Archeological Investigations

ONYX CAVE (9)

Five miles southwest of Arlington, near the Boiling Spring in the Gasconade, is Onyx Cave, so named because much workable stalagmite occurs in it. It has a number of branches, some of which have been explored for several hundred yards without coming to the end. The entrance is 90 feet in width. A pile of talus at the front, lying partly inside the cavern, reaches nearly to the roof; it has a height of 26 to 28 feet above the level of the wet, muddy floor. Drainage is through a small aperture in the north wall, whose outlet is not known. Apparently the bedrock lies at a considerable depth; it is not visible at any point in the steep ravine leading from the mouth of the cave to the river. Formerly a large quantity of ashes covered much of the inner slope of the talus, where it is protected from the weather; but most of them have been hauled away to scatter over the fields. They extend to a greater depth than any digging was ever carried. The cavern has long been a refuge for stock, and this, with the trampling of many visitors, has mingled all the superficial deposits, so that, while ashes may be seen mixed with the débris, no ash beds are now to be found.

There must be a very pronounced cavernous condition in this vicinity. At a number of places, even extending to a distance of 2 miles from Onyx Cave, the passage of a wagon produces a rumbling sound, indicative of a cavity at no great depth. There are also many sink holes, some closed, forming ponds, others with free openings. They are so numerous that no one of them drains any considerable area. The largest of these sinks measures from top to top of its slopes about three-fourths of a mile long and half a mile wide. Around much of its margin are vertical cliffs; there are few places where descent is practicable. It is 300 feet deep, perhaps more; for when the Gasconade, more than a mile away, is at flood stage the water from it, backing through an underground passage, breaks in at two different points not at the same elevation, and covers the nearly level floor of the depression, about 15 acres in area, to a depth or 15 to 20 feet.

Another sink, near this, is conical in form, a fourth of a mile across and more than 200 feet deep.

GOAT BLUFF CAVE (10)

Goat Bluff Cave, 4 miles west of Arlington, on the left bank of the Gasconade, is at the foot of a vertical cliff 50 feet high, the slope above rising about as much higher to the crest of the ridge. A few yards to the west is a slight ravine through which, with a little effort, the top of the hill may be reached. In front, the declivity, while steep as earth will lie, furnishes fairly easy passage to and from the river which lies 200 feet below.

The entrance to the cave is an arch 30 feet high and 75 feet wide, facing a little east of south. The width holds nearly the same for 90 feet, whence it rapidly contracts to 20 feet; the roof meanwhile descending to 10 feet above the floor. The extreme rear of this chamber is nearly filled with large blocks of stone. At the front part the floor is several feet higher along the west wall than at the east; this condition being due to the combined action of accumulation from the ravine above mentioned and erosion by a little rivulet which emerges from a crevice 30 feet within the entrance and flows at the foot of the east wall. Beyond this the floor is practically level across the inclosed space, with a slight and uniform ascent toward the rear. No evidence of rock bottom appears at any point.

A preliminary cut at the outer margin of the cave showed two distinct, sharply separated strata. The lower is a red or yellow clay containing much angular gravel such as usually results from disintegration of limestone in which chert is abundant. Above this is a deposit of very loose fine material. Toward the rear the upper deposit had been disturbed by "curiosity seekers," who reported finding much evidence of prehistoric occupation, such as ashes, charcoal, fragments of pottery, and worked flint, as well as several skeletons, the latter "in a sitting position." The last part of this statement is a mistake. The bodies were closely flexed and placed on the side; the bones settled to the bottom of the grave, while the skull, if intact, is reached first by excavators and the conclusion drawn at once that it is "on top of the other bones." This error of observation is quite common among relic hunters, and is not unknown among student investigators.

In order to dispose of material removed in excavating, it was necessary to start a trench from the slope outside the mouth of the cave. As it progressed the substratum of clay became wetter and more difficult to dig. At 40 feet from the beginning, where the trench was 11 feet deep, the seeping water accumulated until it covered the bottom of the trench, so that no greater depth could be reached. A crowbar forced downward for 18 inches, as far as it could be driven, did not reach solid bottom. Not the slightest trace of human agency was found anywhere below the top of the clay, and from this point excavations were confined to the upper stratum, to which alone the following description is applicable.

