BEAR CREEK.—A very large rock house is on the right bank of Bear Creek, 3 miles above its mouth. It would afford good shelter to a large number of people, except in rainy seasons when they were most in need of it. After heavy storms the creek covers the entire floor.

Other rock-shelters exist along Green River above and below Bear Creek. They are not worth investigating. Some are flooded; others difficult of access; still others become muddy after rains; while in none of them is there any great depth of earth.

WARREN COUNTY

CRUMP'S CAVE.—A mile north of Smith's Grove is a large sink hole, from one side of which extends a cave nearly a mile long. There is abundant room and a good light near the front, and it is reported that quantities of ashes were formerly to be seen on the earth a short distance in. A considerable outside area drains into the cave, and the floor at the present time is everywhere so wet as to be quite muddy. Much water also falls from the roof. A hydraulic ram, not far from the entrance, formerly forced water from one of these falls to the farm residence. A descent of 6 feet, over large rocks and wet earth, brings one to the nearly level floor, 40 feet from the mouth. The amount of flood water running into the cave is indicated by a gully 4 feet deep and the same in width, while trash and driftwood litter the floor from wall to wall for more than a hundred yards.

THOMAS CAVE.—This is a mile north of Bowling Green. The roof of a cavern has fallen in and forms a high mound of rocky debris, down which a path winds on each side, giving access toward either end of the cavern. There is scarcely a possibility that it was ever occupied.

MILL CAVE.—Three miles south of Bowling Green a stream emerges from the foot of a slope, flows a hundred yards through a canyon-like open channel, and disappears under a cliff. This is another instance of an open cave due to a falling roof. The open end is large and forms an excellent shelter for cattle. On either side of the stream, under the cliff, is a shelf or projecting ledge, covered with loose stones. Neither is 2 feet higher than the water level in a wet season.

BARREN COUNTY

PAYNE CAVE.—This, also known as Saltpeter Cave, is near Temple Hill, 9 miles southeast of Glasgow. The bluff in which it is situated is a conglomerate limestone, rising from the waters of Skagg's Creek. The cave has three different entrances, 100 feet or more apart, and each entrance is broken into three or four by columns or masses of stone that have resisted erosion. None of the entrances is large, or opens into spacious chambers within daylight. Flood marks are visible in all, and it is said that after prolonged or heavy spring rains the water covers the floors.

BEN SMITH'S CAVE.—This was discovered while digging out a fox den. It is a tunnel-like cavity, not more than 6 feet high or wide, and not suitable for habitation. It lies a mile and a half south of Temple Hill.

FORD'S CAVE.—This is between Freedom and Mount Hermon, about 14 miles southeast of Glasgow. Originally the entrance was about 8 feet high and 20 feet wide, and opened into a well-lighted chamber probably 40 feet wide and 60 feet long. The floor was of earth and level, with ample space between it and the roof, as shown by marks on the walls, for people to move about readily in any part of the room. The entrance is now artificially closed by earth and stone, except for a space 4 feet square in which a door is hung. Old men in the neighborhood claim they can remember when the floor was 20 feet lower than at present; a manifest impossibility, for that measure would bring it several feet lower than the bed of Mill Creek just in front of the cave. They also claim that blocks of conglomerate and travertine 5 to 10 feet in each dimension have formed from "drip" within their recollection; which, if true, would prove these persons to be almost contemporaneous with the cave men. The more probable statement is also made by them that in early days saltpeter workers dug up and leached all the earth in the cave, filling the entrance and the narrow space before it with the leached earth from the front part of the cave and throwing that from farther back into the cavities and pits left by the prior workings. Inside the cave, near the entrance, is a never-failing spring whose waters flow through a short, narrow crevice at one side. While easily accessible, the water does not reach any of the earth floor.

This would have been an excellent site for aboriginal residence, but there is now no undisturbed earth within daylight nor for some distance beyond, and no one can remember that anything of an artificial nature was ever exhumed.

THE ESMITH CAVES.—Two caves situated on Peters Creek near Dry Fork post office, 14 miles southeast of Glasgow, were reported to be admirably suited for shelter purposes. The smaller is not more than a foot high, from floor to roof, and is filled with flood water after every heavy rain. The larger is above flood line, but the entrance is not over 2 feet high, and the "cave" is scarcely sufficient for a sheep shelter. If the floor were cleared off to a depth of 4 feet from its present level, it would be covered whenever the creek reached high-water mark.

BONE CAVE.—Five miles east of Glasgow human bones were found in a cavern. Particulars could not be obtained. The cave is on a hillside and is entered through a narrow crevice by straddling the walls or going down a ladder. Rocks and trash form a mound in this, the top being 15 feet below the outside surface. On either side of this mound one can make his way continuously downward to darkness, and a rock thrown ahead can be heard going on down some distance over loose stones. If human bones were ever found in here, either they were thrown in or some person fell in and was unable to escape.

SLICK ROCK CAVE.—This is near the post office of Slick Rock, 7 miles east of Glasgow. The entrance is in a narrow crevice at the brow of a low hill. The descent is steep and rugged to beyond daylight.

LOVE'S CAVE.—This is located on Dr. Love's farm, 3 miles north of Slick Rock. It is now used for storing apples and potatoes. The entrance is through a large sink hole, formed by the falling in of the roof of a cave which was at least 50 feet wide at this point. As is usual, the debris has blocked the cave in one direction. Descent is regular, though steep, along the slope into the other end of the cave. The floor is wet and muddy the entire year on account of the drip from roof and overhanging rock at the mouth. The vertical distance from top of the debris to the level floor is about 30 feet, and from the top to the outer surface about 20 feet more. Any attempt at excavation would be difficult and costly, and conditions are such as to make it probably fruitless.

