. If the topographic low now occupied by the Imlay channel stood at the same elevation in Maumee II time as it did at the end of Maumee III time, only intermittent, small- volume discharge during periods of high water could have flowed through that depression from Lake Mau- mee II. However, the floor of the depression pro- bably stood higher during Maumee II time than at the end of Maumee III time, so that even intermittent discharge from Lake Maumee II via the Imlay channel would have been impossible. All lake discharge was probably carried by the unknown Maumee II outlet. 3) The ice readvanced southward along the Marshall sand- stone bedrock band and picked up large amounts of this rock. It overrode the unknown Maumee II outlet, and consequently lake level rose to Maumee III. This sandstone-rich ice flowed over Deanville Mountain, incorporating older kamic sediments as it advanced. At its maximum the ice stood at the western edge of Deanville Mountain, built the Otisville and Imlay moraines, and ponded Lake Silverwood. The Imlay channel divide was low enough relative to lake level for the Imlay channel to have been a main Maumee III outlet. Lake Maumee III drained simultaneously through both the Imlay and Fort Wayne outlets, which have divides with very similar late Maumee III elevations. The Goodland, Deanville, and Yale moraines were de- posited during northward retreat of the sandstone-rich ice. Retreat from the Yale moraine exposed ground lower than the Imlay and Fort Wayne outlets; those channels were permanently abandoned, and lake level fell to Glacial Lake Arkona. The stratigraphic and sedimentological approach to the late Cary history of the Imlay City area requires revisions of some of Leverett and Taylor's (1915) morphological interpretations. Strati- graphy shows that Deanville Mountain is significantly older than 69 the Maumee III ice advance, not of early Maumee III age as Leverett and Taylor claim, and that it is kamic with a thin layer of super- posed till rather than morainic with superposed kames; that the interlobate area west of the Imlay channel is kamic, not morainic; and that the "problematical transverse ridges" are moraines which may provide information about ice lobation, not subglacial or englaciel fluvial deposits related to eskers. These changes in Leverett and Taylor's interpretation of the glacial features of the Imlay City, Michigan, area do not, however, contradict their concept of the outlets used by each of the three Glacial Lake Mau- mee stages. Indeed, the new data presented in this paper point to their 1915 concept of Lake Maumee drainage as the correct one, with the exception of their omission of probable Maumee III discharge through the Fort Wayne outlet. SUGGESTIONS FOR FURTHER STUDY The limestone and sandstone till sheets should be mapped to outline the major late Cary ice advances onto the "thumb" and to determine whether the western portion of the Imlay channel also marks a major contrast in till petrologies. Variations in pebble composition and matrix texture within each till sheet around the "thumb" re-entrant should be examined for characteristics which distinguish Huron-Erie and Saginaw lobe deposits. Such distin- guishing criteria would be useful in estimating the degree and chronology of interaction of the two lobes in the Huron-Erie- Saginaw interlobate area. The Port Huron moraine till should be studied petrologically, and the relationship of the younger bor- der of the sandstone till sheet to the Port Huron moraine should be examined to determine whether the outer border of this moraine marks a strong petrologic contrast in tills. The classical belief that the Port Huron moraine is the product of an ice readvance of sufficient magnitude to subdivide late Wisconsin time suggests that such a contrast in tills should occur at the moraine's outer border. 