Digitized by the Internet Archive in 2018 with funding from Getty Research Institute https://archive.org/details/indianvillagesitOOhoot PEABODY MUSEUM OF AMERICAN ARCHAEOLOGY AND ETHNOLOGY HARVARD UNIVERSITY PAPERS VOLUME VIII CAMBRIDGE, MASSACHUSETTS, U.S.A. PUBLISHED BY THE MUSEUM 1920 - 1921-1922 CONTENTS OF VOLUME VIII NUMBER I Indian Village Site and Cemetery near Madisonville Ohio By earnest a. hooton With Notes on the Artifacts By CHARLES C. WILLOUGHBY NUMBER 2 Basket-maker Caves of Northeastern Arizona By SAMUEL J. GUERNSEY AND ALFRED V. KIDDER NUMBER 3 The Turner Group of Earthworks, Hamilton County Ohio By CHARLES C. WILLOUGHBY With Notes on the Skeletal Remains By EARNEST A. HOOTON PAPERS OF THE PEABODY MUSEUM OF AMERICAN ARCHAEOLOGY AND ETHNOLOGY, HARVARD UNIVERSITY Vol. VIII. —No. 1 INDIAN VILLAGE SITE AND CEMETERY NEAR MADISONVILLE OHIO BY EARNEST A. IIOOTON WITH NOTES ON THE ARTIFACTS BY CHARLES C. WILLOUGHBY THIRTY PLATES AND FIVE ILLUSTRATIONS IN THE TEXT Cambridge, Mass. PAPERS OF THE PEABODY MUSEUM OF AMERICAN ARCHAEOLOGY AND ETHNOLOGY, HARVARD UNIVERSITY Vol. VIII.— No. 1 INDIAN VILLAGE SITE AND CEMETERY NEAR MADISONVILLE OHIO BY EARNEST A. HOOTON WITH NOTES ON THE ARTIFACTS BY CHARLES C. WILLOUGHBY THIRTY PLATES AND FIVE ILLUSTRATIONS IN THE TEXT Cambridge, Mass. Published by the Museum 1920 COPYRIGHT, 1920 BY THE PEABODY MUSEUM OF AMERICAN ARCHAEOLOGY AND ETHNOLOGY, HARVARD UNIVERSITY NOTE The explorations which form the basis of the following paper were carried on intermittently under the auspices of the Peabody Museum from 1882 to 1911. During this period, Dr. Charles L. Metz of Madisonville, the discoverer of the site, had general supervision of the work. By the terms of the will of Miss Phoebe Ferris, the late owner of the land upon which the cemetery is situated, the right to complete the explorations after her death in 1897 was granted the Museum, and the work was then carried on principally by students in the Division of Anthropology, connected with the Museum, under the general oversight of Dr. Metz. The extensive collections of skeletal remains and artifacts, and more than two hundred and fifty photographs taken during the investigations are in the Museum. It is hoped that the study of these remains as set forth in the following pages will prove of special value to students of the archaeology of the Ohio region, as it connects the culture of the occupants of this site definitely with the protohistoric and late prehistoric period of the valley. Cambridge, Massachusetts July 7, 1920 CONTENTS THE DISCOVERY OF THE SITE AND HISTORY OF THE EXPLORATIONS PAGE The Discovery. 1 Excavations by C. L. Metz . 3 Excavations by F. W. Putnam. 5 Excavations by J. R. Swanton. 6 Excavations by R. E. Merwin. 7 Excavations by B. W. Merwin. 9 THE BURIALS Depth of Interments. 10 Forms of Burial . 11 Horizontal or Extended Burials. 11 Contracted Burials. 11 Burials in Sitting Posture. 11 Collective and Disturbed Burials. 12 Burials in Cache-pits. 12 Anomalous Burials. 13 Grouping of Burials. 14 Orientation of Burials. 15 Objects Associated with Burials. 16 Burials of Special Interest. 19 Total Number of Burials . 20 Death Rate as Indicated by Burials.. . 20 Population, and Period of Use of Cemetery . 23 THE CACHE-PITS Their Discovery. 27 Number and Depth. 28 Contents and Stratification. 29 Fauna Represented in the Pits. 32 Artifacts from the Pits . 33 Form of the Pits. 35 Grouping in Relation to Burials . 36 Their Purpose. 36 V VI CONTENTS HEARTHS AND OTHER REMAINS Hearths and Fireplaces. 40 Post-holes. 42 Kitchen-middens. 42 House Circles . 44 ARTIFACTS FROM THE SITE General Distribution of Types. 47 Knives and Projectile Points. 47 Arrowmaker’s Tools. 49 Arrowpoints. 51 Flint Drills. 52 Skin-dressing Tools. 53 Grooveless Stone Axes. 55 Stone Adze Blades. 56 Anvils, Hammer-stones and Grinding Stones. 56 Fishing Implements. 58 Awls and Needles. 58 Antler Blades . 59 Beaver Tooth Chisels. 61 Musical Instruments. 62 Personal Ornaments of Shell. 63 Implements and Utensils of Shell. 65 Combs and Personal Ornaments of Bone and Antler. 67 Personal Ornaments of Copper. 69 Personal Ornaments of Brass. 71 Iron Objects. 72 Glass Beads. 73 Tobacco Pipes... 73 Incised Drawings, etc. 75 Textile Fabrics and Wooden Utensils. 77 Pottery .. 78 THE SKELETAL REMAINS Collection in General. 83 Cranial Deformation. 85 Cranial Vault Indices. 86 Cranial Arcs. 92 Cranial Capacity. 93 Thickness of Left Parietal above Temporo-parietal Suture .... 93 Minimum Frontal Diameter. 93 Facial Index. 94 Mean Orbital Index. 95 CONTENTS vii Nasal Index. 95 Palatal Index. 96 Alveolar Index. 96 Angles Relating to Prognathism . 97 Foramen Magnum. 97 Lower Jaw. 98 Frontal Region. 98 Sagittal Region. 99 Temporal Region.100 Occipital Region.100 Sutures.101 Parietal Foramina .102 Retro-mastoid Foramina.103 Mastoids .103 Facial Portion. 103 Teeth.107 Palate.110 Skull Base.110 Mandible.113 Pathological Features.•..114 Summary of Measurements and Observations upon the Crania . . 115 The Bones in General. 121 Femur.122 Tibia.126 Fibula.128 Humerus .128 Radius .129 Ulna.:.130 Pelvis.130 Pathology. 130 General Summary of Observations.133 CONCLUSION Period of Occupation of the Site.135 Relation of the Madisonville Culture to that of Surrounding Tribes 135 Physical Relation of the Inhabitants to the Neighboring People . . 137 INDIAN VILLAGE SITE AND CEMETERY NEAR MADISONVILLE, OHIO THE DISCOVERY OF THE SITE AND HISTORY OF THE EXPLORATIONS The Discovery. The discoverer of the important Indian village site and cemetery near Madisonville, Ohio, was Dr. Charles L. Metz, a resident of Madisonville, physician by profession, archae¬ ologist by avocation. Prior to his explorations of the site, Dr. Metz published a paper on the prehistoric monuments of the Little Miami Valley in which he described and located on a chart the principal earthworks and mounds in Columbia, Anderson, and Spencer townships, near the Little Miami River. 1 Group A of Dr. Metz’s classification in¬ cludes the site of the prehistoric cemetery and the neighboring monuments. He describes it as follows: 2 Group A is situated partly in section 9, and partly in section 15, Columbia township, Hamilton County, Ohio, one mile west of Plainville station on the Little Miami Railroad, and on the second bottom or plateau of the Little Miami River on a narrow sandy ridge of reddish color. This ridge has an elevation, averaging from 10 to 25 feet above the general level of the plateau on which it is situated. It lies between the Wooster Turn¬ pike and the Little Miami Railroad and River, elevated about 200 feet above the latter. On this ridge the principal work of this group is situated. Commencing at the east end of the ridge, and in a wood known as “ Stites Grove ” we find an earthwork consisting of a circle, central tumulus, and an oval-shaped tumulus impinging on the outer southeast edge of the circle. 3 West and to the south of this tumulus, and on the same continuous sand ridge mentioned above, are four or five elevations or tumuli with an average height of three to four feet, being from two to three hundred feet apart. The ridge is here under cultivation; numbers of relics, flint chips, and broken bowlders are ploughed up on this ridge. 1 Charles L. Metz, M.D., The Prehistoric Monuments of the Little Miami Valley, Journal of ■the Cincinnati Society of Natural History, vol. i, no. 3, October, 1878, p. 119 et seq. 2 Op. cit., loc. cit. 3 Here Dr. Metz quotes the description and account of the excavation of this work by Florien Giauque, The Mound Builders, Harvest Home Magazine, August, 1876. 1 / 2 INDIAN VILLAGE SITE AND CEMETERY Northwest of these tumuli, and on the general level of the plateau, one- fourth mile distant, is a mound which has a circumference at the base of 200 feet, and an elevation of 7 feet; it is as yet unexplored, but cultivated annually. Four hundred yards to the northeast of this mound, and at the junction of the Wooster and Madison turnpikes, can yet be traced a circular work, which has a circumference of 600 feet. . . . Continuing on to the southwestward of the small tumuli and along the previously described sand ridge, we come to what is known as the “ Pottery Field.” Here the ridge slopes gently to the south and southeast, with an elevation of from 60 to 80 feet above the level of the Little Miami River. This field is a plateau of about four acres in extent, sloping back to the higher ground. On this plateau fragments of pottery are found in great abundance. Flint chips, arrowpoints, broken bowlders, burnt limestone, and shells of the freshwater mussels ( Unio) are found all over the surface. Human remains have been found in the adjoining ravines, and on the slopes; the graves were isolated and shallow, and the method of burial was not uniform. Bones of various wild animals are also found. Two hundred yards north of the pottery field are several small tumuli; the largest has a circumference at base of about 100 feet, height 5§ feet; this mound has been dug into, but not yet explored. The pottery field and also the tumulus are situated in section 9, Columbia township, in what is known as Ferris Woods, in Still Home Hollow. The pottery field above described was the site of the cemetery later discovered by Dr. Metz. The circumstances of the discovery are as follows. During the fall and winter of 1878-79, Dr. Metz explored several of the mounds of Group A, in the vicinity of the cemetery site. 1 During the progress of the work on this mound [5, Group A] the laborer, employed by Dr. Metz, had been prospecting by digging holes in the sur¬ rounding forest, until finally, on the 20th of March, in the southwestern sec¬ tion of the plateau, he came upon a human skeleton at the depth of about two feet; these remains were, however, so much decayed that they could be preserved only in fragments. This was the initiatory step toward a most im¬ portant archaeological discovery, as further investigation has revealed the interesting fact that the entire plateau is the site of an ancient cemetery, from which have since been exhumed upward of four hundred skeletons of a pre¬ historic people, accompanied by numerous evidences of their handiwork, in the shape of flint and stone implements, pipes, pottery ware, charred matting and corn, tools and ornaments of bone, shell, and copper, some of which are believed to be unique, all indicating an industrious people, who lived in large communities, and obtained their support by cultivating the soil, as well as by fishing and hunting. 2 1 Charles F. Low, Archaeological Explorations near Madisonville, Ohio, Journal of the Cin¬ cinnati Society of Natural History, vol. iii, no. 1, April, 1880, p. 40 et seq. (Gives an account of the excavation of the mounds. See also American Naturalist, May, 1879, p. 328.) 2 Op. cit., p. 43. NEAR MADISONVILLE, OHIO 3 The above account corresponds accurately with the data in Dr, Metz’s field notes, in which the following entries relative to the discovery occur: 1 March 16, 1879. Britten [the laborer] began digging trenches 4 ft. square and 3 ft. deep and 200 ft. apart, beginning at the Spice Bush Mound (No. 5), and extending in a southwestern direction to the point of the second plain. . . . March 20. Britten dug 3 pits today, the last one on the southwestern point of the plateau. Just before quitting work he found a skeleton at the depth of 2 ft. A brief sketch of the discovery was contributed also to Short’s The North Americans of Antiquity, published in 1879. This ac¬ count is quoted by Low as follows: 2 This cemetery, which is distant about one and one-half miles southeast from Madisonville, occupies the western extremity of a plateau overlooking the Little Miami River, and situated from eighty to one hundred feet above the water-line. It is bounded on the south by the river “ bottom,” on the north and west by a deep ravine, through which flows a small stream known as Whiskey Run; on the east the plateau slopes gradually up to the general level of the surrounding country, of which it is in fact a continuation or spur, its character of an elevated plateau being derived from its position between the eroded river valley and the deep ravine above referred to. The precipitous but well wooded bluff which forms the southern limit of this plateau extends eastward, facing the river, for perhaps half a mile; and distributed along its edge are a number of mounds and other earthworks; at its base are the Cincin¬ nati and Eastern and Little Miami Railways, the nearest station being Batavia Junction, distant about half a mile east of the cemetery. The original forest still covers the site of the cemetery, and measurements of some of the principal trees are recorded by Dr. Metz. The locality has long been known to local collectors and others interested in archaeological matters as the “ Pottery Field,” so called on account of the numerous fragments of earthenware strewn over the surface; and it was until recently supposed to be a place where the manufacture of pottery had been carried on by the ancient inhabitants of the valley, the fragments being considered the debris. A few scattered human remains had also been found in the adjoining ravines, but it was not until some time in March, 1879, that its true character and extent as a cemetery were brought to light. Excavations by Dr. Charles L. Metz. The excavation of the Madisonville cemetery was begun by Dr. Metz on March 20, 1 Charles L. Metz, Superintendent of Explorations, Field Notes, 1878-1882, Archaeological Explorations by the Literary and Scientific Society of Madisonville, O. {Ms. book I, p. 4.) These notes are deposited in the Peabody Museum. 1 Op. cit., p. 43 et seq. quoting, J. T. Short, North Americans of Antiquity. New York, 1879, pp. 524-525. 4 INDIAN VILLAGE SITE AND CEMETERY 1879. The point at which his laborer came upon the cemetery was at the southwestern corner of the plateau bounded on the west by the valley of the little stream called Whiskey Run, and on the south by the bottom lands of the Little Miami River. From this point, Dr. Metz began to trench eastward along the edge of the plateau and his labors were quickly rewarded by the finding of numerous and thickly grouped burials. On the first of April the Madisonville Literary and Historical Society took charge of the work, with Dr. Metz as superintendent of excavations. Successful explorations were conducted on the southwestern slope until June 23, when in accordance with instructions from the society, work was begun on the Stites property known as “ Spice Bush Woods.” This tract of land was on the eastern portion of the same plateau on which the cemetery was located. During the next two weeks Mound 8 in the Spice Bush Woods was explored and four hut circles were trenched. Further excavations yielded but meager results; and on the 7th of July work was resumed in the old trenches of the cemetery on the Ferris property, and car¬ ried on steadily with abundant results up to the end of the year. During this year’s work 360 skeletons were exhumed and 176 cache- pits explored; only 31 of these burials and 16 of the cache-pits are plotted upon the plan (plate 30). About 140 earthenware pots and many other specimens were discovered. Toward the end of the year, Dr. Metz inaugurated the system of dividing the areas to be explored into blocks approximately 40 feet square. The boundaries of the blocks were usually determined by trees en¬ closing roughly square or rectangular spaces, and the burials w r ere located with reference to one or other of these corner trees or other prominent trees within the blocks. All of the excavations prior to the inception of this system were lumped together into one large block. As early as April 12 of the first season, the ground had been staked off by Mr. Low and the principal features sketched. As the excavations in the successive blocks progressed, it is presumed that the burials were plotted on maps, but these maps are not in the possession of the Peabody Museum, nor has the present writer been able to consult them. Throughout the succeeding year (1880) excavations were con¬ tinued in the cemetery up to the 18th of December, when the ex¬ ploration of Block 12 was completed, the local society having NEAR MADISONVILLE, OHIO 5 been aided by an appropriation from the Cincinnati Society of Natural History. During this year 239 skeletons were exhumed, 170 cache-pits examined, and over 75 earthenware pots recovered. Such of the burials and cache-pits as could be plotted with a fair degree of accuracy are shown in the plan (plate 30, n). A good deal of work was also done in two kitchen-middens discovered at the heads of ravines, one on the western slope of the plateau, the other on the western slope of the gully which divides the southern portion of the plateau, running north from its edge. Dr. Metz’s work, in 1881, was commenced on the second of March and continued to August 13, when the entries in his field note-book leave off. Block 17 was completed and 58 skeletons and some 84 cache-pits were discovered. Professor F. W. Putnam visited the site in July of this year and again in September, at which time he made arrangements with the Madisonville Literary and Scientific Society, by which the Pea¬ body Museum, by paying a portion of the expenses of exploration, would receive its share of all that was obtained as the work of ex¬ cavation progressed. Work was continued by Dr. Metz during; the fall, and in February, 1882, Professor Putnam wrote in his report on the Museum: We have already received the first instalment of the material obtained since the cooperation of the Museum, and information that a second lot has been forwarded to the Museum. Our connection with this work has also re¬ sulted in the presentation of many specimens by those who have heretofore sustained the principal cost of the explorations, as will be seen by a reference to the list of additions over the names of Messrs. C. F. Low, P. P. Lane, C. L. Metz, and E. A. Conkling, 1 Excavations by Professor F. W. Putnam. Professor Putnam again visited the cemetery from May 2 to May 16, 1882, during which time the “ Putnam Block ” was excavated for the Museum. There is deposited in the Museum a small sketch map of this block, apparently made by Dr. Metz, and the note-book in which Professor Putnam recorded the excavations. This block yielded 8 skeletons and several cache-pits and was located near the eastern line of the Ferris property (plate 30, 1). During this visit four of the large hut circles on the Stites property northeast of Ferris Wood were explored also. 1 Peabody Museum Reports, vol. iii, p. 67. 6 INDIAN VILLAGE SITE AND CEMETERY Excavations were not carried on in the cemetery except sporadi¬ cally for some time after Professor Putnam’s visit, since he and Dr. Metz began explorations elsewhere in Ohio which continued over a period of several years. In 1891, when Professor Putnam was Chief of the Anthropological Department of the Chicago Exposi¬ tion, work was again resumed at the Madisonville site under the charge of Dr. Metz, with Harlan I. Smith as assistant, and Block M was excavated on the Stites property for this Department. Early in 1897 the Museum received notice from Dr. Metz that Miss Phebe Ferris of Madisonville had bequeathed to the Museum about twenty-five acres of land, on which is situated a large part of the ancient Indian cemetery where, in connection with Dr. Metz, since 1881, the Curator has carried on extensive explorations, and from which place a considerable col¬ lection has been made and exhibited in the Museum. This bequest was con¬ firmed by a notice from the executor of Miss Ferris’ will. It was known that Miss Ferris was much interested in the explorations on her farm to which she always gave her consent and kindly aid; and it was her expressed desire that the Museum should have the right to continue the work, and that after the explorations were completed the land should be used as a public park, thus marking the site of the ancient village and cemetery. 1 As it was deemed desirable for the Museum to carry on the work of exploration during that year, R. B. Dixon and J. R. Swanton, with Ingersoll Bowditch of the class of 1897, assisted by the cooperation of Dr. Metz, explored a portion of the ground. Mr. Bowditch made a survey and plan of the site which forms the basis of the final plan (plate 30). Excavations by J. R. Swanton. The account of the excavations conducted in 1897 is derived from the report of J. R. Swanton, his field notes and map, all of which are in the archives of the Museum. Work was carried on from July 27 to October 20. Mr. Swanton says in his report: The work was under the general supervision of Dr. C. L. Metz, and under the more immediate charge of Mr. R. B. Dixon and myself, Mr. Dixon taking charge during the first five weeks and myself during the remainder of the time. Mr. Ingersoll Bowditch assisted in the work throughout and executed surveys of the bequest to the Museum and the location of the Indian burial place upon it. The general location of the Madisonville cemetery is well known to stu¬ dents. . . . The hilltop on which it is situated sinks south to the lower bottom lands of the Little Miami River. On the west it is bounded by a small Thirty-first Report of the Peabody Museum, p. 10. Peabody Museum Papers Vol. VIII, Plate 1 Portions of Trenches II and III, showing skeletons and positions of pottery vessels found with them, also the distribution of cache-pits, hearths and post-holes. This area is indi¬ cated on plate 30 by the rectangle enclosed by broken lines. NEAR MADISONVILLE, OHIO 7 steep-sided stream known as Whiskey Run and on the other two sides by less considerable declivities. The extreme eastern portion extends beyond the Museum lot into the Stites estate, but this together with the southern part of the cemetery included in the lot itself has already been explored. Our ex¬ plorations this summer were consequently the western part of the hilltop toward Whiskey Run. . . . Our trench lines were run bearing N. 58°E. and stakes were driven every ten feet from which to locate the articles discovered. Five trenches were run in all, four of them of the regulation forty feet width and one of twenty feet. The length was determined at one end by the boun¬ daries of previous explorations, at the other by a straight line two hundred feet from the lower line of Trench I which extended to the edge of the hill. Trenches I, II, and half of III were extended to this line but work had to be discontinued before Trenches IV, V, and the second half of III were carried so far. The soil of this hill was composed of three strata, . . . First came black forest mould two to three feet thick at the edge of the hill and decreasing slowly to one foot and a hah or a foot at the upper end of the trenches. Be¬ neath this was four or five feet of hard yellow clay, and underlying all at a depth of perhaps six feet a very deep deposit of sand. On the Stites estate much of the hill had been removed for this sand. Trench I yielded 67 burials, 71 cache-pits, 7 pots; Trench II, 51 burials, 50 cache-pits, 4 pots; Trench III, 66 burials, 21 cache- pits, 11 pots; Trench IV, 38 burials, 31 cache-pits, 18 pots; Trench V, 3 burials, 4 cache-pits. The total number of skeletons excavated during the season was 230. One hundred and seventy- seven cache-pits and 13 fire places were explored; 42 pots were found. Plate 1 is a copy of a part of Mr. Swanton’s plan showing in detail the position of burials and the distribution of cache- pits. The area illustrated is indicated on plate 30 by the rec¬ tangle enclosed by broken lines in Trenches II and III. Excavations by R. E. Merwin. The next exploration of the site was conducted in 1907 by It. E. Merwin, Hemenway Fellow in the Museum, assisted by Irwin Hayden, also Hemenway Fel¬ low. The work continued from July 6 to November 7, 1907. The results of this season’s excavations may be summarized from Mr. Merwin’s report, maps, and field notes. Mr. Merwin’s Trench A included the tract of land lying be¬ tween Trench I of the 1897 excavations and the northwestern edge of the plateau as delimited by the banks of the creek, Whis¬ key Run. The side of this trench adjoining the previously explored territory is com¬ paratively regular, but the other side is very irregular, due to the meander- 8 INDIAN VILLAGE SITE AND CEMETERY ings of the ravine bank. In size it is 150 feet long, at both ends practically 10 feet wide, its greatest width at the center being approximately 100 feet. The depth necessary to carry the trench to hard-pan varied from 1 to 2 feet. This trench yielded but 4 skeletons, 15 cache-pits and one hearth. It marks the northern boundary of the cemetery. Trench B extended northwest from the head of the ravine which divides the southern part of the plateau into east and west portions. The southern and western sides of this trench adjoined land previously explored by Dr. Metz. Explorations were started in the southeast end of the trench so that the work might be carried forward to a better advantage, for there is a decided elevation in the ground to the northwest. A fifty foot trench was laid out, but owing to the irregularity of the previously explored ground, its width varied from 35 to 75 feet. Its length was 130 feet. Sixty skeletons were found in this trench. Forty cache-pits and two hearths were explored. After carrying Trench B forward about 130 feet, it was thought best to start another trench parallel and adjacent to it. But opening the trench in line with the southeast end of Trench B, it was found that former explorations had not been carried this far. So instead of running a trench to the north¬ west as in the case of Trench B, one was run in an opposite direction, following along a narrow ravine now used as a road. This made an irregular shaped trench; and it was also necessary to vary the width of it in order to take in all the unexplored territory adjoining it. This trench was extended for a distance of 110 feet, the width varying from 30 to 50 feet. There were found in Trench C, 21 skeletons and 31 cache-pits. A portion of this trench was occupied by a large kitchen-midden which had been partially explored by Dr. Metz in earlier years. The total finds of the season’s work were 85 skeletons, 32 pots, 88 cache-pits, and 3 hearths or fire places. In the following year (1908) Mr. Merwin excavated in the site of the cemetery from April 1 until September 15, opening three trenches. The portion explored in 1908 was adjacent to that ex¬ cavated in the latter portion of the season of 1907 (see plate 30). Trench D was a continuation of Trench C northward, and ad¬ joined Trench B on the west. As can be seen by the map, this trench had a width of 80 feet, and in general was 180 feet long, with an irregular strip on the south side. This trench yielded 101 skeletons and 147 cache-pits. Peabody Museum Papers Vol. VIII, Plate 2 Madisonville Site The greater part of Trench D, showing skeletons and the positions of pottery vessels found with them, also the distribution of cache-pits. This area is shown on plate 30 by the rectangle enclosed by broken lines. NEAR MADISONVILLE, OHIO 9 Trench E was merely a continuation of Trench B of the explorations of 1907. It varied in width from 35 to 55 feet and extended toward the bluff of the ravine ; a distance of 170 feet. As can be seen on the map of this trench the finds were very good in the southern portion, but the northern part is practically devoid of skeletons or cache-pits. The reason for the barrenness of this portion of the trench is, as can be seen from the general map, because this tract had already been explored. There were 38 skeletons and 41 cache-pits found in this trench. Trench F was a continuation of Trench D. Perhaps one of the most striking things in connection with its exploration was the depth at which some of the skeletons were found. Many of them were from 3 ft. 6 in. to 4 ft. below the surface and extended 2 ft. 6 in. to 3 ft. in¬ to the hard-pan. There were 37 skeletons and 59 cache-pits found in this trench. The total number of skeletons exhumed in the season was 176 and 247 cache-pits and 3 hearths were explored. Excavations by B. W. Merwin. In 1911 B. W. Merwin, brother of R. E. Merwin, conducted explorations for the Museum in the northeastern portion of the tract. The data in regard to his excavations are derived from his field notes and maps. Trench G was east of Trench E and north of Trench F and ex¬ tended N. 76° E. As laid out, the block was 290 feet long and 30 feet wide. Beginning at the west end the trench was excavated 90 feet eastward and a trial trench was dug at its eastern ex¬ tremity. This trench yielded only 3 skeletons and 7 cache-pits and seems to have marked the northeastern limit of the cemetery. Trench H ran parallel with Trench G. Its southeastern corner • was 200 feet south of the corresponding corner of Trench G. Trench H was 150 feet long and 50 feet wide; it yielded 43 skele¬ tons and 72 cache-pits. Trench I was 80 feet square and ran parallel and adjacent to Trench H on the south. Trench I contained 26 skeletons, 52 cache-pits, and 23 post-holes. Trench J was a plot 30 feet by 50 feet laid off north of the east¬ ern portion of Trench H and adjacent to it, and east of Trench I. Several trial trenches were dug in this tract and two cache-pits were discovered. 10 INDIAN VILLAGE SITE AND CEMETERY Trench Iv was an irregular tract situated on a southern point of the plateau. In this trench were found 12 skeletons and 13 cache-pits. In addition to the above, Mr. Merwin dug several trial trenches in the tract north of Trench G and south of the plateau edge, and was able to establish the fact that the cemetery did not cover this northern portion of the plateau. Six other trial trenches, dug north of Whiskey Run toward the eastern boundary of the Museum tract gave negative results. THE BURIALS Depth of Interments. The depth of the burials varied in the different parts of the cemetery and in individual cases. Usually the graves were dug in the soft leaf mould overlying the hard yellow clay, and did not extend down into the latter. Conse¬ quently, where the leaf mould lay thick, the depth of the inter¬ ments tended to be greater, and where the leaf mould was thin, the graves were likely to be shallow. Naturally, in certain places denudation of the surface soil took place, so that burials were ex¬ posed or their original depth decreased. Mr. Swanton says: The skeletons were usually buried in the lower part of the black mould or the upper part of the hard-pan. The depth was eighteen to twenty-four inches. Tabulation of the depth of interment of 151 burials (all the skeletons excavated by Dr. Metz in 1879 in which depth was re¬ corded) gives the following result: Less than Depth. 12 in. 12-18 in. 19-24 in. 25-36 in. 37-48 in. 4 ft. Number of burials. .. . 8 53 77 12 0 1 Percent. 5.3 35.1 51.0 8.0 0.0 0.6 This random sample agrees well enough with Mr. Swanton’s statement. R. E. Merwin also says: A few were found in the hard-pan, but for the most part they were resting directly upon the hard-pan or above it in the black soil. Mr. Merwin notes, however, that in Trench F some of the burials were of unusual depth: Many of them were 3 ft. 6 in. to 4 ft. below the surface and extended 2 ft. 6 in. to 3 ft. into the hard-pan. Of 37 burials in this trench, 6, or 16.2 per cent r NEAR MADISONVILLE, OHIO 11 were at a depth of more than 3 feet from the highest point of the skull to the surface; 17, or 46 per cent, were between 25 in. and 36 in.; 8, or 21.6 per cent, were between 19 in. and 24 in.; 6, or 16.2 per cent, from 12 in. to 18 in. Forms of Burial. The following table shows the distribution of forms of burial for 650 skeletons exhumed by Dr. Metz: Contracted Horizontal Sitting Collective and disturbed Total Number. . 38 491 32 89 650 Per cent. . 5.8 75.6 4.9 13.7 100.00 Of 231 skeletons found by Mr. Swanton, 164, or 71 per cent, were extended or horizontal burials; 66, or 28.6 per cent, were col¬ lective or disturbed burials; there was but one contracted burial. He says in his report: The position was invariably full length, except in the cache-pits. Not in¬ frequently a skull with a small pile of long bones would be found beside a regular full length burial. [See plate 3, b.] Horizontal or Extended Burials. More than 75 per cent of all the burials in the Madisonville cemetery were horizontal or ex¬ tended burials. In addition to these a large number of the dis¬ turbed or “ bunched ” burials may have been originally extended burials which had been moved to make room for later interments. In most cases the bodies were laid flat upon their backs, with the faces either upward or turned to one side. The arms were usually extended at the sides (plates 3, a, b and 4, b). ■ Contracted Burials. Dr. Metz records 38, or 6.3 per cent, of his burials as contracted or “ in the doubled-up position.” Con¬ tracted burials seem to have been confined to the southern portion of the cemetery excavated by Dr. Metz, since the subsequent excavators found them only in cases where bodies had been buried in cache-pits and were necessarily flexed. Burials in Sitting Posture. These were also confined to that portion of the cemetery which Dr. Metz explored. He records 32, or 5 per cent of the skeletons exhumed by him. Of these, all but one were found during the excavations of the first season, and in the first few weeks of the work. Burials in the sitting posture seem, therefore, to have been confined to the extreme south¬ western portion of the cemetery. 12 INDIAN VILLAGE SITE AND CEMETERY Collective and Disturbed Burials. The collective burials found seem to have been secondary in character. In certain places a number of old burials may have been cleared away to make room for new ones and the disarticulated bones from the old graves buried in a common pit. Dr. Metz records the following find on April 12, 1879: A circular excavation, 3| feet in diameter, was made, and from it were taken twenty-two skeletons. The bottom of this pit was paved with the river mussel shells ( Unio ); over them were the bones of the lower extremities and trunks, and above these the skulls. Another more typical collective burial, consisting of one articu¬ lated skeleton and a miscellaneous heap of bones of several in¬ dividuals, was found by Dr. Metz, September 24, 1879: An adult female with small foetal bones in pelvic region, and a confused mass of bones about her lower extremities, among which were five crania, two facing upwards and three downwards. Finds analogous to the above are reported by all of the exca¬ vators (see plate 3). Very commonly odd portions of a disturbed burial were found in a subsequent interment. In some instances burials were disturbed in order to make room for cache-pits. Dr. Metz on September 3, 1880, reports: Block 10, skeleton 1, head east, adult male; arms flexed on the humeri. The body was disarticulated at the second lumbar vertebra; the lower ex¬ tremities together with the pelvis were buried separately, probably to make way for cache-pit 1 of this block. R. E. Merwin makes the following note on skeleton 37 in Trench D: Adult; ash-pit 71 had been dug through this skeleton, and the portion in the way of the pit had been removed; the parts removed were the right femur, right pelvic bone, sacrum, some of the lower right ribs, right ulna and radius, and about two inches of the distal end of the right humerus, and all of the right finger bones. Burials disturbed by the digging of cache-pits are not infrequent]}" reported throughout the excavations (see plate 4, a, d). Burials in Cache-pits. In rare instances bodies were buried in the cache-pits. The first of these burials was recorded by Dr. Metz on January 20, 1880: Skeleton 3. Block 3, in the bottom of above pit (cache-pit 2), lying on back, face up, head northwest, legs sharply flexed on thighs. The portion of Peabody Museum Papers Vol. VIII, Plate 3 Madisonville Site 1 16 INDIAN VILLAGE SITE AND CEMETERY The following table summarizes the orientation of Mr. Swanton’s burials: N. S. E. W. NE. SE. NW. sw. Total Number 2 4 21 0 62 55 11 14 169 Per cent 1.2 2.4 12.4 0.0 36.7 32.5 6.5 8.2 100.00 The apparent result seems superficially not to agree with that obtained by Dr. Metz since the table shows very few burials with heads south and a larger number with heads northeast. This dif¬ ference, however, is probably due to the fact that Dr. Metz merely estimated his orientations, whereas Mr. Swanton and subsequent excavators took their bearings accurately with a compass. Con¬ sequently very few burials are recorded as due north and due south and none due west. On the other hand 12.4 per cent were due east. Eighty-one and six-tenths per cent of the burials recorded by Mr. Swanton were orientated east, northeast, or southeast. Objects Associated with the Burials. This section deals with the finds in the cemetery only in relation to the burials. The following table shows the number of pots found by the principal excavators and recorded in the field notes deposited in the Peabody Museum. It also indicates the position of the pots with reference to the body, where found in association with burials, and mentioned by the investigators. Number Right Left Between of pots Head Feet side side knees Dr. Metz. 228 137 11 4 11 3 J. E. Swanton. 46 26 0 5 4 3 R. E. Merwin. 72 49 1 6 7 6 B. W. Merwin. 7 7 0 0 0 0 F. W. Putnam. 1 1 0 0 0 0 Total. 354 219 12 15 22 12 Of 354 pots recorded to have been found by the excavators r their position, in relation to the body, is stated in 280 instances. Of the remainder practically all were found with skeletons, but in the first year of his excavations Dr. Metz frequently omitted to note the position of the pots in relation to the associated burials. It will be observed that in the vast majority of cases the pots were found at the head, either above the vertex or to the left or right without apparent distinction. NEAR MADISONVILLE, OHIO 17 About 28 per cent of the entire number of burials recorded had pots associated with them, usually one pot to a burial. The pot¬ tery finds were more numerous by far in the southwestern portion of the cemetery excavated by Dr. Metz in 1879. In the north¬ western portion excavated by the Merwin brothers, pots and arti¬ facts of all kinds were rare. The pots were buried with individuals of both sexes and of all ages without distinction. Mr. Swanton says: As a rule one unio shell was found in each pot, seeming to indicate that it contained — if indeed it contained anything — something of a liquid nature. These pots were frequently such as had been broken around the rims before burial. R. E. Merwin also states that an examination of the contents of the pots usually showed one or more unio shells and a few small fragments of animal bones. All of the investigators found occa¬ sional pots containing traces of red ocher. It is perhaps worthy of note that on April 1, 1879, Dr. Metz found a pit containing four crania and four pots. One of these pots was inverted and rested on a skull. Stone implements were frequently found in the burials. These included adze blades, grooveless axes, hammers, scrapers, arrow and spear points, chisels, knives, and perforated disks. Finds of stone implements in the burials seem to have been far more com¬ mon in the southwestern portion of the cemetery, first excavated by Dr. Metz, than in his later explorations and those of the sub¬ sequent excavators. The majority of the implements seem to have been found in the cache-pits and kitchen-middens, and in the leaf mould, rather than in the graves. Arrowpoints, groove¬ less axes and knives were the commonest implements found with the bodies. In several instances arrowpoints were found em¬ bedded in portions of the skeletons. On April 12, 1879, Dr. Metz found a collective burial consisting of 22 skeletons. A sacrum belonging to one of these skeletons had imbedded in the anterior surface near the upper border a flint arrowhead, which in order to reach that position in the living body must have passed through the abdominal walls and intestines in an oblique direction, from above downwards and to the right. On July 10 of the same year the skeleton of an adult was ex¬ humed, one of the dorsal vertebrae of which had a flint arrow- 18 INDIAN VILLAGE SITE AND CEMETERY point embedded in its dorsal aspect. Mr. Swanton and Mr. Merwin record similar finds. Pipes of limestone or catlinite were often found with the skele¬ tons of males. Dr. Metz mentions 47 of these in his field notes, and it is probable that he found many more than that number. Mr. Swanton says that he found 18 complete pipes, and twice as many broken ones. Pipes were usually found clasped in the hand or laid by the neck. On June 6, 1879, Dr. Metz discovered a burial with which were associated “ two inscribed stones,” one of which seems to have been incised with double-pointed arrow designs, and the other with checker-board pattern. Bone objects commonly found with the burials were beads, cylinders, scrapers or gouges, and awls. Miscellaneous unworked animal bones frequently occurred with the human burials. Aside from unio shells which were commonly found in the pots and loose in the graves, objects of shell were rare in the burials. There occurred, however, shell beads, disks, pendants, perforated box tortoise shells, and at least one ornamented shell spoon. (Metz, skeleton 222, 1879.) Other objects found with the skele¬ tons include points, cylinders, harpoons, arrow straighteners, combs, and arm-bands. Ornaments of copper and pieces of hammered copper were found alike in graves and cache-pits, in the kitchen-middens, and among the surface relics. Dr. Metz records the finding of more than 50 flat pieces of copper, 40 copper rolls or beads, 5 strips of copper, also many beads, rings, and one copper cross. Mr. Swanton and Mr. Merwin also found many copper beads and hammered strips with the burials. Objects of iron in the burials were exceedingly rare. On Decem¬ ber 17, 1880, Dr. Metz discovered the skeleton of an adult female (Block 12, 597), over the chest of which was placed “ a piece of iron resembling a sword hilt” (plate 18, t). Mr. Swanton found a skeleton (I, 40), with which was associated an iron bead or cylinder (plate 18, s) embedded in the deerskin. These seem to be the only cases in which the association of iron with the burials was definite. Several pieces were found in the leaf mould and also in the cache-pits. NEAR MADISONVILLE, OHIO 19 Mr. Swanton states that blue glass beads, apparently of Euro¬ pean origin, were found about the head of skeleton I, 39. A few small shell beads were found with the glass beads. Fragments of the same kind of glass beads were also found by Mr. Swanton in three of the cache-pits. These glass beads were not found by the other excavators, with the exception of Mr. B. W. Merwin, who found one in a cache-pit. Lumps of cannel coal were found with comparative frequency in the burials. In some instances the pottery vessels contained red ocher. Dr. Metz records a burial (Block 7, 1) in which a handful of charred corn was found about the cranium. Bird crania used for pendants, and animal teeth pierced for suspension or for stringing, were also found in the graves. Burials of Special Interest. Certain burials deserve special mention, either because of some anomalous feature, or because of the special interest of the associated objects. Some of these have already been mentioned, and others, which seemed of special im¬ portance to the investigators, will be described briefly here. In Trench I of Mr. Swanton’s excavations, skeleton 60 was as¬ sociated with some interesting finds: These consisted of several copper plates one and a half by two inches square which had been clasped around deer hide still preserved with the hair on it by action of the copper. With these plates were found a number of shell beads made from the inner whorls of univalves, two or three copper beads, etc. A very small sea shell which had been rubbed down in some way and discolored by ocher, was identified as Marginella apicina, a variety of shell from the Gulf Coast and the Carolinas. All these articles were between the thigh bones of the skeleton. Grave 18 of Mr. R. E. Merwin’s Trench D. This grave, primarily constructed for the burial of an adult skeleton, con¬ tains portions of at least three skeletons. The only bones not disarticulated were those below the pelvis of one of the skeletons and these were on the bot¬ tom of the grave. The right femur was slightly burned. Directly above and in contact with this portion of the skeleton was a mass of burnt and unburnt bones of two other skeletons and the upper bones (burnt and unburnt) of the first skeleton. Many of these bones were broken. The mixed earth in con¬ tact with these bones and extending to the surface of the hard-pan was burnt, although the bottom of the grave shows but little action of fire. Of the bones in the lower end of the grave some were burnt, while others in contact with them were unburnt. In this grave were found two pieces of worked antler and thirteen pieces of copper. 20 INDIAN VILLAGE SITE AND CEMETERY Grave 32 of Mr. Menvin’s Trench E. Adult; that portion of the skeleton above the pelvis was in cache-pit 26, and the portion from the distal ends of the femora down was in cache-pit 25. Those portions in the pits had settled to a lower level than the remaihder of the skeleton. Along the inside of the right ulna and radius, resting partially on the right pelvic bone, and at the left of the lower jaw, were the skulls of some species of long-billed birds. Between the pelvic bones were two bird bones; on the breast had evidently been placed a bone with a series of per¬ forations. Total Number of Burials. The following is a summary of the total number of burials exhumed by the principal excavators and recorded in the field notes deposited in the Peabody Museum. Number of Burials C. L. Metz 1879 360 J. R. Swanton 1897 230 u u u 188*0 239 R. E. Merwin 1907 85 U (( u 1881 58 U U (C 1908 176 F. W. Putnam 1882 6 B. W. Merwin 1911 82 Total 1236 To the above total should be added a number of burials sufficient to account for those washed out prior to the discovery of the ceme¬ tery, and for those discovered by Dr. Metz in sporadic excavations subsequent to August 13, 1881, at which date the field entries stop. We know that when Professor Putnam arrived in Madison- ville on May 2, 1882, Dr. Metz was excavating Block 22, whereas his field notes for the preceding year cease during, or at the end of, the excavations of Block 17. Blocks 2-17 inclusive averaged about 18 burials to the block. As the burials were fewer in the portion of the cemetery explored during the latter part of the ex¬ cavations, it is probable that an allowance of 12 burials to the block, for Blocks 18-22, would be sufficient. This would bring the total up to 1296 and allowing for burials destroyed or undis¬ covered it is probable that the total number of burials made in the cemetery was not less than 1350. Death Rate as Indicated by Burials. In order to calculate the population of the village site to which the cemetery belonged we should know the total number of burials, the length of time during which the cemetery was used, and the average death rate per annum. NEAR MAblSONYILLE, OHIO 21 There exists very little reliable data concerning the death rate of modern primitive peoples. However, the following method of approximation seems legitimate in this instance. The propor¬ tions of different ages in the burials of the cemetery will be com¬ pared with the mortality statistics for different ages in various European countries. If we find the constituent age proportions agreeing closely with those of some European country of which the average annual death rate is known, it may reasonably be concluded that the annual death rates are also approximately the same. Fortunately, Dr. Metz carefully recorded the approximate age in the case of more than five-sixths of the burials he found. Sex and Age Adult Adult Adult Adol. Child. Year Sex undeter. Male Female Age 18-11 10-3 Infants 1879 . 114 21 10 8 75 3 1880 . 25 59 53 10 56 30 1881 . 10 15 12 1 10 9 149 95 75 19 141 42 Total 521 Percent. 28.6 18.2 14.4 3.7 27.1 8.1 100.00 In this connection the following table will be instructive. Age at Death Aver. Annual Place Year Age 0-10 10-20 20+ Year Death Rate Italy. 1872-77 52.37 4.22 43.41 1865-78 2.99 France. 1866-77 32.28 4.25 63.47 1865-77 2.46 England. 1860-70 44.23 4.56 51.21 1865-78 2.20 Prussia. 1875-77 52.43 3.51 44.06 1865-78 2.72 Bavaria. 1871-77 52.61 2.22 45.17 1865-78 3.09 Austria. 1865-77 52.38 4.05 43.57 1865-78 3.18 Spain. 1865-70 51.86 4.37 43.77 1865-70 3.12 Russia. 1870-74 62.33 4.13 33.54 1865-75 3.67 Switzerland.. . 1873-77 36.94 3.72 59.33 1870-78 2.38 Madisonville.. 35.2 3.7 61.1 In the above table it should be observed that the average an¬ nual death rate in the various European countries seems to be cor¬ related closely with the number of deaths per hundred under the age of ten years. Thus, France, with a low percentage of infant and child mortality, has a very low average annual death rate; and Russia, with a very high infant and child mortality, has the 22 INDIAN VILLAGE SITE AND CEMETERY highest annual death rate. In this table the average annual death rate seems to vary more or less directly with the child and infant mortality, except in the case of England, which shows a somewhat lower death rate than would be expected from the number of deaths per hundred under the age of 10 years. This may be due to the fact that the average annual death rate given for England includes Wales, whereas the percentage of ages per hundred does not; or that the period, during which the calcula¬ tion of the annual death rate is made, is different from the period of calculation of the ages at death per hundred. On the whole, it appears legitimate to infer the approximate annual death rate from the distribution of ages at death per hundred deaths. Next it appears that the distribution of ages at death in per¬ centage of entire number of burials recorded in Dr. Metz’s sample of the Madisonville cemetery agrees most closely with the dis¬ tribution of ages at death per hundred in Switzerland. The Madi¬ sonville cemetery shows a mortality per hundred under 10 years of 35.2 against 36.94 in Switzerland; from ages of 10-20 years 3.7 against 3.72 in Switzerland; over 20 years, 61.1 in Aladisonville against 59.33 in Switzerland. The average annual death rate in Switzerland was 2.38 per hundred. We might, therefore, conclude that, the death rate per annum in the Aladisonville cemetery would be approximately the same. On the basis of Dr. Aletz’s sample of 521 burials — almost one half of the total burials recorded in the cemetery — the child and infant mortality is lower than in any of the European countries except France during the period recorded. In selecting data from European countries, I have purposely chosen a period some years back, before the recent improvements in sanitation and medical science brought about the modern lowering of the death rate. Conditions in Europe at that time more closely approximated the primitive, and are, therefore, more comparable with the conditions of health in an aboriginal American cemetery. It might be argued that we ought to expect a higher annual death rate in a village of Indians, presumably engaged from time to time in warfare, and living under very primitive conditions; but primitive warfare is not deadly, and it is probable that the number of deaths due to war in pre-Columbian America north of Alexico was proportionately smaller than in Europe at most NEAR MADISONVILLE, OHIO 23 periods of history. Epidemics also were probably lacking in the New World before the coming of the Europeans. We may therefore consider 2.3 per hundred as a reasonable ap¬ proximation for the average annual death rate during the period in which the Madisonville cemetery was used, and if we base our calculations on an average annual death rate of 3 per hundred, we shall be making ample allowance for infant burials overlooked or unrecorded. Population, and Period of Use of Cemetery. We must next con¬ sider the question of the length of time during which this cemetery was used by the Indians. This problem may be approached in two ways — through the historical evidence, and through the evi¬ dence of the burials themselves. The territory at the mouth of the Little Miami River was first colonized in 1788, by a party of 18 persons led by Major Benjamin Stites, within the limits of a tract of 10,000 acres purchased by Major Stites from Judge Symmes. Major Stites constructed a log fort and laid out the town of Columbia. 