MONROE COUNTY

Four caves in this county were represented as being worth investigation. All are north of Tompkinsville, the county seat.

(1) A rock house in the conglomerate sandstone on the land of Dr. E.E. Palmer, 7 miles north of Tompkinsville, shows smoke stains on the ceiling, and some flint chips among the gravel and earth in front where they have been exposed by water dripping over the face of the cliff. There is, however, only 2 to 4 feet of space between the earth floor and the roof, across the cave from side to side, a distance of 20 feet, and from the front to a point 10 feet back. From this rear portion the earth slopes downward, parallel with the roof of the cave, to the wall behind. The amount of descent could not be accurately ascertained owing to the cramped space, but seems to be 5 or 6 feet. At about that level on the outside a ledge was found on both sides of the entrance and appears to continue across. If so, the earth covers the part immediately in front of the cave. Neither tools nor men could be found to do any trenching, but it is not probable the shelter was ever high enough for a man to stand erect in, because most, or all, of the floor earth must have come from the ceiling.

(2) A mile north of Dr. Palmer's is the McCreary Cave. The entrance is from 60 to 70 feet across and the cavern reaches back fully a hundred feet without any diminution of breadth. Two branches then start under the hill. Each has been explored more than a mile. From each branch flows a considerable brook. They unite near the entrance, sink into the floor, and reappear as a strong spring 30 feet lower in the ravine leading from the cave. The earth is not more than 3 feet deep near the front. It becomes greater in amount farther back, but is wet everywhere below the level of the running water, consequently no excavation was practicable. Flood marks show that the whole floor, except in places a strip along the side walls, is completely submerged at times. On one side a rock ledge or shelf above reach of the water is covered with dry loose earth from 1 to 3 feet deep. This has been dug up in nearly every part by treasure seekers, but nothing of human workmanship has ever been found.

(3) The Belcher Cave is 7 miles northwest of Tompkinsville. It is also called Mill Cave, because a gristmill near the foot of the hill below it is run by the outflowing stream. The entrance is wide and high; the front chamber or vault is fully a hundred feet across each way. But the bedrock is exposed in places and the earth is not more than 2 feet thick anywhere. Water from the brook percolating through this keeps the lower portion saturated.

(4) On John Black Tuley's land, on Meshach Creek, 6 miles northeast of Tompkinsville, two human skeletons were found in a small opening, which has since been known as the Bone Cave. It is a room not over 10 feet across at any part, in a limestone conglomerate, and may be of quite recent origin. Being inconvenient of access, it is not in a position for residence purposes. The skeletons, which were less than 2 feet below the surface, were probably those of Indian hunters. The material in which the little cave is formed will crumble easily in cold weather, being rather wet from the soil water soaking through the hill above it.

There are other caves in this county, but from the descriptions they do not seem at all suited even for temporary camping needs.

LOGAN COUNTY

Very little limestone appears in Logan County, the surface rock being mostly conglomerate. A reconnoissance was made here, however, from Russellville to Diamond Springs, to investigate "a broad valley" which was reported to extend in a general north and south direction from the Ohio, near Brandenburg, toward the Cumberland. It was also claimed that beds of drift gravel exist at a considerable elevation above the little creek now flowing through the valley and that rock shelters are numerous at various levels.

As there is an abandoned drainage system, different from the present, somewhere in this part of Kentucky, which has never been traced, the place seemed worth a visit. The result was disappointing.

The valley is due entirely to causes now at work. The gravel beds result from weathering of lower Coal Measure conglomerates. The rock shelters are shallow, or with a thin covering of earth on the floor, or subject to overflow. None was found that offered any incentive for examination.

TODD COUNTY

On the farm of Mr. Robert Glover, 31/2 miles southwest of Trenton, is a cave known generally as "Bell's Cave," from a former owner. This forms the outlet of a large sink hole, all the rainfall of 6 or 8 acres draining out through it. The entrance is wide and deep, with an easy descent to the level floor. It was for a long time a shelter for Indians, for there is a layer of ashes more than 6 feet in depth, 50 or 60 feet long, and about 15 or 20 feet wide. These represent the probable original dimensions, but the top has been leveled for a dancing floor, and the drainage water has cut away a large part of it, depositing the material farther back in the cave. Six feet of vertical face is exposed at one place by the water, but the ashes extend still deeper. It is said that bone needles, animal bones, antlers, mussel shells ("different from any in the creek now"), burnt rock, and much broken pottery were found in leveling the top. A very fine polished flint celt 12 inches or more in length is also reported. One human skeleton has been found, either at the edge of the ash bed or a few feet away from the edge. The floor is covered, where the earth is washed off, with flint nodules and fragments, and the slopes outside have considerable on the surface. The gullies washed along the slope are paved with nodules like a macadamized road, and in a few places the streams have cut into them so as to show a foot or more at the lower part of the bank so filled and packed with nodules that a knife blade could not be thrust in more than 2 or 3 inches. But there is no evidence of aboriginal quarrying. Probably the Indians dug nodules out of the gullies, for chips are found above and on each side of the mouth of the cave.