This deposit was composed partly of fine loose earth, probably carried in by the wind and on the feet of persons and animals; partly of roof dust; and partly of ashes. A considerable portion of it was roughly stratified in layers of varying extent and thickness, though much of it was irregular, and it was mingled throughout with campsite débris. Occasional layers of roof dust several feet across in any direction and of varying thickness, from a faint streak to 6 inches, so closely resembled ashes that many persons could not be convinced of its true character. Its occurrence in this manner indicates that during considerable periods the cave was unoccupied, or at most used only as a temporary refuge. The intermittent character of occupancy is also shown by the distinct segregation of numerous successive layers of kitchen refuse.

About 10 feet within the point where a vertical line from the front edge of the roof would meet the floor the skeleton of a very young infant was found above and in contact with two thick angular blocks of limestone weighing 300 to 400 pounds. These rested on the red clay and had fallen from the roof. The thickness of earth above the bones was about 3 feet.

Ten feet farther in, on the clay floor, under almost exactly 5 feet of undisturbed material, were five flat stones. Three were of sandstone, the largest about 25 pounds in weight, such as can be found in place only on top of the hill. They were carefully arranged for use as a fire bed; on and around them were potsherds, flint chips, animal and bird bones, and a bone awl. This was the greatest depth at which artificial objects were found; and their position shows them to be as ancient as anything discovered.

At 25 feet in an interesting find was made. Eighteen inches below the surface of the floor, in a mass of mingled charcoal, ashes, mussel shells, flint chips, and other aboriginal refuse, was a small piece of glass, apparently part of a bottle, shown in figure 5. Above it and extending for several feet on every side was an unbroken stratum of root dust from 2 to 4 inches thick. Above this, in turn were several thin, undisturbed layers of camp refuse, about 6 inches in all, and then 6 inches of the loose, incoherent surface earth. This discovery is susceptible of two interpretations. One is that between the date when Indians could procure articles from the whites and the date at which they abandoned this fireplace there was time for the accumulation of the given thickness of disintegrated material from the roof, the cave, or at least this part of it, not being used meanwhile for a habitation; then for the accumulation of several distinct layers of camp refuse; and finally for the depositing of the cave earth over it all. This hypothesis is unreasonable. While the rate of formation of either roof dust or stalagmite is extremely variable, so that it is not safe to predicate a definite antiquity for objects found beneath even a considerable thickness of either, at the same time the small area involved precludes the idea that a number of occupants sufficient to account for the volume of débris could have lived here unless we allow a much longer period than would necessarily elapse within the dates indicated. The other, quite plausible, interpretation is that the glass was dragged to the spot by a ground hog or other animal whose runway had become obliterated by settling of the loose material through which it was made.

The only purpose of elaborating this subject is to guard investigators against attaching too much importance to an article found under such or similar conditions, whether it be a "palaeolithic type," or an "object undoubtedly of European origin."

Thirty-five feet in, under three flat slabs whose upper surface was a little more than 3 feet below the floor, was an adult skeleton, on the back, knees flexed to the chest. The body had been laid in a cavity dug in the clay to a depth of 6 inches. The bones were well preserved and fresh looking, but light and fragile.

Forty feet in, 3½ feet down, was a flat stone under which were two skulls. One, shown in plate 6, was perfect, with a full set of sound teeth; from the other, seen in plate 7, the lower jaw was missing. No other bones were found except two cervical vertebræ, belonging to the smaller skull. Undisturbed stratified ashes and roof dust were 30 inches thick above the stone.

To this point the trench was not dug to a greater width than 15 feet; it was now gradually extended to a width of 40 feet to include most of the central portion.

Sixty feet in, in the upper part of the clay, like all the human bones discovered, was a skull with the scapulæ, a few ribs, and one arm bone. The lower jaw was missing, and two phalanges were inside the skull. With the scapulæ was one of a much smaller person. Eighteen inches from these bones, and 6 inches higher, was part of a lower jaw.