1 At the time that Major Stites first settled here there was in all probability no village on the Madisonville site. Cyrus Thomas states that Ohio was not inhabited at the time when it first became known to Europeans. 2 Colonel M. F. Force says: 3 In the latter half of the Seventeenth Century, after the destruction of the Eries by the Five Nations in 1656, what is now the State of Ohio, was un¬ inhabited. The earlier occupants of this region around the mouth of the Little Miami River seem to have been, according to tradition, the Shawnee. The original home and subsequent migrations of this powerful tribe have been a subject that has aroused consider¬ able controversy, into which we shall not enter at this time. It can only be remarked that if we assume the Madisonville Indian village to have been inhabited in the post-Columbian period (and the archaeological evidence supporting this assumption is irre¬ futable), the following considerations point toward the Shawnee as its probable occupants. 1 J. R. Dodge, Red Men of the Ohio Valley , Springfield, 1860, p. 372. 2 Cyrus Thomas, Indians in Historic Times , History of North America, vol. ii, 1903, p. 261 et seq. 3 Cf. Col. M. F. Force, Some Early Notices of the Indians of Ohio , p, 238. 24 INDIAN VILLAGE SITE AND CEMETERY The western Shawnee of the Cumberland basin are first men¬ tioned in the Jesuit relations of 1648 under the name of Ouch- aouanag. In 1670, as Chaouanoan, they are described as living some distance southeast of their neighbours, the Illinois. Dodge says of them: 1 At the period of the settlement of Virginia, they were doubtless the occu¬ pants of what is now the state of Kentucky, from the Ohio River up to the Cumberland Basin, to the country of the Cherokees. The main body, by invitation of the Andastes and Miamis, crossing the Ohio assisted in conflicts with the Five Nations. They are supposed to have shared the defeat of the Andastes and about the year 1672 fled to escape destruction. Early in the Eighteenth Century, the Shawnee, as a result of a war with the Cherokee and Chickasaw in the Cumberland region, again moved into Ohio. According to Dodge, when West Virginia began to be civilized they occupied the Scioto Valley and extended as far west as the Little Miami and Mad Rivers, having been in¬ vited thither by the Wyandot, at the instigation of the French. When the region about the mouth of the Little Miami River was colonized in 1788, the Shawnee were still in the vicinity, ac¬ cording to a quotation from Judge Burnet’s note by Dodge: 2 During these transactions the Judge was visited by a number of Indians from a camp in the neighbourhood of Stite’s settlement. One of them, a Shawnee chief, had many complaints to make of the frauds practised on them by the white traders. The archaeological evidence as to contact with Europeans, yielded by the burials in the cemetery, seems to indicate that the inhabitants were influenced by white culture to a very slight de¬ gree, only at the close of the period during which the site was in¬ habited. A few glass beads and a few bits of iron and brass in the most recent portion of the cemetery are the only traces of Euro¬ pean contact. Traders began to filter through this region about 1750. The inhabitants of the Madisonville site may have been an outlying group of the Shawnee of the Cumberland region who were ex¬ pelled from Ohio about 1705-15. There is further the possibility that this was a Shawnee village, the occupation of which termi- 1 Op. cit., p. 22 et seq. 5 Op. cit., p. 374. NEAR MADISONVILLE, OHIO 25 nated at the time when the tribe was first driven southward out of Ohio after the defeat by the Five Nations in 1672. It scarcely seems probable that this cemetery could have belonged to a Shawnee village of the subsequent migration northward, which took place about 1750, since the site was apparently occupied un- intermittently for generations, and there is not sufficient time be¬ tween 1750 and the period of European colonization of the Little Miami Valley to permit the growth of such a necropolis. Further, it seems that a Shawnee settlement subsequent to 1750 ought to show more traces of European contact than are exhibited here. Moreover we are informed that the Shawnee were entirely no¬ madic bands at the time of the first coming of the traders, but the archaeological evidence points to a sedentary population for this site. Finally, if we are to trust the historical tradition, Ohio was uninhabited in the latter part of the Seventeenth Century. It should be understood quite clearly that the only reason for connecting this cemetery with the Shawnee tribe lies in the fact that the historical traditions mention no other tribe as occupying this general region during the probable period of the occupation of the Madisonville village and cemetery site. As we know practically nothing of the archaeology and physical anthropology of the Shawnee, it is impossible either to verify this tradition or to refute it. It is equally possible that the site under consideration may have been inhabited by some other tribe concerning which local historical records are silent. If we assume that Madisonville was a Shawnee site, there are then two possibilities as to the period of its occupation. Either it was occupied up to about the middle of the Eighteenth Century or for a period anterior to 1672. The latter seems the more prob¬ able. A priori it seems unlikely that, considering the movements of tribes in early post-Columbian times, any Indian village in this region would be occupied continuously for more than one hundred years. But the Madisonville cemetery must have been used for a very long time, since a considerable change in burial customs took place within the period of its utilization. Contracted burials and burials in the sitting posture were found only in the oldest portion of the cemetery — the extreme southwestern portion. There are a number of mounds in the vicinity of the cemetery and these 26 INDIAN VILLAGE SITE AND CEMETERY almost certainly antedate the necropolis, although it cannot be stated with certainty that the same Indian group was responsible for both forms of burial. In the more recent portion of the ceme¬ tery the majority of skeletons were buried unaccompanied by artifacts and there are occasional finds of objects indicating European contact. But the strongest evidence of a long continued use of the ceme¬ tery seems to be afforded by the very large number of burials that have been disturbed in order to make room for later interments. Not only were there found frequent instances in which one body was laid in a grave over some portion of an earlier grave, but even more commonly an old grave was cleared out and its con¬ tents deposited in a heap in a corner of the new grave or nearby. Some of these “ bunched ” burials may be the remains of those who died away from home and whose bones were subsequently interred in the local cemetery, but in most instances they are probably due to the necessity of finding room in the sacred bury¬ ing ground, which brought about the disturbance of the earlier burials by the late dwellers on the site. Now it seems very improbable that people who buried their dead with the comparatively elaborate funeral furniture which is found in many of the earlier graves in this cemetery, would be likely to desecrate the graves of their ancestors within the time when the identity of the individuals who occupied the various graves could be remembered. In other words a man would be loathe to disturb the remains of his great-grandfather, as long as he knew where that ancestor was buried. Allowing twenty years to a generation it is probable that the fifth generation would have been entered upon before all the members of the first were dead and their burial places forgotten. If this argument is valid, it would then seem that the minimum period which could be con¬ sidered to have elapsed between the first burials in the cemetery and the subsequent ones, to make room for which the former were dislodged, must have been between eighty and one hundred years. But, as has been stated above, it seems improbable that any Indian group during this general period of tribal movements would have occupied the same site continuously for much more than one hundred years. Therefore, we may accept one hundred years as the approximate length of time during which the Madison- NEAR MADISONVILLE, OHIO 27 ville village site was inhabited. This may perhaps be considered a minimum estimate. Assuming the total number of burials in the cemetery to have been about 1350 and the annual death rate to have been about 3 per hundred, a village of 450 to 500 inhab¬ itants would have been sufficient to fill this cemetery in a century. We may be able to secure some check on this estimate of the size of the village by considering the house remains on the site ad¬ joining the cemetery. In 1879 Dr. Metz excavated four hut circles, and in 1882 Professor Putnam excavated four more. Those excavated by Professor Putnam varied in diameter from 40 to 60 feet. It is therefore evident that there were at least eight com¬ munal houses in the village, although these may not all have been contemporaneous. Each of these communal houses must have been the home of grandparents, parents, and children, averaging forty to fifty individuals to the house. Taking the lower figure, we should have a total of 320 inhabitants for the 8 houses, if they were occupied at the same time. But it is pretty clear that Pro¬ fessor Putnam and Dr. Metz did not excavate all of the hut circles on the site. There were at least four others in the group north¬ east of the cemetery. Twelve such houses should contain 450 to 500 inhabitants. This group of hut circles to the northeast would probably rep¬ resent the location of the latest village on the site. The sites of the earlier houses are indicated by the hearths found from time to time by the excavators in the various portions of the cemetery. The original settlement probably was on the southwestern corner of the plateau. The dead were buried about the houses, and as rubbish and graves accumulated, the house sites gradually moved northeastward. It thus appears that this cemetery was probably the burying place of a village from 450 to 500 in population for a period of about a century. THE CACHE-PITS Discovery. The cache-pits, called “ ash-pits ” by Dr. Metz and the other excavators after him, are scarcely inferior in interest and importance to the burials. The first of these was explored by Dr. Metz on April 1, 1879. The first cache-pit to be described fully was dug on April 22 of the same year. Of it Dr. Metz says in his field notes: 28 INDIAN VILLAGE SITE AND CEMETERY An ash-pit was excavated to a depth of 4| feet and a diameter of 3§ feet. The sides and bottom showed no evidence of the action of fire. It appears that the pits had been dug and the ashes put in them as ashes to a depth of 2| feet . . . In it were numerous bone implements, shells, and animal remains, none of which bare any traces of the action of fire. A few fragments of charred bones that had probably been in the ashes previous to their deposition here were also found. Hereafter, Dr. Metz carefully explored every cache-pit encount¬ ered, noting, in most cases, the depth, diameter, stratification of contents, and character of associated objects. Number and Depth of Cache-pits. The following table gives the totals of pits recorded by the excavators in their field notes that are in the possession of the Museum. Number of Cache-pits C. L. Metz, 1879-81 429 J. It. Swanton 1897 177 It it tt 1882 12 R. E. Merwin, 1907-08 247 F. W. Putnam, 1882 7 B. W. Merwin, 1911 11S Total 990 In order to give an accurate idea of the main dimensions of these pits, the data in regard to those excavated by Dr. Metz in 1879 have been tabulated. By far the majority were round or oval in shape. The few oblong ones have been omitted from considera¬ tion in the following tables. Diameter of Cache-pits , in Less than 2 ft. 2 ft. 2 ft. 6 in. 3 ft. 3 ft. 6 in. 4 ft. and over Number. 1 6 19 70 36 13 2 Total 147 Per cent. .7 4.1 13.0 47.6 24.5 8.8 1.4 Mode 3 ft. Depth of Cache-pits Depth of Ashes Depth Number Per cent Depth Number Per cent. 2 ft. 6 in.-2 ft. 11 in. i .7 1 ft.-l ft. 5 in. 9 7.9 3 ft.—3 ft. 5 in. 16 10.8 1 ft. 6 in.-l ft. 11 in. 20 17.7 3 ft. 6 in.-3 ft. 11 in. 12 8.1 2 ft.—2 ft. 5 in. 24 21.1 4 ft.-4 ft. 5 in. 22 14.9 2 ft. 6 in.-2 ft. 11 in. 18 15.9 4 ft. 6 in.-4 ft. 11 in. 14 9.4 3 ft.—3 ft. 5 in. 22 19.1 5 ft.—5 ft. 5 in. 32 21.6 3 ft. 6 in.-3 ft. 11 in. 9 7.9 5 ft. 6 in.-5 ft. 11 in. 27 18.2 4 ft.-4 ft. 5 in. 5 4.4 6 ft.—6 ft. 11 in. 22 14.9 4 ft. 6 in-4 ft. 11 in. 4 3.9- 7 ft. 2 1.4 5 ft. 2 1.7 Total.. . 148 Total. . 113 NEAR MADISONVILLE, OHIO 29 It will be observed from the above table that almost three- fourths of the pits were between 3 feet and 3^ feet in diameter; that the depths varied greatly but were in the majority of cases from 4| feet to 6 feet; that the depth of ashes was usually from If feet to 31 feet. Contents and Stratification. Some of the pits were stratified and some were not. In order to give a clear idea of the stratifica¬ tion and contents of typical pits descriptions have been copied from Dr. Metz’s notes. Cache-pit 396 Depth 5 ft. 3 in.; diameter 3 ft. 3 in. 10 inches leaf mould 28 inches sand and gravel 25 inches ashes Contents: animal remains, shells, sherds, burned limestone, flint and bone relics. Cache-pit 400 Depth 5 ft. 10 in.; diameter 3 ft. 2 in. 10 inches leaf mould 12 inches sand 48 inches ashes Contents: burned limestone, animal remains, flints, sherds, shells, one piece of copper. Cache-pit 401 Depth 6 ft. 8 in.; diameter 3 ft. 3 in. 10 inches leaf mould 24 inches sand 46 inches ashes Contents: animal remains, burned limestone, shells, perforated unios, a large quantity of deer skulls, awls, cylinders. Cache-pit 402 Depth 5 ft. 6 in.; diameter 3 ft. 2 in. 10 inches leaf mould 5 inches sand 19 inches yellow sand and ashes 3 inches yellow sand 19 inches sand and ashes 4 inches sand 6 inches ashes 6 inches ashes in depression in middle of bottom Contents: animal remains, burned limestone, large sherds of pottery, shells, elkhorn implement, awls, flints, record stick. 30 INDIAN VILLAGE SITE AND CEMETERY Cache-pit 404 Depth 4 ft. 1 in.; diameter 4 ft. 10 inches leaf mould 40 inches ashes Contents: two bushels of snail shells. Cache-pit 409 Depth 5 ft. 2 in.; diameter 3 ft. 10 in. 10 inches leaf mould 30 inches sand 22 inches ashes Contents: animal remains and fish hook. Cache-pit 411 Depth 5 ft. 6 in.; diameter 4 ft. 10 inches leaf mould 32 inches sand 3 inches yellow sand 10 inches sand and ashes 11 inches ashes 6 inches of ashes in depression in bottom Contents: animal remains, a stone pipe, and an elkhorn implement. The following cache-pits are less typical but present features of special interest: Extract from Dr. Metz’s notes, August 27, 1879: One of the most remarkable “finds” since the prosecution of the work on this cemetery was a large quantity (4 or 5 bushels) of burnt corn, in a pit which was opened and explored August 26 and 27. The shape of the pit was oblong and contracted toward the center, a horizontal section showing some¬ what like a figure 8. Its dimensions were: length 6 ft. 8 in.; diameter 3 feet at each end; total depth 6 ft. 3 in. Beginning at the bottom the contents were as follows: (1) A layer of bowlders, pottery sherds, ashes, with a few animal bones, about 6 inches in thickness. (2) Above this a layer of completely carbonized corn on the cob. (3) Matting, composed of twigs and coarse grasses — two inches. (4) Then about 4 bushels of shelled corn, also completely carbonized. The corn and floor of bowlders occupied one half of the pit or one loop of the figure 8, its superficial area being a circle of about three feet in diameter, its depth 10 inches. (5) Above the corn was a layer of coarsely woven matting composed of grass, cornstalks and twigs, also completely charred, depth 4 inches. (6) Next came a layer of ashes 10 inches in thickness, intermingled with which were bones of deer, elk, bear, racoon, opossum, turkey, mink, woodchuck, squirrel, and beaver, together with unio shells of various species. NEAR MADISONVILLE, OHIO 31 (7) Fifteen inches of gravelly clay also containing numerous anima re¬ mains with occasional implements of flint, bone and stone, anl un¬ finished stone pipe representing a bear sitting on his haunches, and sherds of pottery. (8) Two feet of black leaf mould. FEET Fig. 1. Madisonville Site Cross-sections of abandoned and filled cache-pits from Dr. Metz’s note-books; a, Cache- pit used for burial purposes, opened January 21, 1880; 6, Pit opened August 8, 1881; c, Pit opened in May,' 1881; d, Pit opened August 26-27, 1879. Aside from the materials shown in the cross-sections, the pits contained potsherds, bones of various animals, im¬ plements and other objects of stone, bone and antler. It is quite obvious that this peculiar shaped pit was really two intersecting and independent round pits, one of which had been used for a corn cache; a cross-section of this is given in figure 1, d. Cache-pit 377 Depth 5 ft. 8 in.; diameter 3 ft. 10 inches leaf mould 38 inches ashes containing animal remains 6 inches charcoal and animal remains 4 inches matting 12 inches corn cobs Contents: “ At a depth of 4 ft. 4 in. a layer of charred animal remains was found and below this a layer of matting 4 inches deep, and below this matting carbonized corn cobs. On the matting a layer of plaited grass rope was found [figure 1, c]. Mr. Swanton found a layer of corn cobs in the bottom of pit I, 33, and small quantities of corn in other pits. A layer of corn 2 inches thick and 4 feet from the surface was found in cache-pit II, 23. Two feet, 8 inches from the surface of the hard-pan was a concreted mass of gravel discolored by fire and about 9 inches thick. The mass seems to have become solidified under the influence of a hot fire though whether this fire occurred before the concretion was thrown into the cache-pit or subsequently could not be determined. 32 INDIAN VILLAGE SITE AND CEMETERY Fauna Represented in the Cache-pits. The following is a quota¬ tion from Mr. Swanton’s report: By studying the contents of over one hundred cache-pits, twenty-one species of animals were identified, one species of bird, two of turtles, and two of fish. These are deer, bear, dog, elk, buffalo, racoon, opossum, beaver, porcupine, woodchuck, lynx, fox, puma, wolf, otter, squirrel, skunk, mole, rabbit, marten, badger, turkey, leather-back turtle, turtle, catfish and perch (?). Deer bones overbalanced all others in the proportion of at least three to one. Few badger, squirrel, skunk, wolf, puma, and rabbit bones were ;. Black earth SlACR EARTH - . ; bl;a c k Dearth' ■ : \ - v“A v *•* j' CLAY A H E S W CLAYVno ASHES: V V M IX1 0 “EARTH CLAY t S I-D C " A BCOAiAT SLACK earth b E ■ charcoal..oCORN 0 i a 3 4 5 C a reiT Fio. 2. Madisonville Site Cross-sections of abandoned and filled cache-pits from Mr. Swanton's note-books; a. Pit 6, Trench IV; b, Pit 54, Trench I; c, Pit 15, Trench II; d, Pit 64, Trench I. found. In the case of the rabbit this is very remarkable since rabbits are very plentiful in the region. Of the mole and marten only one skull each was found, the latter asso¬ ciated with a burial; of the otter only two skulls, also associated with burials, and a fragment of a third were found. Below is an analysis of two pits showing approximately the proportion of bones belonging to each species which they contained. •Cache-pit III, 6 Deer. 80 per cent Cache-pit III, 8 Deer. 75 per cent Turkey. 8 Dog. 7.5 Bear. 2.5 “ Elk. 7.5 Elk. 2.5 “ Bear...1. 5.0 Dog. 2.5 Beaver, racoon, tur - Turtle (2 species) 2.5 “ key, turtle, fish 5.0 Racoon. Beaver. 0.5 0.5 Total. 100.00 “ •Opossum, puma, fish, fox, otter, wood¬ chuck, badger. 1.0 “ Total 100.00 NEAR MADISONVILLE, OHIO 33 The finding of the skulls of the marten and otter with burials without other bones of these animals indicates that they formed portions of medicine bags made of the whole skin of these animals with skull attached. Similar bags of otter skin were used in the well-known Shell Society and the presence of these skulls seems to indicate that this ceremony was known to these Indians. The shells found belonged principally to the genus Unio. Most of the species having smooth shells could not be identified, but of the following there can be no doubt: Unio rectus (Lamarck), vermi- cosus, tuberculatus, ovatus, undulatus, cornutus (Barnes), alntus (Say), Margitana rugosa (Barnes). One small species of the uni- » b c d Fig. 3. Madisonville Site Cross-sections of abandoned and filled cache-pits in Trench D, from R. E. Merwin’s note¬ books. Mr. Merwin’s cross-sections were the first to show that probably most of the pits were dug through a stratum of hard-pan, and into the substratum of sand which afforded excellent drainage and served to keep the contents of the caches dry. a valve ( Fulgur carica ) was found in addition to a small sea uni¬ valve Marginella apicina. The latter species is found on the Gulf Coast and the shores of the Carolinas. Artifacts from the Cache-pits. The following extracts from the notes of Mr. B. W. Merwin give a sufficient idea of the number and character of artifacts found in typical cache-pits. Cache-pit 51, Trench H Depth of leaf mould, 8 in. Depth to hard-pan, 1 ft. 8 in. Depth to bottom, 5 ft. 2 in. Diameter at hard-pan, 3 ft. 8 in. Specimens: 1 bone scraper, 3 antler cylinders, 1 antler point, 2 worked antlers, 1 fragment bone scraper, 1 notched rib, 1 celt, 1 fragment celt, 1 perforated shell, 1 flint point, 1 flint scraper, 2 worked stones, 100 bones, 75 unworked stones, 40 shells, 25 flint flakes, 40 potsherds. 34 INDIAN VILLAGE SITE AND CEMETERY Cache-pit 7, Trench K Depth of leaf mould, 5 in. Depth to hard-pan, 1 ft. 1 in. Depth to bottom, 4 ft. 8 in. Diameter at hard-pan, 3 ft. 2 in. Specimens: 1 worked beaver tooth, 1 bone bead, 1 long bone perforator, 1 fragment bone scraper, 1 perforated shell, 1 small whetstone, 1 per¬ forated bone, 150 unworked bones, 30 stones, 25 shells, 30 potsherds, 15 flint flakes. Cache-pit 10, Trench H Depth of leaf mould, 9 in. Depth to hard-pan, 1 ft. 9 in. Depth to bottom, 3 ft. 7 in. Diameter at hard-pan, 2 ft. 8 in. Specimens: 2 antler cylinders, 2 fragments bone scrapers, 1 bone perfora¬ tor, 5 flint points, 8 reject flint points, 1 worked bone, 1 fragment pipe, 2 rough hammerstones, 1 grinder, fragments of bones of two kinds of dog, 100 stones, 175 bones, 50 potsherds, 40 flint flakes, 30 shells, 4 quarts charred wood. Pit filled with black earth and clay with a 5 inch layer of bones and ashes about 2 ft. 9 in. from the surface. Occasionally pieces of copper or ornaments of copper were found in the pits. These finds were more common in the portion of the cemetery excavated by Dr. Metz. Beads were the commonest objects of copper found by Mr. Swanton in the pits. From pit I, 50, were recovered three copper discs made by clinching thin pieces of copper over bone. In pit I, 55, was a small copper orna¬ ment 4 inches long, in shape of a serpent. Iron was of very infrequent occurrence in cache-pits. On Septem¬ ber 19, 1879, Dr. Metz found a piece of an iron hatchet or toma¬ hawk, 13 inches below the surface in the leaf mould between two cache-pits. Mr. Swanton found a flat piece 2 inches long by 1^ inches wide in the first pit he opened. This was 3 feet 1 inch be¬ low the surface and at least 1 foot below the level of the hard-pan. In pit I, 66, another piece was encountered at a depth of 3 feet below the level of the hard-pan and 4§ feet from the surface. R. E. Merwin found an iron ring in pit 98, Trench D. The ring was 27 inches below the surface of the ground, the leaf mould being 4 inches in depth. There was no evidence of its intrusive origin as there was no root hole above it. Skeleton 65 was buried with its head in the cache-pit on the level with the ring, but the ring was 21 inches from the skull. NEAR MADISONVILLE, OHIO 35 B. W. Merwin found a piece of iron in pit 27, Trench I. This was a perforated piece of iron, little oxidized. It came from a depth of one foot eight inches, the soil above it being black earth mixed with potsherds and shells. Merwin thought that it must have worked down a root hole, as there was a root hole 5 inches away, about 3 inches in diameter. r « o z 2 Q < z Q ; p'Tyy mol/loo’;;.'. ^ black . EARTH ; -.' ■ - ■ 8LACK f ftKTff' v;: b lac k ' e a «r h ?. - - a> d Fig. 4. Madisonville Site Cross-section of abandoned and filled cache-pits from B. W. Merwin’s note-books; a, Pit 3, Trench G; b, Pit 24a, Trench H; c, Pit 30a, Trench I; d, Pit 26, Trench I. In common with nearly all pits of this site, potsherds, various implements and the bones of animals were found in the refuse with which they were filled. Mr. Swanton found fragments of blue glass beads in three cache-pits (I, 44; I, 64; IV, 31). Mr. B. W. Merwin also found a blue glass bead in one of the pits that he excavated. Form of the Cache-pits. The great majority of the pits were round or oval in shape, but Dr. Metz discovered some that were oblong. He records about ten of these. The first of this shape, opened May 19, 1879, was 3 ft. 2 in. in depth, 7 ft. 3 in. long, and 4 ft. 6 in. wide. It contained 16 inches of ashes. The follow¬ ing is his description of one of the oblong pits: October 17, 1879 Large oblong cache-pit. Depth 4 ft. 3 in., length 8 ft., width 6 ft. 5 in. Contents: leaf mould 10 in. Ashes and sand 10 in. Pure ashes 15 in. Sand 6 in. Ashes, charcoal, and bowlders, 10 in. Specimens: one grooved stone hammer, one elkhorn implement, one large piece of elkhorn, broken bowlders, potsherds, bear cranium, flints. 36 INDIAN VILLAGE SITE AND CEMETERY It will be observed that the contents of the oblong pits are not different in character from the round ones. Apparently the later investigators did not find any of these oblong pits. The circular and oval cache-pits occasionally had a larger di¬ ameter at the bottom than at the top and often the reverse. Bot¬ tle-shaped cache-pits also occurred. Grouping in Relation to Burials. The cache-pits were often found in groups, as may be seen from the plan of the cemetery (plate 30). Sometimes two or more pits were dug so close together that their sides intersected. Generally speaking the site was divided into portions rich in cache-pits and with very few burials on the one hand, and parts with many burials and few pits on the other. It seems that for the most part the pits are not connected with the burials except fortuitously. In a few instances, which have been dealt with under the consideration of the burials, bodies were buried in the cache-pits, simply because the empty pit offered a convenient ready-made grave. In other cases bodies were buried over these pits, or cache-pits were dug through burials. Yet there is noth¬ ing that points to any special relationship between the graves and the pits. On the other hand there is no doubt at all that the graves and the pits were made by one and the same people during the same period. The identity of artifacts in the pits with those found in the graves establishes this fact beyond question. Purpose of the Cache-pits. The earlier excavators seem to have considered the existence of this large number of pits, laboriously dug down through the hard-pan, as more or less of a mystery. Yet there is no evidence of their having subserved any obscure esoteric purpose. On the contrary it seems that they were dug for a very practical end. All of the investigators note that for the most part the bottoms and walls of the pits show no trace of the action of fire, but that the ashes and other charred debris seem to have been dumped into the pits after having been subjected to the action of fire. In a few instances fires seem to have been made in partially filled pits, or burning embers were emptied into the pits. In the opinion of the present writer the cache-pits were con¬ structed primarily for the storage of corn and other goods. At NEAR MADISONVILLE, OHIO 37 the harvest time the maize crop belonging to each family was de¬ posited in a number of pits grouped together and located con¬ veniently near the house-site or place of habitation. Indeed these pits may often have been within the houses or huts. In the course of the winter the pits were emptied one after another and the empty pits became successively the receptacles into which were swept the accumulating ashes, food remains, and other debris of the hearth and home. When a pit showed stratification of con¬ tents it was because it was filled with rubbish gradually, and when the contents were unstratified it was due to the dumping of a large amount of refuse into the pit at one time. Some of the stratified pits were probably left partially filled for a considerable length of time; fires were occasionally made in them, and in a few instances the empty or partially filled pits were utilized as graves. Ample indication that the pits were used as granaries is afforded by the two pits described above, in which Dr. Metz found charred shelled corn and charred corn on the cob covered with matting (figure 1, c, d). Historical evidence confirms the conclusion that the pits were granaries. Wood says in speaking of the customs of the New England Indians: 1 Their corne being ripe, they gathered it, and drying it hard in the Sunne, conveighed it to their barnes, which be great holes digged in the ground in the form of a brass pot, seeled with rinds of trees, wherein they put their corne. Morgan also states the following in regard to the corn pits of the Iroquois: 2 The Iroquois were accustomed to bury their surplus corn, and also their charred green corn, in caches in which the former would preserve uninjured throughout the year, and the latter for a much longer period. They excavated a pit, made a bark bottom and sides, and having deposited their corn within it, a bark roof, water-tight, was constructed over it, and the whole covered up with earth. Pits of charred corn are still found near their ancient settle¬ ments. Cured venison and other meats were buried in the same manner, except that the bark repository was lined with deer-skins. The above also explains the reason for the carbonized condition of the corn found in the pits by Dr. Metz. 1 William Wood, New England's Prospect, p. 100. 2 L. H. Morgan, League of the Iroquois, p. 319. 38 INDIAN VILLAGE SITE AND CEMETERY Additional evidence is furnished by Catlin, who, in speaking of the Mandan Indians says: 1 The green corn season is one of great festivity with them, and one of much importance^ The greater part of their crop is eaten during these festivals, and the remainder is gathered and dried on the cob, before it has ripened, and packed away in “caches” (as the French call them), holes in the ground, some six or seven feet deep, the insides of which are somewhat in the form of a jug, and tightly closed at the top. The corn, and even dried meat and pem- mican, are placed in these caches, being packed tight around the sides, with prairie grass, and effectually preserved through the severest winters. Alice C. Fletcher, also, in speaking of the expeditions of the Omahas, says: 2 If any were so fortunate as to possess a surplus supply of food or clothing, they would store it in a cache, which they might either conceal or leave un¬ disguised. The cache was dug in a dry place, sometimes lined with poles, but often left with no wall but the hard soil. The goods were covered with skins, the earth was thrown over, and the place marked with piles of stones — meat, corn, clothing, and other personal property were kept for months in this manner, and no one disturbed the hidden store. That caches of this character were used by the Indians over a very wide area in North America is shown by their existence among the Thompson Indians of British Columbia. In regard to their methods of storing food, James Teit says: 3 The most common cache is the Indian cellar. This is used solely for the storing of berries, fish, etc. A circular hole about four feet in depth, and of the necessary diameter, is dug. In it are carefully laid the articles to be stored. If these are berries or roots, they are placed in baskets, and wrapped over with birch bark. The roof is then put on. It consists of small poles laid closely side by side across the excavation. Above them are laid in the same manner, but at right angles, another row of poles. The structure is then covered with pine needles and earth. H. I. Smith states that pits which are supposed to be the re¬ mains of the cellars are found near the ancient underground house-sites in the Thompson River region. 4 In the mound region in general, archaeological confirmation of the identity of the “ ash-pits ” with caches mentioned by the historians is not lacking. 1 Geo. Catlin, North American Indians, vol. i, p. 121 et seq. 2 Alice C. Fletcher, Tribal Life Among the Omahas, Century Magazine, January, 1896, p. 450. 3 James Teit. The Thompson Indians of British Columbia, p. 198 et seq. 1 H. I. Smith, Memoirs American Museum Natural History, vol. ii, no. 6, p. 403. NEAR MADISONVILLE, OHIO 39 Speaking of the enclosures in New York, Squier says: 1 The first feature which attracts notice upon entering them is a number of pits or excavations in the earth, usually at points which are most elevated and dry. These pits are occasionally of considerable size, and are popularly called “ wells,” although nothing is more obvious than that they could never have been designed for any such purpose. They are usually 3 to 4, but some¬ times from 6 to 8, feet in depth, and of proportionate size at the top. Their purpose became sufficiently evident upon excavation. They were the caches in which the former occupants of these works deposited their stores. Parched corn, now completely carbonized by long exposure, is to be discovered in considerable abundance in many of them. Instances fell under my notice where it has been found untouched to the amount of bushels in these primi¬ tive depositories. Traces of the bark and thin slips of wood, by which the deposits were surrounded, are also frequently to be found. Thomas in his work on Mound Explorations identifies pits of this character with caches. 2 Finally Alice C. Fletcher writes of the caches of the Omaha Indians as follows: 3 Each family had outside the lodge a cache, and some of the families would have two. These caches would be used sometimes for two or three years, but after a time the posts would become worm-eaten or the rain get in and if the cache was not repaired, as it occasionally was, a new one would be built close beside it and the old one taken for other uses to be described. In the cache was kept the winter supply of corn, dried meat and other provisions and the gala dresses and ornaments of various kinds. These were kept in parfleche packs; also the sacred articles, such as medicine bags, or sacred bowls, etc. When a village was attacked it was always considered important to try and burn the caches. Fire was put in the cache and the articles consumed in the hole; sometimes they were only charred. A malicious person having a grudge would sometimes take revenge by burning a cache. The old caches were used for ash-pits. The accumulations of ashes in the center fireplace (a circular depression in the center of the lodge) would be cleared, and the ashes thrown in the pit. So also the bones and refuse of eat¬ ing, and of feasts, and the broken implements and weapons, worn-out moccasins, and other articles. When the pit was filled up it was closed over and another one taken. The sites of the old villages are honeycombed with these caches, the Indians say, and I am invited to examine them if I like, the Indians laugh¬ ing heartily at my queer curiosity. 1 E. G. Squier, Smithsonian Contributions to Knowledge, vol. ii, p. 12 et seq. 2 Cf. Thomas, Twelfth Report Bureau American Ethnology, p. 32 et seq. 3 Peabody Museum Reports, vol. iii, pp. 357, 358. 40 INDIAN VILLAGE SITE AND CEMETERY HEARTHS AND OTHER REMAINS Hearths and Fire Places. J. R. Swan ton says in his report: Hearths are formed by a few stones laid side by side. They were usually a foot to a foot and a half in depth and two to three feet in diameter, though the outline was by no means always circular. Mr. Swanton excavated thirteen of these hearths or fire places, R. E. Merwin describes four, and B. W. Merwin two. Dr. Metz apparently did not distinguish the hearths from the shallower cache-pits. From the number of burned stones which he, in con¬ trast with the other excavators, describes as found in the cache- pits, it seems clear that the earlier inhabitants of the village often dumped their hearth stones with their ashes into the cache- pits at the end of the season, and that they often used the depres¬ sions in the earth, caused by partial^ filled cache-pits, as hearths. The following discovery described by Dr. Metz is certainly a hearth: Monday, March 31, 1879. Under the leaf mould I discovered a circular layer of ashes 4 inches deep and below it were burnt limestones and bowlders. Below these was a layer of mussel shells (unio) 3 inches thick. Below the shells a large round bowlder upon which rested an elliptical flint instrument, 3 inches long. The bowlder rested upon a layer of sand 8 inches deep, and on the clay partially covered by the sand and at a total depth of 2\ feet were two large prongs of elkhorn. R. E. Merwin describes as follows certain of the hearths explored by him: Hearth 1, Trench A Depth of top of rocks below surface of ground, 13 in. Total depth of depression, 25 in. Depth of depression below surface of hard-pan, 6 in. The depression or pit was of irregular shape being 5 ft. 2 in. long, and 3 ft. 9 in. at its widest part. Throughout the pit charred wood was found in great abundance. Within the pit and resting on the hard-pan were arranged ir¬ regular pieces of limestone varying in size from 1 inch in diameter to 10 inches, and also a number of pieces of sandstone. These stones showed action of fire as did the floor and walls of the pit. No charred animal bones were found or any artifacts. One potsherd was found above the pile in the loam 5 inches below the surface. NEAR MADISONVILLE, OHIO 41 Hearth I, Trench B Total depth below surface, 2 ft. 7 in. Depth below surface of hard-pan, 5 in. Leaf mould, 3 in. Diameter, 23§ in. This was a saucer-like depression in the hard-pan. The sides were not burnt. Four inches above hard-pan was a layer of large unburnt rocks, below this a stratum of black earth mixed with ashes. In this were found the frag¬ ments of the skull of a bear and a few fragmentary animal bones. Two inches below surface of hard-pan was a layer of burnt clay 1 inch thick, and below it was black earth with a great amount of ashes. All of the following finds in this stratum were burnt: a large potsherd, 6 grains of corn, frag¬ mentary animal bones, and a burnt clay ball. Mr. Swanton distinguishes between hearths and fire places. The latter he defines as “ areas on the surface of the hard-pan where the soil had been reddened by fire.” There was a great variation in the size and shape of these areas. . . . Usu¬ ally such fire places seemed to bear no definite relation to ash-pits or burials, though there may be an exception in the case of skeleton IV, 27, which lay over a stratum of charcoal covering a layer of red earth, probably a fire place. Ash-pit IV, 10 was surrounded by fire place 23. To the uniformity and ap¬ parent absence of importance of these fire places we must except fire places 29, 36, and 37, Trench IV. Fire place 29 from its singular shape and the peculiarity of its contents and surroundings I have ventured to call an altar. It bears some resemblance to the altars found at the Turner Group of mounds higher up the Little Miami River. This altar was a small saucer-like depres¬ sion in the surface of the hard-pan, baked red by fire and filled with very fine white ashes. Its depth below the surface was 2 ft. 10 in. and its diameter 1 ft. 9 in. The depth of the hollow forming this altar was 7 inches. The head of skeleton IV, 30, lay beside this altar, and ash-pit IV, 26, was a short distance off. Fire place 36 was similar to the others except that it was connected with the altar and unlike other fire places, pieces of a pot or of pots were found imbedded in the discolored soil. This discolored soil was about 2 inches in thickness. Fire place 37 also in all probability had something to do with the altar, but it was not directly connected with it. The present writer feels that the evidence upon which Mr. Swanton based his opinion that fire place 36 was an “ altar ” is insufficient. B. W. Merwin found two of these burnt places in Trench E. It seems probable that these fire places and hearths mark the site of habitations or camp-fires. * 42 INDIAN VILLAGE SITE AND CEMETERY Post-holes. Mr. Swanton writes in his report: Under the general name of “ post-holes ” I have catalogued many little excavations too small for cache-pits, and of rather doubtful significance. These were usually one to two feet in diameter, by from two to four feet in depth, running one or two feet into the hard-pan. Their contents were very insignificant — a bed of ashes, some charcoal, a few bones, a little charred wood, a few limestone pebbles, and a celt or two being about all that was dis¬ covered. The pits containing charred wood may really have been post-holes. No one can tell about the others. Dr. Metz probably did not distinguish these “ post-holes ” from cache-pits. Mr. Swanton records about a dozen of them, R. E. Merwin 30, B. W. Merwin 25. They were evidently found for the most part in the northeastern portion of the cemetery, and especially in Trench F and Trench I. The majority of them contained nothing but mixed earth and charred wood. Those re¬ ported by It. E. Merwin varied from 6 inches to 11 inches in diam¬ eter, averaging about 8 inches, the depths being from 2| to 3 feet. They are therefore smaller than those reported by Mr. Swanton. Those excavated by B. W. Merwin were of similar dimensions but with a slightly larger range in diameter and depth. Some¬ times there were a few bones, potsherds, stones, and ashes in them. It seems probable that many of these “ post-holes ” of small diameter, were, as Mr. Swanton suggests, actually what the name implies. Some of the larger ones may have been small cache-pits, but they were usually unstratified. Kitchen-middens. Dr. Metz discovered two kitchen-middens in the course of his work in the cemetery. The first of these he describes in his notes as follows: March 20 to April 1, 18S0 Work in the kitchen-midden was continued until April 1, and having reached the head of the ravine, it was deemed advisable to abandon further excavations at this point for the present. This deposit of kitchen refuse extended 34£ feet across the head of a ravine on the west side of the plateau, averaging in depth 6 to 7 feet, and was made up of irregular deposits of ashes, charcoal and sand, animal remains, unio shells, and sherds. This midden also yielded a few burials and cache-pits, and many artifacts. On April 2, a second kitchen-midden was discovered on the eastern edge of the plateau. This is evidently the one subse¬ quently explored by R. E. Merwin in 1907. NEAR MADIS0NY1LLE, OHIO 43 Mr. Swanton excavated a kitchen-midden in his portion of the cemetery which occupied a depression running east and west. The width of this midden was about thirty feet; its length, as far as traced about seventy, though it ran a considerable distance further east. In the deepest place this kitchen-midden measured about four feet, shallowing gradu¬ ally on each side. The average depth may have been about three feet. The soil composing this kitchen-midden was dark and rich and similar to the forest mould, except that layers of ashes were very common, often extending to the very bottom. A great deal of charcoal was scattered throughout but the proportion of animal bones was comparatively meager. Flints and flint points were common and so many small articles of interest were discovered that it was found advisable to pass all the earth through screens. The kitchen-midden explored by R. E. Merwin near Trench C had been formed by filling up a small ravine with refuse. He states that it was about 40 feet long, 8 to 12 feet wide, and varied in depth from 3 feet 6 inches to 5 feet. The following is a list of the objects found in this refuse heap: Human mandible, 1 Human patella, 1 Perforated shells, 30 Stone celts, 3. Stone celts (fragmentary), 2 Worked stones, 3 Sharpening stones, 21 Flint points, 23 Flint knives, 3 Flint scrapers, 4 Flint rejects, 21 Bone cylinders, 12 Bone bead, 1 Worked antler tips, 23 Worked antlers, 29 Mr. B. W. Merwin also explored a kitchen-midden in the ravine, running north and south. This midden had the following section: Leaf mould, 10 in. Ashes, 4 ft. 3 in. Black dirt, 1 ft. 10 in. Black dirt and ashes, 5 ft. 1 in. Ashes and black dirt, 2 ft. 4 in. Black dirt and clay, 7 ft. 7 in. Clay, 3 ft. 2 in. Most of the specimens in this midden were found in the ashes at depths varying from 3 feet 2 inches to 4 feet 3 inches. Antler blades, 9 Fragments of bone fish-hooks, 2 Bone awls, 12 Bone awls (fragmentary), 12 Notched bone awl, 1 Fragmentary bone scrapers, 48 Perforated bone (4 holes), 1 Notched bones, 2 Notched antler, 1 Grooved hammerstone, 1 Stone implements, 2 Shell object, 1 Piece cannel coal, 1 Reject antler points, 2 44 INDIAN VILLAGE SITE AND CEMETERY House Circles. Before the existence of the cemetery on this site was knpwn, Dr. Metz in his survey of the monuments of the region had recorded the existence of a number of circular depres¬ sions on the Stites property on the northeastern portion of the plateau. On June 24, 25, 26, four of these circles were investigated by him, by running trenches 3 to 5 feet wide through their centers. In one of these was found a bed of ashes 20 inches below the surface, con¬ taining potsherds, animal remains, and charcoal. Near the edges of the de¬ pressions were found an occasional implement or arrowpoint. When Professor Putnam visited the site in May, 1882, he ex¬ cavated four more of these circles (plate 30). The following are extracts from his note-book: May 6, 1882. Circle No. 1 Trench started north and south. Outside to outside of circle 43 feet. At southern end of trench numerous animal bones, burnt limestones, unio shells, and potsherds in the leaf mould. Also a worked antler tip and several flint chips, a rude hoe (?), and on the clay or an inch or two below its surface was a rude grooved axe. In the center of the circle lying on the clay we came to a layer of stones. Took the stones up and found ashes and charcoal and burnt clay under them. This was a fire place. Leaf mould 1$ feet at north end, 2 feet in center and 2| feet at south end of trench. At the north end of the ditch were many small burnt stones, a very few animal bones, and a few unio shells. In the central portion were a number of potsherds, clam shells, two broken flints, a broken flat stone with hole (gorget). The central ditch was run to the east. Finished the trenches but found nothing else of importance. The fire place in the center was on the clay and there is every reason to be¬ lieve that the floor of the house was hard clay. Now the question is how much of the leaf mould was in existence at the time the house was occupied ? The two feet in the center over the fire stones has certainly formed since the house went to decay. How much have the earth worms done here ? In digging the trenches very few worms were found, probably not over a dozen or twenty in the whole day’s work of three men. May S. Circle No. 2 about 60 ft. northwest of No. 1. Trench runs northeast to southwest. In the center under 6 inches of leaf mould was a bed of ashes 4 inches deep, about 5 feet in diameter. Ashes were fresh and comparatively recent. Under the ashes was 17 inches of leaf mould. To the south of the center were several bricks, a few inches in the leaf mould. It was evident that there had been some recent white man’s work on the site of the circle, probably a sugar boiling camp. A few animal bones and a few potsherds, and several flint chips were found. On clearing out under the recent ash-bed we came to about 2 inches of very black old ashes on the NEAR MADISONVILLE, OHIO 45 surface of the clay 17 inches from the surface of the leaf mould. In this bed of ashes were a few flint chips and fragments of pottery. The only thing found in the recent upper bed was a fragment of pressed brick showing its recent origin. Circle 3 was about 40 feet south of 1. It consisted of an embank¬ ment 9 feet in width and had a diameter of 43 feet inside the em¬ bankment. Number 4 consisted of an embankment 12 feet wide which adjoined that of 3 on the southeastern side. The diameter from outside to outside was 58 feet. These circles yielded finds similar to those above described. These circles certainly are the remains of houses or huts. As has been pointed out above it seems probable that these were the habitations of the later dwellers on the site, since the successive villages seem to have moved from the southwestern portion of the plateau in a northeasterly direction, probably owing to the accumulation of debris on the site and the large number of burials. ARTIFACTS FROM THE SITE By Charles C. Willoughby General Distribution of Types. In studying the artifacts from this site, it should be remembered that most of them are of types used at the time of, or immediately preceding, the first intercourse of the northeastern Indians with Europeans. Perhaps the most distinctive stone implements are the mullers or pestles having a flat expanded base, the discoidal stones, and the chipped adze blades. Broadly speaking, these three types are characteristic of an area enclosed by a circle about five hundred miles in diameter, with its center near Louisville, Kentucky, and taking in the greater portion of Ohio, Indiana, Illinois, Kentucky and Tennessee. It is almost wholly within the Algonquian and Iroquoian areas. The distribution of certain types of shell objects from the site is probably about the same. On the other hand, most of the ordi¬ nary forms of stone implements and shell ornaments have, of course, a much wider range. So far as the distribution of the rarer types of antler and bone objects is concerned, but little data outside of Ohio are available for comparison. Probably the most distinctive and least widely ranging group of artifacts is the pottery. The group seems to be confined principally to southern Ohio and certain portions of Indiana and Kentucky. Knives and the Larger Projectile Points. On plate 5 are illus¬ trated most of the forms of flint knives recovered from the cache- pits and general refuse. It is, of course, possible that some of these may be projectile points, but the greater number were probably knife blades which were hafted in short wooden handles and were employed in the manifold uses to which the implement is adapted. In looking over the large number of tools made of antler, and the chips and refuse pieces of this material found everywhere upon this site, abundant evidence is shown of the efficiency of the flint knife as a cutting implement. 