To the west, on top of the hill in which the sink hole occurs, and beginning at its edge, is an aboriginal cemetery. There are two small mounds and numerous graves. Scores of the latter have been opened. They are all alike; flat stones form bottom, ends, sides, and top. Many have only one skeleton; others more. The greatest number yet found in one was six. Few are more than a foot deep or much over 5 feet long. About one in ten contains relics of some sort—in two or three entire pots, beads, arrowheads, and gorgets occurred.

I opened three; two contained one body each. The face of one was down, but all the other bones of this and all the bones in the second grave were so decayed that no statement of their position can be made. In the third grave, which was 21/2 feet deep—the deepest yet found—were three bodies. Two lay with faces north; the other, behind these, with face south. The grave was 24 inches wide and less than 6 feet long. Most skeletons (it is reported) were doubled up; often the graves were not over 3 feet long and 10 to 16 inches wide. In some the bones denoted skeleton burial. One skull had been perforated by a ball; at least there was a round hole on each side exactly such as would have been produced by a bullet.

Another large cemetery is on the farm of Mr. G.S. Wood, next north of Glover's. Mr. Wood has opened 50 or more graves and found some relics.

Flint arrows, spears, knives, drills, hoes, spades, and celts, not to mention unfinished pieces, have been found by the thousand on the surface within a mile radius of these cemeteries.

It would seem useless to make any further examination of the level limestone region of central or southern Kentucky. Nearly all the minor drainage is underground, and most of the caves have inlets through sink holes or in small crevices. The water supply is scanty except along streams, and in such situations the caves are usually, for various reasons, of such character as to preclude a continuous occupation, or one extending to a very ancient date. Search is more likely to be rewarded in the mountains where an ample water supply is always at hand.

* * *

TENNESSEE

MONTGOMERY COUNTY

DUNBAR'S CAVE.—Three miles east of Clarksville a large cave has been fitted up as a summer resort. The earth has been leveled around the entrance, both inside and outside, floors laid for picnics and other gatherings, booths, refreshment stands, and places of amusement erected and the surrounding grounds somewhat improved. On account of all this, the place has become quite noted. At present there is from 15 to 20 feet of loose stones and earth on the solid rock floor, and a strong stream makes its way beneath them. It could never have been occupied in prehistoric times until the debris had practically reached the stage at which it was found by the whites.

INDIAN MOUND CAVE.—A report was received to the effect that the mouth of a cave on the Stewart County line, about 18 miles west of Clarksville, had been closed by a rock wall, and earth piled against the outside of the wall; also, that tool marks are quite distinct in a chamber which is plainly of artificial origin.

The rock wall is the stratified rock, in place; the earth in front has washed down from the hillside; the tool marks are water channelings; and other remarkable things mentioned in the report are equally natural. The entrance is a narrow crevice.

SULLIVAN COUNTY

LINVILLE CAVE.—This is 4 miles almost directly west of Bluff City. Apparently it is of great extent, for large sink holes connected with it are scattered over an area of several hundred acres. There are three principal openings. The largest is near the top of a knoll or low hill, and is due to the falling in of the roof. The sunken part has an area of about 30 by 60 feet. Usually, in such cases, the debris entirely fills one end of the cavity thus made, obscuring that part of the cavern, the other end being kept open by surface drainage. In this case, owing to the dip of the strata—some 8 or 10 degrees—and to a change in direction of the cavern at this point, both ends may be entered from the fallen rocks and earth. At one side the descent is precipitous and winding, over and among large fallen rocks. No level place is reached in daylight. At the other side the descent follows the natural dip of the strata and no level space can be found from which the entrance is visible. This part, also, is filled with rocks, large and small, from the roof and sides, and was never habitable.

Fifty yards from the main entrance is another much smaller cave, on the slope of the knoll. It is at the bottom of a crevice 10 feet deep. The floor is level, but only a few square yards in extent, the sloping roof reaching it within 10 feet. As there is considerable drainage into the cavity from the hillside, it is probable that this floor, at least the upper portion, is of recent origin, and that the earth extends downward indefinitely toward the subterranean stream.

West of the knoll on which these openings are found is a valley 2 or 3 miles long. Timber shuts off the view toward its head. This is drained by a constant stream which after winding from side to side of the little vale flows under the knoll. The hole where it disappears is small, but as no rock floor is visible it may lead into a large cavern, and there is no doubt that all the sink holes in the vicinity as well as the two openings above described eventually have the same outlet. Excavations would be difficult and useless.

THOMAS CAVE.—In the face of a steep hillside, near the south (left) bank of the Holston, 3 miles east of Bluff City, is a room with a nearly level floor 10 by 18 feet in the longest measurements. A narrow passage, high enough for a man to walk in, branches off to the right but soon begins to diminish in size and at 100 feet becomes too small to crawl through. The debris in front of the cave is piled to a height of 16 feet above the present floor, and the highest floods of the river reach to about the same level on the outside. The rapid disappearance of the surface water which finds its way in indicates an underground passage to the river, so that a solid floor would not probably be reached above the ordinary water level.

ARKLOW CAVE.—This is a mile and a half southeast of Bluff City. It was reported to have a level earth floor, not more than 4 feet below the accumulation outside. While this was formerly the case, cultivation of the hills around now causes a great amount of surface water to flow over the little bluff into which the cave opens, and this has carried nearly all of the loose earth away through some underground channel. The descent for upward of 30 feet is steep and rugged; it was not traced farther.