At 50 to 60 feet in, on the clay stratum, lay a slab 10 to 12 feet across and of varying thickness up to 18 inches or more. It fell from the roof so long ago that the latter is worn and smoothed above it in much the same way as at other parts. At the east edge of this slab was a skull so soft and crushed that it could be taken out only in small fragments; the teeth were very slightly worn, though of large size. A few traces of other bones were found; not enough to identify. At the north edge of the slab were two skulls, one of which is shown in plate 8; the other, which belonged to a young person, is given in plate 9. The limb bones, scapulæ, and hip bones, with a few others, were in a small pile at one side; but neither lower jaw, no ribs, and only a few vertebræ were found.

About 65 feet in, near the west side, an inverted pot which shows no marks of use was found in a mass of ashes filling a cavity the size of a half bushel, which had been dug in the upper deposit. Scattered here and there among the ashes were also some mussel shells and broken deer bones; but the presence of these was probably not intentional, as the whole arrangement seemed to have the nature of a votive offering. This was the only perfect vessel found in the entire course of the explorations. It is of the ordinary "cocoanut form," and is represented in figure 6.

Seventy feet in was a skeleton, on the left side; the bones were soft and came out in small fragments. This was fully 6 feet below the present surface, but some of this earth was piled up from earlier excavations.

Beyond this point the ground had been dug over to such an extent that further examination seemed useless, and the work was concluded.

Throughout the deposit of black earth, ashes, and roof dust were scattered irregularly arrowheads and knives of flint, some types of which are seen in plate 10; mussel shells; fragments of bones from food animals; bone perforators, some of which are shown in plates 11 and 12; potsherds; hammers; pestles; two or three mortars; a grooved stone ax of granitic rock, presented in figure 7; and an abundance of flint chips.

There is a small cave near the top of the bluff facing the Gasconade, a short distance above the mouth of Little Piney. Within a few yards of the entrance earth and rock carried in from a sink on top of the hill fill the cavity to the roof. Water runs through after every hard rain.

Three small cairns, built of small stones, stood on the point of the bluff at the junction of Little Piney and the Gasconade. All are destroyed.

On the edge of a high cliff over the Gasconade, 2 miles north of Arlington, are three cairns, destroyed.

In Bryant's Bluff, facing the Gasconade 3 miles below Jerome, are two rock shelters, neither of them more than 20 feet across in any direction. In both are shells, bones, and pottery; a rough stone hammer was found in one. Exposure of bedrock on the outside shows that the earth deposit in either is not over 2 or 3 feet deep.

On top of Bryant's Bluff are four cairns, all of them torn up. The extreme limit of the scattered stone is about 20 feet; so the cairns were probably 12 to 15 feet in diameter.

At the mouth of Turkey-pen Slough, 4 miles north of Arlington, is a terrace with steep banks on two sides, next to the river and to the slough. On this stood a village. Three house sites are plainly marked by the refuse around, and there may be others; vegetation is very dense. Mussel shells and burned stones are abundant, and many flint implements have been picked up.

CAIRNS AT SUGAR TREE CAMP (11)

Six miles north of Arlington is a clubhouse known as Sugar Tree Camp. A short distance from the building is a high vertical cliff rising almost directly from the Gasconade. The top of this cliff, near the front, is of solid rock, almost bare of timber or brush, and in a row along it close to the edge are seven cairns, all now so defaced that any attempt at investigation is useless. The smallest, at one end of the row, is of the common circular form, about 12 feet in diameter. Three others seem to be of the same type; but their appearance may be due to their destruction. One is shown in plate 13, a. The other three are walled vaults. The largest, at the other end of the row, was built up like a foundation wall of sandstone slabs. It is rectangular in form, measuring on the outside 16 by 28 feet. All the walls are more or less destroyed; the small portion of one remaining is shown in plate 13, b. Two "walled-up graves" reported on the first ridge north of Sugar Tree Camp, and one reported on the first ridge south, never existed. There is a small cairn on a high peak half a mile east of the camp.