47 48 INDIAN VILLAGE SITE AND CEMETERY Perhaps the most interesting example in the collection is shown at k. This is made of white flint, and the broad blade, comprising the upper one-fourth of the implement, is thin and beautifully chipped to a sharp cutting edge. Its lower three-fourths forms the handle, which was undoubtedly wrapped with buckskin or similar material as a protection to the hand. The knife or scraper figured in n, is of the uncommon beveled type. It is a little less than half an inch thick, with an abrupt chisel-like bevel at the left of the side shown in the drawing, and it also has the usual corresponding bevel upon the other side of the opposite face. Nearly all knives of this type have the bevel upon the left side when held with the base toward the observer. The reason for this seems obvious, when we remember that most Indians are right- handed, and in using the modern steel curved knife draw the blade toward them in cutting. Experiments show that this form of flint blade with its wide-angled serrated edge is especially useful in working wood. Very few of the larger blades, such as are illustrated in 1, m, o, were found. It is possible that m may have been a spear point, as the shank, which is broken off, seems to have been somewhat narrow for the attachment of a suitable knife handle. The more ordinary forms of knife blades, such as are illustrated in g, h, i, j, with their broad strong shanks for the attachment of hafts, were fairly common, as were also the thin finely chipped leaf-shaped blades (a) which were probably employed in more delicate work, such as flaying and cutting up animals, cutting out buckskin garments, etc., for which they are especially adapted. In the American Naturalist for January, 1902, the present writer figured and described ten prehistoric halted flint knives from the cliff-houses and burial caves of the Southwest. The following forms shown on plate 5 were represented: triangular (d), leaf-shaped (a), stemmed (h, j), and notched (i). These were secured to the handles in most cases with a cement made of gum. In one instance the cement was reinforced with twine wrapping, and in another ex¬ ample the notched blade was fastened with sinew. In two instances the blade was set at a considerable angle to the handle. One of the specimens resembled a pocket knife somewhat more than half Peabody Museum Papers Vol. VIII, Plate 5 Madisonville Site Knives, projectile points and scrapers. (About $.) NEAR MADISONVILLE, OHIO 49 open. In the Mexican collection of the Museum are eight pre¬ historic hafted blades of the leaf-shaped and triangular types, about two and a half to seven inches long, with handles ranging about five to seven inches, which illustrate the manner of hafting the larger chipped implements of these forms. The blades were secured to the handles with gum only. Very few of the sharp edged flaked knives, shown in b, were found during the explorations. The making of these required ex¬ ceptional skill, and they were used necessarily for the most delicate work, such as cutting hair and thin dressed skin and trimming feathers, for the edge is very easily dulled. A considerable number of the ordinary flint rejects and chips were recovered such as are nearly always found upon Indian vil¬ lage sites, but there is no indication that chipped implements were manufactured here in unusual quantities. Arrowmaker’s Tools. On plate 6, a-g, are shown several flint¬ working punches of antler. A considerable number of these were obtained, the majority being about one and a half inches long and three-eighths of an inch in diameter. A few, however, were larger. These were used undoubtedly with a hammer of stone or hard wood, in flaking suitable pieces of flint from large masses, and for the roughing out of blades and projectile points. In a number of specimens, one end is battered or split from repeated blows of the hammer. Unlike the ordinary antler flakers used in finishing blades by pressure, the ends of these punches are nearly always symmetri¬ cally rounded. A piece c f antler from which punches have been cut is shown in h. The antler was worked into a rod several inches long, with an un- worked portion at one end for a handle. The rod was then cut into sections. Many of these refuse pieces were found in the debris. In the accounts of the process of flint chipping by the Indians there are few references to the punch and hammer. It is apparent, however, that their use was wide-spread as a preliminary process to the final pressure flaking by the ordinary antler tools with special working ends. In the collections of the Museum there are examples of these punches from the village sites in other sections of Ohio, from the Iroquoian sites in New York, one good specimen from a Maine shell-heap, and a number from various places in the Mississippi Valley. 50 INDIAN VILLAGE SITE AND CEMETERY Catlin, in describing the process of arrowmaking among the Apache, says that the flint to be worked was placed in the left hand where it was firmly held by two or more fingers. The punch was grasped between the thumb and two fingers of the right hand, and a cooperator sitting in front with a mallet of very hard wood struck the punch upon its upper end. Both the holder and striker sang, and the strokes of the mallet were given in time with the music. 1 This apparently refers to the blocking out of suitable pieces for finishing by pressure flaking. In his illuminating paper on Yahi archery 2 Dr. Pope gives an interesting description of Ishi’s method of making arrowpoints. A boulder of obsidian was shattered by throwing a rock upon it. The chunks thus obtained were broken into smaller size by holding a short segment of deer horn or piece of bone against a projecting surface, and smartly striking it a glancing blow with a stone. The resulting flakes of obsidian best suited for arrow heads were roughly three inches long, an inch and a half wide, and half an inch thick. These were chipped into arrowpoints with the aid of the ordinary pressure flaker of antler. The longer pressure flakers were much less common on this site than the punches. Three of these, made of antler, are shown in plate 6, i, j, k. Their lower ends are specially formed for the kind of work required. They were used in the more delicate flaking necessary in finishing the finer blades and points. These long flakers may have been used without a haft, but some of the shorter ones of the same type from existing tribes are fastened to handles of varying lengths, some of which are long enough to allow the free end to be held between the arm and body of the worker. The relatively large number of the short antler punches found in comparison with the long pressure flakers (the ratio being roughly about one hundred to one) would seem to indicate that the smaller punches may have been used in the final flaking of at least a con¬ siderable portion of the coarser chipped implements. Figures 1, m, plate 6, represent examples of a certain group of stones, showing abrasions and other marks of use, which probably served in preparing sinew and perhaps as hammers for antler 1 George Catlin, Last Rambles, pp. 184-185. 2 Saxton T. Pope, Yahi Archery, Univ. of Calif. Publications, Amer. Arch, and Eth., vol. xiii, pp. 116-117, and plate 27. Peabody Museum Papees You. VIII, Plate 6 Madisonville Site Arrowmaker’s tools; a-g, Flint-working punches of antler; h, worked antler from which a punch has been cut; i-k, Antler tools for pressure flaking; l, m, Arrowmaker’s stones; re, o, Arrowshaft wrenches; p~r, Arrowshaft finisheis. (About i.) Peabody Museum Papers Vol. VIII, Plate 7 Madisonville Site a, Flint points for drills; b, c, Flint arrowpoints; e, Unfinished arrow-points of deer antler; /, i, Finished arrowpoints of antler; d, Tips of antler from which arrowpoints are made; j, Spear point of antler with incised decoration; k. Design incised upon j\ g, Flint arrow- point embedded in human vertebra; h, Antler arrowpoint embedded in human vertebra from a grave at Turpin’s Farm, near Madisonville; l, Antler point on very old arrow from the southeastern Indians. (About i.) NEAR MADISONYILLE, OHIO 51 punches in arrowmaking. Stones of like character were found with three arrowmaking outfits in Iroquoian graves during the Peabody Museum explorations in western New York. On the same plate, n, o, are shown two arrowshaft straighteners of antler. These were used in preparing the twigs out of which the shafts were made. The twig was heated and passed through the hole in the implement which was used as a wrench or lever to straighten its bends and irregularities. Arrowshaft straighteners of wood, horn, and ivory, from the Eskimo, Athapascan, Shosho- nean and other tribes are in the collections of the Museum. In this connection it is interesting to note the resemblance of these Madisonville implements to the so-called “ batons de commande- ment ” of the French caves, some of which were probably used to straighten the shafts of arrows or darts. The end of the small tine of the straightener, shown in n, is rounded and smooth and has evidently been used as a flaker. Specimens of the typical sandstone arrowshaft smoothers or finishers are illustrated in p, q, r. Most of the examples found were fragments. In use, the shaft was drawn through the grooves of a pair of these stones held faces together in the hand. Arrowpoints. Plate 7, f, shows four specimens illustrating the range in size and form of the antler arrowpoints, which are rela¬ tively abundant on this site. Those having a barb are compara¬ tively rare. A number of unfinished points were recovered which illustrate clearly the process of making. In d, are figured two antler tips which have been encircled by a groove and then broken off. These were evidently cast aside for the reason that it would involve more labor to cut away the broken portion below the groove, than to groove another tine and break it squarely off as in the third example. After detachment the tip was trimmed down and shaped as illustrated in e, probably with a beaver tooth chisel or flint knife, then drilled and finished by grinding or scraping. These points have a wide distribution and were used over a con¬ siderable portion of the central and eastern United States. Beau¬ tiful examples are in the collections of the Museum from the Ohio burial mounds, and also on very old historic arrows from the southeastern Indians. One of the latter is figured in 1. An antler point of this type embedded in the vertebra of a human skeleton is shown in h. This was taken from a grave at 52 INDIAN VILLAGE SITE AND CEMETERY Turpin’s Farm not far from Madisonville. The Indian had been shot from behind. The point figured in i, seems to be too large for an arrow and it may have been used on a small spear. Only a very few as large as this were found. The specimen illustrated in j, however, is un¬ doubtedly a spear point. The design shown in k, consisting of crosses, zigzag lines, and what seems to be an inverted spear with ornamented shaft and large head, is cut upon it in incised lines. This is from a cache-pit in Trench H. So far as can be judged from the specimens collected, the triangu¬ lar flint points of the type illustrated in c, seem to have been the favorite among these Indians, although the antler points above described were nearly as common. They are made mostly from a yellowish gray flint of various shades and are quite delicately chipped. This form was also the favorite among the Iroquoian tribes. A vertebra of a skeleton from Trench D with one of these points embedded in the bone is shown in g. Very few notched or stemmed arrowpoints were recovered. Some of these are illustrated in b. One cannot be certain, how¬ ever, that all were used as points for arrows, although most of them probably were, as they seem too small for knives, and the width of the bases between the notches is not sufficient for the attachment of a practical knife handle. Flint Drills. Several flint drills are represented in a, of this plate. A considerable number of these were obtained from the cache-pits and general debris. A few drills of like form, hafted in wooden handles, have been taken from cliff-houses and burial caves in the Southwest, which indicate in general the probable method of hafting the Madison¬ ville specimens. The most common way seems to have been to insert the broad end of the drill into a notch in one end of a handle six or eight inches in length, where it was secured with pitch and twine wrappings. This handle or shaft was twirled between the hands, or with the aid of some mechanical contrivance like the bow, which was in use for this purpose in the Southwest previous to the arrival of the Whites. Another method is shown by a speci¬ men from Colorado. A short stick was split through the center, and the broad face of the drill was placed between the two pieces at about equal distance from the opposite ends, with the drill Peabody Museum Papers Vol. VIII, Plate 8 Madisontille Site a, Beaming tool made from the femur of the puma; b, c, Beaming took made from the cannon bones of the deer; d. Beaming tool made from the spinal process of the buffalo; e-j, Flint scrapers. (About -J.) NEAR MADISONVILLE, OHIO 53 point projecting at right angles to the length of the stick. The two halves of the handle were then bound securely together. In use the handle must have been grasped with the drill projecting between the two middle fingers. Skin-dressing Tools. On plate 8 are illustrated four bone beam- ers for removing the hair from skins which were first made wet and folded, or left in a pile until the hair became loosened by fer¬ mentation. The skin was then thrown hair-side uppermost over a rounded piece of wood or section of tree trunk. The beamer was seized with one end in each hand, and applied to that portion of the skin lying over the beam. A light scraping with the sharp edges of the tool rendered the skin clean and free from hair. By far the greater number of beamers were made from the cannon bones of the deer. Two of these are shown in b, c. Fragments of several hundred of these were obtained. They were usually broken near the middle as this was the thinnest and most fragile part of the tool. Only a comparatively small number of unbroken ones were recovered. Among these are a few which are unfinished and clearly show the process of making. One method was to widen and deepen the natural longitudinal groove on the broad face of the bone, probably with the rounded edge of a flint scraper, until the required sharp edges upon either side were produced. Another was to cut a longitudinal slit through the wall and scrape away the bone upon either side until the proper edges were formed. As the edges became dulled by use they were again sharpened by scraping, and the process was repeated until, in many instances, the middle of the tool became so weakened that it apparently broke in use and was thrown aside. Another favorite bone from which beamers were made was the long spinal process of the buffalo (d). These were less common than those made from deer bones. Fragments of sixty or seventy were recovered, together with a few perfect ones. Their length ranged from about eleven to sixteen inches. Other bones of the buffalo were rarely found in the refuse. It is probable that during the occupation of this site, buffalo were not uncommon in the section of the Little Miami Valley which the village overlooked. If they were hunted to any great extent, how¬ ever, the Indians evidently brought the meat to the village upon the headland, leaving most of the bones, excepting such as were 54 INDIAN VILLAGE SITE AND CEMETERY selected for the making of implements, in the valley. Zeisberger, 1 who wrote in 1780, says the buffalo at that time were found near the mouth of the Muskingum, and were reported in considerable numbers along the Scioto. Judging from the large number of bones of the deer in the refuse of the site, it seems to have been the favorite food animal of these Indians. A very few beamers were obtained wrought from the larger leg bones of the deer and elk. The example shown in a, is made from a femur of the puma. The more essential tools employed in skin-dressing are the beam¬ ers above described, and the scraper or “grainer,” which is used for removing from the inner side of the skin the adherent fat and skin muscles, and also for “ graining ” or softening the skin as it dries. Various kinds of these instruments were in use by Indians in general. They were often made of the tibia of the deer, cut to a chisel-shaped edge which was notched or serrated to render it more efficient. This seems to be a much more practical tool than the adze-like scrapers with smooth edges used for this purpose by some of the tribes of the Plains region. There is no evidence, how¬ ever, that the people of this site used a bone “ grainer ” of this form. Another and equally effective tool of chipped flint seems to have been preferred. One of the most effective skin-scrapers employed in recent times by the Eskimo, but now abandoned among tribes who have lost the art of stone chipping, is the form shown on plate 8, e-j. Among the Eskimo they were usually hafted in short wooden handles elaborately wrought to fit the hand perfectly. They were also sometimes hafted in ordinary straight handles. Flint blades of this form with their scraping edges often beauti¬ fully serrated were recovered in large numbers from the general refuse of this site. They were probably hafted by inserting their narrower ends into straight wooden handles, and were doubtless used largely in skin-dressing, but were probably also employed in working bone and wood. It is not improbable that certain types of antler blades (plate 14, e, f) may also have been used as skin- scrapers. 1 David Zeisberger, History of the Northern American Indians , p. 59. Peabody Museum Papers Vol. VIII, Plate 9 Madisonville Site a-c, Chipped flint adze blades ground near the cutting edge; d, e, Adze blades; g, Hematite blade; h , Small slate blade or chisel; i-k, Grooveless axes. (About $.) The adze blades were probably hafted as shown in/, the grooveless axes as in l. NEAR MADISONVILLE, OHIO 55 Chipped scrapers of erratic shapes, such as occur in most collec¬ tions, examples of which are shown on plate 5, c, were very rare. Only about a dozen were found during the explorations by the Museum. Grooveless Stone Axes. Perhaps the most highly developed stone implements from this site are the grooveless axes made of very compact varieties of stone, the majority being carefully formed and polished. Most of these were obtained from cache- pits. Three are illustrated in plate 9, i, j, k, and the common method of hafting is shown in 1. Sometimes the shorter blades of this type were set into a hole which did not pass through the handle; but in the majority of cases the haft was perforated and the upper end of the blade projected, as is conclusively shown by five prehistoric examples in their original handles which have been taken from the beds of streams in the Iroquoian and Algonquian areas. In the sixth hafted specimen known to the writer, the blade is set into a hole which does not perforate the handle. The larger grooveless axes were probably used principally as im¬ plements for cutting wood, both with and without the aid of fire, while the smaller specimens may have been employed both as implements and weapons. Most references to stone axes among the Indians by the earlier writers probably relate to the grooveless type. In New England a few were in use in the first half of the Seventeenth Century. Johnson in 1654 refers evidently to this form as follows: “ They had a small number of Mawhawks [tomahawks] Hammers, which are made of stone having a long pike on one side and a hole in the handle which they tye about their wrists.” 1 Gookin refers to “ tomahawks made of wood like a poleaxe with a sharpened stone fastened therein.” 2 And Williams says that trees were felled with a “ stone set in a wooden haft.” 3 As the grooveless axe was evidently highly prized by the occu¬ pants of the Madisonville site, it undoubtedly continued in use till replaced by iron blades similar to those illustrated on plate 18, o-r. Very few small stone blades of this class were recovered during the exploration. One of hematite, typical of the miniature axes of this material from Ohio and the neighboring region, is figured (g). This was-found in a cache-pit. 1 Edward Johnson, A History of New England, p. 114. * Daniel Gookin, Historical Collections, Mass. Hist. Coll. 1st S. Repr. 1859, vol. i, p. 152. * Roger Williams, Key into the Language of America, R. I. Hist. Coll., vol. i, p. 130. 56 INDIAN VILLAGE SITE AND CEMETERY So far as the explorations of the Museum show, the grooved axe was unknown to the occupants of this site. Stone Adze Blades. Very few stone adze blades were recovered. They seem to have held a subordinate place among the implements of this people. The antler blades so common here may have served their needs as well. They certainly were more quickly fashioned and there was an abundance of material from which to make them. A few chipped stone adzes polished near the cutting edge were found, nearly all from cache-pits. These belong to a type much more common in the region of the lower Ohio and Mississippi Rivers, than in the north. Three are illustrated on plate 9, a—c. They are made of a cream-colored chert or flint. The one shown in a, was found with a skeleton. Two adze blades of polished stone of the ordinary form are figured in d and e. The larger of the two lay near the right femur of a skeleton in Trench D. The probable method of hafting these implements is shown in f. A small and finely finished blade of compact slate carefully polished is illustrated in h. This was probably used with the aid of fire in making the finer perishable objects of wood which were undoubtedly common among the people, and which the fragment of wooden food bowl, figured in e, plate 21, will serve to illustrate. Anvils, Hammer-stones, and Grinding Stones. Thomas Hariot, writing of the Virginia Indians in 1587, tells us that each house¬ hold had stones for cracking nuts and for grinding shell and other materials. 1 This statement would doubtless be equally true if ap¬ plied at that time to almost any ,tribe inhabiting the section east of the Mississippi. Anvils, hammer-stones, and stones for grinding were very abun¬ dant on the site under consideration, and must have been common objects in all of the houses. The anvils, or pitted stones, as they are more commonly called, were of the type usually abundant on old village sites. Two of them are illustrated on plate 10, d, e. These stones vary considerably, some of them being much larger than the ones illustrated. Their form and size were, of course, determined by the water-worn stones from which they are made, for they are usually unmodified save for the depression pecked 1 Thomas Hariot, A Brief and True Report of the New Found Land of Virginia, Holbein edition, p. 25. Peabody Museum Papers Vol. VIII, Plate 10 Madisonville Site a,6, c, Grooved club heads; d, e, Anvils or pitted stones;/, g, Hammer-stones; k, i, Mortars or stones for grinding; j, Grinding stone and muller; k, Muller. (About §.) NEAR MADISONVILLE, OHIO 57 upon one or both sides. It is generally conceded that these house¬ hold objects are primarily anvils on which acorns, hickory and other nuts were cracked. The bones of food animals were probably crushed on the larger anvils. They also undoubtedly served for other domestic purposes. Some of the smaller ones are battered along their edges, the result of being used as hand hammers. A considerable number of the round hammer-stones were found, of the type illustrated in f, g, such as occur upon many of the vil¬ lage sites east of the Mississippi. These are usually made of chert, quartzite or other hard varieties, and are thought to have been used among other purposes for pecking or roughing out various stone implements. A very large number of fragments of sandstone and limestone were found, with one or more abraded surfaces, which had been used for rasping or grinding in the making of various objects of shell, bone, antler, wood or the softer varieties of stone. In some instances these abrading stones were of special forms, like the grooved fragments of limestone shown on plate 21, f-h, which were evidently designed for finishing cylindrical objects of a nature similar to the shafts of arrows or arrowpoint flakers of antler. Slabs of limestone of various sizes were apparently used for grinding corn, acorns, paint and similar materials, one or both faces being worn smooth, or a depression formed by continued rubbing. Two of these are figured on plate 10, i and j. A shallow stone mortar probably for grinding paint or medicine is represented in h. The pestles or mullers were mostly of the form shown in j, which is the type generally used by Indians of the. Madisonville culture over a large portion of Ohio. Another and rarer form is illustrated in k. Club heads, of the kind figured in a-c, plate 10, were not un¬ common. They usually consist of an oblong pebble unmodified with the exception of the encircling groove, although occasionally one is wrought over nearly the entire surface, and in the case of c, the form is wholly artificial. These, as a rule, show no signs of having been used as mauls; they were probably all club heads, the weapon being similar to the well-known warclub of the Plains tribes. Several well-finished objects of unknown use with carefully ground faces, and made of compact altered slate were obtained 58 INDIAN VILLAGE SITE AND CEMETERY during the exploration. Three of these are figured on plate 21, i-k. They do not seem to have been employed as grinding stones or for preparing sinew. With our present knowledge, no definite use can be assigned to them. Five or six small discoidal stones of the usual type were recovered; two are illustrated in 1 and m of this plate. The upper one is perforated, and accompanied a skeleton. The one shown in m, was obtained from a cache- pit. They are probably gaming stones, and like the shell gorgets with the characteristic eye markings, seem to connect at least a part of the material culture of this people with that of the tribes more to the south. The so-called winged ceremonial stones, which form so attractive a group from Ohio, are represented in the collection from this site by a single fragment. This probably belongs to an older culture and doubtless bears no relation to the people under consideration. Fishing Implements. On plate 11, a-f, are illustrated a number of fish hooks made from fragments of the leg bones of deer or other large animals. These are usually grooved near the end of the shank for the attachment of the line. The one shown at a, however, is perforated at this point. Hooks in the process of making are shown in g, h, i. The fragment of bone was perforated near one end, probably with a flint drill, and grooved or cut through with flint knives. The piece was removed between the grooves which intersect at the perforation, and the hook finished by cutting and grinding. The variation in the size of the hooks perhaps indicates the wide range in the species of fish sought. One broken hook was obtained cut from a unio shell. The harpoon points shown in j-n of this plate are made of antler. Their bases are irregular and do not seem to have been so carefully fitted to the socket in the shaft as is usual among northern tribes. A peculiar feature of the point, m, is the double barb and double perforation for the line. The harpoons, of which these are a part, were probably used principally in taking large fish. Most of these points are from the general refuse of the village site, but one speci¬ men (1) was found with a skeleton. Awls and Needles. Some of the more typical awls and needles are illustrated on plate 12. Mat needles made from deer ribs, such as were probably used principally for sewing together flag leaves in making mats for house coverings, are shown in a-c. Compara- Peabody Museum Papers Vol. VIII, Plate 11 a-f. Bone fish-hooks; ff-i, Unfinished fish-hooks showing process of making; j-ti, Harpoon points of antler. (About J.) NEAR MADISONVILLE, OHIO 59 tively few were found as they are exceedingly fragile and perish¬ able. These needles seem to indicate that mat coverings were used upon the houses of this village. The mats were undoubtedly of the type so widely distributed in Canada and the northern por¬ tion of the United States, which are made of a double layer of flag leaves, so fastened together that the junction of the edges of the leaves in each layer is covered by the central portion of each leaf of the opposing layer. When properly adjusted these mats make a practically waterproof covering. A number of fragments of needles made of antler and having a circular cross-section were found, most of them showing more or less curvature. Two are illustrated at the left in e. In the third example figured, the upper portion has been much reduced in size, and notches were cut near the tip to which the thread was probably tied. Several well-polished thin perforators of bone with sharp points were recovered, two of which are shown in d. A series of antler pins is figured in f. They are well made and vary from one and three-fourths inches to eight inches in length. While they may have been used as bodkins, it seems probable that they were designed for another purpose. The other implements shown upon the plate are mostly bodkins or perforators such as abound in the refuse of this village site. They are made from various bones of birds and quadrupeds, the tarsal bones of the turkey (g, h), and the ulnae of the deer (k), being favorites. They were used in basket making and for many other purposes. A number of the type figured in h, made from the upper portion of the tarsal bone of the turkey, are notched along the upper part, as shown at the left in the figure. This is probably to prevent the fiber or thong wrapping, which sometimes serves as a protection to the hand in similar perforators, from slipping. The specimen shown in i, is a half of a broken beamer made of the cannon bone of the deer. The end has been ground to a thin edge. The implement would be an effective one for use in the construction of coarse splint basketry. Only a few of these were obtained. Antler Blades. Many blades were found, of different forms and sizes, made of elk antler, but it is difficult to determine their various functions with any degree of certainty. There were several of the 60 INDIAN VILLAGE SITE AND CEMETERY type illustrated on plate 13, a, b, each provided with a deep notch on the inner side near the upper end. In a, the upper portion has been broken off through the original notch, and repaired by cutting a second notch below, more shallow than the first and showing little wear. In the type specimen, b, considerable wear is present at the sides and back, near to, as well as within, the notch. The distribution of the worn surfaces seems to indicate hafting in the manner shown in d. All blades of this form have the base of an antler branch at their back, the lower portion of which in some specimens is considerably worn. It is probable that this abrasion was produced by a withe or stout thong arranged somewhat as in the drawing. Another blade of similar shape, but probably hafted in a some¬ what different manner, is illustrated in c. Upon either side, near its upper end, are shallow notches much worn. These indicate that the hafting must have been similar to that indicated in e. The cutting edges of these instruments are sometimes well pre¬ served, but they nearly all show wear. Some have the peculiar striae noticeable on certain stone blades generally supposed to have been used as hoes. This wear occasionally extends upward for a considerable distance above the edge, and it is very likely that the}’’ may have been used as mattocks. Others may have been employed in working charred wood for which their edges were well adapted. On plate 14 are illustrated the more common forms. The type represented i^ d, e, f, was the most abundant. They vary in length from about four to nine inches, and many of them were probably hafted after the manner of adzes. Others, especially those with irregular edges, may have been used as hoes or digging implements in the planting and the care of gardens. One cannot readily understand how implements with edges like those in d and c, could be used advantageously for the ordinary purposes of an adze or scraper, while their employment as digging implements, might tend to produce this irregularity which would in no way im¬ pair their value. Blades with straight or rounded edges, as in e and f, may have served as scrapers in skin-dressing or similar work. The specimen illustrated in d, is especially interesting as it shows the marks of the binding material, probably bark or split roots, which secured the blade to its handle. The tool had evidently Peabody Museum Papers Vol. VIII, Plate 12 Madisonville Site a-c, Bone needles for sewing mats; e. Bone needles; d, g-l, Bodkins; /, Antler pins. (About J.) NEAR MADISONVILLE, OHIO 61 been in or near a fire which consumed the bindings, leaving the blade blackened and charred where the burning wrappings came in contact with it. A number of specimens are perforated as indicated in a, b, c, but the holes show little wear. The perforations probably served in some way for securing the lashing which bound the implement to its handle. The specimen, c, has notches near its upper end to assist in hafting. The example illustrated in b, has less conspicuous notches near its upper extremity and also a small notch upon either side below the perforation. The original length of a few of the blades has apparently been much reduced by repeated sharpening. Beaver Tooth Chisels. The Indians who inhabited the Madison- ville site, as well as nearly all the tribes living in the beaver country, used the incisors of this animal for chisels or cutting implements. Chisels made from the upper and lower incisors are illustrated in plate 15, e. The lower incisors are much more commonly used for this purpose as the curvature is less. The hard outer enamel of these teeth can be ground to a keen edge, and the tool is ad¬ mirably adapted for making the smaller and more delicate objects of wood, bone and antler. These blades were hafted in short handles, usually of wood or antler. Antler hafts, some of which were probably used with beaver tooth blades, are found on certain Ohio village sites, but none which could be definitely attributed to this use were obtained at Madisonville. One handle of antler is shown in f, plate 15. This, however, more probably served in hafting a small flint blade. All of the beaver tooth cutting implements obtained during the Museum exploration were of the chisel type. None of the side cutting knives, such as occur also with this form in New England and certain other sections, were found. Historical references to the use of these implements are rare. Captain John Smith 1 writes of the Virginia Indian: to make the noch of his arrow, he hath the tooth of a Beaver, set in a sticke, wherewith he grateth it by degrees. He also says that bone fish hooks were worked out in the same way — “ grated as they noch their arrows.” In the collections of the Peabody Museum are two old beaver tooth chisels in wooden hafts, which were obtained years ago from 1 Voyages and Discoveries, Arber edition, vol. i, pp. 364-365. 62 INDIAN VILLAGE SITE AND CEMETERY the Eskimo of Bristol Bay. The sharpened tooth is inserted into a curved perforation running from one end through to the side of the handle. As the tooth becomes shortened by regrinding, it is adjusted by being pushed forward from the side of the haft. The remarkable wood cutting habits of the beaver undoubtedly exerted a marked influence in the choice of the incisors of this animal for blades of cutting implements, and it is in keeping with what we know of the workings of the unsophisticated Indian mind if supernatural powers were attributed to these tools. Musical Instruments. About twenty of the flute-like objects of the type shown in plate 15, d, made from the long bones of birds, were found during the explorations by the Museum. They are mostly small, and judging from the unbroken specimens recovered the number of finger holes range from five to nine, the usual num¬ ber being five or six. The holes were commonly about one-half inch apart, but in one specimen the centers of the perforations are placed about one-fourth inch from each other, too near, it would seem, for its successful manipulation by the fingers of an adult. These instruments are, of course, distinct from the bone whistles of the modern Plains tribes which have no finger holes, and they differ materially from the small bone flutes of the Pueblo region which have a large orifice near one end like the modern flute. Similar instruments are still found among the Thompson Indians of British Columbia, where they are used by pubescent girls in their ceremonies. The girl’s mouth must not touch the surface of the water, so she drinks through a bone tube. These are generally plain, but some of them are furnished with holes along one side so as to be used also as a whistle, from which a number of notes are produced in imitation of various birds. Sometimes birds are called with them. There is in the Museum, an instrument of this type collected by Mr. James Teit which is almost a duplicate of some of the Madisonville specimens. A considerable number of fragments of ribs of deer and elk were found with shallow transverse grooves, about one-fourth to one- half an inch apart. Two of these are shown in b and c, platel5, and an unbroken specimen is illustrated in a. In addition to the transverse grooves this example is notched along the convex edge, these notches apparently having no relation to the grooves. A few small ribs were also found without the transverse grooves, Peabody Museum Papers Vol. VIII, Plate 13 « Madisonville Site Blades made of elk antler, about § natural size, probably hafted as shown in d, e. / I * < * Peabody Museum Papers Vol. VIII, Plate 14 Madisonville Site Blades made from elk antler. (About J.) NEAR MADISONVILLE, OHIO 63 but having notches close together along the convex edge. These may have been tally sticks or possibly tools for smoothing pottery, and probably bear no relation to the above specimens. A number of fragments of the former type show considerable wear along the center of the grooved side, and in some instances the grooves at this point have been worn partially or wholly away, as illustrated in b, by the continued rubbing of the scapula or other accessory used in producing the sound. There can be no doubt that these specimens are analogous to the well-known notched sticks used for marking time in the dances of the Pueblo Indians. When in use these sticks usually rest upon a resonator made of a hollow gourd and are scraped along the notches with the scapula of the deer, thus producing a sharp rattling sound. Similar instruments made of the notched cannon bones of the deer and used with scapula scrapers are also found among the Huichol Indians of Mexico; 1 and notched human femora and tibiae, which undoubtedly served a like purpose, were taken from graves by Dr. Lumholtz at Zacapa, Mexico. 2 In the West Indies notched gourds were used for this purpose, and notched sticks, with a plain stick for a scraper, are reported from the Utes and from the Iroquois of New York State. 3 Personal Ornaments of Shell. A considerable number of per¬ sonal ornaments made of shell were obtained from graves. The species most commonly used was probably Fulgur perversa from the coast of Florida, although a number of specimens made ap¬ parently from Fulgur carica were found. The common unio shells of the inland streams, used so exten¬ sively for spoons and implements, do not seem to have been very highly valued for the production of ornaments, although several interesting specimens made of this pearly material were recov¬ ered. On plate 16, a, c, are shown two gorgets cut from these fresh water shells, one of which has some unintelligible marking in incised lines. A few pendants, were also obtained; two of these are represented in j. A number of crudely fashioned discs (s) ap¬ proximately one-half to one inch in diameter, which may have been used as dice, were recovered. One might be disposed to con¬ sider these unfinished discoidal beads, were it not for the fact that 1 Cail Lumholtz, Unknown Mexico, vol. ii, p. 155. 2 Ibid.., p. 429. 3 Frances Morris, Catalogue of the Crosby Brown Collection of Musical Instruments, vol. ii, p. 184. 64 INDIAN VILLAGE SITE AND CEMETERY only a very few finished beads of this type were found during the explorations. A large one, from a grave, is shown in t. The other objects illustrated upon this plate are made from fulgur shells, some of which were undoubtedly obtained by traffic with more southern tribes. The gorget, b, cut from the wall of a large fulgur, was found near the lower jaw of a skeleton. The mask-like gorgets, d-h, with one exception, accompanied skele¬ tons; g was found near the lower jaw of an adult; and d and e are from graves of children. The latter is apparently cut from one of the large circular gorgets having elaborate designs, which are not uncommon in Tennessee and Kentucky. That portion of the original design which remains upon the reverse of the ornament is shown in the drawing. These specimens, all of which represent the conventionalized human face, recall at once the larger gorgets of the same general type from the graves and mounds of Tennessee and Virginia, illustrated and described by Holmes in his Art in Shell. That the same deity or personage is represented by the specimens from Madisonville is evident from the design surround¬ ing the eye-like perforations in g, which apparently embodies a like symbolism to that appearing upon the specimens figured by Holmes. One of a pair of mushroom-shaped ornaments, probably ear plugs, is figured in i. They were found in a grave by Mr. Swanton in 1897. These were cut from large fulgur shells, and are very carefully made, each being a close duplicate of its companion. It is probable that they were obtained by trade from the more south¬ ern Indians, as similar specimens have been taken from the mounds of Arkansas. They are also reported from Georgia. The Arkansas specimens, however, are smaller and relatively thicker and do not have the perforation at the lower end. The specimen illustrated in k, resembles the upper portion of the ornament above described. In place of the long projection with perforated end, however, there is a short and much thinner perforated projection standing out at an angle from near the edge of the disc. This specimen was found near the bones of the left hand of a skeleton. A similar but much smaller ornament, taken from a cache-pit, is shown in the drawing below. In m, n, of the above plate, are figured two of the five pendants made of small fulgur shells. These were found with skeletons. Peabody Musetjm Papers Vol. VIII, Plate 15 Madisonville Site cl—c, Musical instruments made of deer ribs; d, Flute-like instruments; e, Beaver tooth chisels;/* Antlei handle; g, Implement of bone; h, i, Scrapers of unio shell; j, k, Combined spoons and cutting implements. (About |.) NEAR MADISONVILLE, OHIO 65 The shells are unworked, save for a groove or perforation near the smaller end for attaching the cord. Three of these are made from the shells of Fulgur perversa , the others from Fulgur carcia. Only two small “pins” wrought from the columellae of fulgur shells were obtained during the explorations by the Museum. One is represented in r. Both were found with skeletons. Although now somewhat disintegrated, they originally bore a general resem¬ blance to the well-known shell pins from the graves of the Tennes¬ see region, but the heads are much flatter. Considering the number of burials, very few shell beads were recovered. Nearly all were made from the columellae of fulgur shells. As a rule, only a few small beads of the type shown in p, occurred with a skeleton. These were usually found at the wrist or neck. The large massive beads, q, made from the columellae of Fulgur perversa, occurred singly or in groups of two or three. Only one necklace worthy of the name was obtained during the explora¬ tions. This was found with a skeleton by Mr. Swanton and con¬ sisted of fifty-five beads of nearly uniform size, made from the columellae of Fulgur carica, a species common as far north as Cape Cod. Seven of these are illustrated in o. With this skeleton were also found several copper beads and pendants, the copper ornamented belt and the iron bead illustrated on plate 18, g and s, and also two very large shell beads. There were a few pieces of bone or shell without perforations which may have been used for inlaying in wood, an art in which some of the Algonquian tribes were proficient. One of these pieces cut from unio shell is illustrated in 1. A very few of the small marginella shells, with a portion of the shoulder ground away to make a perforation for the passage of sinew thread, were taken from graves. These were probably used in forming designs upon cloth or buckskin, to which they were sewed. Probably the finest example of this work extant is the “ Powhatan ” mantle in the Ashmolean Museum. This is figured on plate 15 of the tenth volume of the Reports of the Bureau of American Ethnology. Implements and Utensils of Shell. Unio shells were found in large numbers in the cache-pits and general refuse of the site. The unworked valves were often used as spoons and ladles. 66 INDIAN VILLAGE SITE AND CEMETERY Many of the pottery vessels contained mush or other food when deposited in the grave. They were frequently accompanied by one or two valves of this shell, placed within the vessel to serve as spoons. In most of the village sites in southern Ohio which belong to the Madisonville culture, a few spoons cut from the valves of this shell are found, having a well-wrought- handle upon one side. Such spoons are common in the graves of Tennessee and Kentucky. They are very rare indeed at this site, only two broken examples being in the Museum collections, one of which is illustrated on plate 15, k. In both of these examples, the handle is furnished with a serrated edge admirably adapted for the cutting of meat and other solid food, which generally makes up a portion of the typical Indian stew. The spoon illustrated has a perforation to receive the suspending cord. A third well-made spoon from this site is preserved in the cabinet of the Literary and Scientific Society of Madisonville and is figured in Holmes’s Art in Shell in the Report of the Bureau of Ethnology for 1880-1881. This has the serrated cutting edge upon one side, and a short rounded handle which is also perforated for suspension. An interesting specimen, probably a combined spoon and food cutter, is shown in plate 15, j. The shell is unmodified except near the edge, which is coarsely serrated — probably also for the pur¬ pose of rendering it more effective in cutting. Such specimens are rare at Madisonville, but are found more commonly on other Ohio sites of this culture. Several hundred implements of the type illustrated on plate 15, h, i, were taken from the cache-pits and general refuse. They con¬ sist of a single valve of unio shell with a perforation about half an inch in diameter near the center. The posterior point of the valve, in some instances, shows little wear, but in the majority of cases is worn or broken, probably in cutting or hacking. In some of the better preserved specimens this portion shows careful grind¬ ing to a well-finished blunt scraping edge, and these were apparently not used for cutting or hacking but for scraping. They may have been employed in skin-dressing. Practically all of these imple¬ ments have the opposite end near the hinge blunted by breaking away the edge of the shell at this point. This was probably done to facilitate hafting. The great majority of the specimens show Peabody Museum Papers Vol. VIII, Plate 16 Madisonville Site Personal ornaments of shell. (About f.) NEAR MADISONVILLE, OHIO 67 no wear near the central perforation. When signs of wear appear it is usually on the convex side of the shell and towards the hinge. It is an interesting fact that some of the antler blades of the types illustrated on plate 14 have a bluntly ground scraping edge similar to that of the better preserved shell implements, and it is very probable the perforation in both the antler and shell blades served a like purpose, which was undoubtedly to assist in securing the blade to the haft. Professor Holmes has discussed the probable manner of hafting these unio blades in his Art in Shell above referred to. Combs and Personal Ornaments of Bone and Antler. On plate 17, a, b, are shown a pair of armlets made from deer ribs. They were found one upon either arm of the skeleton figured at the right in c, plate 4. One is neatly ornamented with incised lines and dots; the other is without ornamentation. A third specimen, illustrated in c, is from a cache-pit. Fragments of several others were found, most of them showing incised markings of a nature similar to the above. These armlets are very neatly made, and each end is perforated for receiving the thong or cord which bound them together and held them in place. DuPratz tells us that the young Louisiana Indian men wore bracelets made of deer ribs softened in boiling water, then bent into the required shape, and finely polished so that they resembled ivory. These ornaments must have been more common than appears from the scant evi¬ dence furnished by archaeological investigations. The use of boil¬ ing water to soften bone and antler for the purpose of rendering them more easily worked with primitive tools, appears to have been common among Indians in general. The pair of neatly made objects of antler, illustrated in d, were found near the jaw of a skeleton. They appear to have been personal ornaments of som a kind, perhaps ear plugs. They are not perforated, the ends being carefully finished by grinding. The central cellular portion of the antler at these points is now somewhat disintegrated, especi¬ ally in the broken one shown in the upper drawing. A few similar specimens were obtained from the general refuse of the site. Several pendants, made from deformed tips of deer antler, which had been shaped into crudely-made effigies of the heads of birds or quadrupeds, were recovered. Three of these are shown at e, plate 17, the lower specimen being found with a skeleton. The upper 68 INDIAN VILLAGE SITE AND CEMETERY one of the three illustrated shows a natural resemblance to the head of a bird. This portion of the pendant is unworked. Several canine teeth of the bear, wolf and dog, perforated for suspension, were secured. Two of these are illustrated in j. A few perforated elk teeth were found, but these were not abundant, and do not seem to have been in general use. Beads of various sizes, consisting of cylinders cut from the large bones of various birds, were found in graves and in the general refuse. A number of these are illustrated in f, g. Most of them were without ornamentation, but a few were decorated with de¬ signs in incised lines. The largest figured in group g, is one of ten found with a skeleton. They extended in a row “ from the mouth to the arms and down them.” Seven of the beads bore the design shown in the illustration. This design appears also upon a number of other objects from the site, as will be seen by referring to plates 16, 19, 20. It is evidently in part a lightning symbol, and is found most commonly surrounding the eyes of the human face in a certain class of shell gorgets, one of which, a small one from this burial place, is illustrated in g, plate 16. Large shell gorgets bearing this general design occur most frequently in Ten¬ nessee and Virginia. 