MORRELL CAVE.—On the south side of the Holston River, 21/2 miles east of Bluff City, lies the farm of E.S. Worley. Except for a narrow strip of river bottom land, the surface is broken and rocky, the highest point being some 400 feet above the stream. Beginning near the brow of the river hill the central portion of the farm is in a depression whose very irregular rim or watershed surrounds an area of more than 100 acres. All the water that falls within this space drains into a sink hole the bottom of which is but little above flood stage of the Holston. On the south side of this sink is a vertical bluff 120 feet high, from whose foot emerges a stream that after a winding course of 50 or 60 yards disappears in a small opening on the east side of the sink hole, and finally comes to the surface at the foot of the hill, near the river. Its volume is sufficient, even in time of severest drought, to turn the undershot wheel of a large mill. The course of the stream above the point where it is first visible is through a cave which has been traced to the foot of the Holston Mountains, 3 miles away, and there are many unexplored branches. Chambers are known with a cross measure of 100 feet or more, and some of them have a height nearly as great. Stalactites and stalagmites, some of them possessing unusual size and beauty, are abundant.

The sink hole is due to the falling in of the roof of the cave, which could no doubt be followed to the river if it were free from obstructions in this direction.

North of west from the mouth of the cave is another opening, partly in the same strata but 40 feet higher, the dip of the rock being 10 or 12 degrees to the southeast. This was so blocked with talus which had fallen from the cliff and washed down the side of the sink hole that it was necessary to creep nearly 40 feet from the entrance, down a moderate slope, before coming to a point where it was possible to stand upright. From here progress to the junction of the two caves, about half a mile from the entrance, is easy except where fallen rocks interfere somewhat.

Early in the Civil War a large amount of saltpeter was manufactured here. A dam was constructed just within the mouth of the main cave, and in the pool thus formed boats were used to transport the material from the interior. The workmen not required for handling the craft usually preferred to walk through the upper cave to the place where the earth was procured.

The combination of natural features at this place is unusually favorable to aboriginal habitation. The main cave is excluded from consideration by reason of the stream filling it from wall to wall after very heavy rains. The upper cave, however, showed, beyond the debris choking the entrance, a level floor, cumbered, it is true, by fallen rocks, but apparently quite suitable for a dwelling place were these removed. Although opening toward the north, its position so far below the summits of the surrounding hills protects it from winter winds. The creek assures an ample supply of clear cold water. Mountains, refuge for game, are in sight in various directions, while the Holston River is less than a quarter of a mile away.

In order to remove the debris a point 3 feet below the lowest spot on the floor was selected on the slope outside. From here a trench was carried in on a level, the additional depth being taken to facilitate clearing away all material that had accumulated inside the cavern in comparatively recent time, and thus lighten the task of deeper excavations should these be required. The trench needed to be only wide enough at the bottom to allow room for running a wheelbarrow, but owing to the great amount of broken rock, loosely held together by a small quantity of earth, the sides continually gave way, so that by the time it was safe to pass through the trench was 25 feet wide at the top and 24 feet deep at the mouth of the cave. The rocks were of every size from small pebbles to blocks weighing more than a ton each.

Nothing whatever of artificial character, not even a flint chip or fragment of charcoal, was unearthed until at a point 4 feet inside the farthest projecting stratum of the roof. Here was found a prehistoric stone wall whose outer side and top had been entirely concealed by debris. On the inner side the upper portion was visible, owing to the fact that the owner had gathered a quantity of loose stones to construct a wall farther down the slope. Previous to this the ancient wall was entirely covered by the detritus, and even after this partial exposure its true nature was not suspected. It was about 6 feet high, built up of rocks of various sizes and shapes loosely fitted together, earth from the outside surface being used to level up in places where the stones would not bind properly. The largest rock in the top layer weighed about 800 pounds.

The horizontal distance between the top of the wall as it was when cleared off and the corresponding portion of the cave roof was 4 feet; to the roof directly above it, about 2 feet. Apparently it had at one time entirely closed the entrance; at the western end where it abutted against the solid rock the upper portion was firmly consolidated by travertine. Directly above it, nearly 2 feet higher, a slab and some small irregular fragments were securely attached to the side and roof by the same agency. A crevice in the bedrock just at the end of the artificial wall contained several wagonloads of small rocks which had been thrown into it. These also were united into a solid mass by the travertine, all of which had been deposited by water flowing through the crevice. It does not follow that the wall was ever higher toward the opposite end than at this time. In the centuries that have elapsed since it was put up, the roof at the front of the cave, being rather thin-bedded, may have disintegrated. It was not possible to uncover the wall in shape for illustrating; portions of it continually crumbled as the looser material piled against it was removed.

From the wall inward the foreign material piled against the west side of the cave was composed almost entirely of small rocks, with scarcely any earth, and so compactly bound with travertine and stalagmite as to resist all attempts to remove it by ordinary means. On the east side—the left as the cave is entered—there was a great variation in the size of the stones; they were intermixed with much loose dry earth, and there was scarcely any "drip-formation" in the mass. The removal of all this disclosed a projection of solid rock forming a shelf from 8 to 12 feet wide, whose top was about 2 feet higher than the bottom of our trench. About 20 feet from the ancient wall the trench reached the original bottom of the cave as the latter was left by the stream to which its origin was due. This was the tough red or yellow clay, filled with water-worn stones such as appear in all gullies or ravines in this region. It contained a small quantity of stalagmitic material here and there and gradually rose until at 20 feet farther, or 40 feet from the old wall, it terminated against solid bedrock, reaching across the cave, the entire width of which at this point was 26 feet. The shelf on the left belonged to the same stratum.