TICK CREEK CAVE

In a ravine which joins Tick Creek about 2 miles from where the latter flows into the Gasconade, and about 12 miles north of Arlington, is a large cave known as the Saltpeter Cave.

The opening is wide and high, but the mouth and floor are much obstructed by large fallen rocks and the bottom is constantly wet from wall to wall with running and seeping water.

There is another entrance to this cavern around a corner of the bluff and much higher up on its face. This opening is small and the sloping passage from it to the cavern is almost closed in places by drip formation.

It was never inhabited.

CAVE IN POOL HOLLOW (12)

A mile east of Newburg a ravine now known as Pool Hollow, but formerly called "Strawhorn's" [Strawhan's] Hollow, opens into the right (north) side of Little Piney. Two miles from the river is a cave at the head of a little cove. The entrance, facing directly south and visible from half a mile down the ravine, is 12 feet high and 75 feet across. The rear wall, where the cave makes a turn at 150 feet from the mouth, is plainly visible from the outside.

At 60 feet within water reaches from wall to wall, and a constant stream flows along the left side. The talus at the mouth is of tough clay with many rocks scattered through it, and much of it has settled back into the cave. Water drips from many places in the roof, so that no part of the floor is ever entirely dry.

Some broken flints and chips were picked up about the mouth and in front of the cave, but nothing else could be found.

In dry weather there might be spots which would afford a resting place for campers, but no continuous occupancy was possible.

HOUSE MOUNDS NEAR ROLLA (13)

Nearly 2 miles northeast of Rolla is the beginning of a little valley which for a short distance is parallel with the Frisco Railway and close to the right of way; it then turns to the southward. Along this "draw" are numerous mounds, starting well toward its upper end and following its course for nearly a mile. They lie along either side, and reach into the tributary widenings. Most of them are on the flats; but they are also scattered along the hillsides, those farthest from the water having an elevation of about 50 feet above it. They vary from 30 to 60 feet in diameter and from 1 to 3 feet high. In all, they are scattered over an area of at least 100 acres.

HOUSE MOUNDS NEAR DILLON

Half a mile west of Dillon a ravine heads at the Frisco track, goes south a short distance, then turns southeastward. Near the track begins a group of mounds which reach for fully a mile along both sides of the little stream.

There are more than 100, most of them small, though at least one is 60 feet across and 3 feet high.

HOUSE MOUNDS NEAR ST. JAMES (14)

At the northern border of St. James is a small shallow valley with a northern and eastern trend, practically parallel with the Frisco Railway, and for 3 miles or more not over a fourth of a mile from it at any point.

Starting near the Soldiers' Home is a group of mounds which extend for fully 2½ miles down both sides of the valley.

Some are partly cut away by the stream, others are on the narrow flat bottoms subject to overflow with every hard rain, still others are built on the slopes to an elevation of 40 feet. They are somewhat larger than the average, a diameter of about 60 feet and a height of 3 feet being not uncommon.

PULASKI COUNTY

MCWILLIAMS CAVE (15)

A cave on the McWilliams farm, near Jack Hinshaw's, at the upper end of the Big Eddy, near the south line of Pulaski County, has an entrance 8 feet high and 15 feet wide. There is a good light for 150 feet, at which distance the cavern turns. It is an excellent location for an Indian home, having a floor of dry earth, and a small amount of refuse was found; but the earth has been thoroughly dug over in the search for missing residents, some human bones rooted out by hogs having given rise to a belief that these may have been murdered and concealed here.

DAVIS CAVES (15)

Facing Roubidoux Creek, on the farm of J.W. Davis, 3 miles north of Cookville, are three caves. The largest is 40 or 50 feet above the foot of the bluff. It has an entrance 30 feet wide, the roof being 8 feet high. It is well lighted to a depth of 120 feet, where it curves. No refuse was observed, but the situation is favorable for habitation.

Another cave, near this, has an entrance 30 feet wide and 10 feet high; it is well lighted for 40 feet back.

The third cave of this series is a rock shelter a short distance south of the second, and higher up in the bluff.

All these appear to deserve an examination.