1 The same figure occurs surrounding the eyes of birds, serpents and human beings in numerous drawings and sculptures from various sections of the Mississippi Valley. A fragment of the bone base of a holder for an eagle feather is illustrated on plate 17, h. The broken lines indicate its original form, which may be duplicated by recent examples in almost any large ethnological collection from the Indians of the Interior Basin. One of the cylinders which holds the feather and fits over the large hole in the narrow end of the base of these ornaments, is figured in i. There is a hole upon either side near the lower end for the thong, which passes also through the perforations upon either side of the large hole in the base to which it is tied. This thong also serves to fasten the ornament to the lock of hair on the crown of the warrior. The base, supporting an upright bone cylinder and feather, and worn nearly in a horizontal position, with the broad end projecting backward, is placed within an ornament of upright deer hair dyed 1 See W. H. Holmes, Art in Shell, Rept. Bureau of American Ethnology, vol. ii, plates 67-69. Peabody Museum Papers Vol. VIII, Plate 17 Madisonville Site a-c, Antler arm-bands; d, Cylinders of antler, perhaps parts of ear-ornaments; e, Pendants in form of bird heads made from diseased tips of deer antler; f, g, Bone beads; h, i, Bone base and cylinder for supporting eagle feather in head-dress; j, Canine teeth pendants; k-o, Antler combs. (About \.) NEAR MADISONVILLE, OHIO 69 red and fringed with the black “'beard hair ” of the turkey or long stiff hair of the porcupine. Taken altogether it forms a very hand¬ some and striking head ornament, and was worn by warriors of various tribes over a considerable area. The fragment of the base was found in a cache-pit. Only two or three of the cylinders were recovered during explorations by the Museum. Five antler combs are shown on plate 17, k-o. These are all from cache-pits, with the exception of k, which lay near one of the skeletons illustrated in c, plate 4. Upon the upper portion of this comb is engraved a rattlesnake. Each of the combs is perforated near its upper end for the suspending cord. In two of the specimens, k and n, horizontal striae are noticeable upon the teeth, which are evidently caused by long continued use of the implement, for dress¬ ing the hair or for a similar purpose. Combs of like form, but with more pointed teeth, are often used by the Eskimo for combing and renovating the hair of their skin garments. Personal Ornaments of Copper. A considerable number of metal objects were found with skeletons, and in the cache-pits and general refuse. Most of these were in the form of cylindrical cop¬ per beads of various sizes, made by rolling small pieces of sheet metal into tubular form. The majority of these are probably wrought from native copper, but some are undoubtedly made of European copper. It would be difficult, if not impossible, to de¬ termine even approximately the number belonging to each group without a chemical analysis of the metal. Most of the smaller beads of the longer type, however, appear to be of native metal, for they are, as a rule, more crudely fashioned than the larger specimens. No single individual appears to have been the pos¬ sessor of more than a few copper beads. In the graves they were usually found singly or in groups of only a few, sometimes in com¬ pany with beads of shell. Plate 18, a, shows the various forms. The six small beads in a row of nearly uniform size, were taken from near the wrist of a skeleton, and probably formed part of a bracelet. Those in the lower row were found with the skeleton of a child, and still retain their position on the original string. A few bell-shaped tinklers are illustrated in f. They were ob¬ tained from both cache-pits and graves. In recent times they are made of tin, and are used extensively by the Indians as ornamental 70 INDIAN VILLAGE SITE AND CEMETERY pendants. The smaller end is clasped to a thong, and a bunch of hair, usually dyed a brilliant color, often projects from the larger end. The band bent into a circle and figured in b, was probably for the finger, for four similar rings made of brass and still clasping the phalanges of the fingers of a skeleton were taken from a grave by Professor Putnam. One of these is shown in i. A few rings made of native copper beaten into a small rod and bent into shape (c) were recovered from the cache-pits, as were also the serpent or lightning symbols illustrated in d. The largest amount of copper found with a burial was taken from Trench I, grave 60, by Mr. Swanton. There were several pieces of sheet copper with one or more of their edges clinched over pieces of buckskin. Two of the larger are figured in g. They lay near the thighs, and seem to have ornamented a girdle. With this burial were also found seven copper beads and pendants, the iron bead illustrated in s, and fifty-five shell beads. Taking into con¬ sideration the iron bead, which appears to have been made from a piece of sheet iron, it is not improbable that the copper also may be of European origin, as there seems to be no evidence that the proto-historic Indians of southern Ohio used meteoric iron in fashioning their ornaments, as did the builders of the great earth¬ works of this region. Small copper ornaments with two of their edges bent over, apparently for securing to thongs, are figured in h. These are from graves. Only a few examples of overlaying with thin copper sheets were recovered. The first of these were two small pear-shaped objects about three-eighths of an inch in diameter, made of wood and cov¬ ered with thin copper, which is now badly corroded. The salts of copper have preserved one of the wooden forms perfectly. These were found with the skeleton of a child, in company with several beads and a few coils of copper wire, evidently of native make. The other examples of overlaying were thin discs of bone, one- half to three-fourths of an inch in diameter, covered on one side with thin copper, the edges of which were turned over and clinched upon the opposite side (e). Vegetal fiber was placed over the discs before the copper sheet was applied. One of these discs was found beneath the lower jaw of a skeleton. The other three came from a cache-pit. In the Journal of the Peabody Museum Papers Vol. VIII, Plate 18 Madisonville Site a-h, l, m, Personal ornaments and other objects of copper; i-k, n. Personal ornaments of brass; o-r, Iron blades; s, Iron bead; t. Iron sword guard from grave. (About §.) NEAR MADISONVILLE, OHIO 71 Cincinnati Society of Natural History, July, 1880, p. 131, is figured a crude bell or rattle from this site. It was made of a single piece of copper of irregular shape, the edges of which have been brought together so as to form a ball, or rather like a sleigh bell, having an irregular opening on one side. A small hole was punched through the top and a strip of copper doubled up and the ends pushed through the opening from the in¬ side, forming a handle. Inside this bell is a fragment of copper about the size of a large pea and when the ornament is shaken it produces a rattling or tinkling sound. Perhaps the most interesting objects of this metal are the two double crosses illustrated in 1, m. The first of these was found with a skeleton by Dr. Metz. The account of the discovery follows: 1 On Thursday, October 28, Skeleton No. 8 was found, an adult male, with head southeast, length 5 feet 8 inches, depth 15 inches. A broken vessel was found at the right of the head, and on the left side a pipe made of limestone, well finished, and carved to represent the head of some animal. A copper ornament was also found at the right of the neck. This relic, which has two bars or cross arms, is made of a very thin piece of copper, rolled or beaten evenly, with a small perforation at one end, doubtless for suspensory purposes. The other double cross was taken from a skeleton by Mr. B. W. Merwin while conducting work for the Museum in 1911. It lay just to the left of the skull. With the skeleton were also a stone pipe, a flint point on ribs of the right side, and an antler harpoon point near left humerus, (plate 11,1). A flint arrowpoint was em¬ bedded in the frontal bone of the skull. These copper- specimens at once recall to mind the double crosses of silver obtained by the Indians from the Catholic missionaries. The copper crosses, how¬ ever, are probably of Indian make. The first example has the appearance of having been roughly cut from European sheet cop¬ per, then the surface and edges carefully finished by grinding. The second specimen has more the appearance of being made of native copper. Only a chemical analysis, however, can settle these points. Personal Ornaments of Brass. A cross cut from sheet brass, and of the same general form as those above described, but having one cross-bar instead of two, is illustrated in n, plate 18. This was found with the skeleton of a child by Professor Putnam in 1882, 1 Journal of the Cincinnati Society of Natural History , July, 1880, p.-133. 72 INDIAN VILLAGE SITE AND CEMETERY together with a few copper beads and spirals of brass wire, and a pottery vessel of the usual type having four ears or handles. One of the beads is shown at the left in the upper row in a, and two of the spirals are illustrated in k. This is one of the most instructive burials found during the explorations, as it definitely connects this form of cross with the proto-historic period, and also proves that the pottery vessels of the type so characteristic of this ceme¬ tery were in use at this time. An interesting copper cross of somewhat similar design, from a stone grave near Nashville, Tennessee, is in the Museum collec¬ tion. This is illustrated on page 98 of the third volume of the Reports of the Peabody Museum. No other object from this group of Tennessee graves shows indication of contact with Europeans, and perhaps this specimen bears no relation to the Christian cross. A few beads made of European sheet brass and of the same type as those of copper were obtained from different cache-pits. Two of these are shown in j. In 1882, Professor Putnam opened a grave in which were the skeletons of a woman and child. Accompanying them were twenty- two copper beads, varying from one to two and one-quarter inches in length, and about one-quarter of an inch in diameter. On the first phalanges of the first and second fingers of each hand of the woman, were rings made of bands of sheet brass, about three-fourths of an inch wide and two and one-half inches long, each bent into a circle with the ends of the strip overlapping. One of these is illustrated in i. A number of fragments of both sheet brass and sheet copper were found in the general refuse of the site. Iron Objects. On plate 18, p, is shown what is apparently a badly corroded iron adze blade. It was found four feet beneath the surface in a cache-pit. A second blade is illustrated in r, from three feet one inch below the surface in another pit. Both of these were recovered by Mr. Swanton. Two more of the same general form are figured in o and q. These blades wei’e obtained by the Indians either directly or indirectly from white traders. In the leaf mould, and in one instance twenty inches beneath the surface of a cache-pit, fragments of iron objects were found which were probably lost by the early white settlers of the region. The only iron objects obtained from graves are the bead illus¬ trated in s, already referred to on page 70, and the sword guard Peabody Museum Papers Vol. VIII, Plate 19 Madisonville Site Tobacco pipes. (About §.) NEAR MADISONVILLE, OHIO 73 shown in t. The latter was found over the chest of the skeleton of an adult female. There were also three bone beads at the neck and six copper beads near the hands of the skeleton. I have been unable to ascertain with any degree of certainty the probable origin of the European sword guard or the approximate date of its manu¬ facture. Glass Beads. A few globular blue glass beads, about three- sixteenths of an inch in diameter, were found near the skull of a skeleton of a child, by Mr. Swanton. These were of the same kind as those obtained during the exploration by the Museum of the proto-historic Iroquoian cemetery in Erie County, New York. Mr. Swanton also found fragments of blue beads in three cache- pits, and Mr. B. W. Merwin reports the finding of one such bead in a fourth pit. The recovery of iron objects or brass or glass beads, from at least nine cache-pits and five graves, seems to indi¬ cate beyond question that this site was not abandoned by the. Indians till after their contact with Europeans either directly or through their Indian neighbors; but the scarcity of such objects, as well as their nature, also indicates that the site was deserted long before the arrival of English settlers. Tobacco Pipes. A large number of tobacco pipes were found, the majority being obtained from cache-pits or by general digging. A few occurred with skeletons. Most of them were made from the limestone of the region. Some were of sandstone, while a small number were of other varieties, including serpentine and red pipe- stone. Two or three rude pipe bowls were found made from sections of deer antler, also fragments of four or five pottery pipes, fash¬ ioned of fine clay of a kind very different from that used by the inhabitants in making pottery. Judging by the fragments, these clay pipes resembled those of the Iroquoian and Algonquian tribes of the East, and were probably imported from that section. A considerable number of stone pipes in the process of making were recovered, which indicates that most of the finished specimens secured were made by the inhabitants and were not obtained from the neighboring tribes. The majority of the specimens were of the simple designs and forms figured in plate 19, a-i. They were usually without orna¬ mentation. A few were ornamented with notches or with designs in incised lines, as shown in a, b, e, and f. A very good figure of 74 INDIAN VILLAGE SITE AND CEMETERY the thunderbird is engraved upon both sides of b; while the peculiar eye designs, also occurring on certain shell gorgets, bone beads, and pebbles (plates 16, g, 17, g, 20, a, b), are engraved upon either side of e and f. A small animal figure, probably the totem mark of the owner, is shown on the narrow side of e. Fig. 5. Madisonville Site Tobacco pipe of clay-stone, representing an owl. Found with a skeleton in Trench I in 1911. (About $.) Lightning symbols appear upon either side of the specimens figured in i and 1. The pipe shown in j, has two horn-like append¬ ages projecting backward from either side of the opening forming the bowl. The pipe figured in k is made of nearly black serpentine and is highly polished. It is evidently intended to represent the turkey, for the “ beard,” peculiar to this bird, appears upon the breast. It was taken from a cache-pit by Dr. Metz in 1881. A fish-shaped pipe of limestone, also found in a cache-pit, is fig¬ ured in m. It differs from most suecimens by having the hole for the reception of the stem at the side. Two interesting effigy pipes are illustrated in o, p. Apparently swimming birds are represented, both of them being in the position usually taken while oiling the feathers. The first of these is shown with two heads. This was foimd beside the skull of a skeleton in Trench IV, by Mr. Swanton, as will be seen by referring to plate 3, a. The other is from a cache-pit. Peabody Museum Papers Yol. VIII, Plate 20 Mabisonville Site a-c, Pebbles with incised drawings upon both sides; d, e, Pebbles with incised drawings; /, h, i. Pebbles but slightly modified by art, these are probably fetishes; a, k, Pendants of lignite, the first of these represents a buffalo; j, Effigy of a turkey cut from antler. (About §.) NEAR MADISONVILLE, OHIO 75 The specimen figured in q, of limestone, was taken from near the head of a skeleton by Professor Putnam in 1882. The largest and finest pipe obtained during the explorations is illustrated in figure 5. It represents an owl, and was found with skeleton 9 in Trench I, by B. W. Merwin in 1911. It measures about five and three-fourths inches in length, including tail, which has been broken off. The material is a variety of fire-clay or clay- stone of mottled gray color. It is well modeled and carefully fin¬ ished. The hole for the stem is in the center of the back, while the cavity for the tobacco is in the neck. Incised Drawings, etc. If the art of the occupants of the Mad- isonville site is indicated at all clearly by the few crude drawings and sculptures which were recovered, it seems to have reached about the same stage of development as that of the surrounding Algonquian tribes at the time of their first intercourse with Euro¬ peans. It was far inferior to the art of the builders of the great earthworks, who preceded them in southern Ohio. It is highly probable that the painted designs and quill work patterns of this people on dressed skin and other perishable objects were of a higher grade than one would be led to believe, judging from the crude efforts illustrated on plate 20. A thin waterworn pebble of lignite, from a cache-pit, with de¬ signs upon either side is shown in a. The edges are ornamented by a series of short lines, and upon one side is incised the picture of a horned serpent, a small quadruped, and what appears to be a mythical monster with open mouth and prominent teeth. The characteristic triple eye design occurring on certain shell gorgets and pipes from this site is also shown in the figure, which makes it probable that the creature represented belonged to the mythology of the people. A representation of this same being is also etched upon the opposite side of the pebble. The same figure, more crudely drawn, is scratched upon either side of the flat clay stone pebble shown in b. Subsequently, prob¬ ably on account of the resemblance of the pebble to the outline of a bird’s head, two large circles were cut on opposite sides to rep¬ resent the eyes of the bird, and in front of each a shallow hole was drilled to indicate the nostrils. It is a fact well known to ethnologists that natural objects re¬ sembling human or animal forms, or any of their parts, were sup- 76 INDIAN VILLAGE SITE AND CEMETERY posed to possess supernatural powers, and were often used as fetishes. A number of such objects were recovered. In most instances their forms were but slightly modified by the Indians. Usually only a mouth or eyes were added to accentuate the realistic appearance. In addition to the one above referred to, (b), good examples are illustrated on the plate. In the head shown in h, the pebble is unmodified with the exception of a hole for the eye upon either side, the groove for the mouth, a small notch near the end opposite the head, and the beginning of a perforation for suspension. The form of the clay concretion (f), is also wholly natural, with the exception of the groove for the mouth which has been cut upon both sides. In the fragment of another clay object, which somewhat resembles the human profile (i), the grooves which may represent the mouth, the scratches near the nose and surrounding the eye, and the etched band on the forehead, are artificial. The rest of the specimen, including the perforation, is natural. Other fetishes of this class, cut from diseased antler tips, are figured on plate 17, e. The small object shown in g, plate 15, is made of bone and is carefully polished. While it resembles some of the fetishes above referred to, it is probably an implement, consisting of a handle terminating in what was evidently intended to represent the head of a bird. The eyes and the opening of the beak are indicated on either side by a dot and an incised line. A beautifully finished tool, similar in form but much larger, which was obtained from a Mandan site in North Dakota, is illustrated in the third volume of the Peabody Museum Papers, page 172. One end was in the form of a “ crane’s head.” This portion was serrated along its upper and lower edges, and the implement quite closely resembles a modeler’s tool. The lignite pendant illustrated in g, plate 20, is probably in¬ tended to represent a buffalo. Another, of the same material, showing a simple arrangement of incised lines is figured in k. These two were found by Mr. Swanton during general excavations. A figure cut from antler and resembling a turkey cock with its tail spread, appears in j. A part of a nearly black pebble, illus¬ trated in c, has an etched design upon either side. The drawing at the right seems to represent a quadruped, with the head, tail, Peabody Museum Papers Vol. VIII, Plate 21 Madisonville Site a-d. Charred fragments of bags made of bast; e, Charred fragment of wooden bowl; f-h, Grooved stone finishers; i-k, Stone tablets, use unknown; l, m, Discoidal stones. (About J.) NEAR MADISONVILLE, OHIO 77 and the greater part of the body appearing on the fragment. The lower portion is missing. Upon the opposite side is etched what appears to be a bird. A number of small stones, bearing designs which are to us unin¬ telligible, were found in the general refuse; one is shown in d. The scene represented on the small pebble illustrated in e, shows the upper portion of three human figures with arms extended and wave-like lines below. These lines may indicate water. The objects illustrated upon this plate are from cache-pits or were found while excavating. None of them are from graves. Textile Fabrics and Wooden Utensils. Almost nothing remains to indicate the types of basketry, bags and other fabrics, or the various forms of wooden objects used by the inhabitants of this site. Pottery, stone and bone implements formed, of course, but a very small portion of the artifacts of the people. The few charred remains of the more perishable objects, how¬ ever, indicate that their bags and wooden utensils, and probably their basketry also, were similar to those of the surrounding country. On plate 21, d, is illustrated a charred fragment of the border of a bag made of coarse bast, which was taken from a cache-pit con¬ taining about three bushels of carbonized corn. Another fragment of probably the same bag, is shown in a. The latter drawing is from a sketch made by Dr. Metz, at the time the fragment was un¬ covered. In removing the “ matting,” however, it crumbled to pieces. Dr. Metz’s sketch, together with the fragment of the border (d), show that the bag was technically identical with others of coarse bast, which were, until very recently, common among the Ojibwa and the people of the region south of the Great Lakes. In the collections of the Museum is a bag of the same weave and probably of the same material, obtained from the Sauk and Fox. It is strongly made of coarse bast and is well adapted for holding and transporting shelled corn. Fragments of the bag found in the pit by Dr. Metz were mixed with the charred corn, and at least a portion of this corn had doubtless been deposited in the bag. A lower corner of another bag of bast is illustrated in c. This was found with charred corn in a pit by Mr. Swanton. In the Ojibwa exhibit in the Peabody Museum is a bag of the same 78 INDIAN VILLAGE SITE AND CEMETERY technic and material, which has the border at the opening also finished as in d. The fragments of braided cordage shown in b were found with charred material in another pit. They are probably the remains of cords for closing the mouths of bags similar to those above described. In common with most of the tribes of the eastern section of the United States, this people undoubtedly were well supplied with wooden bowls and platters. Only one small carbonized fragment, however, was brought to fight during the Museum’s explorations. This was taken from a cache-pit near the southern edge of Trench K by B. W. Merwin, and is figured in e. The broken lines show the approximate form of the bowl. Wooden bowls and platters, especially among the Algonquian tribes, were of excellent design and workmanship. They were usually made of the knotty or burly portions of maple or other hard wood, and the rim was often ornamented with a representa¬ tion of the head of a bird or animal after the manner shown in the illustration. 1 Pottery. The pottery recovered from this site shows a consider¬ able variety in form, as will be seen by referring to the plates. Most of the vessels are of the cooking-pot type, although a few food bowls were found, and also several erratic forms which do not seem to belong to either of these groups. The ware is as a rule well-made and fairly durable. The material is the clay of the region tempered principally with ground or pulverized unio shells. Approximately half of the vessels show the impression of cord- wrapped paddles used in their making. The texture of the ware is well brought out in the photographs, plate 22. Nearly all of the pots are furnished with ears, usually four but sometimes two only, which were used for holding in place bands or thongs to which sus¬ pending cords were attached. It does not seem to have been cus¬ tomary to fasten suspending cords directly to the ears, although this may have been done occasionally. To facilitate the carrying of the smaller pots of this type, the Mandan sometimes tied a thong around the neck of the vessel, passing it through the ears, and a single looped handle a few inches long was made by again 1 For examples of other bowls, see Willoughby, Wooden Bowls of the Algonquian Indians , American Anthropologist, vol. x, p. 423, 1908. Peabody Museum Papers Vol. VIII, Plate 22 Madisonville|Site Typical pottery vessels from graves. (About p) NEAR MADISONVILLE, OHIO 79 tying the ends of the thong together. Most of the smaller un¬ broken pots, holding approximately a pint to a gallon, such as are illustrated on plate 23, with the exception of n, and s, are from graves. They were more commonly placed near the head, in the manner illustrated in plates 3 and 4. See also table on page 16. These pots evidently contained food, in some cases probably samp or maize porridge or a similar substance, shown by a deposit on the inner side of many of the vessels, which marked the original level of the semi-liquid contents. As the water slowly evaporated, the bran or other material near the surface was deposited on the inner side of the pot in a narrow horizontal band or line more clearly marked at the top, which, although not very conspicuous, can be easily traced. Many of the vessels contained a single valve of a unio shell which served as a spoon, and sometimes bones of food animals were also noted. The majority of the vessels found with skeletons were of the style shown in plate 23, h, i. They varied considerably in size and contour, but were usually furnished with four ears; and were, as a rule, without ornamentation with the exception of slight inden¬ tations around the rim and the usual cord-wrapped paddle marks on the body of the pot as shown in plate 22. Few vessels without ears were taken from graves; two are illustrated in f, g, plate 23. Cups with a single handle or ear (a) are rare. Quite a number of pots having two plain ears and two in the form of small quadrupeds with the usual passage between the body of the animal and the neck of the vessel, as shown in r, t, accompanied skeletons. In a few instances a pair of lugs curving downward replace the animals on either side (q), and in one example four pairs of lugs were sub¬ stituted for the usual looped ears as shown in e. These of course were intended to assist in holding the neck band in place. Bowls in the form of birds or quadrupeds were uncommon. Besides the one illustrated in d, fragments of two or three similar ones were found in the refuse, also several heads broken from bowls of the same general kind, including one which may have been intended for that of a dog. The vessel with the human face modeled upon one side, shown in j, plate 23, is the only example of this type obtained. This was found near the head of a skeleton. But one vessel modeled to represent a pot set within another (plate 24, j) was obtained during the exploration by the Museum. One or two 80 INDIAN VILLAGE SITE AND CEMETERY others, however, were previously taken from this cemetery. On plate 24, k, and 1, are figured two curious vessels with solid bases, one of which has been broken off. These are now in the collection of the Cincinnati Museum. The unbroken specimen was found with a skeleton by Dr. Metz many years ago. The base of a third vessel of this type was obtained during explorations of the Museum. Food bowls of forms illustrated in m, n, and o, were not plentiful. Fragments were found in the cache-pits and during general exca¬ vations. Those illustrated are restored from large pieces. The only example of a vessel with painted decoration known to the writer from this site is in the Cincinnati Museum. It is illustrated in b. The design is in black on an unpainted ground and represents the primitive cosmic symbol, the cross within a circle, and a dot for each quarter. It is highly probable that this was a ceremonial bowl. Fragments of cooking pots of the larger sizes were found throughout the cache-pits and in the general refuse. One of the largest, measuring nineteen inches in height and about the same in diameter, restored from fragments, is shown on plate 23, s. These large pots were quite common, judging from the numerous sherds. They were not found with burials, as they were evidently not in¬ tended for individual food containers, but were used in preparing food for families or a considerable number of individuals. Many of these larger vessels were ornamented with incised decoration ap¬ plied principally to the neck. The prevailing designs are variations of the guilloche as indicated in plate 24, e, f, i, but many of the patterns consist of a herring-bone arrangement of straight lines or lines and dots as in c, d, g. In some instances the ears were quite elaborately decorated with straight or curved lines. A few miniature cups were found and a small number of little clay effigies of animals all crudely executed, probably the work of children or made by their elders for toys. The large number of fragments of vessels of all sizes found in the cache-pits, refuse piles, and during general excavations, indicate the abundance of vessels in use, as well as the long occupation of the ground. Perhaps no group of artifacts from this site is more distinctive than the pottery. A comparison with that from other village sites in the Little Miami Valley, however, notably those near Newtown, and certain others at or near Fort Ancient, as well as from such widely separated localities as the Feurt site near Peabody Museum Papers Vol. VIII, Plate 23 Madisonville Site With the exception of the two largest pots (n and s) which are restored from fragments found in cache-pits, the specimens here shown are typical vessels from graves. ( 3 * 5 .) / Peabody Museum Papers Vol. VIII, Plate 24 a, j, k, l, Pottery vessels from graves; b, Small bowl with primitive cosmic symbol in black; c-i, Pots with incised decorations, restored from large fragments from cache-pits; m, n, o. Bowls restored from large fragments; b, k, l, Cincinnati Museum. OyV) NEAR MADISONVILLE, OHIO 81 Portsmouth and the Gartner site near Chillicothe in the Scioto Valley, show that it all belongs to one group. This is true also of other forms of artifacts in general. Madisonville, however, seems to have been the most important center yet known of this culture, and the pottery in general had reached a somewhat higher degree of technical excellence here than in the other localities noted. THE SKELETAL REMAINS The Collection in General. In the various collections of human remains gathered from the cemetery near Madisonville by the ex¬ cavators of different periods, and deposited in the Peabody Museum, 86 crania were found which belonged to full-grown individuals and were in condition to be measured. Two of these were of doubtful sex, and for this reason the measurements have not been included in the tables. In the other cases sexual differentiation according to the usual criteria was not difficult. The majority of the better preserved crania are separated from their bones because of lack of proper storage facilities in the Museum, and so, in many instances, the bones and skull of the same individual were sexed independ¬ ently. Upon comparison of the separately sexed skulls and bones agreement as to the sex assigned by the observer in the two in¬ stances was found in every case except one, where there is un¬ questionably a mixture of the bones of two individuals of different sexes under the same catalogue number. The skeletal material was catalogued according to graves and in some instances the graves contained mixed burials. It was not possible to examine and make a full report upon all of the skeletal material from Madisonville, and the total number of individuals represented in the material studied is 217. Of these 111 were adult males, 60 adult females, 16 adults of undetermined sex, 22 children and adolescents, and 8 infants. Many of the skulls measured were sent in from the early excavations by Metz and are unaccompanied by their bones. Most of the bones studied are from skeletons of which the skulls are fragmentary. Apart from the crania, the bones of 112 skeletons were studied, including 90 adults, but only 17 of 86 crania measured belong with these adult skeletons of which the bones were studied. About half of the total number of crania measured, including the majority of the best preserved, are from the early excavations 83 84 INDIAN VILLAGE SITE AND CEMETERY of Dr. Metz and Professor Putnam. Of the remainder about half are from J. R. Swanton’s excavations and half from R. E. Mer- win’s. Most of the bones studied are from the excavations of Mr. Swanton in 1897. The measurements, with the exception of the cranial capacities, were taken by the writer with accurate instruments, and, unless otherwise stated, conform to the national agreements of Monaco and Geneva. The cranial capacities were measured by Mr. George Schwab, Associate in African Archaeology and Ethnology in the Peabody Museum, according to the method of Dr. Hrdlicka, and under the supervision of the writer. Mr. Schwab also rendered valuable assistance in the preparation of this paper by recording the measurements and observations. The observations conform to the system devised by Dr. Ales Hrdlicka, and in use in the Department of Physical Anthropology in the United States National Museum. For comparative material in the study of the crania an unpub¬ lished doctoral dissertation by the late R. G. Fuller, on the crania from the Tennessee Stone Graves in the Peabody Museum, has been utilized. This study of a series of 188 crania from Tennessee was prepared by Mr. Fuller under the direction of the writer and according to the methods in use in the Peabody Museum and in the National Museum. Frequent reference has been made also to an article by F. W. Langdon, M.D., entitled The Madisonville Prehistoric Cemetery Anthropological Notes, Journal of the Cincinnati Society of Nat¬ ural History, vol. iv, no. 3, October, 1881, pp. 237-257. This article includes an examination of 83 crania from the early exca¬ vations in the Madisonville cemetery and notes on the pathology of the bones. Constant use has been made of two studies in the physical anthropology of the American Indians by Dr. Ales Hrdlicka: Re¬ port on an Additional Collection of Skeletal Remains, from Arkansas and Louisiana, a reprint from the Journal of the Academy of Natural Sciences of Philadelphia, vol. xiv, 1909; and Physical Anthropology of the Lenape or Delewares, and of the Eastern Indians in General, Bulletin 62, Bureau of American Ethnology, 1916. NEAR MADISONVILLE, OHIO 85 Cranial Deformation. Occipital Deformation None Slight Medium Pronounced Total Males: number. 14 34 1 4 53 “ percent. 26.41 64.14 1.88 7.54 73.5 Females: number. 5 20 2 2 29 “ per cent. 17.25 68.96 6.89 6.89 82.7 This series includes comparatively few skulls with pronounced cranial deformation, as the table above shows. A considerable proportion (26 per cent in males and 17 per cent in females), shows no artificial deformation whatever, and in the majority of cases deformation, where present, is very slight. The total percentage of male crania exhibiting some degree of deformation is 73.5, while the corresponding figure in the case of the females is 82.7. In 115 male crania from the Tennessee Stone Graves, Fuller found 63 or 54.7 per cent with very pronounced occipital flattening, and a series of 70 female crania from the same graves included 34.2 per cent of markedly deformed crania. This does not take into consideration the slight degrees of deformation. Deformation is of the occipital type only — a fact noted by Langdon also — and is probably due to a slight cradle-board flattening. It is perhaps of interest to note in this connection that no cases were observed in this series in which the auditory meatus seemed compressed in an antero-posterior direction, nor were bony exostoses found in the meatus in any case, although they occur in 13 per cent of Fuller’s Tennessee series. It is of some importance in connection with the utilization of artificially deformed crania for purposes of craniometric study to attempt to ascertain to what extent the original form has been altered, since it is obvious that very radical changes in the head- form brought about by this cause render the study of the propor¬ tions of the cranial vault useless, unless some method of correcting for deformation is available. No exact method exists, and with a good deal of reason some anthropologists exclude from their series for measurement all crania showing pronounced artificial deforma¬ tion. In the case of American Indian crania the debarring of specimens showing artificial deformation would mean the exclusion of some of the largest and most important groups. 86 INDIAN VILLAGE SITE AND CEMETERY In the present investigation an attempt has been made to as¬ certain the differences in cranial indices between skulls showing no deformation and skulls showing medium or pronounced defor¬ mation. Cranial Vault Indices. Cranial Indices of Undeformed Crania Dolicho. Meso. Brachy. Total Min. Max. Mean Males: number. . . I 7 6 14 74.46 85.80 79.61 “ per cent.. . . . 7.14 50.0 42.85 Females: number. . . 0 2 3 5 77.09 83.04 79.78 Cranial Indices of Crania with Medium or Pronounced Deformation Dolicho. Meso. Brachy. Total Min. Max. Mean Males: number. . . .. 0 0 5 5 82.49 90.9 86.72 Females: number. . . 0 0 4 4 84.21 92.55 87.56 Cranial Indices of Crania with Slight Deformation Dolicho. Meso. Brachy. Total Min. Max. Mean Males: Number. .. .. . 0 7 26 33 76.24 92.31 83.03 “ per cent... . . 0 21.21 78.78 Females: number. .. 1 1 1 18 20 70.39 1 88.62 84.25 per cent. . . 5.0 5.0 90.0 Cranial Indices of all Crania Dolicho. Meso. . Brachy. Total Min. Max. Males: number. . . . 1 14 37 52 74.46 92.31 “ per cent. . . 1.92 26.92 71.15 Females: number. . 1 3 25 29 70.39 92.55 per cent. . 3.44 10.34 86.20 Both sexes: number. 2 17 62 81 70.39 92.55 per cent. 2.47 20.98 76.54 It is apparent from the consideration of the tables given above that the nine crania of our series exhibiting medium or pronounced artificial deformation are all brachycephalic skulls, and all ex¬ tremely brachycephalic, with the exception of two male skulls, which have indices of 82.58 and 82.49 respectively. If these last had not been artificially deformed they might have been meso- cephalic, but not dolichocephalic. The others belong without ques¬ tion of deformation in the brachycephalic class. On the other hand, of 19 undeformed crania 9 are brachycephalic, so that it ap- 1 This cranium showed a slight flattening in the lambdoid region which had the effect of increasing its length. NEAR MADISONVILLE, OHIO 87 pears that a minimum of about half of our series would be brachy- cephalic if all were undeformed. But the high percentage of brachycephaly in the largest sub-group, that of crania with slight deformation (78*78 per cent of males and 90 per cent of females), indicates that a higher original proportion of brachycephaly than 50 per cent is to be expected in the whole series. For in the sub¬ group with slight deformation there has probably been very little alteration in the indices through this cause, and many of the crania exhibit high degrees of brachycephaly. The writer is inclined to believe therefore that the percentages given for brachycephaly for the whole series are only slightly in excess of the figure to be expected if there had been no deformation. Of 17 crania yielding mesocephalic indices, 9 are undeformed and 8 show slight occipital deformation. The slight occipital de¬ formation has perhaps shifted some of these from the dolichocepha¬ lic class to the mesocephalic class. Of the 7 mesocephalic male skulls with slight occipital deformation, 4 have indices of 76, one of 77, and 2 of 78. Of the undeformed crania, 50 per cent are mesocephalic, and of the slightly deformed 21.21 per cent. When we consider the extreme rarety of dolichocephaly in the unde¬ formed sub-group (one case only), it will appear that the decrease in mesocephaly in the slightly deformed sub-group is in accordance with what we should be lead to expect if more crania were shifted from the mesocephalic to the brachycephalic class through defor¬ mation than from the dolichocephalic to the mesocephalic class. In order to check still further the effect of unintentional arti¬ ficial deformation on cranial form, the writer has made use of the series in the Peabody Museum from the Tennessee Stone Graves, measured by R. G. Fuller, to show the cranial indices in the unde¬ formed sub-group, the deformed sub-group, and in both combined. In this series occipital deformation is much more pronounced, and Fuller states that a slight degree of frontal deformation may often be observed. Cranial Indices of Undeformed Tennessee Crania Dolicho. Meso. Brachy. Total Min. Max. Males: number. 4 15 24 43 70 92 “ per cent. ,. 9.3 34.88 55.81 Females: number. 1 5 22 28 74 93 “ percent. 3.57 17.85 78.57 88 INDIAN VILLAGE SITE AND CEMETERY Cranial Indices of Deformed Tennessee Crania Dolicho. Meso. Brachy. Total Min. Max. Males: number. 0 1 62 63 79 105 “ percent. 0 1.58 98.41 Females: number. 0 0 36 36 81 106 “ percent. 0 0 100.00 Cranial Indices of all Tennessee Crania Dolicho. Meso. Brachy. Total Min. Max. Males: number 1 . 5 18 92 115 70 105 “ percent. 4.34 15.65 80.0 Females: number 1 . 1 6 63 70 74 106 “ percent. 1.42 8.56 90.0 Both sexes: number 1 . 6 24 155 185 70 106 “ percent. 3.24 12.97 83.78 If we compare the above tables with the similar tables for the sub-groups according to deformation in our Madisonville series, it will be observed first of all that the undeformed Tennessee crania include a larger percentage of brachycephals than the cor¬ responding Madisonville sub-group — an excess amounting to about 13 per cent in the males. In the entire series the percentage of brachycephals for both males and females combined is 83.78 against 76.54 in our Madisonville series. The deformed Tennessee crania include no dolichocephals and but one mesocephal, whereas in our series one dolichocephal and 8 mesocephals show slight de¬ formation. As the percentage of mesocephalic crania in the Ten¬ nessee series is smaller in the undeformed sub-group than in the corresponding Madisonville sub-group, there were probably fewer Tennessee individuals of hereditarily mesocephalic crania who could be shifted into the brachycephalic class through accidental deformation. This may partly account for the smaller excess in proportions of brachycephals in the entire Tennessee series over that in the entire Madisonville series as compared with the cor¬ responding proportions in the undeformed sub-group. But on the other hand, only 61.6 per cent of the Tennessee series show defor¬ mation as against 73.5 in our series. Of the Tennessee brachy¬ cephals, 29 per cent show no deformation, against 11 per cent of our brachycephals. This difference may be due to the somewhat more rugged character of the Tennessee crania, which might render 1 Includes several crania in which measurements were approximate. NEAR MADISONVILLE, OHIO 89 them more resistant against deformation; it may be due to the personal equation of the two different observers in recording defor¬ mations; it may be due to error incident to the sampling process, since both series are comparatively small, and our entire series in¬ cludes 81 cranial indices as against 185 in the Tennessee series. The examination of the Tennessee series confirms our conclu¬ sion that accidental occipital deformation affects dolichocephalic crania very little, mesocephalic crania slightly more, and brachy- cephalic crania most of all. Assuming artificial deformation to have been caused by pressure of the occiput on a hard cradle board, it seems clear that this cause would naturally affect round-headed infants to a greater extent than long-headed infants, for if the head of the child is free to turn from side to side, the tendency for the dolichocephalic child with the protruding occiput is to rest the head on one side or other rather than on the back. On the other hand, a brachycephalic child with an occiput more or less flat is likely to rest on the back of the head rather than on the side Again, if the head is fixed so that it must rest on the occiput the greater convexity of the long¬ headed occiput presents less surface for deformation than the relatively flat occiput of the round head. On the whole the present writer is inclined to believe that radical changes in head form were not usually brought about by uninten¬ tional occipital deformation, but that the most of crania affected thereby were originally brachycephalic or sub-brachycephalic. Some few dolichocephalic crania may have been transformed into mesocephaly; more were probably transferred from the higher grades of mesocephaly to brachycephaly; still more brachycephalic crania were made increasingly so. It remains to compare the cranial indices in our series with those of the group studied by Langdon, and with other groups. In rear¬ ranging his seriations in accordance with the modern divisions of the index, only crania which have been sexed have been included. Cranial Indices of Madisonville Crania (Langdon) Dolicho. Meso. Brachy. Total Males: number. 1 12 23 36 “ per cent. 2.77 33.33 63.88 Females: number. 2 6 14 22 “ per cent. 9.09 27.27 63.63 90 INDIAN VILLAGE SITE AND CEMETERY As compared with our series, that of Langdon shows an excess of mesocephalic crania (33.33 per cent against 26.92 in males) and an excess of brachycephalic crania. These differences are even more pronounced in the case of the females (10.34 per cent of meso¬ cephalic crania in our series as against 27.27 in his series). That these differences, however, are due to the shortness of the series may be seen from the results of including in Langdon’s series 14 other crania which he was unable to sex. The following table shows the fundamental similarity: Cranial Indices of Madisonville Crania (Male and Female) Dolicho. Meso. Brachy. Total Peabody Museum: number. . 2 17 62 81 “ “ per cent . 2.47 20.98 76.54 Langdon series: number. . . . 5 15 52 72 “ “ per cent... . 6.94 20.83 72.22 The average length (diameter antero-posterior maximum) of 52 male crania of our Madisonville series is 177.4 mm. and the range from 161-195 mm. The mean length of 29 female crania is 169 mm. and the range from 161 -179 mm. The average width (diameter lateral maximum) of males is 146.1 mm. and the range 133- 160 mm. In the females the average width is 141.7 mm. and the range 126-150 mm. Length-Height Index Chamaecephalic Orthocephalic Hypsicephalic x-69.9 70-74.9 75-x Total Peabody Museum series: Males: number. . 0 10 32 42 “ per cent. . 0 23.81 76.19 Females: number. . 0 6 21 27 u per cent. . 0 23.33 77.77 Both sexes: number. . 0 16 53 69 “ “ per cent.... . 0 23.19 76.81 Langdon series: Both sexes: number. . 1 2 55 58 “ “ per cent.... . 1.72 3.44 94.82 As the series given above includes all crania, deformed and un¬ deformed, it is subject to the same reservations as the cranial index discussed above. It will be noted in the analysis of the above table that there is a considerable discrepancy in results between our Peabody Museum Papers Vol. VIII, Plate 25 Madisonville Site Male cranium, brachycephalic type Norma lateralis Norma facialis Norma verticalis Norma occipitalis NEAR MADISONVILLE, OHIO 91 series and that measured by Langdon in 1879. Our series shows a range of the index from 70 to 83, whereas Langdon’s series ranges from 68 to 89. Langdon’s series has an excess of hypsicephals amounting to about 18 per cent. This difference isd istinctly puz¬ zling in view of the similarity as regards the seriation of the cranial index in the two series. Langdon does not state how he measured cranial height. He may have taken maximum height instead of basion-bregma height. Langdon’s series comes from the first year’s excavations in the cemetery, whereas the majority of our crania are from later excavations in other portions of the cemetery, but if there had been a marked difference in headform we should have expected it to show in the cranial index. Of 53 undeformed Tennessee Stone Grave crania of both sexes measured by Fuller, 96.22 per cent were hypsicephalic and 3.77 per cent orthocephalic. Breadth-Height Index Tapeinocephalic Metriocephalic Akrocephalic x-91.9 92-97.9 98-x Total Males: number. 