This brought the work to the terminus that had been the aim from the first, namely, the lowest level of the floor, which was thus shown to be only a foot above the solid rock instead of at least 10 or 12 feet as the general appearance of the entrance and its surroundings had indicated. It was completely cleaned off as far as this was possible, but within 3 feet of the end of the trench began a mass several feet in thickness of fragmentary rocks of every size up to 20 tons or more which had fallen from the roof and were bound together by stalagmite.

Altogether, more than 300 cubic yards of material were removed. The workmen had been carefully instructed as to what the search was for, and kept a close lookout, as evidenced by the very small objects they were continually offering for inspection. It is safe to say that not a spadeful of earth missed scrutiny; but, aside from the artificial wall, the only traces of human presence were three valves of mussels, a turkey bone rudely pointed for use as a perforator, and three or four bones which seem to have been subjected to fire. Not a chip of flint or other stone showing work, no ashes or charcoal, not a piece of pottery, were discovered. If aboriginal burials were made in the cave—and the wall is almost definite proof of such fact—they are either on the floor under stalagmite or in crevices now concealed by fallen rocks.

Numerous small fragments of animal bones were found scattered singly at all depths in the material removed. Nearly every one showed marks of the teeth of rodents. According to Prof. F.A. Lucas, of the National Museum, they all belong to modern species except one tooth, which is that of the cave tapir, and (possibly) the jaw of an otter.

BLEDSOE COUNTY

COLLEGE CAVE.—About three-fourths of a mile west from the old Sequatchie College is a cave which was described as the largest in the county, and as the only one in which people might ever have lived. The opening is about 5 feet wide and 4 feet high; and from it comes a stream sufficient to run a mill.

No other caves could be located in this county or in the Sequatchie Valley north of it.

SEQUATCHIE COUNTY

LAKEY'S CAVE.—In the foothills of the Cumberland Plateau, about 5 miles southeast of Dunlap, the county seat, is the largest cave in the county. A great quantity of earth and rock has accumulated in front of the entrance, washed from the mountain side over an area of several acres. Formerly most of the surface drainage carrying this down flowed into the cave, thus keeping a passageway open through which a man could crawl. Ditches have recently been cut to turn away the water, the entrance walled up, a solid door hung, and the cave is now used for a storeroom. It was never habitable.

A mile north of the above-mentioned cave, toward Dunlap, is a cave with a very large entrance: a sort of rock-house or half dome. The floor is covered with huge rocks and a constant stream flows out. It is said that a party once entered Lakey's Cave and emerged at this one. There is no dry place in it.

PICKETT'S CAVE.—Seven miles southwest of Dunlap is a cave, described as having an ample entrance, with much room inside, perfectly dry, and opening in a cliff 20 or 30 feet above a large, never-failing spring. The description is correct as to location, but not as to size. The opening is about 4 feet across each way, with a slight covering of earth on the floor. The cave winds like a flattened corkscrew. At no place near enough to the mouth for a glimmer of light to penetrate is the roof more than 5 feet above the floor or the side walls more than 5 feet apart.

There are two recesses in the cliff on the opposite side of the little creek formed by the spring. They are 40 to 50 feet above the water, each with an irregular floor of 20 by 30 feet under shelter of the rock. No solid rock is visible in front of them, but a projecting ledge, which seems continuous, appears on either side about 6 feet below the present average level of the floor; and this is probably the depth of accumulation at the front. It may be less toward the rear. The cavities are in a stratum which is somewhat shelly and crumbles easily.

HIXSON'S CAVE.—Six miles northeast of Dunlap is a cave said to be large, accessible, dry, and well suited for occupancy. It is on the side of Walden's ridge, 400 feet or more above the base, a mile from water, and with an opening in the solid rock that can not be entered except on hands and knees. By the time one can straighten up he is in absolute darkness.

LAND COMPANY'S CAVE.—This is 7 miles northeast of Dunlap. To enter, one must crawl between the rock front and the detritus, descending 10 or 12 feet. The floor is fairly level, where it can be found, but is nearly hidden from sight by rocks of all sizes, over and between which it is necessary to scramble almost from the starting point.

HENSON'S CAVE.—This cave, 9 or 10 miles northeast from Dunlap, and perhaps in Bledsoe County, is somewhere on Raccoon Mountains, near the head of a valley up which a mountain road winds along in the bed of a stream. It is said to have a dry dirt floor, with an entrance through which one must crawl. After driving until the horses were tired out and being assured at several scattered cabins that it was "jest a leetle mite furder up thar," search for it was abandoned.

GRUNDY COUNTY

HUBLIN'S OR BAT CAVE.—Numerous caves and rock-shelters are reported in the region about Beersheba Springs. The shelters seem to be shallow with comparatively little earth on the floor. Of the caves, the description given of all but the one named was such as to show them not worth visiting. It is about 10 miles northwest of the springs. Its course is approximately parallel with the mountain ridge, passing under two low foothills or spurs separated by a ravine. When the stream flowing through the latter had cut its channel down to the top of the cave it poured into the hole it had worn. Frost and the natural erosion have made an opening more than 60 feet long. Both parts of the cave remain open, being too large at this point to become choked by the small amount of material which the brook had left as a roof. In some places, so far as it was examined, the ceiling is 50 feet or more above the rocks covering the floor; and one end, that into which the ravine drains, has a continuous and rather steep descent, along the natural dip, as far as it could be followed. Where the exploration ended logs, drift, brush, etc., piled 10 or 12 feet high against huge rocks that had tumbled down, proved a current strong enough to wash away any deposits that may ever have existed; consequently the only earth in this end was that brought by floods.