70 APPENDIX Student's t-Test Used in This Paper Definitions: ii - Mean of first group ‘i 2 a Mean of second group 812 - variance of first group, larger mean square s2 - variance of second group, smaller mean square n1 - sample size of first group“ n2 a sample size of second group F-Test for Equal Variance: 2 1 F8 2 , df-(nl-l), (112-1) E‘:2 Equal Variances: , ‘Y -'Y t . l 2 , 2 2 ;L’+ AL 81 (n1 - l) + 82 (n2 - 1)‘) n1 n2 D‘ where D n max (nl - l, 0) + max (n2 - l, O) and df - D. Unequal Variances: 35-3? t,...._____.L 2 , 82 2 1 82 + 1{"1 n2 71 where at - ‘ 2 1 81 /n1; 72 2 2 s1, /nl + s2 /n2 round to nearest integer. + m2 l REFERENCES CITED Bay, J. W., 1936, Glacial-lake levels indicated by terraces of the Huron, Rouge, and Clinton Rivers, Michigan: Michigan Acad. 801., Arts, and Letters, v. 22, p. 411-419 1958, Glacial history of the streams of southeastern Michi- gan: Cranbrook Inst. Sci. Bull. 12, 59 p. Bergquist, S. G., and MacLachlan, D. C., 1951, Pleistocene fea- tures of the Huron-Saginaw ice lobes in Michigan: Geol. Soc. America Guidebook, Detroit mtg., Glacial field trip, 56 p. Bretz, J. H., 1955, Glacial Grand River, Michigan: Michigan Acad. 801., Arts, and Letters, v. 58, p. 559-582 Dreimanis, Aleksis, and Karrow, Paul F., 1965, Southern Ontario, p. 90-110 in_Goldthwait, R. P., and others, Guidebook for field conference G, Great Lakes - Ohio River Valley: Internet. Assoc. Quat. Research, VII Cong., 110 p. Farrand, W. R., 1962, Postglacial uplift in North America: Am. JOur. 301., V. 260, pa 181-199 Fidlar, M. M., 19h8, Physiography of the lower Wabash valley: Indiana Div. Geol. Bull. 2, 112 p. Hough, J. L., 1958, Geology of the Great Lakes: Urbana, Univ. Illinois Press, 515 p. 1965, The prehistoric Great Lakes of North America: Am. Scientist, v. 51, p. 8Q-109 Kelly, T. E., and Baker, Claud H., Jr., 1966, Color variations within glacial till, east-central North Dakota - a prelim- inary investigation: Jour. Sed. Petrology, v. 56, p. 75- 80 Leverett, Frank, 1902, Glacial formations and drainage features of the Erie and Ohio basins: U. S. Geol. Survey Mon. #1, Po Leverett, Frank, and Taylor, F. B., 1915, The Pleistocene of Indiana and Michigan and the history of the Great Lakes: U. S. Geol. Survey Mon. 55, 529 p. Malott, C. A., 1922, Physiography of Indiana, p. 59-256 in Logan, W. N., Handbook of Indiana geology: Indiana Dept. Conserv. Pub. 21, Pt. 2, 1120 p. Malott, C. A., and Shrock, R. R., 1929, Features of the Wabash sluiceway of northern Indiana: Geol. Soc. America Bu11., V. ‘+0’ p. 73 74 Russell, 1. C., and Leverett, Frank, 1908, Ann Arbor folio, Michi- gan, Geol. Atlas of the U. 3.: U. 8. Cool. Survey, 20 p. Taylor, F. B., 1897, Correlation of Erie-Huron beaches with out- lets and moraines in southeastern Michigan: Geol. Soc. America Bull., v. 8, p. 51-58 Thornbury, W. D., 1950, Glacial sluioeways and 1acustrine plains of southern Indiana: Indiana Div. Geol. Bull. 4, 21 p. 1958, The geomorphic history of the upper Wabash valley: Am. Jour. 801., v. 256, p. hh9-Q69 Wayne, W. J., and Thornbury, W. D., 1951, Glacial geology of Wabash County, Indiana: Indiana Div. Geol. Bull. 5, 39 p. Wayne, W. J., and Zumberge, J. H., 1965, Pleistocene geology of Indiana and Michigan, p. 65-84 in,wright, H. E., Jr., and Frey, D. 0., Editors, The Quaternary of the United States: Princeton, N. J., Princeton Univ. Press, 922 p. UNIVERSITY OF ICHIGAN Illmuygnp I11 M m 5 O 327 4241