15 20 7 42 “ percent. 35.71 47.62 16.66 Females: number. 10 15 2 27 “ percent. 37.03 55.55 7.40 Both sexes: number. 25 35 9 69 “ “ per cent. 36.23 50.72 13.04 The range in this series is from 82 to 100 in the males and from 86 to 100 in the females. I have not calculated the mean as it is of doubtful significance in so heterogeneous a group of crania. Comparison with the Langdon group of Madisonville crania is omitted because the indices are not given. From the above table it will be seen that the relative height of this series is somewhat small. In 53 undeformed Tennessee crania of both sexes Fuller found 3.77 per cent tapeinocephalic, 28.30 per cent metriocephalic, and 67.92 per cent akrocephalic. This brings out sharply an important difference in these neighboring groups. The basion-bregma height is absolutely lower in our series (average for males 136.9 mm., females 131.3 mm.; against 144 mm. for males and 139 mm. for females in the Tennessee series). 92 INDIAN VILLAGE SITE AND CEMETERY Cranial Arcs. Cranial Arcs and Circumferences Mean Range Number Male Female Male Female Male Female Horizontal circumference: Madisonville. 513 493 482-540 462-515 44 26 Tennessee (Fuller).... 512 481 480-530 462-495 Nasion-opisthion arc: Madisonville. 361 345 334-379 328-365 41 21 Tennessee (Fuller).... 359 349 331-400 326-368 Transverse arc: Madisonville. 316 306 300-350 281-325 42 26 Tennessee (Fuller).... 310 302 293-348 292-321 In the table above the cranial circumferences and arcs of the Madisonville series are compared with those of the Tennessee Stone Grave series measured by Fuller, including in the latter case only undeformed crania. In the case of the horizontal circumfer¬ ence (taken above the brow ridges), the males of the two series show very similar values, while the Madisonville females surpass the Tennessee females in the mean value of the circumference by 12 mpri. The mean values of the same measurement for Arkansas and Louisiana series studied by Dr. Hrdlicka are as follows: Arkan¬ sas males 500, Louisiana males 504; Arkansas females 485, Louisi¬ ana females 488. This measurement and the following were not taken by Langdon on his Madisonville series. In the case of the nasion-opisthion arc the mean value for Madi¬ sonville males again exceeds slightly that of the Tennessee males while the Madisonville females fall below the Tennessee females. The corresponding values in the cases of the Arkansas and Louisi¬ ana series are: Arkansas males 352, Louisiana males 355; Arkan¬ sas females 342, Louisiana females 348. The excess of the Madisonville crania over the Tennessee crania in the transverse arc may be due partially to the fact that Fuller measured to the supra-mastoid crest, while the writer measured to porion. NEAR MADISONVILLE, OHIO 93 Cranial Capacity. Madisonville series: males. “ “ females. Tennessee series (Fuller): males. . . “ “ females. . Arkansas series (Hrdlicka): males. . “ “ “ females Mean Range Number 1435 cc. 1265-1630 cc. 38 1287 1150-1380 21 1410 1175-1680 ? 1276 1110-1460 ? 1455 1310-1670 19 1255 1140-1395 14 The mean capacities of the series compared above are similar when it is considered that the range is great and the series are comparatively small. Langdon found the mean capacity of 48 Madisonville crania of both sexes to be 1338, but these were measured with dried peas and the writer ascertained by experimen¬ tation that this method yields results 50 cc. below the correct capac¬ ity of a bronze control skull. If we then add 50 cc. to Langdon’s mean we find that it corresponds closely enough with the mean of both sexes in our series, 1382 cc. Thickness of Left Parietal Above Temporo-parietal Suture. It is very difficult to obtain accurate results in this measurement. In the crania of 46 Madisonville males the mean value of this measure¬ ment was 5.8 mm. The range in the males was from 3 mm. to 9 mm. except in the case of one extraordinarily thick cranium which yielded a measurement of 14 mm. on the left parietal just above the temporo-parietal suture. In 28 females the mean thick¬ ness at this point was 5.78 mm., very little less than in the males. The range was from 4 mm. to 7 mm. Dr. Hrdlicka found the aver¬ age thickness of the left parietal above the squamous suture in 13 male Arkansas crania 5 mm., and in 9 females 5.3 mm. In another Arkansas group of 22 male crania the average was 5 mm. and in 13 female crania 4.5. In the Louisiana group of 17 males and 18 females the averages were 5.5 mm. and 4.85 mm. respectively. Minimum Frontal Diameter. Mean Range Number Males Females Males Females Males Females Madisonville. 94.97 92.71 87-103 85-103 48 28 Tennessee (Fuller). . . . 93 90 83-108 82-99 ? ? Louisiana (Hrdlicka). . 96 94 85-102 83-99 17 17 Arkansas (Hrdlicka).. . 96 93 87-102 86-102 13 10 Langdon gives as the mean of this measurement for 69 Madison¬ ville crania of both sexes 93, with a range of 84-107. The corre- 94 INDIAN VILLAGE SITE AND CEMETERY sponding mean in our series is 94.1. The frontal breadth in this series is therefore rather small, though it exceeds that of the Ten¬ nessee Stone Grave group. Facial Index. Upper Facial Index Hypereuryene Euryene Mesene Leptene x-44.9 45-49.8 50-54.9 55-x Total Males: number. 1 6 18 2 27 “ per cent. 3.7 22.22 66.66 7.4 Females: number. 0 6 6 1 13 “ percent. 0 46.14 46.14 7.69 While the majority of the Madisonville crania fall into the mesene group as regards the upper facial index, it will be observed there also exists a strong tendency toward the euryene which is naturally accentuated in the female group. Quite markedly dif¬ ferent is the seriat.ion of the upper facial index in the undeformed crania from the Tennessee Stone Graves measured by Fuller. Of 97 males he found 34 per cent leptene, 56.7 per cent mesene, and 9.2 per cent euryene. In 64 Tennessee females the seriation is 34.3 per cent leptene, 54.6 per cent mesene, and 10.9 per cent euryene. In our series bizygomatic diameter in 28 males averages 141 mm., and in 14 females 131.3 mm. Upper facial height averages 72 mm. in 34 males and 66.1 mm. in 18 females. Total Facial Index Hypereuryprosopic Euryprosopic Mesoprosopic x-79.9 80-84.9 85-89.9 Total Males: number. 1 12 5 18 “ per cent. 5.55 66.66 27.77 Females: number. 0 4 2 6 percent. 0 66.66 33.33 The number of specimens upon which the measurements neces¬ sary for the calculation of the total facial index can be taken is unfortunately small. There is a strong tendency toward eury- prosopism in this group, due to large bizygomatic diameters and somewhat small total facial height (averages of 117.9 mm. in 21 males and 110.2 mm. in 7 females). The average total facial index in male crania from Louisiana and Arkansas, as given by Dr. Hrdlicka, is between 85 and 86, while the average for the males in our series is about 83. Fuller found the average total facial index NEAR MADISONVILLE, OHIO 95 in 92 male crania from the Tennessee Stone Graves 86, and in 59 females 85.2. Fuller found 19.5 of leptoprosopic crania in his males and 10 per cent in his females. The percentages of euryprosopic crania in the Tennessee group were 25 for males and 44 for females. It therefore appears that our Madisonville crania are sharply differentiated from the Tennessee, Arkansas, and Louisiana groups in respect to facial proportions. Mean Orbital Index. Chamaeconch Mesoconch Hypsiconch x-82.8 83-88.9 89-x Total Males: number. 20 13 3 36 “ percent. 55.55 36.11 8.33 Females: number. 12 7 1 20 “ per cent. 60.0 35.0 5.0 The orbital index is extremely variable. The mean index of the two orbits ranges in the males from 67 to 101, and in the females from 71 to 90. Often the measurements of the two orbits differ considerably and in such cases the index of the left orbit is usually higher, a fact previously observed by Dr. Hrdlicka in regard to the orbits of crania of Indians from Louisiana and Arkansas. In the Tennessee Stone Grave series of 95 males, 33.6 per cent were hypsiconch, 49.4 per cent mesoconch, and 16.8 per cent chamae¬ conch. Of 69 females 44.9 per cent were hypsiconch, 47.8 per cent mesoconch, and 7.2 per cent chamaeconch. Dr. Hrdlicka states that the majority of the Arkansas and Louisiana crania measured by him were megaseme (hypsiconch). It may be observed from the tables above that the Madisonville crania are prevailingly chamae¬ conch and mesoconch. This is to be expected in view of the large breadth and inconsiderable height of the facial skeleton. Nasal Index. Leptorrhine Mesorrhine Platyrrhine Hyperplatyrrhine x-46.9 47-50.9 51-57.9 58-x Total Males: number 7 11 15 2 35 “ per cent 20.0 31.42 42.85 5.71 Females: number 1 3 10 6 20 ' per cent 5.0 15.0 50.0 30.0 Almost half of the male crania and four-fifths of the female crania are platyrrhine. This is to be expected in broad, short-faced people, and the sexual difference is usual. In 102 male crania from the Tennessee Stone Graves, Fuller found 45 per cent platyrrhine, 96 INDIAN VILLAGE SITE AND CEMETERY 37.3 per cent mesorrhine, and 17.6 per cent leptorrhine. In 73 female crania of the same group the seriation is 63 per cent platyr- rhine, 26 per cent mesorrhine, 11 per cent leptorrhine. The Madi- sonville crania show a higher percentage of platyrrhine indices than the Tennessee group, especially in the females. The mean nasal index in males is 51.6 and in females 55.9. The mean of the Ten¬ nessee males is 50.1 and of the Tennessee females 51.5. Palatal Index (Maxillo-alveolar). Dolichuranic Mesuranic Brachyuranic x-109.9 110-114.9 115-x Total Males: number. 2 7 23 32 “ percent. 6.25 21.87 71.87 Females: number. 2 4 13 19 “ percent. 10.52 21.05 68.42 The palate in our group is usually brachyuranic with a range in males of 106-129 and a mean of 117.6. The range in the female crania is 108-136 and the mean is 118.7. The means in the Tennes¬ see groups are 122.6 for males and 120.9 for females. Of 84 Ten¬ nessee males 89.2 per cent were brachyuranic, 8.3 per cent mesur¬ anic and 2.3 dolichuranic. The corresponding figures for 55 females are: brachyuranic 78.1 per cent, mesuranic 7.2 per cent, dolichu¬ ranic 14.5 per cent. Dr. Hrdlicka gives as the average palatal index for Louisiana males 116, and for females 122, but the number of specimens included is small. On the whole it may be said that the degree of braehyurany exhibited by the Madisonville group is somewhat less than would be expected in view of the facial pro¬ portions and that it is evident that the palates in this group have undergone reduction, which often has the effect of decreasing the index. Alveolar Index. Orthognathous Mesognathous Prognathous x-97.9 98-102.9 103-x Total Males: number. 19 9 0 28 “ percent. 67.85 32.14 0 Females: number. 8 8 1 17 “ per cent. 47.05 47.05 5.88 While the alveolar index has been discarded by Dr. Hrdlicka and other authorities, and is much influenced by the facial length, it is still of interest as an indication of prognathism. According to this index more than two-thirds of our males are orthognathous Peabody Museum Papers Vol. VIII, Plate 26 Madisonville Site Female cranium, braehycephalic type Norma lateralis Norma facialis Norma verticalis Norma occipitalis NEAR MADISONVILLE, OHIO 97 and the rest mesognathous, while the females show a somewhat greater tendency toward mesognathism. In the Tennessee group of 94 males and 56 females the distribution of index classes is as follows: orthognathous, males 70.2 per cent, females 48.2 per cent; mesognathous, males 26.5 per cent, females 41 per cent; progna¬ thous, males 3.2 per cent, females 10.7 per cent. This is quite similar to the Madisonville seriation, but the Tennessee group in¬ cludes a few more prognathous skulls. Angles Relating to Prognathism. The facial angle according to Rivet and Hrdlicka, included between nasion, alveon, and basion, and measured according to the direct or graphic method, gives an average of 73.1° in 28 males and 72.1° in 17 females. This puts both sexes in the orthognathous group according to Rivet’s division of the index which markes the lower limit of that group at 73. The range in the males was from 68° to 79° and in the females from 67.5° to 80°. Dr. Hrdlicka found an average of 73° in Munsee or Lenape males and 74° in the females of that group. The following is the seriation of the index. Naso-alveolo-basilar Angle Prognathous Mesognathous Orthognathous x-59.9 70-72.9 73-x Total Males: number. 4 6 18 28 “ percent. 14.28 21.42 64.28 Females: number. 1 9 7 17 per cent. 5.88 52.94 41.17 If we compare the angle with the results of the alveolar index given above we find that the classification according to the angle removes 4 male crania from the mesognathous to the prognathous class and one orthognathous cranium into the mesognathous class. In the case of the females it changes one cranium from orthogna¬ thous to mesognathous. This is a better indication of prognathism than the alveolar index. Foramen Magnum. The mean diameter of the foramen magnum in 36 male Madisonville crania is 32.8 mm. and in 22 female crania 31.8 mm. This is decidedly below the average for Indian males as recorded by Dr. Hrdlicka, but in the case of the females there is a substantial agreement. The following figures are of interest by way of comparison: Munsee, 7 males, 35 mm., 8 fe¬ males, 32 mm.; Louisiana, 10 males, 34.5 mm., 14 females, 31.8 mm.; 98 INDIAN VILLAGE SITE AND CEMETERY Arkansas, 22 males, 33 mm., 16 females, 31.4 mm. It remains to be seen whether this ratio has any particular significance in relation to stature or other bodily characters. Lower Jaw. Males Females Number of specimens Average Number of specimens Average Height at symphysis . .. 25 36.5 9 32.4 Minimum breadth of ramus.. . . 30 35.5 9 34.2 Bicondylar width. . . 22 129.0 7 122.7 Condylo-symphysial length. . . . 24 104.6 7 104.5 Bigonial diameter . . . 24 103.4 8 99.1 Mean angle. . . 25 125.4 7 126.4 The dimensions of the lower jaws in this series are moderate. The mandibles are especially broad in proportion to their length. The female jaws seem relatively longer than the male but this is possibly due to the shortness of the female series. The mean angle of the ascending rami is unusually high in the male group, but otherwise the measurements do not diverge markedly from those observed in the case of other Indian crania. Frontal Region. Height Very low Low Medium High Very high Total Males: number... . 2 18 30 2 i 53 “ per cent... . 3.75 34.0 56.6 3.75 1.9 Females: number. . 1 12 12 6 0 31 “ per cent.. 3.2 38.7 38.7 19.3 0 Doubtful: number. 0 1 2 0 0 3 Breadth Narrow Medium Broad Very broad Total Males: number. . . . 13 27 10 3 53 “ per cent. . . . . 24.5 51.0 18.9 5.6 Females: number. . 6 17 4 4 31 per cent. . . 19.3 54.9 12.9 12.9 Doubtful: number. 0 1 2 0 3 Slope Bulging Submedium Medium Receding Very receding Total Males: number.... 2 4 37 7 3 53 “ per cent.... 3.7 7.5 70.0 13.0 5.6 Females: number. . 7 0 20 4 0 31 “ per cent.. 22.6 0 64.5 12.9 0 Doubtful: number. 0 0 3 0 0 3 NEAR MADISONVILLE, OHIO 99 An analysis of the above tabulations of observed characters in the frontal region of the crania shows that the majority of males (56.6 per cent) had brows of medium height, while a considerable proportion of the remainder (34 per cent) had low foreheads. Only a few crania exhibited high frontals, or very low frontals. In comparison with the males the females show a lesser proportion of frontals of medium height and a decided increase in the num¬ bers of skulls with high frontals (19.3 per cent). In the matter of frontal breadth, as observed in relation to height and slope, about one half of the male skulls are medium and the other half almost equally divided between narrow and broad. The females show a similar distribution, except for a somewhat larger proportion of relatively very broad frontals. In the large majority of the male skulls the slope of the frontal bone is medium (70 per cent). The remainder show more with receding frontals than with steep or bulging frontals. The female crania differ from the males in the high percentage of bulging frontals (22.6 per cent). This is, of course, a common sex difference. Sagittal Region. Breadth Submedium Medium Broad Very broad Total Males: number. . 3 26 15 9 53 “ per cent. . 5.6 49.0 28.5 17.3 Females: number. . . . 2 15 11 3 31 “ per cent. . . . 6.5 48.4 35.5 9.6 Doubtful: number. . . . 0 2 1 0 3 Elevation Absent Submedium Medium Marked Very marked Total Males: number .... 10 19 18 4 2 53 “ per cent. 18.9 36.0 34.0 7.5 5.6 Females: number. . . 7 14 9 1 0 31 “ per cent.. . 22.6 45.1 29.0 3.2 0 Doubtful: number. . 1 0 2 0 0 3 The tables above clearly show a predominance of medium and broad sagittal regions both in males and in females, with a some¬ what larger proportion of the broad categories in the latter sex. A large majority of the male crania show a varying development of the sagittal elevation, but it is very pronounced in a few cases only. Naturally the females show a lesser development of this character. 100 INDIAN VILLAGE SITE AND CEMETERY In 9 male crania, or 17 per cent, a slight post-coronoid depression was observed. This character also was present in a slight degree in 13 female crania, and markedly in one female cranium, — a total of 45.2 per cent. This again is a sexual difference of common observation. Temporal Region. Flat or depressed Medium Bulging Total Males: number. 19 18 16 53 “ per cent. 36.0 34.0 30.0 Females: number. 11 10 10 31 “ per cent. 35.5 32.2 32.2 Doubtful: number. 1 1 1 3 The table above shows an almost equal distribution of flat, medium, and bulging temporal regions in both sexes. A pronounced depression of the wing of the sphenoid is very common in this series. Occipital Region. Flat or Medium Occipital steep convex Protuberant torus Total Males: number.. . . 38 14 1 16 53 “ per cent. . . 71.7 26.4 1.9 30.0 Females: number.. 21 8 2 1 31 “ per cent.. 67.7 25.8 6.5 3.2 Doubtful: number. 2 1 0 1 3 The high percentage of flat occipital regions in this series, as shown in the table above, is partially due to artificial occipital deformation, but in a greater degree to the natural shape of the skull. The percentage of male crania exhibiting flat occipital regions is 71.7, while the percentage of all male crania showing occipital deformation is 73.5. On the other hand 67.7 per cent of female crania have flat occiputs, whereas 82.7 per cent show artificial deformation. A slight occipital torus was observed in 11 male skulls, an oc¬ cipital torus of medium development in 3 skulls, and of pro¬ nounced development in 2 skulls, making a total of 30 per cent of the male crania exhibiting this feature. One female cranium and one cranium of doubtful sex also show a slight development of this feature. NEAR MADISONVILLE, OHIO 101 Sutures. Occlusion of Coronal, Sagittal, and Lambdoid All open Beginning in all Beginning in sagittal Beginning in coronal Males: number. . . .. 26 3 4 0 “ per cent.. . . . 49.05 Females: number. . . . 25 1 0 2 “ per cent. .. 80.06 Beginning in sagittal and lambdoid Beginning in sagittal and coronal Beginning at pterion Advanced in sagittal,others open or beginning Males: number . . . . . . 3 2 0 8 Females: number. . . . . 0 1 1 0 Advanced in sagittal, and in coronal, lambdoid open Advanced in sagittal, and in lambdoid, coronal open Advanced in sagittal, coronal, and lambdoid Total Males: number. . . . 2 2 3 53 Females: number. . . 1 0 0 31 As may be seen from the above tabulations, about half of the male crania in our series and four-fifths of the female crania showed no external traces of obliteration of the sutures. Contrary to what has usually been observed in the case of American crania, points of obliteration appear first in the sagittal suture and obliteration pro¬ ceeds more rapidly in this suture. Hrdlicka observed in his Arkan¬ sas and Louisiana series that synostosis began dorsally in the coronal suture, and Fuller makes a similar statement in regard to his Tennessee Stone Grave series. The difference may be due, on the one hand, to the less pronounced occipital deformation of the Madisonville series, or, on the other, to an error conse¬ quent upon the small number of crania. It is of some interest to note that in two of the female crania synostosis was recorded to have begun in the coronal and in no case in the sagittal, while of the males no example of priority of coronal obliteration was observed. The form of pterion found in this series is almost invariably the broad H type, but one male cranium and one female cranium ex¬ hibited the K or X type. 102 INDIAN VILLAGE SITE AND CEMETERY Serration Simple Medium Complex ? Total Males: number. 33 18 1 1 53 “ percent. 62.23 33.94 1.88 1.88 Females: number. 21 10 0 0 31 “ percent. 67.72 32.25 0 0 Doubtful: number. 1 2 0 0 3 As indicated above, the conformation of the sutures in this series of crania is simple in about two-thirds of the cases and of medium complexity in the remainder. Wormian Bones Temporo- Temporo- ? None Lambdoid parietal occipital Others Total Males: number. 4 11 32 4 13 9 53 “ per cent. 7.54 20.75 60.37 7.54 24.52 16.98 71.69 Females: ntlmber.. . 0 13 15 0 7 2 31 percent... 0 41.93 48.38 0 22.58 6.45 58.06 Doubtful: number. . 0 1 2 0 0 0 3 The table above shows that sutural bones occur in 71.69 per cent of the male crania and in 58.06 per cent of the female crania. They are found oftenest in the lambdoid suture and next in the temporo-parietal. Nine epipteric bones were observed in the male crania, one in the coronal suture, one in the sagittal suture, and one os apicuin or triangular Wormian bone at lambda. Two of the male crania showed traces of the transverse occipital suture, but it was not complete in either case. There were two epipteric bones among the crania of females and one complete transverse occipital suture. The so-called Inca bone, then, occurs in but one of the 87 crania of our collection, whereas Fuller found it in about 16 per cent of his Tennessee Stone Grave series. Parietal Foramina. Number None One Two Total Males: number.. 24 16 13 53 “ percent. 45.26 30.17 24.51 Females: number. 18 7 6 31 per cent. 58.05 22.57 19.35 Fuller found parietal foramina absent in about 40 per cent of his Tennessee Stone Grave crania. This corresponds fairly closely with the results shown in the above table. Peabody Museum Papebs Yol. VIII, Plate 27 Madisonville Site Male cranium, subdolichocephalic type Norma lateralis Norma facialis Norma verticaiis Norma occipitalis NEAR MADISONVILLE, OHIO 103 Retro-mastoid Foramina. These are extremely variable in num¬ ber, size, and position. Mastoids. Size Submedium Males: number. 22 “ percent. 41.49 Females: number. 7 “ per cent. 23.33 Medium Large Very large Total 24 6 1 53 45.26 11.31 1.88 16 7 0 30 53.33 23.33 0 The mastoid processes are usually medium or somewhat below medium in size in our Madisonville crania as compared with aver¬ age Europeans. Fuller found 55 per cent of his Tennessee Stone Grave males with mastoids of medium size, and the remainder equally divided between sub-medium and large. On the whole our series falls somewhat below the Tennessee group in the develop¬ ment of this process. Hrdlicka observed in his examination of crania from Arkansas and Louisiana that, while the mastoid proc¬ esses often showed only a moderate development in the males, in females they frequently exceeded the average determined in the same sex in Whites and in Indians from other regions. He related this development in females to the growth of the sternocleido¬ mastoid muscles favored by the habitual carrying by these women of heavy jars and other burdens. From the table above a similar superiority of mastoid development on the part of females may be observed in our series, and doubtless for the same reason. Facial Portion. SlJPRA-ORBITAL RlDGES Absent Submedium Medium Large Very large Total Males: number. . 4 13 25 7 4 53 “ per cent.. 7.54 24.51 47.14 13.16 7.54 Females: number 11 13 7 0 0 31 1 per cent 35.48 41.93 22.58 0 0 It is of interest to compare with the above table the results obtained by Fuller in the observation of the same character on 148 male crania and 72 female crania of Tennessee Indians. In the males he found no case in which the brow ridges were undeveloped, 29.1 per cent were of submedium development, 44.6 per cent me¬ dium and 26.3 per cent above medium. This shows a distribution of the grades of percentages in various development closely sim- 104 INDIAN VILLAGE SITE AND CEMETERY ilar to that of our series, but the Madisonville crania, on the whole, show more instances of the lesser degrees of development. (Ten¬ nessee females: absent 31.9 per cent, submedium 33.3 per cent, medium 29.1 per cent, large 5.5 per cent.) The types of supra-orbital ridges most commonly found in this series are: (a) in which the ridges are over the median portions of the orbits and limited laterally, ( b ) in which the lateral portions of the superior orbital margins are thickened to form ridgeS which are separated from the median ridges, (c) in which the lateral and median ridges are continuous forming a torus from one external angular process to the other. Type a is by far the most common in our series. Type b is frequently found and type c only occasion¬ ally. The orbits in this series present no unusual features. There is considerable variation in the dimensions of right and left orbits in the same individual, as remarked above. Practically all of the orbits are of the oblong shape with rounded corners. In case of the male crania 57.5 per cent of the orbits have their long axes horizontal and 42.5 per cent have long axes inclined downward and outward. In the female crania these figures are 54.54 per cent and 45.54 per cent respectively. Infra-orbital Suture Absent Right Left Both Total Males: number. 21 2 3 6 32 “ percent. 65.62 6.25 9.37 18.75 Females: number. 13 2 3 3 21 “ percent. 61.88 4.76 14.28 14.28 In the above tabulation of the occurrence of the infra-orbital suture the writer has not distinguished between cases in which the suture was present, only on the facial aspect and cases in which it was complete on both facial and orbital aspects. ^Tiere found it was usually complete. Fuller records the complete absence of this feature in (34 per cent of his male crania from the Tennessee Stone Graves, which agrees closely enough with our figures. But in Tennessee females he finds the suture absent in only 30 per cent of cases as against 61.88 per cent in our series. Nasion Depression. The nasion depression, which is due in large part to the protuberance of glabella, is very slight or entirely NEAR MADISONYILLE, OHIO 105 absent in most of the male skulls in our Madisonville series, and does not occur at all in the crania of females. The type of depressed nasion that is often seen in Australian skulls (i.e. where the fronto¬ nasal suture is not only depressed below the level of the glabella eminence, but is sunken deep below the level of ophryon) does not occur at all in this series. Nasal Bridge Breadth Height Submedium Medium Broad ? Low Medium High ? Total Males: number 14 24 5 7 ii 28 4 7 50 “ per cent 28.0 48.0 10.0 14.0 22.0 56.0 8.0 14.0 Females: number 7 10 5 9 11 10 1 9 31 “ per cent 22.57 32.25 16.12 29.03 35.48 32.25 3.22 29.03 Shape of Nasal Bridge Males: number. . .. “ per cent.. .. Females: number 1 . ? Straight 22 4 48.84 8.88 21 2 Convex Concave Concavo-convex Total 8 0 11 55 17.76 0 24.42 1 0 8 . 31 From the analysis of the above tables it will be observed that the nasal bridge in males of our series is prevailingly of medium breadth, often of submedium breadth, and sometimes broad. Usually it is of medium height (56.0 per cent), and often low (22.0 per cent). The nasal bridge in the females tends to be lower and broader, a generally observed sex difference. Convex and con¬ cavo-convex nasal bridges are the prevailing forms. Nasal Spine ? Submedium Medium Pronounced Total Males: number. .. . 1 20 14 0 35 “ per cent .. . 2.85 57.14 39.99 0 Females: number. . 0 10 9 0 19 per cent.. 0 52.63 47.34 0 Lower Borders of Nasal Aperture Indistinct Medium Sharp Total Males: number. . . . . 20 5 13 38 “ per cent. . . . 52.63 13.15 34.21 Females: number. . . 8 7 5 20 per cent. . 40.0 35.0 25.0 1 Percentages omitted because of inadequate series. 106 INDIAN VILLAGE SITE AND CEMETERY Nasal Grooves Absent Submedium Medium Pronounced Total Males: number. . . 26 7 2 3 38 “ per cent . . 68.38 18.31 5.26 7.89 Females: number . 16 3 1 0 20 “ per cent . 80.0 15.0 5.0 0 The above tables show a prevailingly poor development of the nasal spine and indistinct lower borders of the nasal aperture as characteristics of both sexes in our series. This is quite the reverse of what Fuller states to be true of the Tennessee Stone Grave series of which he says that the lower borders are “ almost uni¬ versally sharp ” and the nasal spines usually well developed The observations on our series, however, accord with the results of Dr. Hrdlicka’s findings on Arkansas and Louisiana skulls in re¬ spect to these characters. Sub-orbital Fossae Absent Submedium Medium Pronounced Total Males: number. 0 9 14 16 *39 “ percent. 0 23.07 35.89 41.02 Females: number. 1 2 10 9 22 per cent. 4.54 9.08 45.45 40.86 Doubtful: number. 0 1 1 1 3 The depth of the sub-orbital fossae depends very largely upon age, to some extent upon sex, and in an indeterminate degree upon racial and individual variation. The distribution in the series under consideration, as shown above, seems to have no unusual significance. Malars Submedium Medium Large Very large Total Males: number. 16 17 2 40 “ per cent. . . . . . 12.5 40.0 42.5 5.0 Females: number.. . 11 8 0 0 19 per cent. . 57.86 42.08 0 0 Doubtful: number.. 1 0 1 0 2 Zygomae Submedium Medium Large Very large Total Males: number.... 4 11 18 5 38 “ per cent. ... . . 10.52 28.94 47.35 13.15 Females: number.. . 5 10 1 0 16 per cent. . . . 31.25 62.50 6.25 0 Doubtful: number.. 1 0 1 0 2 NEAR MADISONVILLE, OHIO 107 As shown in the table, the malars are largfe in nearly half of the male crania, but not “ very large,” except in two cases. Rugged and massive malar bones are not nearly as common in the group from this cemetery as they are, for example, in crania from the Tennessee Stone Graves. Large malars are not found in the few female skulls in the series in which observations could be made on the facial skeleton. The marginal process on the malar bone, which Langdon found well developed in 76.4 per cent of 68 Madi- sonville crania, is also of common occurrence in our series. The zygomatic arches are naturally stronger and more rugged in males than in females. Alveolae Prognathism Absent Submedium Males: number.... 1 16 “ per cent.... 2.94 47.04 Females: number. . 1 7 per cent.. 5.0 35.0 Doubtful: number 0 1 Medium Very- Pronounced pronounced Total 12 5 0 34 35.28 14.70 0 7 4 1 20 35.0 20.0 5.0 0 0 0 1 The above tabulation shows a somewhat greater development of alveolar prognathism among female crania than among the male. Langdon remarks that prognathism “ is a generally well-marked, though not constant feature of these crania.” 1 Fuller, in his ex¬ amination of Tennessee Stone Grace crania found a more marked tendency to alveolar prognathism among females than among males. Teeth. As in most series of crania from old graves, so many teeth have been lost post-mortem in the present group that observations on the dentition are very unsatisfactory. Dentition Complete Incomplete Total Sub- None medium Wear Medium Pronounced Total Males: number . 38 3 41 i 27 9 5 42 “ per cent . 92.68 7.32 2.38 64.26 21.42 11.90 Females: number 20 6 26 2 11 6 1 20 per cent 76.92 23.07 10.0 55.0 30.0 5.0 Doubtful: number 2 1 3 1 1 0 1 3 The first division of the table above shows that of 41 male crania on which observations concerning the dentition could be 1 Op. cit., p. 241. 108 INDIAN VILLAGE SITE AND CEMETERY made, 38 or 92.68 per cent showed completed dentitions. By this it is meant that all of the permanent teeth had erupted. In the female series, 20 of a total of 26 crania showed completed dentitions. In every instance in the crania of both sexes in which the dentition was incomplete, one or more of the third molars had not erupted. Most of these, according to the age estimates, were young adults between the ages of twenty and thirty years, but several were obviously middle aged. There is clearly shown here a very marked tendency toward the suppression of third molars in the females (amounting to 23 per cent of the cases observed), while in a lesser degree (7.3 per cent) the same is true of the males. The doubtful skull in which the dentition was incomplete was that of an adolescent. In this connection it is interesting to note that Fuller found the suppression of one or more third molars in 24 of the 185 Tennessee Stone Grave crania he examined—13.9 per cent of male crania, and 11.4 per cent of female crania. This percentage doubtless would have been higher if it had been possible to estimate it on the basis of the number of crania in which dental observa¬ tions were possible, instead of the entire number studied. But the author did not give the former figures. Analysis of the table showing the degree of wear of the teeth in¬ dicates that the majority of the individuals represented were young adults and that the females included fewer aged persons than the males. Teeth Lost in Life Few Many All Total crania Males: number. 1 8 2 41 “ per cent. 2.44 19.51 4.88 Females: number. 1 5 1 20 “ per cent. 5.0 25.0 5.0 Quality Very poor Submedium Medium Good Total Males: number. 3 5 12 18 38 “ percent. 7.89 13.15 31.57 47.35 Females: number. 3 4 8 4 19 “ percent_ 15.78 21.04 42.08 21.04 Doubtful: number.... 1 0 0 2 3 Peabody Museum Papers Vol. VIII, Plate 28 Madisonvillf, Site Female cranium, subbrachycephalic type Norma lateralis Norma facialis Norma verticalis Norma occipitalis no INDIAN VILLAGE SITE AND CEMETERY Shovel-shaped incisors, identified by Hrdlicka as a characteristic of Indian teeth, were observed in 12 of the 13 male crania in which examination was possible, and in all of the 6 female crania. Palate. Shape U-shaped Parabolic Hyperbolic Elliptical Total Males: number.. . 7 20 2 4 33 “ per cent. . 21.21 60.60 6.06 12.12 Females: number 3 14 2 1 20 “ per cent 15.0 70.0 10.0 5.0 The distribution of palate form is shown in the table above. It agrees very closely with Fuller’s figures for Tennessee Stone Grave crania, in which of 70 observations on male crania, 65.71 per cent were parabolic and hyperbolic, as against 66.66 per cent in our series; 22.85 per cent of Tennessee Stone Grave males had U- shaped palates, and 21.21 per cent of Madisonville males; 11.42 per cent of Tennessee crania and 12.12 per cent of Madisonville crania had elliptical palates. In the case of females the comparison shows almost as striking a similarity, for of 45 female crania examined in the Tennessee Stone Grave series, 68.88 per cent were parabolic and hyperbolic, 15.55 U-shaped, and 15.55 elliptical. The roof of the palate was observed to be high in 9 of the male crania of our series, or 27.27 per cent, as against 26.37 per cent of the Tennessee Stone Grave series observed by Fuller. In female crania from Madisonville, 2 of 20, or 10 per cent, had high roofs, as against 22.38 per cent of the Tennessee series. A slight development of the palatine torus occurred in 4 male crania of our series or 12.12 per cent, and in 4 of the female crania or 20 per cent. Fuller estimated the occurrence of the palatine torus in Tennessee Stone Grave males at 15 per cent, and in females at 10 per cent. There is little doubt that the shortness of our female series has fortuitously increased beyond normal the occurrence of this feature. Skull Base. Styloids Absent Submedium Medium Large Very large Total Males: number. .. 1 14 19 • 13 3 50 “ per cent... 2.0 28.0 38.0 26.0 6.0 Females: number. 2 19 5 3 0 29 “ per cent 6.9 65.3 17.2 10.3 0 NEAR MADISONVILLE, OHIO 111 The styloid processes are frequently broken off short, but it is usually possible to judge their development by the stumps. Glenoid Fossa —Depth Submedium Medium Deep Very deep Total Males: number.. . . . 1 26 24 3 54 “ per cent. . . . 1.85 48.14 44.44 5.55 Females: number. . . 2 19 8 1 30 per cent . . 6.66 63.33 26.66 3.33 The table above shows that the series is characterized as a whole by the prevalence of glenoid fossae of medium or greater depth. Females exhibit fewer instances of the latter. Postglenoid Process Submedium Medium Large Very large Total Males: number. 6 6 5 0 17 “ per cent. 29.62 Females: number. 1 0 2 0 3 “ per cent. 10.0 The postglenoid process is not usually to be observed in this series. It occurs much more frequently in male crania than in female crania. Foramen Lacerum Medium Small Medium Large Total Males: number. . 24 19 1 44 “ per cent. . 54.52 43.15 2.27 Females: number. . . . . 18 9 0 27 per cent. . . . . 66.6 33.3 0 Depression of Petrous Parts of Temporal Bones Absent Slight Medium Pronounced Total Males: number 2 13 23 6 44 “ per cent 4.54 29.51 52.25 3.62 Females: number 2 15 10 0 27 per cent 7.4 55.5 37.0 0 In large-brained races the lacerate foramina are usually large and the petrous portions of the temporal bones are depressed well below the level of the basilar process of the occipital bone. A well- developed brain pushes the basilar process outward beyond the level of the rigid petrous parts. An analysis of the above table shows that our Madisonville crania do not exhibit high develop¬ ment in these characters. 112 INDIAN VILLAGE SITE AND CEMETERY Foramen Lace rum Posterior Small Medium Large Right larger Left larger Equal Total Males: number 9 30 3 17 7 18 42 “ per cent 21.42 71.40 7.14 40.46 16.66 42.84 Females: number 5 19 1 8 3 14 25 “ per cent 20.0 76.0 4.0 32.0 12.0 56.0 Where there is a difference in the size of the two posterior lacer¬ ate foramina, the right foramen is usually the larger, as may be seen from the above tabulations. Fuller states that the right fora¬ men is usually much larger in Tennessee Stone Grave crania. Post-condyloid Foramina Absent Small Medium Large Right only Left only Total Males: number.... 2 5 28 1 3 1 40 “ per cent... 5.0 12.5 70.0 2.5 7.5 2.5 Females: number.. 0 0 18 4 4 0 26 per cent. 0 0 78.24 4.34 17.38 0 PARA-MASTOID PROCESSES. These were observed in 3 male and 2 female crania. Shape of For.amen Magnum Irregular Oval Half diamond Round Diamond Hexagon Total Males: number 4 20 8 6 2 0 40 “ percent 10.0 50.0 20.0 15.0 5.0 0 Females: number 2 8 8 4 2 2 26 “ per cent 7.69 30.76 30.76 15.38 7.69 7.69 There is considerable variation in the shape of the foramen magnum. In the males the oval form predominates. In the table above “ half diamond ” means that the anterior half of the fora¬ men is shaped like two sides of an equilateral triangle, while the posterior half is semi-circular or semi-oval. Pterygo-spinous Foramina Right Indicated Left Both Total Right Complete Left Both Total Males: number.. .. 3 1 4 8 1 3 1 5 “ per cent. . . Females: number.. 1 3 1 17.0 5 1 1 1 10.63 3 “ per cent. 20.0 12.0 Pterygo-spinous foramina, complete or indicated to have been completed by a fibrous bridge, are relatively common in this series. NEAR MADISONVILLE, OHIO 113 Fuller seems to have found a higher frequency of occurrence in his Tennessee series, for he reports the foramen, or indications of it, on one or both sides in 45.71 per cent of males, and 47.54 per cent of females. Dehiscences in the Floor of the Auditory Meatus Right Left Both Total Males: number. 2 3 4 9 “ per cent. 20.0 Females: number. 3 2 4 9 “ per cent. 33.33 Defects in the floor of the auditory meatus are common in the crania of American Indians. Fuller found them in 14 per cent of 144 Tennessee Stone Grave crania; Dr. Hrdlcika found them in 14 per cent of crania from Arkansas, and in 29 per cent of crania from Louisiana. Both of these authors also found them to occur much more commonly in the crania of females. Mandible. Mandibles were associated with the skulls in the case of 29 males and 12 females. The more important observations on these are tabulated below. Size Submedium Medium Large Very large Total Males: number. . . , . . . 7 13 7 2 29 “ per cent. .. ... 24.1 44.8 24.1 6.90 Females: number.. 2 9 1 0 12 per cent. . .. 16.6 75.0 8.3 0 Males: number.. “ per cent.. Females: number per cent Neutral 1 8 3.4 27.6 2 5 16.6 41.6 Large 5 17.2 5 0 41.6 0 Very large Total 0 29 0 0 12 0 Mental Prominence S ubmedium Medium 15 51.7 Mylo-hyoid Ridge Submedium Males: number. 11 “ per cent.... 37.9 Females: number.. . 7 percent.. 58.3 Medium Prononuced Very pronounced Total 17 1 0 29 58.6 3.4 0 5 0 0 12 41.6 0 0 114 INDIAN VILLAGE SITE AND CEMETERY Genial Tubercles Absent Submedium Medium Large Very large Total Males: number... . 1 16 10 1 1 29 “ per cent... . 3.4 55.2 34.5 3.4 3.4 Females: number. . 0 9 3 0 0 12 “ per cent. . 0 75.0 25.0 0 0 From the above tables it may be seen that the characteristic mandible is of medium size with a medium development of the mental prominence and a medium or submedium development of the mylo-hyoid ridge. The genial tubercles are poorly developed. Pathological Features. In the following discussion alveolar abscesses and dental caries have not been considered, as these pathological features have already been dealt with above. Lesions Probably of Traumatic Origin Depressed scars Linear fractures Miscellaneous Males: number. 9 2 4 Females: number. 4 0 0 About 22 per cent of the male crania in our series show lesions that are probably due to wounds. Most common are depressed scars on various portions of the cranial vault. No. 57512 has an old perforated lesion on the left parietal near bregma, oval in shape, and 9 mm. in its longest diameter. The tip of a small flint arrowpoint is also embedded in the outer table of the left parietal, 27 mm. behind the coronal suture and 42 mm. from bregma. No. 35529 has a large depressed fracture on the right parietal, 2 cm. in diameter. No. 35527 also has an arrowpoint embedded in the occiput. No. 57056 has a healed linear fracture of the occipital bone extend¬ ing from opisthion to lambda. No. 25126 has a linear fracture of the left parietal extending diagonally from lambda to a point 3 cm. behind the coronal suture. Near the middle of this fracture and extending over an oval area for a distance of 45 mm . in its long axis is an irregular cicatrized scar with a maximum breadth of 35 mm . This is the area of impact of the blow which was apparently de¬ livered with some blunt weapon. No. 58733 presents a cut in the right supra-orbital ridge and a fracture of the left zygomatic arch. No. 58058 has a healed fracture of the right horizontal ramus of the mandible. Peabody Museum Papers Vol. VIII, Plate 29 Madisonville Site Male cranium, intermediate type Norma lateralis Norma facialis Norma verticalis Norma occipitalis NEAR MADISONVILLE, OHIO 115 Langdon found eleven cases of fracture in 141 Madisonville crania examined by him, and one skull with the fragment of a small flint arrowpoint imbedded in the occiput. Arthritic conditions were observed about the occipital or man¬ dibular condyles or in the glenoid fossae of three male crania and one female cranium. In addition to this two male crania presented ankylosis of the atlas to the occiput, probably of arthritic origin. Langdon found this condition in 3 of 141 Madisonville crania ex¬ amined by him. Three female crania presented slight exostoes of doubtful origin in various regions. No. 25128, the skull of a female, presented evi¬ dence of an abscess in the left ear, which had involved extensive ab¬ sorption of the bony tissue and a perforation of the tympanic plate. No. 35528 presented a small bony protuberance on the left parietal near bregma corresponding to a deep circular impression internally, circular in shape and about 5 mm. in diameter, connected with the groove for the meningeal artery, which was unusually large and deep. Summary of Measurements and Observations upon the Crania. A slight degree of unintentional occipital deformation is very com¬ mon in the Madisonville series, but this is pronounced in a few cases only. Apparently it has affected brachycephalic crania much more than dolichocephalic crania, and probably no crania have been changed from the dolichocephalic class to the brachycephalic class through his agency. No frontal deformation occurs. ’About three-fourths of the crania are brachycephalic (plates 25, 26), and the remainder are mesocephalic (plates 27, 28, 29), with the exception of two, which are dolichocephalic. Three-fourths or more are hypsicephalic and the rest orthocephalic. The breadth- height index is medium in about half of the series and low in the majority of the remainder. The cranial capacity is well up to the average for Indians. The skull walls are not extraordinarily thick. The forehead is a little narrower than average for Indians. The faces are broad and quite short; the majority being eury- prosopic and the rest mesoprosopic. The orbits are variable, but the majority are chamaeconch or microseme. The nose is short and broad, and platyrrhine indices are in the majority. There is some alveolar prognathism but practically no facial prognathism. The palate is brachyuranic and the lower jaw is short and broad. 116 MEASUREMENTS OF CRANIA FROM THE Catalogue No. &% e«.». ^ 1 N v 1 A*i so ' IJI ,**' i* ^ , < . ^.«\\\\^ A i\\\> V \ \ I I I \ I i\ \ ' 11 I . \W \\\w /I j \\ l ' \' ; ||\\lH'V^ /, ! * \ ..ill 1 ,\V I 1 ..'. hlA'jft- HOUSE CIRCLES STITES PROPERTY 3 if ' AiWn Z',. L /£>? - v#, ., * 1 *.»* f - s ' ' '•-• i jsvwii il^ x £ p // i/|\V INDIAN VILLA6E SITE AND CEMETERY near MADISONVILLE OHIO SHOWING EXPLORATIONS of 1873*1911 &CALC >H EIET I-V Peabody Museum Explorations, 1897, J. R. Swanton, Field Director A-C “ “ “ 1907, R- E. Merwin, “ “ D-F “ “ “ 1908, “ “ “ “ “ G-K “ “ “ 1911, B. W. Merwin, “ “ L “ “ “ 1882, F. W. Putnam, “ “ M Dept. Anthropology Exp., 1891, World’s Columbian Exposition, C. L. Metz, Field Director N Madisonville Literary and Historical Society, Cincinnati Hist. Soc. and Peabody Museum Explorations; 1S80, C. L. Metz, Field Director O Madisonville Literary and Historical Society, 1879, C. L. Metz, Field Director 1-4 House Sites P Trial Trenches, 18S2 Owing to the small scale of this plan, some of the minor burial smaller cache-pits, post-holes, etc., are om.tted, to add clearness to the grouping of the more .mportant interments an<1 Detail Plans on a larger scale are given on plates 1 and 2 ortions of Trenches II and III and the greater part Ot ot P h D The positions of these enlarged areas are indicated plan bv rectangles enclosed by broken lines. PAPERS OF THE PEABODY MUSEUM OF AMERICAN ARCHAEOLOGY AND ETHNOLOGY, HARVARD UNIVERSITY Vol. VIII. — No. 2 BASKET-MAKER CAVES OF NORTHEASTERN ARIZONA Report on the Explorations, 1916-17 BY SAMUEL JAMES GUERNSEY AND ALFRED VINCENT KIDDER FORTY-FOUR PLATES AND SIXTEEN ILLUSTRATIONS IN THE TEXT CAMBRIDGE, MASSACHUSETTS, U. S. A. PUBLISHED BY THE MUSEUM 1921 PAPERS OF THE PEABODY MUSEUM OF AMERICAN ARCHAEOLOGY AND ETHNOLOGY, HARVARD UNIVERSITY Vol. VIII.— No. 2 BASKET-MAKER CAVES OF NORTHEASTERN ARIZONA Report on the Explorations, 1916-17 BY SAMUEL JAMES GUERNSEY AND ALFRED VINCENT KIDDER FORTY-FOUR PLATES AND SIXTEEN ILLUSTRATIONS IN THE TEXT CAMBRIDGE, MASSACHUSETTS, U. S. A. PUBLISHED BY THE MUSEUM 1921 COPYRIGHT, 1921 BY THE PEABODY MUSEUM OF AMERICAN ARCHAEOLOGY AND ETHNOLOGY, HARVARD UNIVERSITY INTRODUCTION In the summer of 1914 the Peabody Museum of Harvard Uni¬ versity sent an expedition to northeastern Arizona under the joint leadership of the present authors for the purpose of studying the relations between the cliff-houses of that district and those of the north side of the San Juan River. In the course of this trip, evi¬ dence was found of the presence of the Basket-maker culture. This culture had hitherto only been reported from a single rather re¬ stricted area in southeastern Utah. 1 Furthermore, no Basket- maker remains had ever been taken out by trained investigators; so that the claims, put forward by the commercial collectors who discovered and named the culture, that it was a distinct one, ante¬ dating that of the Cliff-dwellers, had been received by archaeol¬ ogists with more or less incredulity. We felt, therefore, that the opportunity for studying these little known remains in a region untouched by earlier diggers, was one which should not be neglected; all our subsequent work has accordingly been directed toward the finding and excavation of Basket-maker sites. In 1915 the junior author regretfully gave up field work in this region to undertake other excavations, and the expeditions of that and the following years were conducted by Mr. Guernsey. The results of 1914 and 1915 have already been published, 2 the present report deals with the explorations of 1916 and 1917; at the close of the latter season field work was temporarily discontinued because of the war. In each year the expeditions were carried on under permits granted by the Secretary of the Interior. The Museum wishes to make grateful acknowledgment to the fol¬ lowing persons whose generous contributions, supplementing the Museum appropriation, served greatly to enlarge the scope of the work: Mrs. S. K. Lothrop, and Messrs. Bronson Cutting, Lawrence Grinnell, F. E. Guernsey, Augustus Hemenway, Henry Horn- 1 Pepper, 1902. The existence of the Basket-makers was first pointed out in print by Dr. T. Mitchell Prudden in An Elder Brother to the Cliff-dwellers (Prudden, 1897). 2 Kidder-Guemsey, 1919. IV INTRODUCTION blower, J. M. Longyear, D. L. Pickman, and John E. Thayer. It wishes also to tender its thanks to Professor Byron Cummings of the University of Arizona, who unselfishly shared with it the field in which he was the pioneer; to Clayton Wetherill for his enthusiastic and faithful services as guide and interpreter; and to Mr. and Mrs. John Wetherill and Mr. Clyde Colville of Kayenta for their unfailing hospitality and constant helpfulness. In the two seasons covered by this report, the party outfitted at Farmington, New Mexico, and proceeded by wagon and horse¬ back to the trading post of Wetherill and Colville at Kayenta, the base from which further explorations were conducted. Kayenta, which may be found on the more recent Government maps, is reached from Farmington by a journey of four to five days, depending on the condition of the stock, and the abundance of grass and water. The caves and ruins described all lie in Arizona within a radius of one day’s ride from Kayenta. The country exerts a charm which the authors confess their in¬ ability to describe. Its physical aspect has already been noted by more competent writers; 1 it is sufficient for the purpose of this paper to say, that although essentially a semi-desert region, there is no difficulty now, nor was there ever, apparently, in earlier times for the dweller here who understood the environment, to obtain sufficient sustenance for simple requirements. The wastes of the valleys and mesa tops that once supplied the wild game with which the early people supplemented the fruits of their agriculture, now furnish ample grazing grounds for the Navajo’s flocks of sheep and goats; these Indians also succeed on selected sites in producing good crops of corn, under conditions that to a white farmer would seem quite impossible. Cambridge, Massachusetts March 5, 1921 1 Prudden, 1903, pp. 282-285; and 1907; Gregory, 1916, pp. 45-67. CONTENTS FIELD WORK, SEASON OF 1916 The South Comb page Sunflower Cave Revisited. 