BERRY CAVE

A cave on George Berry's land, in a ravine opening into the east side of Roubidoux Creek, 3 miles from Hanna post office, has a small entrance which is nearly closed by "drip rock," the roof, walls, and floor being thickly incrusted. These deposits, which it is said are even more abundant farther in, seem to be rather rapidly increasing in volume.

MAXEY CAVE (16)

What is known as Maxey's Cave is 7 miles south of Waynesville, on the west side of Roubidoux Creek. It is by far the largest open cave in this region, the entrance being 40 feet high and 100 feet wide. It extends across the head of a ravine, and if the loose earth at the sides were cleared away it would be found still wider. The entire floor is covered with a mass of rocks of every size up to several tons, except at one side of the entrance where there is a small amount of loose earth. The front chamber is 300 feet long to where the cavern forks; in one of these forks daylight extends for 100 feet farther, or 400 feet from the mouth. Marks on the walls show that the entire floor is sometimes covered 2 or 3 feet deep with running water.

A survey made some years ago disclosed a mass of earth and rock "a long ways back in the hill;" definite figures could not be obtained. Beyond this point it was impossible to proceed. By running corresponding angles and lines on the surface outside the surveyors came to a very large sink hole, into which flowed the drainage of several farms. This explains the flood marks. Clearly the roof of the cave had fallen in at this point.

YOARK CAVE

Yoark Cave, a fourth of a mile east from Maxey's in a bluff facing south on the left bank of Roubidoux Creek, has an entrance 40 feet wide, 30 feet high, and is in daylight for 150 feet. Cave earth extends for 100 feet from the entrance, and apparently continues from this point under the gravel and clay which have washed from the interior.

It is on the land of A.L. Foote, having been in his family continuously since it was secured by Government patent. The name is derived from "Grandma Martha Yoark," who was among the earliest white settlers in the region. Her home was on the opposite side of the creek, in a pioneer log cabin, the last vestige of which, except the stones of the chimney, disappeared before the Civil War.

In the front portion many large rocks are lying on the surface of the clay floor and others are imbedded in it; probably still others are entirely covered. Farther back the clay is mixed with gravel washed from the interior. This deposit is never entirely dry and in rainy seasons is quite muddy. The difficulty of removing or digging under the rocks, added to the certainty that water would be encountered before the bottom is reached, render useless any effort at complete excavation. The amount of refuse on the surface, however, is a good indication that such researches as would be possible in the upper layers, among the rocks, would disclose a large quantity of aboriginal remains of comparatively modern date.

GRAVES AT LAUGHLIN'S (17)

On the Laughlin goat ranch, 6 miles southeast of Waynesville, a high narrow ridge level along the top and sloping abruptly on each side extends northward from the hills on the right side of Roubidoux Creek and terminates in a vertical cliff. Bedrock projects on the top and on both sides, and vegetation is so scanty that the crest is almost a "bald."

On the summit of this ridge are seven cairns, the first one only a few feet from the edge of the cliff, the last one about 300 feet back, near where the ground begins to ascend toward the plateau. They are small, none more than 3 feet high, and all have a depression in the top where the stones have been thrown out from the center toward the outside by relic seekers and rabbit hunters.

In three of them flat stones remaining in place at parts of the margin indicate that an irregular square inclosure was constructed around the bodies, as in those examined at Gourd Creek. Possibly this feature existed in all of them at the time of their construction, but there was no evidence that any of them had been walled up like those at Sugar Tree Camp or the Devil's Elbow. Views of their present conditions are shown in plate 14.

KERR CAVE (17)

Near the site of Kerr's Mill, on Roubidoux Creek, 5 miles south-east of Waynesville, is a cave at the foot of a bluff, the entrance 60 feet above the bottom of the hill. Viewed from the outside it has the appearance of a rock shelter 40 feet wide and 45 feet deep. Above most of it the stratum forming the roof is 15 feet high; near the front the successive overlying strata project in a hollow curve until at the face of the bluff the drop from the ledge to the talus immediately beneath it is fully 50 feet.