The other end of the cave is large, with an entrance of such size that small print could easily be read 100 feet from the front if the broad fence across it were removed. This fence was made to close the cave against changes of temperature and also against marauders, it having been used until lately as a storage room for fruit, potatoes, etc.

During the Civil War it was worked for saltpeter. All the earth, down to the rock floor, was removed, even in crevices only wide enough for a man to squeeze through. An incline was built so that horses could be brought into the cave, and no earth now remains within reach of daylight. The rock floor is almost as clean as if swept.

Their exhaustive digging extended for about 200 yards from the entrance. The "face" of the earth is here about 15 feet high; for some reason, which could not be learned, the miners continued their work from here by means of a tunnel 4 or 5 feet high and wide, leaving a floor of earth, and a covering of the same nearly 6 feet thick. This tunnel was not followed.

Near the entrance a crevice barely wide enough for a man to walk in and in some places only 4 feet high turns off toward the left and holds practically the same size for about 100 yards. Here it becomes larger and higher. Earth has been carried out of this and its narrow branches wherever there is room to use a shovel. In a large chamber 200 yards from the front, at the end of the crevice, much digging was done; the "face" left is 13 or 14 feet high.

As far as the diggers went, there is nothing left to explore. Beyond that it is not probable any remains can be found, as it is totally dark long before any remaining earth is reached.

FRANKLIN COUNTY

Several caves were reported in the vicinity of Sewanee and Monteagle. They are objects of curiosity to students and summer residents who frequently visit and make tours through them. They have thus acquired a fame much beyond what is justified by their real interest. They seem to be wet, or with contracted entrances and front chambers, or difficult of access, and, so far as could be judged by the descriptions given, none of them is worth examining.

MARION COUNTY

ACCOUNT'S CAVES.—There are two of these, both with high and large openings, on the right bank of the Tennessee, 2 miles above Shellmound or Nickajack. One is in the face of the bluff, the entrance 50 feet above the river bottom land. Huge rocks lie in front and over nearly all the floor. Surface water flows in at the entrance and after winding its crooked way among the rocks sinks at a point 25 or 30 feet below the top of the debris in front of the entrance. This indicates an open way to the river, so the bottom of the cave is probably down nearly or quite to the water level.

The second cave is 100 yards above the first. A little stream, whose head is in a valley, nearly a mile away, flows around the foot of the bluff and into the mouth of the cave. When the Tennessee rises to flood height the backwater comes into the bed of this stream through the cave before submerging the low ridge between it and the river.

CALDWELL'S CAVE.—This is on the right bank of the Sequatchie River, a mile above its junction with the Tennessee. It is said that formerly a man could walk into it easily for 20 or 30 feet and then crawl 50 or 60 feet farther. This is probably an error of memory. By stooping one can now go in about 10 feet from the edge of the roof, and with a pole feel where the floor and roof come together, nowhere more than 10 or 12 feet beyond. It is said, also, that this accumulation results from throwing in earth to prevent foxes from having a den in the cave. A small hole might thus be closed, but it is too much to believe that the people now living around here would carry in many hundred cubic yards of earth for any such purpose.

Human bones are reported unearthed near the surface; at least bones of some sort were found which the discoverers supposed were human.

The entrance to the cave is more than 25 feet in width, and about 25 feet above the flood plain of the Sequatchie, or only 15 feet above extreme high water. It is in the only exposure of rock for nearly half a mile along the bluff. On either side of the opening the walls are solid, down to the alluvial earth, but in front of the cavity only detritus can be seen from top to bottom. For this reason it is improbable that any solid bottom could be found above the level of the river. Much of the stone weathers out in small fragments, and the process of disintegration is going on continually, as shown by the fresh appearance of the sheltered fragments. How rapid or how regular it may have been in former time is impossible to guess, so that excavation, to be of any value, would have to begin at the bottom of the slope, with the knowledge that the original floor of the cave may be still lower.

NICKAJACK CAVE.—This is the largest and most widely known cave in Tennessee. It is half a mile from and within plain sight of the railway station of Shellmound, 20 miles west of Chattanooga. The entrance is fully 100 feet wide and 40 feet high; a short distance within the cave enlarges, a little farther it contracts somewhat. Daylight penetrates, in spite of curves and immense piles of debris, for more than 500 feet. It has been a resort from time out of mind; first, for Indians and pioneers, then for refugees, now for various social gatherings.

All the earth in sight has been worked for saltpeter, leached, and thrown aside. A vastly greater quantity than now remains has been washed out of the cave by Nickajack Creek, which always has some flowing water and in wet weather rises 5 or 6 feet. Long bridges are required where the highway and railroad cross it.

It takes its name from the Nickajack Indians, who once dwelt here. The field in front is strewn with flint chips and other indications of aboriginal settlement.

There is nothing in the cave to dig for. The saltpeter miners moved all the earth they could reach, while the immense rocks and the creek make any further excavations impossible.

HAMILTON COUNTY

There are many caves in the vicinity of Chattanooga, but all that were visited possess some feature which makes examination appear useless. Most of them have small, inconvenient entrances; others are subject to overflow or have running water in them. None could be heard of in which conditions were better.