3 Goat Cave. 7 White Dog Cave. 10 Kiva. 22 FIELD WORK, SEASON OF 1917 Sayodneechee Canyon Cave 3 28 Cave 4 29 Cave 5 29 South Comb Revisited Cave 6 30 Cave 7 33 Sagiotsosi Canyon Cave 8 34 Cave 9 35 Cave 10. 35 Cave 11. 36 Cave 12. 37 Cave 13.... 38 Cave 14. 38 MATERIAL CULTURE Food Vegetal Food. 41 Animal Food. 44 Dress and Personal Ornaments Body Clothing. 45 Sandals. 47 Necklaces. 47 V VI CONTENTS Beads. 48 Pendants . 49 Feathered Pendant. 49 Ornament of Mountain-sheep Horn. 50 Deer-hoof Rattles . 50 Unfinished Ornament . 50 Tablet. 50 Head Ornaments. 51 Hair-dressing . 52 Cradles and Accessories Rigid Cradles .. . 54 Flexible Cradles . 58 Umbilical Pads. 58 Basketry Coiled Basketry . 59 Twilled Basketry. 63 Textiles Plain Weaving. 63 Twined Weaving. 65 Narrow Fabrics. 75 Netting and Cordage Coiled getting. 77 Rabbit Net . 77 Snares. 79 Objects of Wood Atlatl or Spear-thrower. SO Darts. 83 Foreshafts. 85 Dart Points. 87 Atlatl Stones. 87 Grooved Clubs. 88 Planting Sticks. 89 Scoop-like Objects . 90 Curved Wooden Tools. 91 Other Objects of Wood.. . . 92 Objects of Stone Manos. 93 Metate. 93 Chipped Knife Blades. 93 CONTENTS vii Hafted Knife. 94 Pipe Drill. 95 Graver. 95 Flaking Tool. 96 Flaking Stone. 97 Objects of Clay, Bone, Etc. Pottery. 98 Bone Objects. 98 Dressed Skin. 99 Sinew. 99 Feathers . 99 Ceremonial Objects Ceremonial Whip.100 Problematical Objects.100 Ceremonial Wand.101 Ceremonial Bundle.102 Ceremonial Bone Objects.103 Medicine Pouches of Skin Bag and Contents.108 Bag with Colored Minerals.108 Dice Bags.108 Sack with Beads and Feathers.108 Pouch and Small Articles.109 Summary and Conclusions Summary of Material Culture.109 Conclusions.113 Bibliography.119 BASKET-MAKER CAVES OF NORTH¬ EASTERN ARIZONA REPORT ON THE EXPLORATIONS OF 1916-17 FIELD WORK, SEASON OF 1916 The plans of the 1916 expedition included the investigation of a Cliff-dweller ruin discovered the previous year on the west bank of the Chinlee, one day’s journey east of Kayenta. 1 A week was spent here. After reprovisioning at Kayenta, camp was made near the mouth of Yellow Head Canyon, about 10 miles to the west, where two days were occupied in examining a small cave and in studying cliff-dwellings that had been cleared by Professor Cummings in 1914. Sunflower Cave (see map, figure 1) a site left unfinished in 1915, was then visited with the object of further investigations. 2 The remainder of the season was occupied in exploring the South Comb and in excavating two caves some 5 miles north of Sunflower Cave. THE SOUTH COMB 3 The South Comb is a great sandstone monocline that extends from Marsh Pass in a generally northeastern direction as far as the San Juan River. About 16 miles from Marsh Pass its con¬ tinuity is broken by a narrow valley which leads through it from Kayenta to the Agathla Rock. Our work was confined to that section lying between the break and Marsh Pass. Hereabouts the course of the Comb is sinuous and its appear¬ ance constantly changing; some stretches are tilted steeply toward the sheer walls of Skeleton Mesa, whose top at those points rises higher than the jagged summit of the Comb itself, which is shown in plate 1, b. Other stretches show gentle inclines that seem to lead to the Mesa, but on reaching the crests the way is invariably blocked by deep intervening chasms. It is hard to imagine more i To be described in a separate article. 1 For the location of this and other sites, see map, figure 1. 3 For the geology of the region, see Gregory, 1916, p. 47. 2 BASKET-MAKER CAVES rugged rock formations than those to be found in this part of the Comb. Frequently, and with little strain on the imagination, one can make out along its crests weird forms in natural sculpture: the outlines of colossal animals, faces, solitary spires and minarets, whose silent grandeur at nightfall intensifies the brooding gloom of the desert . In the walls of the tortuous gorges that wind up among the cliffs are countless caves, large and small, many of them so well hidden among the contorted rocks that they can be found only by working one’s way on foot along the ledges. Before exploring for new sites, the expedition occupied itself with two caves found in the Comb during the previous year. OF NORTHEASTERN ARIZONA 3 Sunflower Cave Revisited. While work at this site was still in progress in 1915, a sudden flood in Laguna Creek cut off com¬ munication between the camp, which lay on the east bank, and the ruin. As time was very limited, it was thought best not to wait the several days that it would probably take for the water to subside; and the party moved on, leaving a section at the rear of the cave unexplored. Sunflower Cave was occupied by a small cliff-house in which was found the remarkable cache of ceremonial objects that gave the place its name. 1 Of even greater interest, however, was the presence of certain remains which led us to suspect that in this cave might be found evidence as to the relative age of the Basket-maker and Cliff-dweller cultures. Cist 4, sunk into the hard-pan behind the cliff-house rooms, had given the most positive indications of this; it is described as follows in the previous report (p. 96): The outlines of this cist could be traced by a disturbed area showing in the face of the trench. It had originally been a stone enclosure, though but two of the slabs were still in place. A few bones of a child were found in the upper part; near the bottom at the side nearest the back of the cave were two decorated bone tubes. Imprints of coiled basketry could be seen in hard lumps of the adobe filling, but nothing of the basket itself remained. The cist gave us the impression that it had been a Basket-maker burial chamber which had been pulled to pieces, partly emptied and then filled in with rubbish dur¬ ing the cliff-house period. There was also found in the loose rubbish a typical Basket- maker sandal, the presence of which, in what was a purely cliff- house site to all outward appearance, required some explanation. We were accordingly very anxious to examine the still undug portions at the rear of the cave. The results of the second visit amply repaid the effort, for we discovered unmistakable strati¬ graphic evidence of a sequence of occupation. The new excavations revealed Basket-maker burials, some of them entirely undisturbed, below a stratum of typical Cliff-dweller debris. The location of the finds is shown on the plan (figure 2); their relation to the Cliff- dweller remains is clearly brought out in the diagrammatic cross- section (figure 3). Cist 5 (cists 1 to 4 opened in 1915) was a shallow bowl-shaped hole dug in the hard-pan. In it were parts of the skeletons of a young i For a general description of this cave and of the finds made there in 1915, see Kidder- Guernsey, 1919, pp. 92-96. 4 BASKET-MAKER CAVES child and an adult, while scattered through the loose dirt about the top were portions of the skeleton of a second child, which had prob¬ ably originally been deposited with the other remains. The bones Figpbe 2 Plan of Sunflower Cave, South Comb. of the adult had been carefully disposed at the bottom of the hole, in a manner to make the most of the limited space. They con¬ sisted of an undeformed skull in good preservation, the long bones of the arms, the scapulae, and a few ribs and vertebrae. The arm bones were placed on either side of the skull, the other bones OF NORTHEASTERN ARIZONA 5 being packed close about it. Lying across the arm bones was a section of a femur which showed a long splintered post-mortem break. The lower jaw was found in the loose rubbish some fifteen inches from the edge of the cist. It had probably been dragged out by rats, a thing we found to be not uncommon in caves. A small white chipped point lay among the bones. Above these remains was the disarranged skele¬ ton of the young child. The second child’s skeleton as before stated, was scattered through the loose earth about the cist. We Figure 3 Sunflower Cave, Cross-section. are at a loss to account for the neat arrangement of the adult bones. It is clearly a case of secondary burial, but we have never found any instance of this practice in undisturbed Basket-maker sites, and the people who looted Basket-maker graves did not, as far as we are aware, ever trouble themselves to restore anything to place. Cist 6 was 2 feet 6 inches in diameter and was cut 3 feet deep into the hard-pan. It lay 4 feet east of Cist 5, and contained only a quantity of loose cedar bark and shredded grass piled in the bottom. It is possible that the bones found in Cist 5 came from here, though no positive evidence remained that it had been used for burial. Cist 7 was an untouched Basket-maker grave; the original filling passed unbroken above it, and was in turn overlaid by Cliff-dweller rubbish (figure 3). It was 4 feet in diameter, 3 feet deep, and held the well-preserved skeletons of two adults with undeformed crania. They lay flexed on their left sides, hands between the lower thighs (plate 10, c); over the head of each was inverted a small coiled basket, one of which can be seen in the photograph. The 6 BASKET-MAKER CAVES earth about the skeletons showed traces of decayed organic matter, probably from fur-string robes and other wrappings; rotted cedar bark was found at the bottom. The only object besides the de¬ composed baskets was a small strip of bark with one end neatly trimmed off. Cists 8, 9 and 10 had all been plundered in early times and con¬ tained only fragmentary skeletons; a number of cylindrical seed beads accompanied the remains of a child in Cist 10. Cists 11 and 12 were within 3 feet of the rear wall of the cave. Although very close under the surface they had not been molested. Cist 11 was a shallow bowl-shaped scoop in the hard-pan, and held two infants. One of these had been wrapped in a fur-string blanket and lay on what seemed to be a twined-woven cedar-bark mat, be¬ neath which was a reed-backed cradle too badly rotted to preserve. Infant 2 was also wrapped in a fur-string blanket and lay on a de¬ cayed reed-backed cradle; near the head were remains of a coiled basket inverted over traces of a substance resembling meal. Both cradles were of the rigid type shown in plate 20. Accompanying the bodies were two bark objects covered with prairie-dog skin, which we have since been able to identify as umbilical pads. Cist 12 was a small hole in the hard-pan. In it was an infant wrapped in a fur-string robe and encased in a twined-woven bag. The robe had been destroyed by insects, but the bag was in a fair state of preservation. All the above Basket-maker cists lay below a layer of cliff- house rubbish from 6 to 8 inches deep, made up of ashes, turkey droppings, bits of straw and many potsherds of the same wares as those found on other cliff-house sites in this region. Beneath this rubbish, the surface of the hard-pan above the cists gave no in¬ dication of their presence, being as compact and of the same ap¬ pearance as the surrounding hard-pan. If, therefore, we had fol¬ lowed the 1915 method of clearing and examining the Cliff-dweller rubbish down to the hard-pan, and not cutting into it except where the tops of cists were encountered or other surface indications excited interest, these burials would have escaped notice altogether. Fortunately, however, the trench was run much deeper than usual and entered Cist 7 from the side. The section thus exposed showed the top to be filled to a depth of 1 foot with a compactly tamped mass exactly like the hard-pan in which the cist itself was Peabody Museum Papers Vol. VIII, No. 2, Plate 1 South Comb a, White Dog Cliff and Navajo Hogan; b, South Comb, near White Dog Cave. OF NORTHEASTERN ARIZONA 7 excavated (figure 3). That the infant burials in Cists 11 and 12 remained undiscovered through the period of Cliff-dweller occu¬ pancy is remarkable, since they were covered by hardly more than 3 inches of the cave earth; the Cliff-dweller rubbish here was also very thin. A possible explanation may be that this part of the cave was used by the Cliff-dwellers for storage or for sleeping places, and was thus in a measure protected from the random dig¬ ging to which the more open portions were exposed. Had the Cliff-dwellers, the final tenants of the cave, been more persistent in their search, there would have remained no trace of the Basket-maker period except the cists, empty or refilled with Cliff-dweller rubbish. Attention is called to this for the sake of emphasis, as further on in this report, caves are described where all evidence of Basket-maker occupancy other than the empty cists has been effaced. Goat Cave. This site was located by the expedition of 1915. It lies about two miles north of Sunflower Cave at the foot of a steep incline leading to the top of the Comb (see figure 1). The approach is through a narrow ravine choked with great rocks, among which a thick growth of large old cedars has found root. These trees screen the place from view except at a few points in the ravine. The cave is a deep shelter at the west end of which is an even deeper recess. As shown in the plan (figure 4) there are two levels: a front or lower one, extending the entire length of the cave; and a higher rear level, consisting of the whole floor of the inner recess and of a narrow gallery running all along the back of the more open part of the cave. The whole upper level is formed of the original hard-pan fill; along the gallery or terrace this breaks away in a vertical bank. The walls and roof of the cave are much blackened by smoke. At one point in the rear of the cave the floor is covered by a thick layer of ashes and charcoal. In the recess and on the end of the gallery next to it, are a number of partly fallen walls (plate 2, a, b). Room 1, five feet in diameter, the walls 2 feet 4 inches high, is built of upright slabs of stone. Room 2, from the foundations that remain, appears to have been oval in shape. From front to back it measured 8 feet, its length could not be determined as the end wall had disappeared. The foundation is of thick stone slabs of uniform size set on end, on 8 BASKET-MAKER CAVES these small stones were laid flat (plate 2, a), but little of the upper course remained in place. Joints between the foundation slabs were closed with adobe mortar. The upper courses appear to have been chinked with the same material. Back of this room are re¬ mains of two curved walls built of coursed masonry in the usual Cliff-dweller manner. Stone apparently from these walls w T as used to construct a small cairn on the opposite side of the recess. It Figure 4 Plan and Cross-section of Goat Cave, South Comb. resembles monuments built by the Navajo to mark water or trails; nothing was found beneath it. Directly in front of the cairn is a heap of rocks fallen from the roof of the cave. On the lower level in front of the gallery are two roughly circular rooms which we at first wrongly thought to be Cliff-dweller kivas, but they were found to contain none of the special features of ceremonial rooms. Both were built against the steep bank of the terrace which had been cut away to form their rear walls. Room 3, the less well-preserved of the two, measured 15 feet across its greatest diameter; the wall stood 4 feet at its highest Peabody Museum Papers Vol. VIII, No. 2, Plate 2 Goat Cave a, Slab foundation of Room 2; b, General view. Room 3 in foreground. OF NORTHEASTERN ARIZONA 9 point. The masonry is interesting and unusual; medium-sized flat stones are laid up without any mortar in such a way as to produce an even surface on the interior (plate 2, b), the exterior being left irregular and rough. So carefully are the stones placed that in spite of the absence of mortar the construction is firm and solid. In clearing this room a slab cist was uncovered, measuring 4 feet in diameter at the top, 3 feet at the bottom, and 2 feet deep; in the bottom was a 2-inch layer of ashes and charcoal and over this 2 inches of cedar bark. It was very similar to Basket-maker slab cists found in Cave 1, 1915. 1 The original floor of Room 3 was so ill-defined that we could not determine exactly the relation of the cist to the floor, but as near as could be judged the upright slabs had been sunk into it a depth of about 8 inches. Cached in the loose filling of the room, at the point indicated in the plan (figure 4; note also its position in the cross-section) was a black corrugated olla. It was covered with a thin flat stone, but contained only drift sand. Room 4- The general shape of this room is shown in the plan. Its greatest diameter, measured inside, is fourteen feet, from back to front eleven feet. The highest point in the wall, five feet, is probably the original height, as no loose building stones were noticed here. No trace of roofing remains. The masonry wall has no sharp corners. The back wall is cut in the face of the gallery and has a slight bend or angle. The stones are laid to produce a smooth face on the inside as in Room 3, and with considerable skill, since they are still firmly in place though there is no trace of adobe mortar in the joints. In excavating the room we found quantities of charcoal and scattered bundles of cedar bark, but no artifacts. Two rude cists lined with cedar bark were also opened. As in Room 3 the floor was not well-defined. In the floor of the gallery were several jar-shaped cists dug in the hard-pan (see figure 4). These were exactly like the burial cists found in the Sayodneechee burial cave, 1914. 2 At a point back of Room 3 where the terrace wall had caved off carrying with it one half a cist (see section in figure 4) the exposed cross-section showed plainly the marks of digging sticks in the side of the cist thus brought to view. Two of the cists contained a few human bones; 1 Kidder-Guernsey, 1919, p. 77 and plate 27. 2 Ibid., p. 28 and figure 8. 10 BASKET-MAKER CAVES while other portions of skeletons, some bleached by long exposure, were found in the loose sand covering the floor of the terrace. These were, no doubt, plundered Basket-maker burials. The authors wish to call particular attention to the rooms un¬ covered in this cave. Their masonry, with the exception of the single wall in the recess, is quite different from that of the cliff- dwellings. White Dog Cave. This was by far the most prolific site dis¬ covered by the Museum’s expeditions to northeastern Arizona. Its position is most inconspicuous and the first view of it was ob¬ tained during a climb high up among the rocks of the Comb, the only place in fact, from which it could be seen from any dis¬ tance. It might easily have escaped notice altogether, for a rider passing along the valley below would not be tempted to explore the narrow ravine leading up to it, particularly as the cliff in which it is located is apparently in full view and seems to be entirely un¬ broken (see plate 1, a). One short section of the cliff is, however, out of sight from the flat land, and just there is tucked away the cave. The above conditions are described thus at length in order to show tiie absolute necessity of a careful search on foot among all the little side canyons of this broken country. The approach is up a tortuous ravine. Arriving below it the visitor is astonished that so great a cavern should be so effectively hidden. It occupies a commanding position in the rounded front of a buttress-like swell of the cliff. The huge portal, 120 feet across the base and at least 125 feet high, seems carved by nature to conform to the dome-shaped top of the cliff above it. The accom¬ panying photograph (plate 3), aside from having in it no familiar objects by which relative proportions may be judged, shows so clearly the process of formation and general aspect that further description is unnecessary. Reaching the cave after a stiff climb of 100 feet up a steep talus, one enters a spacious chamber measuring approximately 70 feet from back wall to line of shelter and 120 feet across the opening. The ceiling is high and arched, the floor rises at an easy grade from front to back. Somewhat more than half the floor space is covered by large rocks fallen from the roof, one of which measures 20 feet in length, 12 feet in width and 10 feet thick (figure 5 and plate 11, a). This and other rocks near it we found later had fallen since the Peabody Museum Papers Vol. VIII, No. 2, Plate 3 White Dog Cave. OF NORTHEASTERN ARIZONA 11 cave was occupied. The unencumbered portion of the floor was composed of clean sand and small broken stones. Although we subsequently unearthed considerable accumulations of ashes and charcoal in different parts of the cave, the walls and ceilings showed not a trace of soot, having been scoured clean by wind¬ blown sand. A demonstration of this process was furnished one day when a high wind from the proper quarter created a veritable whirlwind in the cave, gathering up the surface sand and swirling it about in such quantities that we were forced to abandon work while it continued. A piece of paper released at the back would sometimes make as many as three complete circuits of the cave clinging close to the wall except as it passed across the front. On mentioning this to Mrs. Wetherill we were told by her that the place was known to the Navajo as the Cave of Winds. The first examination of the cave for traces of occupation showed at the back against the wall the tops of several sand-filled cists, dug in the hard-pan. Searching the surface, a few bleached 12 BASKET-MAKER CAVES human bones were seen and a small handful of Cliff-dweller pot¬ sherds was picked up. Digging at random with a trowel, a few fragments of basketry and some bone beads were found. Near the center of the cave the ends of two upright stakes were noticed, projecting from 2 to 3 inches above the surface. Not until our second and more thorough examination did we discover on the west side a low foundation wall mudded on to the sloping rock floor of the cave. This was apparently the beginning of a small Cliff- dweller storage room or bin. As a “prospect” the cave fulfilled every requirement. Its exploration yielded a collection which fully represents most phases of the material culture of the Basket- makers. Across the front of the cave where work was commenced there was found a natural ridge of coarse debris, back of which the sand fill had accumulated above the hard-pan floor to a depth of from 5 to 7 feet. Toward the back this deposit grew shallower until along the rear wall the hard-pan cropped to the surface. The fill carried no refuse pockets or well-defined rubbish layers such as are found marking floor levels in Cliff-dweller caves. In general it was made up of a surface layer G inches to 1 foot deep of drift sand, below which it was composed of sand and bits of stone mixed with straw, pieces of bark, and particles of charcoal. Occasionally there appeared thin strata of coarse charcoal and in certain areas there were encountered quite extensive accumulations of ashes and charcoal. In the general digging a number of speci¬ mens were found at various depths. They consisted mainly of basket sherds, fragments of fur-string blankets and tattered bits of woven bags; a mummified foot and other fragments of human remains were also recovered. All other objects were taken from cists. In the plan, figure 5, are indicated a large number of cists grouped along the east wall; there were no cists on the west and north sides. The majority of these were jar-shaped excavations in the hard-pan ranging in size from small pot-holes 1 foot in diameter and of about the same depth, to examples 5 feet deep and 4 feet G inches in diameter. Some burials were found in this type of cist but for the most part they were empty, save for sand or sometimes cedar bark and grass at the bottom. Most of the burials were in the front half of the cist area. A few, as was just stated, were in Peabody Museum Papers Vol. VIII, No. 2, Plate 4 White Dog Cave g, Cradle bundle as found. The other figures show cradle and contents unwrapped, a, Woven cloth; b, f, Fur cloth blankets; c, Mummy of child; d, Umbilical pad; e, Absorbent bark; i, Cradle. (About 1/12.) OF NORTHEASTERN ARIZONA 13 cists completely excavated in the hard-pan, others were in shallow excavations in the hard-pan with one or two stone slabs so placed as to hold back the loose sand, and a single burial was in a cist (51) of the stone slab type described in the previous report. 1 Some of the burials had been previously disturbed, 2 but a number were found intact, the remains and mortuary offerings in a remarkably fine state of preservation. In the account of the excavations which follows, certain cists and burials are described in detail. The intention is to present the salient features of the more typical ones, hence many small objects found in the cists or concealed among the wrappings of the mum¬ mies are not enumerated. They are, however, described in detail in another section. Cist 6 (figure 6, a). The first burial cist to be encountered measured 3 feet in diameter, 2 feet in depth and was 4 feet below the surface. It represents a type that was evidently constructed primarily for sepulchre. At one side was an upright stone slab. Although the cist had been relieved of a good portion of its con¬ tents by ancient diggers we obtained from it a collection which required 51 catalogue numbers to record. In the upper part were the scattered bones of three infants; at the bottom a few bones from the skeleton of an adult. In the loose fill were several bunches of human hair (plate 32, c, d). A quantity of human hair evidently from the head of a mummy 3 that had been pulled from the cist was also found in the loose fill. One small strand was wrapped about with a leather thong. Later we found in another cist a mummy with coiffure intact, having a queue-like strand wrapped in the same manner. These were practically all the human remains that were left. At the bottom against one side were a quantity of pinon nuts, the rotted remains of woven bags, loose beads, basket sherds, pieces of woven bags and fur-string robes. Cist 13, a shallow bowl-shaped excavation, contained the re¬ mains of two infants. One, a very young child wrapped in two fur-string blankets and a fragment of woven cloth, was lashed 1 Kidder-Guernsey, 1919, p. 77 and plate 27. 2 This grave looting so commonly found in Basket-maker cave cemeteries is not modern. Although we have no direct evidence in its support, our theory is that it was the work of the Cliff-dwellers. See Kidder-Guernsey, 1919, p. 84. 3 The mummies were, of course, not artificially preserved in any way; they are merely desiccated bodies. 14 BASKET-MAKER CAVES tightly to a small reed-backecl cradle; an umbilical pad was in place and the dried umbilical cord was tied to one of the blankets. This mummy bundle as found, and also unwrapped so that all its parts can be seen, is shown in plate 4. The second body, that of a child about 4 years of age, was completely encased in a woven bag 1 (plate 30, f). It was also shrouded in a fur-string robe. Beneath this bundle were pieces of a cedar-bark mat, and over it was spread a fur-string inverted tray basket. At one side of the cist was a bowl-shaped basket also inverted. In the fill some 8 inches above the tray basket was a skin bag containing shelled corn (plate 15). At one side of the cist lay an atlatl in perfect condition save that before being placed in the cist it had been bent nearly double. This and the baskets are illustrated in situ in plate 10, e. Cist 22 contained the bodies of three individuals. Its shape was roughly circular, the greatest diameter being 5 feet 2 inches, depth 2 feet 10 inches; the top was 5 feet 6 inches below the surface. Each body occupied a shallow depression scooped out of the bottom of the cist as shown in figure 6, b. The remains were partly mum¬ mified though not in a good state of preservation. The heads, i The design on this bag is shown in color in plate 28. OF NORTHEASTERN ARIZONA 15 however, retained their hair and much of the dried tissue of the face. Each body had been wrapped in a fur-string blanket and sewed up in woven bags, all of which were in an advanced state of decay. Number 1, the body of a young female, lay on its right side, knees drawn up and hands between the thighs. A skein-like rope of human hair was wound around the left forearm, passed between the thighs and made fast about the right leg below the knee. At the waist were fragments of a string apron. Some portions of bags that had been used to cover the body remained. A fragment at the feet was of very fine weave while pieces adhering to the knees were much coarser. Covering the whole were two tray baskets. Number 2 was a female. Three baskets were used to cover the body. It rested on its back with head and legs inclined to the left; the feet were drawn up close to the body; the upper legs, bent at the hips, were at right angles to the torso. The hands were in front of the lap, and were bound together at the wrists by fourteen turns of a tightly twisted cord of human hair. This cord was then knotted to a skein-like rope of human hair and both rope and cord passed through between the thighs and about the lower legs above the ankles. At the waist were remains of a string apron and on the breast lay a disk-shaped pendant of shell, ornamented with incised lines. About the neck were beads of olivella shells and thin disk-beads cut from shell, together with part of the leather string by which they had been suspended. In the bottom of the cist under the body were a number of dice-like stones and a single corn cob. Number 3 (male, 20 to 25 years of age) rested on its left side, limbs loosely flexed, hands between thighs. Two tray baskets covered the body. At the right side lay a grooved club, at the feet were a pair of badly rotted square-toed sandals with leather tie- strings and a quantity of small deer or antelope hoofs. Near the hoofs were two handle-like bone objects with small stones attached to their ends. About the neck was a string of shell beads. Among the objects found under the body was a fine chipped knife blade (plate 35, k) and its shrunken wooden haft. Cist held the mummies of two adults, one male and one female, each accompanied by the remains of a dog, and an unusual number of mortuary offerings. The remarkably fine state of preservation of everything in this cist is due to the fact that the burials were surrounded by dry sand. The excavation in the hard-pan made 16 BASKET-MAKER CAVES to receive the bodies was a shallow hole just deep enough to hold them. As in Cist 22, each individual occupied a scooped-out place in the bottom of the cist. At the back was an upright stone slab; as none were used at the front or sides, its purpose was evidently to hold back the loose sand while the hard-pan was being excavated. Just in front of the slab was a stout log 3 feet in length, the ends and sides charred by fire. This reached to the surface and was one of the stakes observed when the cave was entered (see upper right center, plate 6, a); whether or not it was so placed at the time the burials were made we were unable to tell. It may have been a marker, but we have found no other burials indicated in this way. Mummy 1 (female) lay on its right side, limbs loosely flexed. Two large woven bags split down the side encased the remains, one drawn over the head, the other over the feet; the tops met at the middle of the body and were sewn together with yucca leaves (plate 7, a). As usual the corpse was wrapped in a fur- string robe. Over it were inverted two baskets, a bowl-shaped one covering the feet; the other a large carrying basket with tump¬ line attached covering the head and upper part of the body. The baskets and the manner in which a number of digging sticks were disposed in the grave is shown in plate 6, a. The planting stick at the front with one end resting on the edge of the cist was evi¬ dently placed to hold the basket upright. The cedar bark that appears in the upper left hand corner is from another cist. On re¬ moving the carrying basket, a small dog was found lying below it on the left side of the mummy. Under the bowl-shaped basket was a substance resembling meal. On lifting the body from the cist there was found beneath it a thick bed of fur and feathers compacted by decay into a mass that was taken out unbroken. On examination at the Museum this proved to have embedded in it bundles of feathers, skin containers and skin bags; these and their contents are described under Material Culture. On the bottom of the cist was a badly shrunken, but complete atlatl and near it, but not in contact with it, was a roughly chipped piece of quartzite which may originally have been tied to its back. At one side on the bottom was a wand with a yucca braid and twigs attached to one end. Quantities of grass seed, pinon nuts and squash seeds were also found at the bottom of the cist. Peabody Museum Papers Vol. VIII, No. 2, Plate 5 a, Cradle in situ. Cist 54, White Dog Cave; b, Cave 10, Sagiotsosi Canyon. OF NORTHEASTERN ARIZONA 17 Mummy 2 (male about 35 years of age) lay on its left side with feet drawn up tight against the body; head east and facing south. It was wrapped in the same manner as mummy 1 (see plate 8). Inverted over the body was a large pannier basket which is shown behind the front basket in the photograph (plate 6, a); over the head was a bowl-shaped basket. A second basket of the same shape lay just to one side, covering the fragments of a squash shell vessel. Removing the pannier, three tray-shaped baskets gradu¬ ated in size with the smallest at the bottom were found beneath. The pannier also partly covered the remains of a large long haired and nearly white dog, which in turn lay across the two bowl-shaped baskets (see plate 6, b). There was also found under the pannier a large quantity of flies, the dog having apparently been already fly blown when placed in the cist. The eggs evidently hatched and the flies died in the space under the carrying basket without ever seeing the light of day. We thought that the flies might serve to fix the time of year in which these burials were made, but Mr. N. Banks of the Museum of Comparative Zoology, to whom we are indebted for their identification, informs us that they are Caliphora coloradensis, a very hardy species which flourishes from early spring to late fall, so it is not possible to fix a very definite date by them. The digging sticks might indicate that the spring planting was in progress, but this is of course mere conjecture. Extending from the edge of this cist on the east side was a shallow hole just deep enough and of sufficient size to contain the remains of a young infant. Only the bones, and part of a badly rotted fur-string robe were left. Cist 27. The unusual plan of this cist is shown in figure 5. It was dug in hard-pan to a depth of 2 feet 10 inches, measured 4 feet 9 inches in length and 2 feet 6 inches at its widest point. The sand and fill above had a depth of about 1 foot. One side of the cist was formed by the face of a vertical break in the rock floor of the cave, the ledge nearly cropping through the hard-pan at this point, a circumstance which probably accounts for the elongated shape, as the rounded end seems to indicate an original intention to dig the conventional circular cist. In it were found the partly mummified bodies of two adults placed one above the other, facing in opposite directions (plate 9, a). 18 BASKET-MAKER CAVES Number 1, the uppermost, an adult, probably male, lay face clown, knees drawn up and crushed against the chest, feet under hips, left arm extended at full length along the side; the right forearm was bent across the waist. Number 2, a male of about 25 years, lay on the bottom of the cist directly under mummy 1 and with head in the opposite direction. The limbs were arranged in practically the same manner as those of the upper mummy, the feet of which rested on the face of this one. Accompanying these remains was a large number of specimens some in a good state of preservation, though objects at the bot¬ tom of the cist and baskets at the top and sides had suffered from decay. We were, however, able to determine that there had been at least seven baskets, mostly medium sized trays. In preparing the cist to receive the bodies, a number of atlatl spear-shafts had been broken into various lengths and placed crisscross on the bot¬ tom. On the upper side a few inches out from the rock there stood on edge a rectangular frame of sunflower stalks and broken atlatl spear-shafts tied at right angles to each other. Back of it, also on edge, were placed several tray baskets. On the opposite side next to mummy 2 were bundles of sticks or reeds so badly shrunken that their nature could not be made out with certainty; they were probably atlatl spear-shafts. Placed over mummy 2 were more spear-shafts and the bundled fragments of a wooden device, part of which is figured in plate 30, d, e. At one side of mummy 1 were two grooved clubs. Quantities of grass and squash seed were found in much decayed skin containers; also a number of small objects, among them a fine chipped knife blade, beads of seed and stone, pendants of shell and stone, a comb-like head-ornament and a bone handle with leather strings attached. Cist 80 was a jar-shaped excavation in the hard-pan, 15 inches in diameter at the top, 23 inches in diameter, 1 foot below the rim, and 24 inches in depth. In it were the skeletons of six infants. Four were found in woven bags. Of other wrappings there re¬ mained tattered pieces of dressed skin and bits of fur-string. Five umbilical pads, similar to those from Cist 11, Sunflower Cave, were taken from various parts of this cist. These could not be assigned to individual burials as the skeletons were more or less mixed as if the cist had been partly rifled in early times. At the bottom were two cradles in excellent condition. A few inches Peabody Museum Papers Vol. VIII, No. 2, Plate 6 Cist 24, partly cleared, showing baskets in situ; b, Pannier baskets removed, showing small baskets, mummy of white dog, and many dead flies. OF NORTHEASTERN ARIZONA 19 above these were about 8 quarts of shelled corn; no trace of a container could be found. Scattered through the fill were beads of seed, stone, and olivella shell, a green stone pendant, a small grinding stone, and two strips of bark, like the piece found in Cist 7, Sunflower Cave. Joined to this cist by a small funnel-like hole was a second cist, the same diameter but not so deep, while cutting the rim of this was a third and larger one (Cist 33, figure 5). These were empty; they form a good example of a number of similar arrangements found in the course of the excavation (see Cist 52, figure 5). All are characterized by one or more small flue¬ like holes dug down from the surface and penetrating the sides of the cist, or, as in the case above noted, connecting small potholes to the cist (plate 9, d, and plate 14, a). Sometimes these holes, instead of entering the large cist obliquely, were dug at nearly right angles from the pothole to the side of the larger cist. As a rule cists of this type were empty save for bark or grass stalks. They strikingly resemble the field pit-ovens used by the Hopi for roasting corn; 1 there are no indications, however, that these had ever had fires built in them. Cist 31 as shown in the plan, figure 5, was partly under one end of a large rock. In order to reach it we were obliged to remove from the surface many others, some so large that they had first to be broken up. The top of the cist was 3 feet 6 inches below the surface, its greatest diameter 4 feet, depth 1 foot 10 inches. At one side was a single stone slab. In the cist was the partly mummi¬ fied body of an adult, the bones of the skeleton held together by dried tissue and caked adobe (plate 7, b). The remains rested on the left side, knees drawn up level with chin, hands palms to¬ gether under left cheek and supporting head. A woven bag cov¬ ered the head and shoulders. It had been split down the side before drawing on, then sewn together again with yucca leaves. A portion of the bag was in good condition. Over the mouth of the mummy outside the bag, was tied a sandal of the square-toed type. About the feet and lower part of the body were the remains of a fur-string blanket. The bag and wrappings were held in place by a binding of yucca leaves. About the neck were seed beads. Inverted over the middle of the body was a coarse bowl¬ shaped basket; under it lay a quantity of plant stalks, apparently 1 Hough, 1919, figure 3. 20 BASKET-MAKER CAVES of Brigham tea, also an animal bone and a pointed twig with a string attached. In the lap, as shown in plate 7, b, was a bundle made up of two wooden implements, a foreshaft with stone point, a wand-like stick with a bunch of reddish fiber tied to the end, and a small woven object, the whole wrapped about with a feather headdress and a number of turns of fine string (plate 40). The fill about the body was caked and discolored. Nothing was found in the cist under the body. Cist 32 gave indications of previous disturbance. It was dug in the hard-pan against the side of the cave and showed more than usual care in the smoothing of its walls. It was oval in shape, 3 feet 6 inches in length, 2 feet in width, and 2 feet 6 inches deep. In the edge of the end opposite the cave wall was a shallow groove perhaps made to seat a cover. In the upper part of the cist was the skeleton of an infant and remains of a small reed-backed cradle, both too far gone to collect. In a sub-excavation at the bottom was the skeleton of a child about six years of age, knees drawn up to chin, head north, face southeast. About the remains were traces of fur-string wrappings and coiled basketry; under them a small quantity of green powder. This cist was probably originally a storage cist and perhaps had a stone slab cover which fitted into the groove at the end. It may have contained at one time other remains than those found, for it would hardly have been dug for them alone, as it was of much greater size than necessary. Cist 35 was not dug straight into the hard-pan, but was slightly undercut. It measured 1 foot 3 inches across at the top and 2 feet 6 inches in greatest diameter; the bottom was rounded. In it was the mummy of a baby on a reed-backed cradle; the body was enclosed in a bag and lay on a twined-woven cedar-bark mat (plate 21, d). All were in good condition. The mat appears to be part of an old cedar-bark cradle like the ones found in Caves 1 and 2 by the 1915 expedition. 1 Cist 40 was a large jar-shaped storage cist excavated in the hard- pan. It was very symmetrical in shape and measured 2 feet in diameter at the top, 4 feet in diameter 2 feet below the rim, and 4 feet 6 inches in depth (plate 9, b). The rim was 2 feet below the surface. In the top was found a rabbit net tied in a compact bundle, together with a quantity of apocynum bark done up in 1 Kidder-Guernsey, 1915, p. 165 and plate 72. Peabody Museum Papers Vol. VIII, No. 2, Plate 7 White Dog Cave t, Wrapped mummy of woman from Cist 24; b, Mummy of man from Cist 31. OF NORTHEASTERN ARIZONA 21 bundles. The net had evidently been cached here after the cist was abandoned and filled up, since the hole in which it rested was partly dug in the hard-pan at the edge of the cist, and partly in the fill of the cist itself. In clearing the cist a thick layer of cedar bark was found 1 foot from the bottom; below it was clean sand. One foot from the rim on the side opposite the net there was a pot¬ hole, 1 foot in diameter and the same in depth. The rabbit net, a remarkable specimen, is described in detail in another place. Its lack of definite relation to the cist or to other Basket-maker remains at first raised a doubt in the authors’ minds as to whether it might not have belonged to a later period. On the other hand it will be remembered that a very similar excava¬ tion at the side of Cist 24 contained the remains of a Basket- maker infant. Cist 4-1 gave evidence of previous disturbance. On clearing it a small niche was found in one side that contained the remains of an infant, a small basket, a skin covered object (umbilical pad) and the usual fur-string robe. Cist 51, 3 feet 6 inches deep, and 2 feet 6 inches in diameter, was constructed of slabs set about the sides of a shallow excavation in the hard-pan. It contained the skeletons of an adult and an infant. The former lay on its right side, head south. The infant rested across the breast and left arm of the adult and had been wrapped in a fur-string blanket and placed in a skin bag. Both blanket and bag were in an advanced state of decay. There were traces of a woven bag that had once covered the remains of the adult. At one side of the cist near the head of the adult was a small bowl- basket containing beads and a variety of small objects, which are described elsewhere. There were also in the cist food offerings of corn and pinon nuts. Cist 54- After removing from the surface a large number of rocks, the fill under the end of the great rock in the center of the cave was explored. Here, 2 feet below the under side of the rock in what appeared to be a rude cist, there was found a cradle in ex¬ cellent condition. With it were fragments of fur-string blankets and pieces of woven bags, but no trace of a body. The photo¬ graph, plate 5, a, shows the cradle in situ. The thin edge of the rock had been broken off somewhat before the picture was taken; it originally extended nearly a foot further than is shown. The 22 BASKET-MAKER CAVES rock may have broken from the roof centuries ago or in very recent times. The cradle, however, must have been in the position in which it was found when the fall occurred. Summing up the evidence as to mortuary customs contained in the foregoing descriptions, we see that the bodies were placed in cists of three sorts: jar-shaped excavations, whose primary pur¬ pose seems to have been for storage; larger, shallower pits appar¬ ently dug expressly to contain burials; and slab cists of the tj r pe illustrated in plate 9, c. Almost every cist held more than one individual and all the indications pointed to the interments having been made simultaneously. 1 The bodies of adults were always wrapped in fur-string blankets and at the loins of most females were small string aprons. The limbs were flexed to occupy the least possible space and oc¬ casionally held in that position by cords. The bundles thus pre¬ pared were encased in large woven bags, which were cut down one side for greater ease in drawing on, and then stitched together again with yucca leaves. Babies were sometimes placed in bags, but were more commonly buried on their cradles with their blankets, umbilical pads and “diapers” of bast in place as in life. No fixed manner of orienting the remains was adhered to, this detail having been decided, apparently, by the manner in which the body best accommodated itself to the shape and size of the cist. Mortuary offerings were numerous and varied and seem fairly representative of the food, implements, weapons and ornaments of daily life together with some objects of a ceremonial nature. The standard gift to the dead was basketry; tray baskets were practically always inverted over the heads of adults, often over children; large panniers also served as covers; and smaller baskets, empty or filled with trinkets, were generously piled into the graves. Kiva (?). There remains to describe a peculiar and puzzling room found at the front of the cave (see figure 5). The first inti¬ mation of its existence came when, in clearing the surface above what proved later to be the ventilator shaft, the wall of the main structure was exposed. The room, as shown in the plan, lies at the foot of the great rock pile which rises at a sharp angle to the 1 The same thing was noted in Sunflower Cave (Cists 7 and 11); in Cave 1 Kinboko (Kidder-Guernsey, 1919, p. S3) and in the Sayodneechee burial cave (Ibid. p. 29); at the latter site there were more individuals per grave than in any of the others, one cist holding no less than 19 bodies; all, apparently, buried at one time. Peabody Museum Papers Vol. VIII, No. 2, Plate 8 White Dog Cave Mummy of man, Cist 24: a, Wrapped; b, With coverings removed. OF NORTHEASTERN ARIZONA 23 back of the cave. It was owing to the imminent danger of rock slides from this source that we were unable to excavate the room completely, either in 1916 or on a second visit to the cave in 1917, when another attempt was made to do so. This chamber is, and apparently always was, entirely subter¬ ranean. The part that we were able to clear is irregularly circular. The room is sunk through the surface sand and into the hard-pan, which, standing as dug, forms the lower part of the wall (see figure 7, b). The upper wall is masonry of rough and irregular Figure 7 White Dog Cave: Plan and Cross-section of Kiva. stones laid with little attempt to preserve a smooth face either within or without. At one point on the east side two upright slabs were set in and the wall was built on them. The top courses are somewhat more carefully constructed. Adobe mortar is used, sparingly below, more abundantly above. The whole structure is thickly “spalled” with small fragments of stone wedged into the cracks. The general shape of the wall, partly straight, partly curved, can best be seen in the plan and section. The southern offset, which in the plan has the appearance of a bench or banquette, we are inclined to think was not a part of the original design of the build¬ ers, but was made necessary by the occurrence here of an outcrop¬ ping of the ledge, the upper surface of which slanted inward at too 24 BASKET-MAKER CAVES great an angle to furnish a stable foundation for a wall along the inner edge. At any rate, the offset overcame this difficulty, though for some reason, instead of continuing the wall as before, of laid-up stones, stone slabs set on end were used. We do not know whether or not this method of construction is continued under the rock pile. Placed across the top of the slabs was a stout log, one end resting on the top of the offset, the other passing out of sight under the rock heap. It is possible that the entrance to the room was at this point, as the sloping surface of the ledge here is very smooth as if from wear. South of the offset and outside the room we found slabs, set at right angles to the wall, and three up¬ right stakes burned off close to the adobe in which they were em¬ bedded. There was a large amount of charcoal in this area. The slabs of the offset wall and those outside were much blackened by smoke. On the east side of the room 2 feet above the floor, there is a small opening leading through the wall into a ventilating shaft. This orifice is five and one-half inches high by eight inches wide; it has jtwo slender, round lintel sticks running across its top, their ends embedded in the masonry at either side (figure 7, b). All the edges of the opening are neatly finished off with adobe, the corners carefully rounded. On the floor of the room, nearly in front of this hole, lay a thin slab of rock measuring 11 by 12 inches; on trial it was found to fit exactly into grooves around the hole that had obviously been made for it (plate 10, b). The horizontal shaft, to which the opening gave access, extended out from the wall for a distance of 3 feet 6 inches. It was built of flat stones set on either side with their bases together and their tops slanting outward, making a V-shaped trough 2 feet 6 inches wide across the top. This was roofed over with short stout logs covered with cedar bark, brush and coarse grass, the whole held down by flat rocks. The photograph, plate 11, b, shows the east end of the shaft with its log roofing. Behind and above may be seen the outside of the top courses of the wall of the main room, the position of which is also indicated by the dotted line in plate 11, a. There is no trace of a vertical flue connecting this horizontal pas¬ sage with the surface. The pitch of the deposit is so steep here that it is probable that such a shaft was unnecessary, and that the horizontal passage ran straight through to the outer air. Peabody Museum Papers Vol. VIII, No. 2, Plate 9 Types of Basket-maker cists: a, b, d, White Dog Cave; c, Cave 6; e, f, Cave 14. 