At one side, near the rear, is a passage 5 or 6 feet wide, not visible from the front, extending back into the hill. Although the cave is usually dry, clean gravel in this passage shows that sufficient water flows through at times to prevent earth from accumulating; further evidence of which fact is found in the mud cracks of the floor and the ferns growing amid the rocks, large and small, which cover it.

The place could never have been occupied except for temporary shelter, and there is no evidence that even this use was made of it.

SELL CAVE (18)

Half a mile directly south of Waynesville, on the farm of Dr. W.J. Sell, is a cave located in the northern end of a ridge entirely detached from the surrounding hills. The entrance, facing northeast, is halfway up the point of the ridge, overlooking a fertile bottom along Roubidoux Creek. From the top of the ledge over the entrance the hill has an easy upgrade for a fourth of a mile to the summit, which is at an elevation of 250 feet above the creek. On top of the hill is the site of an Indian village where some mortars, grinding stones, and numerous flints have been found.

The roof of the cave has partially fallen in at the entrance, forming a re-entrant curve 30 feet across and extending 11 feet inward; the large blocks from this, and from the stratum described later, were lying on and in the talus at the present front but did not extend to the red clay beneath. Some of the blocks could be reduced with a heavy sledge hammer to an extent that made it possible to roll them out of the way; but 24 of them had to be broken up with dynamite.

The talus at its thickest part has a depth of 6 feet; it extends down the hill on the outside and has washed back into the cave, gradually decreasing in quantity, to a distance of 50 feet. The roof, at the front, is 5 feet above the talus; the thickness of the ledge forming it is only 8 feet, the slope of the hill starting from this line. Owing to the restricted width of the ridge, on top, the entire area draining over the ledge measures only 70 feet in width above the entrance, and narrows irregularly to a breadth of 30 feet at an outcrop 120 feet up the hill, or with an approximate space of 6,000 square feet. On this small tract more than half the rock is bare, with scanty patches of soil and humus in the crevices and on flat places. At the present time the water which flows over the ledge during hard rains is scarcely turbid; consequently a period of several centuries was required for the débris to accumulate.

Fourteen feet back from the farthest-receding part of the curve of the roof at the front is the edge of a stratum 3 feet thick; the bottom of this was 3 feet above the talus immediately beneath it. This stratum is continuous, with a perceptible dip to the interior, as far as it can be seen.

The width of the cave at the mouth is 44 feet; 30 feet within it widens to 51 feet. A small amount of water making its way from the interior over the level floor collects in a little basin scooped out to receive it, and sinks into the floor near the inner foot of the talus 55 feet from the entrance. At this point the width of the cave is 36 feet; the height to the roof is 4½ feet. As the floor beyond here is soft mud, the cavern was not followed farther.

Owing to the limited space between the floor and the roof it was necessary to remove the excavated earth to the outside. The water which flows from the hill and falls upon the talus during rains also had to be provided against. A trench 4 feet wide at the bottom, with sufficient slant to the sides to prevent them from falling in, was started 25 feet out from the entrance, on a level which gave it a depth of 6½ feet at the highest point of the talus, thus carrying it a few inches into the clay which was the original floor of the cave. This depth also brought it well below the level of the little pool inside. When its greatest depth was reached the excavation was at once widened to 25 feet, thus reaching well toward the cliff on either side. Growing trees and large rocks made a greater width here impracticable.

In the talus were flint implements, none small enough for arrowheads, some well finished, others roughly made, a few being shown in plate 15; three sandstone mortars and fragments of four others; probably 100 cobblestones used as hammers and pestles, some of them pitted on the sides, a few showing marks of much use (pl. 16, A); a small, very solid piece of hematite worn round by use as a hammer; a small, imperfect tomahawk made of quartzite (pl. 16, B, a); many mussel shells, some used as knives and scrapers; animal bones, some of them worked into implements, including a perfect skiver (pl. 16, B, b); several pieces of hematite and limonite used as paint stones (pl. 16, B, c); many fragments of pottery, some of them worked into disks and perforated (pl. 16, B, d); occasionally small deposits of charcoal, ashes, and burned earth. The meager amount of artificial material, and its random distribution, as if one piece was lost here, another thrown there, throughout the talus from the present surface to the underlying clay would appear good evidence that the cave was never used as a place of permanent abode, but merely provided temporary refuge at intervals extending over a prolonged period.