* * *

ALABAMA

LAUDERDALE COUNTY

SMITHSONIA.—There is a noted cave at Smithsonia, near Cheatham's Ferry, 15 miles west of Florence. It was reported as suitable for a dwelling, but at the entrance the roof is not more than 4 feet high, and a stream a foot deep reaches to the wall on either side.

KEY'S CAVE.—On the Buck Key farm, 6 miles west of Florence, is a cave which may have afforded shelter to the earliest man in the region. There are two entrances or antechambers, separated by a solid rock partition a few yards thick. One is partially filled with huge solid blocks, some of them several hundred cubic feet in size; the other has in it and in front of it a mass of earth and loose rock whose crest is fully 20 feet above the highest part of the inside floor a few feet back from the front margin of the roof. From here an additional descent of 10 feet leads to the floor behind the first-mentioned entrance, and there is about the same descent to a nearly level floor in the cave a short distance beyond. The way is partially blocked by large rocks which, it is said, have fallen within a few years. For this reason persons in the neighborhood are afraid to venture in. There is a rumor that the corpse of a woman, coated with stalagmite, can be seen in this cave; also several bodies (sex apparently indeterminate) lying like spokes in a wheel, with heads at the center. No one could be persuaded to go in and point out the place where they lie.

From its position, high in a bluff but easy to reach, not more than one-fourth of a mile from the Tennessee River and the same distance from a clear creek, with a strip of bottom land between it and the streams, this cave seems worthy of exploration. At least a month of work by several laborers would be required to clean away the fallen material so that excavations would be practicable.

COLYER'S CAVE.—This is about 5 miles west of Florence. It faces a ravine that leads into the creek discharging near Key's Cave. Human bones were found in it many years ago. The entrance is a round hole, through which one must creep a few yards, then by means of a pole or ladder descend 6 feet. From here the cave is nearly level, with several branches. In some places the floor is solid rock; in other parts it is covered with a thin layer of earth. The "human bones" consisted of one skeleton, lying on a rock floor, fully a fourth of a mile from the mouth of the cave.

COFFEE CAVE.—This cave, 4 miles west of Florence, is said to be "like the Colyer cave, but smaller in every way." It was not visited.

SHOAL CREEK.—A cave is reported on Shoal Creek "3 or 4 miles above its mouth." No one could be found who knew its location more definitely or was able to give a clear description of it.

BLUEWATER CAVE.—Bluewater Creek comes in several miles above Lock No. 6 of the Mussel Shoals Canal. A cave is reported to be near its mouth, but the only caves anywhere in that vicinity, so far as anyone living or working there knows, are a small hole a mile below on the canal, into which a man can crawl, and one some 3 miles up the creek, reached by climbing down a sink hole in a field. The opening to the latter results from fallen rock.

COLBERT COUNTY

NEWSOM SPRINGS.—Numerous caves, most of them small, are reported in the county. The best known is at Newsom Springs, 8 miles south of Barton, on the Southern Railway. It is locally known as the "three-story cave." The lower "story" is a cave from which water always flows. The second "story" is directly above the first. The two have no connection, unless far back in the hill. The floor of the upper cave is mostly rock. It is now fitted up by some people in the neighborhood as a camping place, where they spend a part of each summer. The third "story" is an excavation for a cellar under a house recently erected.

MURRELL'S CAVE.—Tradition has it that this cave was one of the hiding places of a famous desperado and horse thief whose gang operated over all this country in early days. The only entry is by means of a ladder in a narrow crevice 20 feet deep. The place may have been a refuge, but never a residence. It is one-fourth of a mile from Bear Creek, not far above the mouth.

Two other holes or crevices within a few hundred yards, difficult to crawl through, reach small caves. Possibly all these are connected.

BAT CAVE.—One-fourth of a mile from Murrell's Cave is a small cavern, the roof not more than 4 feet above the floor. It has been inhabited from time immemorial by myriads of bats. Several tons of guano have been taken out for fertilizing purposes, but no evidence has been discovered that it was ever a habitation for humans.

PRIDE'S CAVE.—In the river bluff a mile from Pride Station is a cave in which a fisherman has made his home for several years. There is a rather thin deposit of earth on the floor which may have recently accumulated.

CHEATHAM'S FERRY.—Near the landing some boys, while hunting a few years ago, discovered a stone wall across the mouth of a small cave. Tearing it away, they found within some human bones, flints, pipes, including one "with a lot of stem holes," and fragments of pottery. All these were on top of the earth or only a few inches below it. Various excavators or relic hunters have failed to find anything more. The cavity is quite small and difficult to reach, and is undoubtedly a burial place for modern Indians.

On both sides of the river here are immense shell heaps. The shell is mingled with earth near the top, but below 2 or 3 feet the mass is of clean shell to a depth, as exposed by the river, of at least 10 feet. The bottom of the deposit is not visible, being concealed by mud piled against it in high water. The old ferryman says it is 20 feet deep. Although the shell piles are built up higher than the bottom lands to the rear or on either side, they are submerged several feet in great freshets. It is impossible to explain this fact otherwise than by the assumption that the bed of the river has been elevated in recent times, although there are no other indications apparent that such is the case.