4 OF NORTHEASTERN ARIZONA 25 The floor of the room itself, as far as we were able to lay it bare, was of hard packed adobe with a smooth but uneven surface. At what seems to have been a little east of the middle of the room there is a firepit, a saucer-shaped depression in the floor with a neatly made coping or rim of hard baked adobe (plate 10 a). It was filled to the brim with clean white ashes. In outline the pit is a perfect circle, 2 feet in diameter; the rim is raised 3 inches above the floor, and the bottom is somewhat scooped out giving a depth of 5 inches to the center of the pit. At the floor level in the back of the room is an oval niche dug horizontally 12 inches into the hard-pan of the wall, and measuring 18 inches across the front (see figure 7, b). There are two holes five and one-half inches in diameter and twelve inches deep, dug in the floor, one at the angle of the back and east wall, the other at the front directly opposite. So close are these holes set to the wall that at the back the sides of the holes are continued up through the adobe of the wall for some 6 inches. For this reason we are quite sure they are intended for post-holes though no post ends were found in them. The filling of the room was entirely free from rocks, showing that the great pile that now covers its rear portions and its northeast wall must have fallen after the place had already been deserted for a long time. On the floor was a 3-inch bed of pure sand; above this was an equal amount of coarse brush and charcoal, topped by a layer of cedar bark. The remaining 4 feet 6 inches to the sur¬ face was a homogeneous deposit composed of equal parts of rat dung and sand, laid down in perfectly regular, thread-like hori¬ zontal strata, separated from each other by thin layers of clean wind-blown sand. The peculiar make-up of this fill has been a matter of much dis¬ cussion between the authors. A plausible history of the fill might be that the room, with roof still intact, was abandoned for a period sufficient to allow the three-inch layer of clean sand to sift in and accumulate on the original floor, after which it was retenanted for a short time, the brush and bark brought in, and fires built, then vacated finally by man to become the rendezvous of rats through the long period which must have been required to build up the deep deposit of rat dung and sand found in it. During this latter period the roof remained; otherwise, instead of thin regular layers of ap- 26 BASKET-MAKER CAVES parently sifted sand, there would have been sand deposits of vary¬ ing thickness, marking the occurrence of high winds such as we experienced while at work in the cave. Finally, and prior to the falling of the rocks from the ceiling of the cave, there came other visitors who found the roof a convenient source of fuel supply thus accounting for its complete disappearance. Such a long discussion on the foregoing may appear unnecessary'-, but any condition which marks the lapse of time seems worthy of careful consideration. It is unfortunate that we were unable to clear this room com¬ pletely as there may be concealed beneath the debris which still covers the unexplored portion some evidence that would settle definitely the question of whether it is the work of the people who excavated the cists and buried their dead here, or of the Cliff- dwellers who came after. Such artifacts as were found in it are of little assistance in identifying the builders since they are either devoid of character or of such a nature as might easily have been dragged into it by rats. Outside the wall on the northeast and cast Sides we found some evidence of disturbance, such as might have been made in excavating for the foundation of the room, and in this disturbed area, close against the wall, lay two sandals with side-loops, of a type quite common in cliff-dwellings but which we have not yet found directly associated with Basket-maker remains. One of these was touching the wall at a depth of about 3 feet below the surface. Had the chamber just described been found in a pueblo or cliff¬ dwelling, it would have occasioned no particular surprise, for while its ventilator opening is smaller and higher set than usual and the V-shaped horizontal passage is of unfamiliar construction, yet the mere presence of a ventilating apparatus, the adobe rimmed fire¬ place full of white ashes, and the subterranean situation of the room itself are all features perfectly normal in Cliff-dweller kivas. Furthermore the kivas of this particular district are typically vari¬ able and unspecialized. 1 The sandals seem to be Cliff-dweller and to have been left where found while the wall was under construc¬ tion. All these things point to an origin subsequent to that of the Basket-maker cists. On the other hand we have never seen, nor have we read of, a kiva built as is this room all by itself 1 Ividder-Guemsey, 1919, p. 201. Peabody Museum Papers Vol. VIII, No. 2, Plate 10 White Dog Cave: a, Interior of kiva; b, Ventilator cover in kiva; d, Baskets in Cist 22; e. Objects in Cist 13. Sunflower Cave: c, Skeletons in Cist 7. OF NORTHEASTERN ARIZONA 27 with no living-chambers in the vicinity. All kivas with which we are fa mili ar form integral parts of house-clusters. The only surely identifiable Cliff-dweller remains found in the cave are enumerated as follows: A storage room foundation was built on the sloping rock floor against the west side of the cave (see figure 5); it measured 5 feet in length, 2 in width and consisted of a low wall, 8 to 10 inches high, the stones mudded in with adobe mortar. In the enclosure was a bed of plant stalks, “Brigham tea”; the floor is bare uneven rock. We collected in the top sand of the cave a few handfuls of Cliff-dweller potsherds, for the most part plain gray and black-and- white ware, and a few pieces of feather string. A small corrugated pot covered by a flat stone was found cached in the sand 1 foot 6 inches below the surface; the mouth had been sealed with adobe mudded on to corn cobs, but this had crumbled and was found at the bottom of the jar. About the jar was a harness, made, with the exception of one short section, of Cliff-dweller feather string. The short piece is apparently Basket-maker fur-string and was probably a stray bit picked up from the surface. The above is not an imposing fist and leads us to doubt that the place was ever regularly used as a dwelling by the Cliff-house people. As to the identity of the kiva-like room, the writers them¬ selves are not agreed; the senior author believes that it may pos¬ sibly be of Basket-maker origin, the junior considers it surely Cliff-dweller, but can offer no explanation for its isolated situation. 28 BASKET-MAKER CAVES FIELD WORK, SEASON OF 1917 Reaching Kayenta by the usual route via Farmington, New Mexico, and the Chinlee, the party first attempted explorations near Sayodneechee Canyon in Monument Valley, but was forced by lack of water to abandon the work after a few caves had been examined. Returning to Kayenta the exploration of the South Comb was resumed. White Dog Cave was revisited and an unsuc¬ cessful attempt was made to move the rocks from above the lciva- like room. Two new caves were discovered and investigated. Again forced to move by lack of water, the remainder of the season was spent in Sagiotsosi Canyon, where nine caves were either wholly or partly explored (see map, figure 1). SAYODNEECHEE CANYON This is one of the numerous short canyons which head near the Agathla rock and run northward into Monument Valley. Although it is without living water, the Navajo are able to culti¬ vate corn in certain places. In the winter, rain and melting snow furnish sufficient drinking water for the Indians and their flocks; and in some years enough of this is held in pockets among the rocks to last until the showers of July and August. Generally, however, these natural reservoirs go dry in June and the Navajo must move away for a month or so to some more favored locality, returning after the rains to harvest their crops. Aside from its dryness, Sayodneechee is a most attractive place; the scenery is magnificent, grass and firewood are abundant, and the cliffs contain many caves to tempt the archaeologist’s shovel. Caves 3, 4 and 5 are in a break of the rock ridge that forms the west wall of Sayodneechee Canyon, and are nearly opposite the Basket-maker burial cave in the above cam r on excavated by the 1914 expedition. 1 Cave 3 is a mere shelter measuring 15 feet in depth by 30 feet in width. The wash of a small canyon has cut away the floor at the front. On the back wall are a number of pictographs done in white, red, and yellow paint; some of these are reproduced in plate 13, a. We found several slab cists buried beneath the sand floor. They contained nothing except cedar bark. 1 Kidder-Guernsey, 1919, p. 27 and figure 1. Peabody Museum Papers Vol. VIII, No. 2, Plate 11 White Dog Cave a, Rock pile in center of cave; b, Southern wall and ventilator in kiva. OF NORTHEASTERN ARIZONA 29 Cave 4, a short distance up the canyon, is 20 feet above the wash. It has a depth of 12 feet and measures about 24 feet across the front. The floor is of hard-pan free from surface sand. In it are a number of small cists or pot-holes. At the front the hard-pan formation has a vertical break, in which is dug a small cubby hole measuring 4 feet in depth by 3 feet 6 inches in width. At the entrance to this little room, shown at the left in plate 12, a, are a number of flat slabs arranged like steps, a single slab 2 feet long and 8 inches wide serving for a sill. There are several small holes dug through the top of the room to the surface above. The largest of these holes is plugged with a rock. A little further along the cliff is a rectangular Cliff-dweller room, the dimensions of which are, length 12 feet, width 7 feet, height of wall 6 feet 6 inches. In the center of the front wall is a door¬ way 29 inches high, 16 inches wide. At the top is a flat stone slab lintel supported by two round sticks built into the wall, another slab serves as a sill. The edges have grooves or jambs for the re¬ ception of a slab door. The masonry of this room is good. There were no pictographs. Potsherds were plentiful and along the cliff near the room there was some rubbish and a number of ash beds. Cave 6 is still further up the canyon. It measures 45 feet across the front, and 15 feet in depth. At the back are the foundations of a room 10 feet long by 6 feet wide built out from the cliff. The masonry is of stones laid flat in adobe mortar. Two slab cists and two cists dug in the hard-pan floor were found in the cave, but no specimens. Near the sites just described is a small shelter on the ground level of such insignificant size that no number was assigned to it in our field notes. We dug here, however, and at a depth of one foot below the surface found two slab cists partly filled with cedar bark. These were undoubtedly storage cists, as near by is a Navajo cornfield, located in a small basin which collects and retains such water as in time of rain runs off the surrounding cliffs, an advantage probably recognized by the early occupants of the region as readily as by the present day farmers. The principal structures in these caves are of course Cliff- dweller. The slab cists and possibly some of those excavated in the hard-pan we are inclined to think are Basket-maker. No great amount of work was done at any of the sites, as we were on such 30 BASKET-MAKER CAVES short rations of water that our examination really only amounted to a reconnoissance. Continued drought finally drove us away, and we returned to the South Comb. SOUTH COMB REVISITED Cave 6. This site is in the next break in the Comb north of White Dog Cave, a distance of about one mile in an air line. It consists of a small alcove at the back of a huge crescent-shaped bay or cove in the cliff wall. Filling the open end of the crescent and hiding the cave from view in front is a high sand hill covered by a growth of thick brush and tall pines. The cliff on either side of the cave overhangs, sheltering a wide strip along the wall some fifteen feet lower than the floor level of the cave proper. On this level to the left of the entrance there is part of a roughly laid wall, built against the cliff. It forms a small enclosure and is probably the work of Navajo herders or possibly Ute, as on the smooth cave wall back of it are a number of drawings in charcoal (plate 13, f), one of which, a female figure, is shown wearing a dress that has characteristic features of the Ute woman’s dress. Inside this en¬ closure were traces of recent fires and on the surface was a small mudded-up fire pit, which gave us the impression of having been the work of children. The walls and ceiling of the inner cave are much blackened by smoke. It had been used as a sheep shelter and the old floor was covered by a thick layer of dung. The most careful search of the surface on the first level and the bank leading up to the cave proper failed to produce a single Cliff-dweller artifact and our excavations later showed not a vestige of Cliff-dweller occupation. Here for the first time we had a cave containing only Basket- maker remains, and while but a few specimens were found they were for the most part very true to type, the exceptions being en¬ tirely new material. A single burial was encountered. This was in a stone slab cist (plate 9, c), exactly like those found in such numbers in Cave 1, Kinboko (1915). Unfortunately, however, it had not only been plundered at some early date, but what re¬ mained of its contents had been partly destroyed by fire. The top of the cist was 18 inches below the surface. It measured 3 feet 4 inches in diameter at the top, 2 feet 6 inches in diameter at the Peabody Museum Papers Vol. VIII, No. 2, Plate 12 a, Structures in Cave 4 Sayodneechee Canyon; b, Cists in Cave 14, Sagiotsosi Canyon. OF NORTHEASTERN ARIZONA 31 bottom, and was 2 feet, 4 inches deep. Ten slabs were used in its construction. In the upper part was a quantity of cedar bark and a few bones from the skeleton of a child, then a mass of charcoal and charred wood in which were fragments of human bones. On the bottom at one side was a partly burned cradle frame, and the mummified foot of an adult. Other objects found scattered in the fill are as follows: fragments of fur-string robe, dressed skin robe, twined-woven grass mat, string apron, a sandal, an atlatl, a grooved club, a skin-covered umbilical pad, the bark core of another, a skin bag, a bunch of human hair, a fragment of squash shell, and many small bits that could not be identified. All these specimens were more or less charred. But one other slab cist was encountered. Its only unusual fea¬ ture was a bottom lining of thin slabs of spruce bark. Nearly all the level portion of the cave floor was occupied by a deep ash bed in which only a few minor specimens were found. Just outside this area at a depth of 1 foot 6 inches was a tray basket, and buried in the loose fill near it at about the same depth was the small woven bag in which was the little skin pouch shown with its contents in plate 44. On the right of the cave the floor rises and narrows until it gives place to a mere bank of debris piled up against the back wall. At the highest point of this bank and next to the wall, three deer or possibly mountain-sheep snares were found. They had been cached in a shallow hole scooped out of the fill, and were covered with cedar bark and a thin layer of dirt. These snares are new items in our Basket-maker fist. They are described and figured in another place (plate 32). A few feet from where the tray basket was found, and at the same depth, were three sherds of a substance resembling pottery of unbaked clay, tempered with shredded cedar bark, and bearing on one side the imprint of coiled basketry (plate 25, a). This may really be a primitive form of pottery or may represent only some left-over material for smearing joints in a slab cist, which was prepared outside the cave where both water and clay could be had, and then brought into the cave, while soft, in a convenient tray basket, from which it was not removed until it had hardened by drying. It is the nearest approach to pottery we have yet encountered under circumstances that would free it from suspicion of Cliff-dweller origin. Mr. John Wetherill, 32 BASKET-MAKER CAVES to whom it was shown, said it recalled the pottery found in the Basket-maker caves of Grand Gulch. This, according to McLloyd and Graham’s description as quoted by Pepper, 1 was “ a very crude, unglazed ware, some of the bowls showing the imprint of the baskets in which they were formed.’’ As stated before, all our work in this cave brought to light not one trace of Cliff-dweller occupation, which includes not only potsherds, but also turkey droppings and turkey feathers, beans i 1902, p. 9. OF NORTHEASTERN ARIZONA 33 and rubbish layers. Hence the collection obtained here, though not extensive, is important as it supplies unmixed material with which to check our previous identifications. Cave 7. About one mile north of Cave 6, we found another shelter very similar to it, except that it lacked the alcove room at the back. A steep hill rises directly in front of it. The slope of the hill next to the cliff lies almost wholly inside the line of shelter and its base at that point is cut away by an arroyo which continues along the wall for some distance. This seemed a very unpromising site, but on investigation we found a number of slab storage cists filled with cedar bark or grass, located as shown in the cross- section, figure 8. No Cliff-dweller remains were found here and only two Basket-maker specimens. These were the digging sticks shown in plate 37, e, f. This shelter seemed never to have been used as a place of abode for any great period as we found no exten¬ sive ash bed. Perhaps it was conveniently near some cornfield and was used only for storage purposes or as a temporary dwelling place while farming was in progress. By the time that the work in Cave 7 was completed, the water in this section had become so bad that we were again forced to move. SAGIOTSOSI CANYON Sagiotsosi Canyon, though small in size compared with many others in this region, exceeds all that the writers have visited in the number of caves to be found in it and its branches. Its scenery is exceedingly picturesque, and it is rendered doubly attractive in this parched land by a stream of clear cold water fed by numerous springs that emerge from the base of the cliffs on either side at the upper end. This stream flows the entire length of the canyon finally to disappear in the thirsty sands just outside the entrance. In one place where it has cut a deep arroyo, a dark peat-like stratum can be seen in the vertical sides of the cut, marking an old lake bottom that probably once provided a natural reservoir for the ancient inhabitants. Today a number of well-irrigated Navajo cornfields and thrifty peach orchards show the water supply to be still ample for the requirements of primitive farming. The caves in the main canyon are for the most part high up under the rim-rock and are perhaps more properly described as shelters. Some are of huge size with high arched openings, but of 34 BASKET-MAKER CAVES no great depth. Occasionally they occur in groups of three or four, quite close together. To enter them one must first climb over huge fallen rocks to the first bench of the cliff, then up a steep talus of finer detritus to the caves, the bottoms or floors of which are really nothing but the truncated apex of the talus. Several of these caves have in them small Cliff-dweller structures. A number have already been explored by Professor Cummings. 1 On the right about half way up the canyon and high in the cliff is a fair sized cliff-dwelling which to date has not been excavated. An interesting feature of this ruin is a tower that commands every approach to the cave. A cursory examination indicated that the roof had been destroyed by fire. On the back wall of the cave is a pictograph similar to the one illustrated in plate 13, e. Cave 8. This cave is in the first branch-canyon leading out of Sagiotsosi to the west. It is in reality a shelter under the over¬ hang of the cliff, 30 feet in width, some 70 feet in length and about 25 feet above the bed of the wash. There is in it ample evidence of Cliff-dweller occupation, consisting of some foundation walls, a good depth of rubbish, with man} r potsherds, and a number of Cliff-dweller pictographs (plate 13, d, e); there is also a square¬ shouldered human figure done in white and yellow paint. This shows very faintly and a small Cliff-dweller painting of a snake overlaps it in one place (d). It was this square-shouldered picto¬ graph that induced us to dig here, as our previous experience had shown these figures to be of Basket-maker origin. Our excavations disclosed considerable Cliff-dweller rubbish with hard-pan below it in which we found a number of cists, empty except for cedar bark or coarse grass. These cists and the square¬ shouldered figure are the only remaining evidences of Basket- maker occupation. From the general digging we obtained a num¬ ber of Cliff-dweller specimens including the skeleton of a young child on a perfectly preserved cradle which had been buried under the rocks at the top of the bank at the front. This shelter seems insignificant in comparison to the huge caves in the main canyon. It provides, however, a further illustration of the fact that no cave or shelter in this region is so small that it has not at some time attracted tenants who have left traces of their occupancy. • 1910, pp. 9-18. OF NORTHEASTERN ARIZONA 35 Cave 9. Across the canyon from Cave 8 is a small Cliff-dweller ruin in a low cave that shows signs of previous investigation. Rooms along the back wall have been reroofed by the Navajo and used for storage purposes. This cave in the writers’ opinion gives evidence of two occupations. This belief is, however, based wholly on the presence of typical Basket-maker cists excavated in the hard-pan floor (plate 14, c, d), for we found here no objects that could be classed as Basket-maker. The cists occurred in a small unoccupied area in the center and were completely filled with Cliff-dweller rubbish. There is, nevertheless, evidence at one place Figure 9 Plan of Cave 9, Sagiotsosi Canyon. that the cists were here when the Cliff-dweller structures were erected, for the side wall of one room is built partly across a cist (see figure 9). The latter could hardly have been made by the Cliff-dwellers, since they could have easily avoided weakening the foundation of their wall by digging the cist a little to one side. In objection to the foregoing it may be said that the cists are of Cliff-dweller origin; they are, however, exactly like ones found in other caves containing Basket-maker burials, and since all Basket- maker cists have a certain unity of design and a certain “ look,” hard to describe but at once apparent to anyone who has opened a number of them, the authors are satisfied that their identifica¬ tion of the present examples is correct. Compare c and d, plate 14 with a and b of the same plate; the latter are from photographs of Basket-maker cists in White Dog Cave. Cave 10. Just below Cave 8 there is a narrow break in the canyon wall with a length of perhaps 400 feet. About half way up this 36 BASKET-MAKER CAVES gulch is a shelter 20 feet in depth and 40 feet across the front (plate 5, b). The only sign of occupation noticed on entering was the top of a stone slab cist which just showed above the sur¬ face sand and a number of hand-prints in red on the back wall at one side. Excavation proved, however, that the place had been occupied by both the Basket-makers and the Cliff-dwellers. The Cliff-dweller remains consisted of a few potsherds, several bone scrapers of a typical Cliff-dweller form, 1 and a quantity of corn¬ cobs which we think are Cliff-dweller because thejr are much longer and larger than the Basket-maker corncobs we have found. The Basket-maker remains were empty storage cists, both slab and excavated, with cedar bark in their bottoms. There was also one Basket-maker burial cist containing the partly mummified and headless body of a child, wrapped in a fur-string robe. With the body was part of a large dressed skin bag and at the feet lay badly rotted square-toed sandals. This burial was identical with those found in other Basket-maker caves. Evidence appeared that this or other cists had been plundered, as in the general digging there were found a number of fragments of Basket-maker basketry and a small piece of rabbit net made of human hair and fiber- string combined. To gain entrance to the gully in which this cave is located one must cross a smooth, waterworn ledge. Up this is pecked a series of tracks representing the hoof-marks of a horse. They are very neatly executed and are the first instance that has come to our notice of pecked pictographs of recent (Navajo or Paiute) origin. Cave 11. This cave is in the east wall of the main canyon near its head. Tt is some 200 feet above the wash and consists of a nar¬ row shelter with a frontage of about 150 feet. On the back wall are a number of hand-prints and some nearly obliterated human figures all in white. On the surface were scattered a few bleached human bones. Large flat rocks along the front show deep axe- grinding grooves. We were only able to spend a half day here. Our limited digging showed that for a considerable period the cave had been used bjr Cliff-dwellers and we recovered a number of their characteristic ■ See Morris, 1919, figure 23, e. We found none of this variety in our cliff-house excava¬ tions in 1914. Peabody Museum Papers Vol. VIII, No. 2, Plate 13 Sayodneechee Canyon: a, Pictographs in white paint, Cave 3. Sagiotsosi Canyon: b, Pieto- graph in red paint, Cave 12; c, Pictographs in white paint, Cave 14; d, e, In white paint, Cave 8; f, In charcoal, Cave 6. ✓ OF NORTHEASTERN ARIZONA 37 artifacts from the rubbish. At one point we found a loom-anchor in place. This consisted of a smooth pole one and one-half inches in diameter and six feet long, having loops of braided yucca and heavy fiber cord strung on it at regular intervals. It was buried several inches below the floor and held down by flat rocks, the tops of the loops just protruding above the surface. Under some large rocks at the front of the cave, we uncovered a small Basket-maker pannier basket in a poor state of preservation, inverted over a quantity of corncobs; probably the corn had been stripped by rodents. Attached to it was part of a carrying-strap of human hair string. In a narrow part of the shelter and under what must have been the path ordinarily used in entering it, we found a disturbed Basket-maker burial. Some of the bones including the skull were missing. There were with the remains fragments of a coiled basket, square-toed sandals and a piece of finely woven cloth. Cave 12. This is a deep cavern a short distance down the canyon from Cave 11 and on the same side. It is about 90 feet above the wash and has a fairly level floor area 40 feet deep by 70 feet across the front. The walls and ceiling are much blackened by smoke, and the floor is thick with charcoal. At one point the top of a rude enclosure of stone slabs shows just above the surface. This is circular in shape and has a diameter of 12 feet. At one place in the back wall are a group of hand-prints in red placed as near together as possible and covering a space of 6 feet or more; the only other pictograph noticed is the small figure shown in plate 13, b, also done in red. On a flat rock at the front are a number of axe-grinding grooves. Our digging here was confined to test holes, as it was obvious that it would be too much of an undertaking for our small party to clear the cave completely. We found rubbish along the back wall to a depth of a little more than one foot. It was very com¬ pact and contained a large amount of broken sticks and twigs, straw and charcoal. There were two or three slab cists partly filled with cedar bark but holding no specimens. We do not think any great returns would reward further work at this site. It had apparently been used by Basket-makers and Cliff-dwellers in turn, but did not appeal to the latter strongly 38 BASKET-MAKER CAVES enough to warrant the erection of any structures. It is set verjr deep in the cliff and gets but little sun; it may have been con¬ sidered undesirable on this account. Cave 13. This is a very long shallow shelter high up in the cliff near the head of the branch canyon in which Caves 8, 9 and 10 are located. At some not very remote time a great quantity of the roof had scaled off, burying almost the entire floor beneath tons and tons of rock. At one end of the cave is a series of small cliff- house rooms, some of which still retain roofs; others are crushed and the walls partly buried beneath the fallen rocks. Along the whole front of the cave can be traced a low roughly built wall. It seems probable that beneath the rocks are structures similar to those in the end of the cave, but to reach them would be a very large undertaking. We noticed no pictographs here. Cave 14. This cave, the last to be explored, is but a short dis¬ tance from Cave 13. It consists of a shallow shelter 200 feet above the canyon bottom, and has a usable floor space 20 feet deep by 70 feet in length. The line of shelter extends some 20 feet beyond the point where the floor breaks away at the front. At one end is a small niche in the back wall 7 or 8 feet above the floor. Leading up to it are a number of pecked toe-holes. The ceiling and some parts of the walls of the cave are blackened by smoke. On a smooth area of the wall near the center is a group of square-shouldered human figures painted in white, while other similar figures show' faintly at other points (plate 13, c). These are distinctly Basket- maker. Built against the back wall of the cave is a series of seven stone slab structures, six of which are in a fair state of preservation. These will be given a more detailed description further on. In our excavations here we found below the surface several slab cists of the usual Basket-maker type. From one w^e obtained a small skin pouch, which with its contents is showm in plate 38, a-c; also, in the loose fill, a wooden implement plate 36, a; and the bundle of human hair wrapped with string illustrated in plate 32, e. At the extreme right of the cave a single square-toed sandal was found in the general digging, and several ears of corn cached in the loose dirt against a large flat rock. So near is this cave to Cave 13 that it is inconceivable that it had not been frequented by Cliff- dwellers to some extent, yet careful search of the surface, and Peabody Museum Papers Vol. VIII, No. 2, Plate 14 a, b, Cists dug in hard-pan, White Dog Cave; c, d, Cists, Cave 9. OF NORTHEASTERN ARIZONA 39 watchfulness throughout the digging failed to produce a trace of their handiwork with the possible exception of the corn which may be Cliff-dweller, as it is unlike the characteristic Basket- maker corn. It was found in a part of the cave quite remote from the cists. There were no potsherds, twilled sandals, feather cloth or even axe-grinding grooves. The latter are seldom absent from caves in which the Cliff-dwellers have lived. The most interesting things in the cave are the slab structures along the back wall (plate 12, b). They average about 5 feet in diameter, the best preserved standing three and one-half feet above the surface. Large stone slabs are used in their construction, in most cases overlapping. The space between the joints is filled with adobe mortar which in some instances has been plastered all over the slabs both outside and in. Small stones are set in to fill holes between the slabs and the cave wall to reinforce the slabs at their bases. In the structures and on the surface about them were a number of timbers from 4 feet 6 inches to 6 feet in length and 4 to 6 inches in diameter, probably roof timbers. Other shorter sticks were found which had once formed a part of a rim molded on to the top of the slabs. These pieces had traces of adobe on one side; there were also found large lumps of adobe tempered with cedar bark with one side moulded round, the other bearing imprints obviously made by the short timbers just mentioned. These sections of stick and adobe are important because they show that the present above-ground cists are identical in rim construc¬ tion with a subterranean Basket-maker storage place (Cist 14) found in Cave 2, Kinboko during the 1915 season. 1 Another larger cist (12) in the same cave had a similar rounded adobe coping strengthened with stones instead of sticks. The drawing, plate 9, e, represents one of the Cave 14 cists with a short section of the rim restored. The slabs are shown partly denuded of the adobe plaster, while on the wall behind the cist a line of adobe is indicated which probably marks the outline of the roof. This structure more fully restored appears in f, of the same plate. Why so much care should have been taken to finish the r im , if the roof timbers were to rest directly on it, we are unable to say, though it is evident that a rim made in this way would greatly 1 Kidder-Guernsey, 1919, p. 88. 40 BASKET-MAKER CAVES strengthen the whole structure. These slab cists seem hardly large enough for living rooms or even for sleeping places. It seems more probable that they were storage cists. We do not hesitate to identify them as Basket-maker, because they are exactly like the Basket-maker structures in Cave 2, Ivinboko. OF NORTHEASTERN ARIZONA 41 MATERIAL CULTURE 1 FOOD Vegetal Food. Maize. In 1914 and 1915 we found indications that the Basket-makers cultivated but a single and rather primitive type of corn, while that grown by the Cliff-dwellers seemed to have been more highly developed and more varied in character. Our evidence was not, however, absolutely conclusive, for certain speci¬ mens of the advanced corn were taken from Basket-maker caves, though from so near the surface that we regarded them as probably intrusive. The expeditions of 1916 and 1917 supply us, fortunately, with enough new finds to settle the question beyond any reasonable doubt. A number of Basket-maker caves were thoroughly investi¬ gated and many samples of corn were recovered from undisturbed and surely identifiable burials and storage cists; among all this material there is not a single kernel of any of the parti-colored flour or large white flint corns that are so common in the cliff-houses. 2 On specimens submitted to him for examination Mr. G. W. Collins of the United States Bureau of Plant Industry has kindly given us the following report: The collection of maize samples from the Basket-maker caves is of unusual interest. The specimens all appear to belong to one general type, a type we have called Tropical Flint. This type resembles the New England flint varieties in having a large part of the endosperm hard or corneous. It differs from New England flint in having a larger number of rows and smaller seeds. Tropical flint varieties are common in Central and South America but are rare among the types grown by the Indians of the United States. So far as our collections show the Papago is the only tribe with varieties uniformly of this type. The cobs of the specimens from the Basket-maker caves are all light brown in color. The pericarp is either red or colorless. The endosperm is either light yellow or white. The aleurone or layer of cells just beneath the pericarp in all the specimens is a yellowish red. This is a color entirely unknown in the aleurone of existing varieties. If this color is not the result of some slow dis¬ integration, it constitutes the first clearly marked distinction between pre¬ historic maize and present day varieties. Most of the specimens are remarkably well-preserved. The embryos have of course disintegrated but the colors are much brighter than is usual with old specimens. 1 Only objects believed by us to be of Basket-maker origin are included. Specimens re¬ covered from the cliff-houses will be treated in a later paper. 2 See Kidder-Guernsey, 1919, p. 154. 42 BASKET-MAKER CAVES The specimens cannot be referred to any existing variety with which I am familiar but with the possible exception of the unusual aleurone color they present no new characters. Here then is an undifferentiated, and judging from its distribu¬ tion, a primitive form of corn grown by a people whom the purely stratigraphic evidence shows to have antedated the highly de¬ veloped agriculturists of the region. This agrees very well with the other manifestations of Basket-maker culture, and particularly with its lack of true pottery, stone architecture, and cotton weav¬ ing, all of which traits are characteristic of the perfected puebloan civilizations. We have thus good evidence that the Basket-makers were the pioneer corn growers of the district. To what degree these people depended upon maize is uncertain, but quantities of it were found in the burial cists and cached for future use as food or for seed. There were also recovered agricul¬ tural implements such as would be needed for its cultivation, and the large number of storage cists in the caves would indicate by their capacity that a considerable harvest was obtained. The sites explored by us were all within easy reach of tillable land and this is also true of the Grand Gulch Basket-maker caves. Of the actual finds of corn the best example is the skin bag full of shelled kernels from Cist 13, White Dog Cave (plate 15); there are about four quarts, every grain in perfect preservation. This may represent a food offering deposited with the dead, or perhaps it is carefully selected seed cached unknowingly in the same cist with the burials (it was found some 8 inches above the remains shown in plate 10, e). Other interments, however, were accompanied by corn and the remains of rotted hide containers, so that it may indeed be a food offering. A selection of the more perfect ears of Basket- maker corn is shown in plate 15. Squash. This seems to have been the only other cultivated crop of the Basket-makers. 1 We unearthed with the burials varying quantities of squash seed, Cucurbita pepo, and many pieces of rind, as well as the complete vessel made from a squash shell that is shown on plate 31, b. Seeds. In a number of the burial cists in White Dog Cave, large quantities of coarse grass seed were found. We saw growing in the 1 Though we were constantly on the watch for beans in the Basket-maker sites, none were found. This strengthens our belief that they were not grown by the Basket-makers. OF NORTHEASTERN ARIZONA 43 vicinity, the same variety of grass from which it was obtained. Mr. W. E. Safford of the Bureau of Plant Industry identifies this as follows: Oryzopsis hymenioides, commonly called Indian Mountain Rice, is used by several Indian tribes for food; by some only in times of scarcity, by others as a regular food staple. Mr. F. V. Coville states that the squaws of the Panamint Indians of southern California gather it by means of a wicker paddle resem¬ bling a small tennis racket with which they beat the seeds from the standing grass into wicker baskets, after which they are winnowed and sifted, and parched and ground into pinolli. The late Dr. Edward Palmer found this seed in use among the Paiute and Pueblo Indians, who store it for winter use. Cummings 1 found caches of seed in Sagiotsosi (“ coarse bunch grass ”), which may be the same. No doubt other seeds were gathered and stored for food, as we found in 1915 several quarts of Coreocarpus seeds in a burial cist in Cave 1. Powell in his ex¬ plorations of the Colorado found a tribe which subsisted chiefly on wild fruits, nuts and native grains. In our own explorations we came upon an old Navajo squaw in the vicinity of Sagiotsosi who was gathering the small seeds of a low weed. She told us that these were cooked and made into a kind of mush by mixing with goat’s milk, also that they were now (1917) being used again for the first time since the “great war” (Navajo war, 1863). These are identified by Mr. Safford as Chinopodium sp., who writes as follows regarding them: They are perhaps the most interesting of the collection. It has been im¬ possible to determine their specific identity. They are much larger than the seeds of Chenopodium fremontii, gathered for food by the Klamath Indians, and those of Chenopodium leptophyllum eaten by the Zuni. In shape they bear a close resemblance to the seeds of Chenopodium quenua, the well-known food staple of the Peruvian and Bolivian Plateau, but they are of smaller size and of a much darker color than the latter. These seeds have been carefully compared with those of the species growing commonly in the southwestern United States; they bear a closer resemblance to Chenopodium petiolare than to any other species in the herbarium, but they do not seem to be identical with the seeds of that species. They are evidently rich in starch and would undoubtedly form a nutritious article of food. Pinon Nuts. These were also an important item of diet and were found with other food offerings in many of the graves. 1 Cummings, 1910, p. 14. 44 BASKET-MAKER CAVES Unidentified Food. Small quantities of plant stalks, shriveled beyond recognition, accompanied some burials. These are prob¬ ably from certain edible plants that grow in the region, and which are eaten today by the Navajo. Animal Food. The bones of mammals and birds, generally so common about the dwelling places of primitive people, were en¬ tirely lacking in the group of Basket-maker caves examined. We do not believe that this indicates a preponderatingly vegetarian diet, but rather that it proves the caves to have been used merely as temporary shelters and as burial places for the dead. That these people killed a great deal of large game is evidenced by the abun¬ dance of articles made from the hides of deer and mountain-sheep; while quantities of the pelts of badgers, rabbits, prairie-dogs, and other small animals were employed for bags, pouches, and in fur¬ string robes. It is probable that the flesh of all the above was eaten. As to the birds we have less evidence. Such feathers as were found came principally from hawks and owls, species not com¬ monly relished as food by any people; or from very small birds of bright plumage such as warblers, bluebirds, and woodpeckers. As we have never come across a single identifiable turkey feather, it is reasonably certain that the turkey was not domesticated, nor indeed does it appear to have been commonly hunted. Although there is no evidence that the Basket-makers used the dog for food, it may be well to refer here to the finding of two re¬ markably well-preserved dog mummies in White Dog Cave. They represent different types, formerly of wide distribution in the warmer parts of America (plate 15). Dr. Glover M. Allen of the Museum of Comparative Zoology, who has made an exhaustive study of the native Indian dog, has kindly contributed the follow¬ ing regarding these specimens: The larger is a long-haired animal the size of a small collie, with erect ears and long bushy tail. The hair is still in good condition and though now a light golden color, with cloudings of dark brown, it may in life have been darker. It is, apparently, a breed very similar to the long-haired Inca dog found at Ancon, Peru, in a mummified condition and described by Nebring (Sitzb. Ges. Naturf. Freunde, Berlin, 1887, pages 139-141). The latter specimen is also described as yellowish in color, though this may have been in part due to fading. A more detailed comparison of the two specimens is not possible without removing and cleaning the bones and so injuring the present example for exhibition purposes. Peabody Museum Papers Vol. VIII, No. 2, Plate 15 White Dog Cave Mummies of two varieties of dogs, ears of corn, and skin bag containing shelled corn. OF NORTHEASTERN ARIZONA 45 The other dog is a much smaller, black-and-white individual, about the size of a terrier, with short, but not close, shaggy coat, erect ears, and long full¬ haired tail. Its muzzle is rather short and stubby in contrast to the fine slender muzzle of other Indian dogs of about the same size. In common with many skulls of American Indian dogs, the first premolar is lacking in the adult den¬ tition of the lower jaw. This specimen is of especial interest as establishing beyond doubt the identity of certain dog bones from Ely Cave, Virginia, described as Pachycyon robustus, for they agree perfectly with corresponding parts of the Arizona dog. An identical breed is represented among the mummi¬ fied remains of dogs from the necropolis of Ancon, Peru, and has been figured by Nebring as Canis ingae vertagus in the folio report of Reiss and Steubel, plate 118, figure 1. Evidently it had a wide distribution in our south and southwest, and was known also to the Peruvians. I have called this the short nosed Indian Dog. These and other dog remains, are true dogs, in no way derived from Coyotes or other native dog-like animals of America. Their forebears probably reached America with their human masters, but their Old World ancestors still remain to be determined. 1 DRESS AND PERSONAL ORNAMENTS Body Clothing. We have few data on this subject; it is probable, indeed, that the Basket-makers wore very little clothing except robes of fur-string or hide, 2 and “gee strings” or cord aprons. It so happens that all the robes found in sufficiently good preservation to permit of measurement had been interred with babies; the largest of these (plate 16, a) is only 25 by 23 inches. About an adult mummy (A-2939) from Cist 22, White Dog Cave, however, there is wrapped what appears to be a very large blanket of fur¬ string; and we have fragments from deer and mountain-sheep hides which seem to have been originally of ample size for use as mantles by grown people. Nothing resembling fitted garments of leather or cloth has so far come to light; it is possible, however, that certain woven fabrics, bits of which were recovered from the caves 3 may have been used as ponchos. This guess is based on the resemblance between a zigzag decoration on one of the cloth specimens (plate 26, c) and similar patterns painted on the chests of Basket-maker human pictographs from the Monument country. 4 It must be admitted, 1 For a discussion of the types of prehistoric American dogs, see Allen, 1920. 2 For details of the weave of these robes, see p. 65. 3 See plate 26, b, c. 4 Kidder-Guernsey, 1919, figures 100, 101. 46 BASKET-MAKER CAVES however, that the zigzag was a favorite Basket-maker design, and that the marks on the pictographs may perfectly well represent body-painting. A string apron recovered by the 1915 expedition still remains our best specimen of this type. Although it was illustrated in our former report (plate 66, a), we have since succeeded in unraveling it for a somewhat clearer photograph; this, with a picture of a second example from the general digging in White Dog Cave, are here reproduced (plate 16, c, d). It will be seen that in both cases there is a waist cord to which is attached a fringe of pendent strings. In the 1915 specimen the strings are of apocynum and are looped over the human hair waist cord and gathered in bunches of about three hundred; the fringe is 12 inches long. 1 The apron from White Dog Cave (plate 16, c) is more fragmentary; the yucca-fiber waist cord is double; over it are hung yucca strings which are gathered together in pairs and held, close under the waist cord, by a row of twined weaving, one strand yucca, the other human hair. Although somewhat longer than the first apron this garment is much thinner and contains fewer strings. Platd 16, b, shows part of a similar skirt made of cedar bark. The pendent strands are about 12 inches long and are held to¬ gether by a twining of twisted cedar-bark string, the prolongations of which once formed the waist cord. As the term apron implies, the fringes of these articles did not extend all the way around the body, but merely covered the front of the waist; it is probable that they hung loose, for the strings are too short to have been pulled between the legs and fastened over the waist cord behind. They are evidently a woman’s garment, as in every case where they were discovered in place on a mummy, the body proved to be that of a female. Though we have never found any covering at the loins of a male, there are in the collection two objects that may well have been the ties of “gee strings.” One is a loose twist of thirty animal wool threads (plate 16, f); it is nearly 7 feet long and its ends are tapered as if for knotting. The other is 5 feet 2 inches long and made of fifty to sixty thin strings of human hair; the ends are seized with fiber thread to prevent raveling. 1 For a fuller description, see Kidder-Guernsey, 1919, p. 157. Peabody Museum Papers Vol. VIII, No. 2, Plate 16 Clothing: a, Fur cloth blanket; b, Apron of shredded bark; c, d, Aprons of fiber string; e, f, String belts. All from White Dog Cave except d, f, which are from Kinboko Canyon, Marsh Pass. (About g.) OF NORTHEASTERN ARIZONA 47 Sandals. As most of the specimens recovered by the 1916 and 1917 expeditions are very badly rotted and as no new types appear, the reader is referred to the classification and descriptions of the 1914, 1915 material given in the previous report. 