None of the pottery was decorated in any way, though most of it was cord-marked; no piece was found which had a handle or a foot. Nearly half a bushel of pieces was found, fragments of many different vessels, with a range in thickness from one-eighth to three-fourths of an inch.

If all this talus were examined, much material might be found, but the result would not justify the labor.

Fifteen feet west from the east corner of the cave, 8 feet within the edge of the roof, 3½ feet under the surface of the débris, which was a foot lower here than at the highest point, was a bundled or bunched skeleton; only small fragments of arm and leg bones, most of the lower jaw, a little of the upper jaw, and traces of skull were remaining. The bones were small but solid. They were packed tightly in the dark, wax-like clay, but there were no indications of a grave; the earth in contact with them could not be distinguished from that lying around them. The body had been crowded into the smallest possible space, with the head against a large stone. All the teeth were well preserved, some of them not at all worn. Small fragments of deer bones were found among the remains; these, also, were very soft and decayed.

In fact, all bones found, whether human or other, in this wet, tough, heavy earth were nearly destroyed, and such portions as remained had but little more consistency than the mud in which they were imbedded. Much care was necessary in order to get them out.

Sixteen feet from the entrance, 13 feet from the east wall, 4½ feet down, 18 inches above bottom, were part of a large femur and a few fragments of other bones too small and crushed to identify.

Seven feet southwest of this femur, 14 inches lower, was a closely folded skeleton, the skull nearly north, the other bones toward the east wall. Some mussel shells, fragments of deer bones, and two flint knives were near the head. The body had been placed in a shallow hole dug in the talus as it existed at that time, some earth thrown over it, and small rocks piled on. The covering rocks were under 3 feet of detritus, washed in since they were placed there. Near the knees was a piece of antler, neatly perforated, with rounded ends, giving it the shape of a reniform bannerstone (fig. 8). This may have been an ornament, an arrow-shaft straightener, or the holder for a drill or a fire-stick. Near it was a polishing stone deeply worn on both sides (fig. 9).

Twenty-two feet within the reentrant curve at the front, 20 feet from the west wall, at the bottom of the talus, was a skeleton, the skull in small fragments, which, however, were held in place by the tough clay. The teeth were worn below the enamel in places; two well-worked flint knives and one rough one (fig. 10) were near it. The bones looked as if they had been thrown in, occupying only a small space; but probably a folded body had been laid in on the left side.

At 24 feet from the entrance, 17 feet from the west wall, in a hole dug to 20 inches below the present surface of the talus, were broken and spongy bones of an adult. Pelvis, feet, and leg bones were in confusion; the tibiæ were reversed in position, but it may be that the body was laid on the back with the knees flexed and that the bones had fallen as they were found. This is probable, as each patella was where it belonged, and the body lay extended toward the southeast, as shown by the position of the skull. The humerus was about 12 inches long; all the bones were in small pieces. There were many mussel shells among and above the remains, over which earth and small rocks had been piled.

Two feet south of this skeleton and a few inches lower were the crushed and decayed bones of an old person with the head lying toward the east. The one tooth found (a molar) was worn entirely below the enamel except for a small space at the front; the dentine was polished until it resembled a piece of agate. Mr. De Lancey Gill first remarked the fact that wear of this character denotes that the individual did not gnaw bones, crack nuts, or indeed bite hard on any substance. If he had done so this thin shred of enamel would have broken off. Two large rocks which lay on the head and body seem to have been thus placed before the grave was filled with earth.

Near these bones were fragments indicating three other interments; the humerus of the last was perforated.

Other arm bones found showed the same olecranal perforation.

Twenty-one feet from the entrance, 19 feet from the east wall, was a skeleton, closely folded, on left side, head toward rear of cave. The teeth were worn flat. The bones were crushed by rocks laid on or above the body at the time of burial, as was the case with all the skeletons found in this part of the cave; probably timbers had been interposed.