1 Necklaces. These were of two sorts: strings of beads; and twisted skin or fiber cords, to the middle of which were attached a few pend¬ ants or extra handsome beads. Of the latter class there was re¬ covered only one fragmentary example (plate 17, b); it bears two very beautifully polished lignite discs strung on a fiber cord, which is itself attached to a sinew-bound thong; the whole was probably fastened to a longer neck cord as was done with a similar specimen found in 1915. 2 The second type is more fully represented, several strings of beads having been taken from the necks of skeletons in White Dog Cave. A selection is given in plate 25, e-h. The most interesting of these is composed of seventy-one thick discoidal black lignite and white limestone beads strung alternately on a narrow thong. They are graduated in size from a maximum diameter of f of an inch at the center of the string, to f of an inch at the ends. An unusual refinement of technic was practised by cutting several of the beads to a wedge-shape (see figure 10, c, and plate 25, h) and introducing them here and there throughout the set in order that it might hang evenly. Loose behind the neck of the mummy who wore this string were fourteen olivella shells that apparently had once been fastened together to form a sort of “ dangler ” at¬ tached to the tie-strings of the necklace. Another string (plate 25, f), which was recovered in order, is made of one hundred little saucer-shaped shell beads (figure 10, g); seventy-five thin, roughly discoidal shell beads (figure 10, f); and eighteen olivella shells, one of which bears an incised zigzag decor¬ ation (figure 10, i). These different kinds of beads were grouped together. Plate 25, e, shows a third necklace composed of ninety- five beads arranged as follows: one of lignite, seven olivella shells, one of seed, one of bone, one of red shale, one of green shale, one of red shale, eighty-one of white limestone. Plain strings of olivellas designed to go once or twice around the neck are not uncommon. 1 Kidder-Guernsey, 1919, pp. 157-160. 2 Kidder-Guernsey, 1919, p. 161 and figure 72, a. A full description of this type of necklace is there given. 48 BASKET-MAKER CAVES Beads. Under this head are considered all the beads found, whether strung into necklaces, discovered loose in the cists, or in¬ cluded in “ medicine outfits.” The commonest of all are little cylinders averaging ys of an inch long (figure 10, e, and plate 25, g); some are of albatite, a phase of asphaltic shale, but the great major¬ ity (hardly distinguishable from the above except under a magni¬ fying glass) are made from some hard black seed so cut down in manufacture as to be unidentifiable. Other seeds were used un- 9 h i FiGunE 10 Beads from White Dog Cave. (Full size.) worked except for a narrow bore. 1 Two varieties of these seed beads are identified by Mr. Safford: The first is the polished white nutlet of Onosmodium occidental !e, a plant of the Borage family, belonging to a genus not far removed from Lithospermum. These beautiful little nutlets may well be called pearl-seeds, since when strung they must bear a close resemblance to small seed-pearls. Accompanying these is a small longitudinally grooved dull brown seed, somewhat resembling the seeds of the bead tree (Mclia azederach) in form. The terminal scar is removed by the perforation, and it has been impossible to identify this, or even to de¬ termine to what botanical family it belongs. Stone beads are of fine-grained white limestone, lignite, ser¬ pentine, quartz, hematite and alabaster. Most of them are large, no minute beads, such as those from Aztec 2 or the Upper Gila, 3 occurring. In shape they run from the flattened spherical type (figure 10, a) 4 to the more or less thickened discoidal form (figure 10, c). 1 See also Kidder-Guernsey, 1919, plate 70, k, a string of acorn cups. 2 Morris, 1919, p. 99. 3 Hough, 1914, p. 24. « Wrongly called “hemispherical” in our former report (p. 163). OF NORTHEASTERN ARIZONA 49 Most of the shell beads were made from olivellas simply by cut¬ ting off the end of the spire. There are in one of the strings (plate 25, f) seventy-five very thin disc-shaped beads, A of an inch in di¬ ameter cut, apparently, from the shell of a fresh-water clam (figure 10, f). The same necklace contains one hundred shell beads made from the curving wall of the large olivella (figure 10, g). The saucer-like form of these allows them to fit closely over each other when strung. Enormous quantities of identical beads are in the Museum’s collection from the Channel Islands, California. There are a few small bone beads (figure 10, h) apparently made in imi¬ tation of these. Pendants. These were less common in the burial cists of White Dog Cave than they were in the mortuary cave of Sayodneechee. 1 The single stone specimen (plate 17, h) is of a hard brown stone mottled with brownish green; the surface is highly polished and has a waxy texture. Four shell pendants were found, all of abalone; three are illus¬ trated in plate 17, c, d, e; the fourth is attached to a ceremonial object (plate 39, b). The largest (plate 17, c) is round and 2 inches in diameter. Jt has two perforations in the center from which radiate the four arms of an incised cross figure. Along the edge are two other round holes and three pairs of minute perforations. At the bottom of this disc there is a drilled hole which has been stopped up by inlaying a little piece of abalone shell carefully shaped to fit the aperture. The second abalone pendant (plate 17, d) is the reused half of a disc similar to the above; it fractured, apparently, along an incised median line. Traces of the favorite Basket-maker zigzag may be seen along the upper edge of the old break. The third specimen (plate 17, e) is a bit of the thickened rim of an abalone, the edges ground down and polished. Feathered Pendant. This object (plate 18, f) is described under the head of personal ornaments although it may have served some other, possibly ceremonial, function. It consists of nine two-ply twists of rawhide thong, seized with sinew to a loop of the same material. Small feathers, whose butts alone remain, were once fastened to the ends of the streamers. 2 1 See Kidder-Guernsey, 1919, p. 164. 2 Compare Kidder-Guernsey, 1919, figure 77. 50 BASKET-MAKER CAVES Ornament of Mountain-sheep Horn. This object (plate 17, i) is 3 inches long by 2\ wide. The convex side shown in the drawing bears, besides two pairs of drilled perforations, a double series of small holes which do not run through. Incised lines drawn be¬ tween the two series, seem to show a start at a zigzag decoration. The toothed ends of the specimen were produced by sawing broad notches along the upper and lower edges. The bottoms of the notches are well worn and smooth, but whether from general use or from friction of threads (supposing the object to have served as a weaving comb), we do not know. Deer-hoof Rattles. As in the preceding two cases, the identi¬ fication of these specimens (plate 17, j, k) as ornaments is open to question; a ceremonial use is quite as likely. One of them consists of the horny outer coverings of two large hoofs, attached to the ends of a buckskin thong. 1 The other shown in j is made of much smaller hoofs; these are fastened to the ends of thongs which themselves are looped over a slim pliable twig and held to it by a twining of fine cords. This is an incomplete specimen, as is an¬ other similar one (not figured, A-2930) which had, in place or de¬ tached, nearly a hundred hoofs. There is little doubt that the stringing together of these dry resonant hoofs was done to produce a rattling sound, but whether the assemblages were employed as belts, as fringes, or fastened to handles to form true rattles we have no means of telling. Unfinished Ornament. This object (plate 35, h, i), found in the general digging in White Dog Cave, is a neat example of two proc¬ esses in working stone: flaking and grinding. The specimen is a disc of grey flint, convex on both sides. It was first chipped roughly to its present form, then ground to efface the chipped surface. The grinding process was, however, not completed and there remain on either side marks of chipping, as well as numerous grinding facets. Tablet. Plate 17, a, shows, partly restored, a tablet-like object of compact white limestone found in Cist 6, White Dog Cave. The pieces fitted together have a length of 7 inches, but a number of fragments that could not be joined show that the original length was considerably more; the greatest width is 3 inches, the thickness i Modern Hopi hoof rattles are figured by Hough (1919, plate 22). Peabody Museum Papers Vol. VIII, No. 2, Plate 17 White Dog Cave a, Tablet-like object of stone; b. Neck ornament; c, d, e, Shell pendants; f, g, Object of stone; h, Stone pendant; i, Object of mountain-sheep horn; j, k, Hoof rattles. (About J.) OF NORTHEASTERN ARIZONA 51 uniformly A of an inch. The edges are rounded and all surfaces very smoothly worked down by grinding. The fine finish and the fragile nature of this object seem to indicate that it was used as an ornament. Head Ornaments. An object, of whose function we are not positive, but which was probably used to decorate the hair, was found on the breast of mummy 2, Cist 27 (plate 18, b). It con¬ sists of five neatly made bone pins, each inches long and a little less than f of an inch in diameter, fastened together side by side. The bindings are of sinew; the upper set is overwrapped with fine fiber cord evidently as a finish, since the string, though badly de¬ cayed, shows traces of a central red band. Projecting from the top, and held by the wrappings just described, were bundles of small feathers, of which only the butts of the quills and traces of the pile now remain. 1 Figure a, plate 18, shows a similar ornament from Cist 6, made up of three wooden pins each 10 inches long and J of an inch in diameter. A bundle of six wooden pins, each 8 inches in length and | of an inch thick, possibly ready to be made into a pair of ornaments like the ones just described, is figured in c. A number of finely fashioned but broken bone objects, of about the same size and shape as large knitting needles, some tied up in bundles, others loose, were found in the course of the excavations in White Dog Cave; most of them show signs of long use. These no doubt are also unassembled parts of head ornaments. There are in the 1915 collection similar broken bone pins. 2 Just how these contrivances were worn we do not know, but from their comb-like structure we judge that they were probably stuck in the hair, singly or in pairs. Some basis for this belief is found in certain Basket-maker square-shouldered pictographs de¬ picted with objects which may represent ornaments such as these protruding from their heads. 3 In the Peabody Museum there is a Paiute “warrior’s plume,” made of five wooden pins placed side by side and held together by colored strings woven about them in such a way as to produce a simple pattern; this specimen is not feathered, but is otherwise much like those from White Dog Cave. 1 A fairly well-preserved example from Grand Gulch is in the American Museum of Natural History, New York (cat. no. H-13375). 2 Kidder-Guernsey, 1919, plate 86, e. 3 Ibid., figure 101. 52 BASKET-MAKER CAVES In the Coahuila, Mexico, cave collection in the Museum there is an arrangement of six wooden pins which may be either a head ornament or a comb; we are inclined to think the former, as the same collection contains an object that is surely a comb, con¬ structed in an entirely different manner. The object shown on plate 18, d, may be an ornament, a pro¬ jectile for a dart game, or possibly a ceremonial object ; it is a thin twig with three small feathers seized to it at their butts and tips by sinew; the ends of the stick are broken off, so that its original length is unknown. Hair-dressing. Several of the mummies from White Dog Cave are in so good a state of preservation that their heads still retain the hair, dressed, probably, as in life On plate 19 are illustrated the various methods; figures a, b, c are drawn from mummies, and d, is restored from a scalp found in the same district in 1915. 1 Figure a, shows the simplest manner of wearing the hair, which in this case is cropped to an average length of 2 inches. The raggedness of this haircut is apparently the result of gathering to¬ gether and hacking off a single lock at a time. The individual in question was a female about twenty years of age found in Cist 22 (mummy 2). Figure b, shows the arrangement of the hair of an adult male from Cist 24. It is parted in the center from forehead to crown and falls loose on either side; that of the back of the head is gathered into a queue, the end of which is turned back on itself and wrapped for a space of 2 inches with a fine string. From the crown there hangs a lock the thickness of a pencil closely wound with string for nearly its entire length. 2 The end of this tress is bound up with the end of the queue. Where this lock grows from the scalp, the surrounding hair is clipped away for a little space. Figure c, is drawn from the head of a male about twenty-five years old, from Cist 22. The hair is arranged as follows: from a strip 1| inches wide straight back from the middle of the forehead the hair has been cut off close to the scalp. This exaggerated “ part ” terminates at the crown in a circular tonsure in the center of which there is a thin lock of long hair. The hair on either side 1 For pictures of this interesting specimen, and for a description of its preparation, see Kidder- Guernsey, 1919, plates S7, S8, and pp. 190-192. 3 As was noted on p. 13, a section of a similar lock wound spirally with a leather cord was found in Cist 6, White Dog Cave. Peabody Museum Papers Vol. VIII, No. 2, Plate 18 White Dog Cave a, b, d-g, Feather ornaments; c, Package of wooden pins, probably used in making feather ornaments. (About J.) OF NORTHEASTERN ARIZONA 53 of the “ part ” is gathered together and tightly bound 3f inches from the ends with fine human hair string; these tresses hang in front of the ears. The back hair, which is about 14 inches long, is similarly gathered together and bound near the end for a space of 2 inches. The lock from the center of the tonsure is included in this binding. The following description of the scalp shown in d, is quoted from our previous report: 1 “ A ‘ part ’ 1 inch wide, from which the hair has been clipped, runs up to a large semilunar tonsure at the crown. The brow tresses on either side are gathered together in ‘ bobs ’ that fall in front of or over the ears, and are tied up with wrappings of apocynum (?) string. The long hair from just behind the ton- sure is braided into a thin plait, the lower end of which is doubled back on itself and bound with hair string. The remainder of the back hair is made into a single short thick ‘ bob,’ string-wrapped, that falls to the nape of the neck.” As shown in the drawing this specimen combines features of both figures b and c, but is more elaborate than either. It seems to have been preserved as a trophy and for this reason, when discussing it in the earlier report, we were in doubt as to whether it represented a method of hair¬ dressing practised by the Basket-makers, or that of some tribe of which we had no knowledge. The side-bobs inclined us to the belief that it was a Basket-maker style, as Basket-maker picto- graphs are often shown with “ bobs ” on either side of the head. The finds from White Dog Cave serve of course to confirm this idea. Although many tribes shaved one portion or another of the head, and the thin scalp-lock was not an unusual thing, we can find no reference to analogous coiffures ancient or modern with the ex¬ ception of those of the Maya thus described by Bishop Landa: They wore their hair long, like women. On the top they burned a sort of tonsure; they let the hair grow around it, while the hair of the tonsure re¬ mained short. They bound the hair in braids about the head with the excep¬ tion of one lock, which they allowed to hang down behind like a tassel. 2 Judging from our material it would seem that the men dressed their hair more elaborately than did the women. 1 Kidder-Guernsey, 1919, p. 191. s ScheUhas, 1904, p. 617. 54 BASKET-MAKER CAVES CRADLES AND ACCESSORIES Rigid Cradles. It seems well, before taking up the several empty cradles in the collection, to describe the one case in which we have the baby with all its wrappings still in place. The bundle is shown as found in plate 4, g, and plate 21, c; its different parts are separated and spread out in the other figures of the former plate. The infant, enveloped in robes, is tied in by means of a criss-cross lashing. The binding cord is of human hair, four-ply and 5 feet long (plate 4, i); it is rove through a series of string loops that are attached to the sides of the cradle. The seven stout cords that may be seen hanging loose on the left side of the unwrapped bundle (plate 4, g and plate 21, c), and laid out separately in h, had prob¬ ably been used for hanging up or transporting the cradle; if the baby had not died so soon (it can hardly be more than a few days old), these cords would undoubtedly have been woven into a regular carrying strap like those shown in plate 23, k, 1. The outermost wrapping is a much tattered remnant of woven cloth (plate 4, a); it is described on page 63. The second cover is a fur-string baby blanket, measuring 17 by 17 inches. The body of the robe is of cords overlaid with strips of rabbit skin, its outer sides have a border, two strands in width, made of string, between the plies of which are caught bunches of long, coarse hair, probably dog. AYe have called coverings of this sort baby blankets because they were obviously woven to their peculiar bifurcated shape for the special purpose of leaving an opening at the place where they would otherwise constantly have been wet and soiled. Inside this blanket there was another of exactly the same size and shape; (plate 4, f) but, because it was to hold the baby itself, much softer and more carefully made. It is also of string, wound with strips of Huffy white fur from the bellies of rabbits. In handling this speci¬ men, one is so impressed by the freshness of the fur that it is diffi¬ cult to reconcile its perfect condition to its great antiquity. The mummy of the infant (plate 4, c) lay on this inner blanket with the lower side-pieces folded over its legs. It was provided with a loose bundle of shredded cedar bark to serve as a diaper (c). On the abdomen, covering the navel, was a pad (d), made of cedar bark sewed up in prairie-dog skin, the hair side out. This obviously acted as a binder to prevent rupture. The umbilical Peabody Museum Papers Vol. VIII, No. 2, Plate 19 Styles of hair-dressing as shown by the remains from Basket-maker caves. OF NORTHEASTERN ARIZONA 55 cord itself had been dried and was attached by a string to one corner of the outer baby blanket, so that it hung directly before the face of the infant; 1 it may be seen at the upper right-hand edge of the blanket (b). The cradle (i) is 14 inches long and 10 inches wide. The frame is a single unpeeled withe, f inch in diameter, bent into an approxi¬ mate oval. The body is made of fifty straight, unpeeled twigs placed close together; these run transversely and are fastened underneath the frame by a continuous lashing of fiber string. Along each side of the cradle there extends a stout cord, fastened to the hoop at intervals and forming loose loops for the attachment of the binder that held the baby and its wrappings in place. This cradle is much the smallest in the collection and is crudely made. It shows none of the careful finish and ornamental features of the specimens about to be described. The uncompleted carrying strap, the roughly put-together umbilical pad and the small size of the baby itself all point to the probability of birth having taken place before the usual elaborate “ layette ” was ready. There are five other more or less complete cradles in the collec¬ tion, all of which were found in White Dog Cave. Four had been buried with babies upon them but disturbance in some cases and decay in others rendered it impossible to recover the “ mummy bundles” in their original condition; the fifth cradle was found in rude Cist 54 (plate 5, a) that contained no bones. While these specimens are all much alike in general make-up, they differ con¬ siderably in details. As no account of a rigid Basket-maker cradle has yet been published, it seems worth while to describe each one of this exceptionally well-preserved lot. The handsomest cradle is the one illustrated, front and back, in plate 20, a, b. It is 23| inches long, by 14| inches wide at the broadest part. The rim is composed of two trimmed and peeled hardwood sticks \ inch in diameter, each bent into a U; the open ends of the two U-shaped pieces are spliced together with their sides overlapping a little; tight ligatures hold them in that position, and so envelop the joined ends that they cannot be seen. The 1 As recorded by Catlin in 1842, Vol. II, p. 133. The custom of preserving the cord as a charm was practised by many tribes, particularly those of the plains. The Ute, Dakota, Ara- paho, and Gros Ventre enclosed the dried cord in more or less elaborate coverings of skin orna¬ mented with quill or bead work and fashioned usually to represent reptiles. These were hung on the front of the cradle (see Kroeber, 1908, pp. 166,167). 56 BASKET-MAKER CAVES body of the cradle is made of two series of slim willow twigs, from which the bark has been scraped. The transverse rods are ninety- nine in number; they are laid as close together as they will fit and are fastened at their ends to the under side of the frame by a con¬ tinuous figure-eight lashing of yucca string. This binding is over¬ wrapped with soft fiber, until the slightly protruding ends of the rods are entirely hidden, and each side of the cradle is built up into a soft, bolster-like roll an inch in thickness; this in turn is sewed up in a cover of deer or mountain-sheep hide dressed with the hair on. The hard sides of the hoop and the sharp projecting rod ends are thus completely padded and form a sort of rim along the two edges of the cradle on its upper surface. The second, or longitudinal, set of rods consists of five twigs running up the middle of the transverse rods and attached to them by a lashing of heavy sinew, so arranged as to produce the zigzag design seen in the photograph. The ends of the longitudinal twigs are fastened to the head and foot of the hoop in some manner which cannot be made out, because the attachment is padded and tightly sewed up in a hide covering. Tied around the bottom of the hoop there is a horse-shoe shaped roll of cedar bark, which must have formed a kind of soft platform for the baby’s feet to rest against when the cradle was held up¬ right. A series of human hair strings are caught into the “bolsters” along the sides of the cradle; these, like the loops on the specimens first described, were to hold the laced binding cord. At the head and foot are much longer loops, designed, apparently, for suspend¬ ing the cradle in a horizontal position. 1 A double yucca string is tightly stretched across the upper surface of the cradle about 8 inches above the foot. From just below this string to the foot, the cradle is much discolored by the excreta of the baby. The purpose of the string was probably to hold in place the rather inefficient diaper-bundles of cedar bark or fiber. Plate 20, c, d, illustrates a cradle very similar in shape to the above; its measurements, 23£ by 14 \ inches, are almost identical; the hoop is also made of two pieces tied together at the sides. The backing is of reeds instead of twigs; there are eighty-three in the transverse series and twenty-two in the longitudinal, the latter is secured to the former by narrow rawhide thongs whose emergences 1 See Saunders, 1912, photograph facing p. 86. Peabody Museum Papers Vol. VIII, No. 2, Plate 20 White Dog Cave a, b, Front and back of cradle, Cist 35; c, d, Front and back of cradle, Cist 54. (About 1/9.) OF NORTHEASTERN ARIZONA 57 produce a pattern of diamond figures. The longitudinal reeds were once attached to the head and foot of the bow, but their ends are now missing. The sides are padded with fiber and covered with hide, and there are the remnants of a cedar-bark foot rest. The ends of a diaper string are present, but there are no side loops for the laced binding cord. The remaining three specimens are more nearly oval than the two preceding. The largest one (plate 21, b) is 25 inches long by 12 inches wide. Viewed from the side it is rocker-shaped, but this curve is probably due to warping. The frame and its side-padding (mostly decayed) offer no new features, nor does the method of attachment of the seventy-nine transverse willow backing-rods. As will be seen in the plate, the longitudinal rods are differently arranged; they are in two sets of six each, spaced well apart and curving away from each other as they approach the head of the cradle where each set is bent about the side of the frame and tied back on itself; the lower attachments are gone. The diamond- pattern lashings that hold the longitudinal to the transverse rods are of strips of rawhide. Between the two longitudinal sets, and also alongside them, the transverse rods are bound together by a sort of over-eight-under-eight twilling of leather thongs painted red. Side loops and diaper string have disappeared; the mark of the latter, however, can be made out on the backing, and below it there are as usual heavy stains and caked mud. The cradle shown in plate 21, a, is from the same cist as the fore¬ going. It is an elongated oval, 19 § by 10 \ inches. Of the two sticks bent to form its frame, the upper one is peeled, the lower unpeeled. The sides are padded into the usual long rolls, but there is no evidence that they were ever encased in skin; no loops or diaper string remain. The transverse twigs are ninety-eight in number; the first seventeen, counting from the top, are in natural color; then comes a row of eight rods dyed black, then eight in natural color, eight black, eight natural, eight black, eight natural, and eight black; the last twenty-five to the bottom are undyed. The eight longitudinal twigs are not attached to the transverse ones by the usual ornamental bindings. They are turned about the frame at the head of the cradle and tied back on themselves; at the bottom they are cut off at the level of the last transverse element and their ends are made fast to it by a row of twined yucca string. 58 BASKET-MAKER CAVES The last of the three oval cradles is 21 § inches long, and 11^ inches across. The two sticks of its frame are unpeeled. There are seventy-seven transverse rods (willow twigs, scraped and trimmed as usual) and seven longitudinal ones, bound to the former with the conventional diamond pattern of thong-emergences; their attachments to the top and bottom of the frame have been broken off. The frame padding along the sides is of string and yucca fiber, and was once encased in hide. There are no side-loops, but the diaper string is still in place, stretched tightly across the upper surface of the cradle at a point one-third of the distance from the head to the foot. Flexible Cradles. These are of two types. The first has a rim made of a long thin bundle of grass rolled tight, tied with yucca leaves and bent to the same shape as the wooden hoop of the rigid cradle. The body or filling is a rough mesh of yucca leaves. The second type is a sort of mat made from long strips of cedar bark held together by twined-woven rows of yucca leaves; the edges of the mat are turned up and fastened together by a yucca network. Both types are illustrated and more fully described in the report on the 1914-1915 expeditions; 1 all the specimens recovered in 1916-1917 were very fragmentary, but enough of them were found to show that these cradles were in common use. Umbilical Pads. During the early part of the 1916 season there were taken from the graves of infants a number of flat pads, made by sewing up various substances in covers of prairie-dog hide. Their use, at first doubtful, was made clear when the well-preserved baby burial from Cist 13 was examined, and a similar pad (plate 4, d) was found lying against the navel of the infant; a second case (infant from Cist 35) was discovered later. It was then obvious that all these specimens had been used as are our modern “ binders” to prevent umbilical hernia by exerting pressure on the navel of the new-born child. Each of these pads has a light but rigid or semi-rigid core, most commonly made of five or six corncobs cut to equal length and bound together side by side; several examples are whittled from slabs of yellow-pine bark (plate 22, c); 2 still others consist of a rope or tight twist of cedar bark, coiled and sewed to itself to form 1 Kidder-Guernsey, 1919, pp. 165, 166; plates 71, b; 72, a, b. 2 The piece of bark figured in our first report (Kidder-Guernsey, 1919, plate 85, b), and classed as problematical is one of these. Peabody Museum Papers Vol. VIII, No. 2, Plate 21 White Dog Cave a, b, Cradles; c, Cradle containing mummy of child, Cist 13; d, Package containing mummy of child, Cist 35. (About 1/10.) OF NORTHEASTERN ARIZONA 59 a small oval mat (plate 22, b); in one case a thin slab of sandstone is used. 1 The crudest were wads of cedar bark or grass. The cores were wrapped and padded with shredded cedar bark, more or less thickly according to their hardness, and were finally enclosed in prairie-dog skin covers prepared as follows (plate 22, a); the com¬ plete hide was trimmed by cutting away the feet and tail, and shaped into a long bag with the fur outside. The padded core was placed in the bottom of this, the upper part folded down, and the whole neatly sewed up with sinew or fine fiber thread. There is one specimen (plate 31, a) to which is still attached the narrow human hair string band that formerly held it in place against the abdomen of the infant. BASKETRY Coiled Basketry. The Basket-maker culture was so named by the Wetherill brothers because of the abundance of baskets found in the graves. The burials of this people excavated by the Pea¬ body Museum expeditions in Marsh Pass ran true to type in this respect as in all others; and, wherever the cists were protected from moisture and undisturbed by ancient looters, fine specimens were always to be found, while throughout the general digging in the caves fragments of worn-out baskets were encountered in great abundance. All the specimens recovered were of the coiled variety, no case of twining, checkerwork, or wickerwork having been found; a single twilled example, in reality more like a flexible pouch than a true basket, will be described later. In weave the coiled baskets form a very homogeneous group; they are made over a foundation consisting of two slim osiers laid side by side, with a padding or welt of yucca fiber or shredded roots. The sewing elements are wooden splints averaging a little less than | inch wide; they enclose the rods and the fibrous padding bundle and also pass through about half of the bundle of the coil below. It is this gripping of the bundle of the lower coil which alone holds the fabric together, as the stitches of one coil never interlock with those of the coil below them. 2 While the weave is so solid and compact that many of the 1 For a description of this specimen see Kidder-Guernsey, 1919, p. 192; its use was then unknown to us. 2 For a diagram of the weave, see Kidder-Guernsey, 1919, figure 80. 60 BASKET-MAKER CAVES better pieces must have been watertight, it never attains the fine¬ ness of texture seen in many California coiled baskets. These ancient weavers strove, apparently, for strength and serviceability rather than for refinement of technic. No more stitches than necessary were used; hence the relatively great width of the in¬ dividual sewing splints and their broad spacing, which allows the foundation to appear between them. The average tray basket has five coils to the inch and nine to eleven stitches along each inch of coil; the finest specimen has eight coils and twelve stitches; the coarsest, a fragment from a large pannier, has coils \ inch wide and six to seven stitches to the inch of coil. The edge bindings of all the baskets save one are in simple wrapping; the exception is a bowl-shaped piece (plate 23, i) in which the entire rim is finished in “ false-braid ” as in Navajo baskets. 1 Our specimens fall into the following five classes: 1. Trays 4. Water baskets 2. Bowls 5. Trinket baskets 3. Carrying baskets Trays. This is by far the commonest type. The examples are very flat, and run from 12 to 24 inches in diameter. They were probably used for the serving of food, and perhaps in gambling. One tray (plate 23, j) obviously had another purpose; about its rim at equal distances apart were four loops, two of which remain (the others were in place when found, but soon crumbled away). Each loop is made of a twig tied into a circle 2 inches in diameter and is attached to the rim of the basket by a short buckskin thong. The whole interior of the tray shows much wear, particularly severe at the bottom where, indeed, it had begun to give out and was re¬ enforced by overstitching with new splints, which themselves were partly worn through. The outside and bottom exhibit no wear at all. It seems likely that this basket was suspended by the loops and used for the simultaneous hulling and winnowing of seeds too delicate to be shelled in a mortar. The process might have been to keep a stone rolling among the seeds by shaking the suspended tray, and to blow off the hulls as they were detached by 1 For details of this stitch, see Mason, 1904, figure 197. A Basket-maker basket from Grand Gulch, in which the last inch of the terminal coil is done in “false-braid” is mentioned by Pepper (1902, p. 16); exactly the same treatment appears in a basket from Step House, Mesa Verde (Nordenskiold, 1893. plate XLIV, 4); Diegueno and Kawia (southern California) tray baskets also have the last inch of coil in ‘ false-braid” (Peabody Museum Collections). Peabody Museum Papers Vol. VIII, No. 2, Plate 22 White Dog Cave a, Covering for umbilical pad; b, c. Umbilical pads. (About f.) OF NORTHEASTERN ARIZONA 61 the bruising of the stone. This explanation is, of course, pure guesswork, but it seems to account satisfactorily for the presence of the loops and for the excessive wear on the inside. Bowls. As will be seen in the illustrations (plate 23, a, c, f) these baskets are of lesser diameter than the trays and of much greater depth; their bottoms are flat and the sides rise more or less steeply. The largest is 14 inches wide at the mouth, by 8 inches deep. We believe that some of the larger bowls were used for boiling by the hot stone method, as two examples are heavily daubed with a mixture of mud and ashes applied, apparently, to render them watertight; they also have a soiled and battered look and many patches that indicate hard use. Carrying Baskets. These are the largest of the coiled baskets, measuring 28 to 30 inches in diameter at the top, by 17 to 20 inches deep. They have pointed bottoms, oval in cross-section; and widely flaring upper parts (plate 23, k, 1). By actual count of coils and stitches to the inch these are the coarsest of the baskets, yet they are as carefully and regularly woven as the finest; are very strong, but flexible enough to adapt themselves to the curves of the neck and shoulders of their bearers. There is no doubt that they served as panniers for carrying loads on the back; their shape and the use of similar forms by modern tribes are sufficient indica¬ tions. The identification, however, is rendered certain by the fact that they all have pairs of loops, usually of human hair string, worked into their sides at the proper height for the attachment of head bands. In two specimens these bands are still in place. The common use of these panniers to cover interments is, of course, a secondary one. Water Baskets. The excavations of 1916-1917 produced no whole specimen of this type, yet fragments of oval bottoms of a finer weave than is usual in panniers seem to indicate that such baskets were not rare. A fine example from Cave II, Kinboko, is figured in our former report. Dimensions: total height 17 inches, greatest diameter 14f inches, orifice 4| inches. It has an elongated base, oval in cross-section. The upper part flares out and becomes round; it is constricted again at the top, and the orifice is small. There does not seem to have been a neck, but there is some evi¬ dence that there was once a string-hinged cover. On opposite sides, just below the point of greatest diameter, are pairs of carry- 62 BASKET-MAKER CAVES ing loops made by twisting into a heavy cord eight or ten two- strand human hair strings. The entire inner surface of the basket is thickly pitched with pinon gum, and the same material has been daubed on such parts of the exterior as had begun to wear through. A design of small stepped units may be faintly made out on the upper curve. 1 Trinket Baskets. These are neatly made little receptacles with round bodies and small orifices. The range of sizes and shapes is shown in the illustrations (plates 23, h, and 24, d). It is prob¬ able that these baskets were put to a variety of uses; many of those found in the graves contained small trinkets of one sort or another. Decoration. Baskets of all the above types were ornamented with designs in black. Red elements, reported by Pepper 2 in Grand Gulch baskets, are not found in our collection. The designs are of great interest because they are without much doubt the oldest examples of basketry ornamentation that have yet come to light in the United States. Furthermore, they illustrate the deco¬ rative art of a people who preceded the pottery-making tribes of the region, and so may eventually be expected to throw light on the vexed question of whether or not southwestern pottery designs developed from those of basketry. We give, accordingly, all the decorations that are sufficiently well-preserved to copy (plate 24). These, together with the fine series of baskets figured by Pepper, 3 will give the reader a very good idea of the make-up of the designs. Descriptions of the patterns tell no more than do the pictures, and any attempt to supply symbolical meanings to designs as old as these would naturally be pure guesswork. We have made notes towards a comparative study of these and the designs of the baskets from the Plateau and Pacific Coast areas, but they are as yet far from complete, nor have we space in this publication to present the mass of data which has already accumulated. It may be said, however, that the art as a whole seems to find its nearest parallel in that of the central and northern California tribes. In technic, on the other hand, the baskets most closely resemble those of the Paiute. 1 Kidder-Guernsey, 1919, p. 170 and plate 78. 2 1902, p. 15. 3 Ibid., the same pictures may also be found in Mason, 1904, a more accessible publication, plates 84, 104, and 205 to 211 inclusive. Peabody Museum Papers Vol. VIII, No. 2, Plate 23 Baskets: All from White Dog Cave with the exception of h, which is from Cave 1, Kinboko Canyon, Marsh Pass. (About 1/16.) OF NORTHEASTERN ARIZONA 63 Twilled Basketry. The only specimen in this weave is a flexible bag-like basket of yucca leaves with flattened spherical body and small mouth. Although it is fragmentary, the following measure¬ ments are approximately correct: width 8| inches; depth 4§ inches; diameter of aperture 4 inches. It is made of entire leaves of Yucca angustifolia; the butts of the leaves are turned outward over a heavy fiber cord that rings the mouth of the basket, and are fastened by twined strings. The long ends of the leaves are then plaited together, over-two-under-two, to form the body. The bottom is not woven, the last couple of inches of the leaves being simply laid across each other and tied in that position with string (plate 23, b). Although the over-two-under-two weave is the same, this speci¬ men is entirely different from the twilled ring baskets so abundantly found in cliff-houses. 1 The latter are always bowl-shaped and have a wooden hoop at the edge. They are fabricated upwards from the bottom; not, as in this case, downwards from the rim. No trace of ring baskets has yet come to light in our excavations in Basket- maker caves; a bit of twilled work found in Cave 1, 1915, 2 was probably part of a flexible bag-basket like the present one. TEXTILES Plain Weaving. As the collection of Basket-maker textiles described in our first report contained no example of straight over- and-under weaving, we believed that the Basket-makers practised but two technics, namely twining and coiled-netting (coil without foundation). Among the material collected in 1916-1917 there are, however, three pieces of plain over-and-under weave. The largest of these is the cloth outer wrapping of the infant from Cist 13, White Dog Cave. Though much torn and showing long use, enough remains so that by arranging tattered ends of selvage in their proper positions one dimension is shown to be 27§ inches. The other, based on extending the design to a symmetrical termination, would be 26 inches. It is probable that allowing for error in these measurements the original piece was square. The general appear- 1 See Kidder-Guernsey, 1919, p. 108 and plate 43. The specimens figured by Pepper (1902, p. 23) are probably not Basket-maker, particularly as one of them was found filled with beans; the basket shown on p. 25, however, seems to be identical with the one under discussion. 2 Kidder-Guernsey, 1919, p. 167. 64 BASKET-MAKER CAVES ance of the fabric is the same as that of the twined-woven bags both in color and design, the difference in technic not being ap¬ parent at first sight. The weave is rather coarse, having nine warp and fifteen weft strands to the inch. Both warp and weft are of a uniform sized two-strand twist of rather coarse vegetal fiber pre¬ sumably yucca. As far as it is possible to work it out from the scant material at hand the weave is as shown in the diagrammatic draw¬ ing, figure 11, b. Details as to the manner in which the warp edge is finished appear in figure 11, b, and plate 25, c. The warp ends are cut close and the weft kept from unraveling by a buttonhole stitch. The edge running parallel to the warp is finished by twining two fine strands of human hair through the loops that result from turning back the weft for a new start; this also is illustrated in figure 11, b. In the photograph, plate 4, a, there is seen at one point a circular hole, cut in the fabric, and finished all around by overcasting with fiber thread. The design (plate 20, b) consists of a series of large rectangles arranged in three rows, the two outside rows red, the center one black. The units average 2\ inches long by inches wide. Separately dyed elements were not introduced to produce the design; but apparently, when the weaving reached a point where a change of color was desired, the weft strand was thoroughly rubbed with color for the required length and then woven in. The warp cords show little color, such as appears on them probably re¬ sulting from contact with the weft. It is possible that the finished piece may have been treated with some mordant to fix the dye. The second example of this weave is a fragment 12 inches long by 2 inches wide in very bad condition, one end showing darning. It is also from White Dog Cave. There are traces of a broad design in red, the exact character of which cannot be determined. The piece appears to be a part of a blanket very similar to the one just described. There remains a short section of one edge finished with a thread of human hair twined through the weft loops. The third piece, from Cave 11, Sagiotsosi, was found with the disturbed burial described on page 37. It is very even]}' woven with fourteen warp and twenty-one weft strands to the inch. The fragment has a length one way of 12 inches, and is a part of one corner of the original piece, so that two edges remain. Both warp and weft edges are finished in the same manner as the one first de- Peabody Museum Papers Vol. VIII, No. 2, Plate 24 Baskets: All from White Dog Cave with the exception of d, which is from Cave 2, Kinboko Canyon, Marsh Pass. (About 1/16.) OF NORTHEASTERN ARIZONA 65 scribed: a buttonhole stitch of fine string, and human hair twining thread respectively. The design is in red and black, and so far as it can be traced is shown in plate 26, c. It is painted, not woven, and the color was applied only to one side of the cloth; the red pigment has soaked through the fabric and the red parts of the design appear faintly on the back. The black paint has not soaked through at all. To the corner is tied a dressed leather thong, which leads us to think that it may have been part of a garment. These fabrics remind one strongly of the Coahuila cave textiles, many of which are large poncho-like blankets woven in the same a, Detail of weave, fur cloth blankets; b, Plain woven cloth, detail of weave and selvage. way as these, and also have one edge finished with the buttonhole stitch. The latter resemblance seems significant, since we have not been able to find in the Museum collection textiles from any other region so finished. The designs, it is true, are different, though some of the elements seen in the Basket-maker twined- woven bags are also found in the Coahuila blankets. The zigzag lines seen in the second specimen (plate 26, c) are very similar to the zigzags painted on the breasts of certain square¬ shouldered Basket-maker pictographs from the Monuments. 1 This resemblance has suggested to us that these woven fabrics may have been used as shirts. Twined Weaving. The bags illustrated on plates 26, 28, and 30 form one of the most interesting groups in the collection, not only 1 Kidder-Guernsey, 1919, p. 197, figures 100, 101. 66 BASKET-MAKER CAVES because of the excellence of their manufacture and the variety and beauty of their decoration, but also because they are so peculiarly characteristic of the Basket-maker culture. We have, fortunately, a large amount of material: complete bags to illustrate size, shape, and design; and great numbers of rags and fragments to make clear the details of technic. The bags are flexible seamless sacks with full, round bodies and long, gradually constricted necks (plate 26, a, d). They range from \\ inches to 2 feet or more in length. All are made in the same way, of close twined weaving; the majority of specimens have both warp and weft of two-ply apocynum string, though some have yucca warp and apocynum weft. The combination of apocy¬ num warp and yucca weft is rare. Our study of the weave was begun by examining the bottoms of the bags in order to make out how the preliminary “ set-up ” of the warp cords was accomplished. By dissecting several frag¬ mentary specimens we found that there were two methods, one common, the other rare. The former was as follows: six long strands were laid across each other, three above and three below (figure, 12, a); the middle strand of each set of three runs out straight, the others are bent so that their ends radiate from the common center. There are thus produced twelve original warps. The second method consists of twisting three strands about each other and then bending their ends so that they radiate and form six warp cords (figure 12, b). The above systems are very simple and practical, and avoid the ugly lump and the potential weakness in the fabric which would have been the result of knotting the warps together at the base. The method of inserting the weft also obviates knotting: a single long string is worked over and under the radiating warp cords close about their common center; this is shown slack in figure 12, a, b; in reality it is pulled up very tight and holds the warp firmly to¬ gether. When a circuit of the spoke-like warps has been made, the two ends of the weft string of course come together; they are then combined into a single strand of twined weaving, which con¬ tinues spirally around and around to form the body of the bag fabric. To return to the warp-skeleton. Many large bags have as many as three hundred and fifty warps at their point of greatest diameter. Peabody Museum Papers Vol. VIII, No. 2, Plate 25 a, Pottery, Cave 6; b, Twined-woven fabric, White Dog Cave; c, Plain woven fabric, Sagiotsosi Canyon; d, Coiled netted fabric, White Dog Cave; e-h, Necklaces, White Dog Cave. OF NORTHEASTERN ARIZONA 67 It is obvious that these could not all come together at the bottom of the bag; hence the base begins with six or twelve warps only (as described above) and sets of new cords are introduced as the original ones radiate away from each other. Upon the number of new warps depends the size of the finished bag; and upon the rapidity of their insertion depends the degree of flare imparted to the base. If many new warps are added close to the bottom, the latter will naturally be very flat; if they are put in more gradually the bag will have an egg-shaped base. Figures 13, a, b, illustrate Methods of arranging and binding warp cords when beginning the construction of twined-woven bags. The weft cords are shown in solid black. this; each one represents, diagrammatically, a circle about 1} inches in diameter at the bottom of a bag. In figure 13, a, the original twelve warp cords are multiplied to forty-eight by two series of insertions, the first or inner series consisting of twelve new cords, the second of twenty-four. In figure 13, b, the same total is arrived at, but there are three series of insertions; the first of six, the next of twelve and an outer one of twenty-four. Figure 13, c, shows an area of bottom no greater than in the former specimens, but containing seventy-six warps, set in as follows: original series twelve, first insertion series twelve, second series fourteen, third thirty-eight. The weft in all three cases is woven in with approximately the same degree of tightness; hence the warps of a and b are pulled close to each other and the bags have 68 BASKET-MAKER CAVES narrower bottoms than in c, where the quicker insertion of warps allows the base to grow rapidly broader. We have not yet mentioned the actual method of inserting new warps. Two ways were employed. In one (plate 27, b) the string to be added was looped and laid between two of the old warps (b, b') thus forming two new ones (a, a'); the first two or three turns of the weft (c. o') attach the new strands to the old warps on either side of them holding all firmly in place; the next turn of weft (d) takes in each new element separately and the weaving continues normally. In the second method (plate 27, a), the strand to be added was doubled into a loop, making, as before, two new warps; the string Methods of inserting new warp cords to increase diameter of bottom of bags. at the bend of the loop was twisted apart into its two component plies and one of the old warps (bi was threaded through the re¬ sultant opening; the loop (a. a') was then slid up the old warp and brought close against the last woven turn of the weft (c), thus producing a pair of new warps (a, a') one on each side of the original one