SHEFFIELDS.—In the river bluff 2 miles above the Sheffield end of the railway bridge is a crevice or joint which has been widened to 10 feet at the outlet by water percolating from the top of the bluff. When discovered, a rock wall was piled across it near the entrance. Behind this human bones were found with "pieces of pottery and other things." They were close to the surface. Subsequent explorations have revealed nothing below them. It is plainly a burial cave for Indians. The river now reaches at flood tide to within 10 feet of the floor. The earth covering the bones may have washed over them, as there is some evidence farther back in the crevice that surface material is still carried in from the rear, in very small amounts, during rainy seasons.

ROCK SHELTERS.—Several very large rock houses exist on the southern slope of the hill or "mountain" lying a mile to 2 miles south of Pride, 7 miles west of Tuscumbia. Water drips from the roofs, keeping the floors wet all the year and collecting in pools to which stock resorts when the little creeks or brooks in the ravines become dry.

It is useless to search in this part of Alabama for caves presenting indications that they may have been habitable, or the reverse, in ages past. The native rock is a cherty or flinty limestone, crumbling easily, and readily susceptible to changes from atmospheric influences, and especially so to the action of water. New subterranean channels are continually developing, with consequent changes in the interior of any cavern near them.

JACKSON COUNTY

ISBOLL CAVES.—It was reported that habitable caves with spacious rooms occur on the Isboll farms, near Limrock. They have entrances and front chambers of ample size to move about in, though not more than 15 feet wide. There are broader expansions back some distance beyond daylight. In both caves rocks up to 15 or 20 tons in weight strew the floor, until only narrow passageways exist between them. In addition, water flows from them in rainy seasons, being frequently 2 feet or more in depth.

BLOWING CAVE.—This takes its name from an outward current of cold air which is so strong as to distinctly modify the temperature of the atmosphere at least 100 yards from the entrance. The opening and the front chamber are nearly 40 feet across, but the distance from the roof to the muddy floor strewn with large rocks is not more than 5 feet at any point. A creek flows across the cave 200 or 300 yards from the mouth, and there is evidence in the way of drift and mud to prove the statement by the owner that after very heavy rains the overflow comes out the front of the cave in such amount as to fill it to the ceiling, and with a velocity that will roll stones larger than a man can lift.

CULVER'S CAVE.—This is somewhere on the side of a mountain about 4 miles from the station of Limrock. Owing to destruction of forests and subsequent growth of brush, the guide was unable to locate it. He described it as a room in which a man could walk about and reached by going in through an opening like a sink hole, which, however, is only about 5 feet deep. The locality, a rugged, barren hillside, near the head of a cove, is not one in which it is probable a cave would be used for any purpose.

HARRISON'S CAVE.—This is 21/2 miles west of Limrock. It has a large, high opening, an easy approach, and is quite accessible, being at the foot of a mountain with level bottom land in front. A stream flows directly across it some 30 feet from the entrance, emerging at the foot of one wall and disappearing under the other. The earth bank on each side of the stream is about 5 feet high, indicating at least that depth of deposit on the rock floor; as the latter is not visible the amount may be much greater. This earth is soft and wet. In rainy weather water from the interior flows along the floor into the little stream. Sometimes this can not dispose of the surplus, and the overflow rises until it makes its exit through the mouth of the cave. When this happens all the earth within is covered from 2 to 5 feet deep.

SALTPETER CAVE.—This lies 4 miles south of the railway, between Limrock and Larkinsville. It is described as being dry, with a large, high entrance, and "plenty of room inside right at the front." But it was thoroughly worked during the war by saltpeter miners who took out all the dirt they could easily reach, going back "200 or 300 yards." For this reason it was not visited.

DEKALB COUNTY

FORT PAYNE CAVE.—A mile south of Fort Payne is a cave in Lookout Mountain, which, a "boom" company some years ago converted into a summer resort. The detritus in front of the entrance was leveled off, steps constructed to the top, and a heavy stone wall built across the mouth, leaving an entrance a little less than 7 feet in width which was closed by gates. Inside the barrier the floor, now made tolerably level, extends about 30 feet toward the rear, to the natural rock wall, and is 50 feet from side to side, with a roof from 6 to 15 feet high. In the wall at the rear are two small openings through which explorers can pass to large chambers farther within. To the right of the front chamber is a branch cave which is high and wide at the beginning but soon becomes impassable from the accumulated rocks and earth rising to the roof. The left side of the front chamber is continued in another branch going directly back into the mountain. The roof and floor have an equal slope downward to a point some rods from the beginning, the clear space between them being not more than 4 feet. Beyond here the roof is high and there are some large expansions. A creek flows from the rear of the cave to a point estimated as 200 yards from the doorway, where it sinks into the earth. The noise of its fall is distinct throughout the front part of the cavern. There is considerable drip, and though dry stalactites and stalagmites occur in some places, over most of the front chamber their formation is still in progress. Outside of the doorway the solid rock walls show on each side, nowhere less than 25 feet apart. At a depth of 30 feet water flows from the rock and earth between these side walls, but there is no sign of solid bottom, so the depth of the cave is probably more than 30 feet below the present floor.

Under existing conditions the cave would form an excellent shelter, being accessible, roomy, and with an abundant supply of fresh water. The drip from the ceiling could be avoided. But it does not follow that such was the case in the remote past. It is apparent that at one time the creek had its outlet through the mouth and down the gorge in front, the right branch of the cave being then open. From some cause, probably the formation of a sink hole above, water from the surface or near the surface found a way through this branch, carrying mud and rocks sufficient to fill the front chamber to its present floor, diverting the flow of the stream, and finally filling the cave through which it came. While the creek was flowing, occupation would be impossible, or at least inconvenient. When the mud began to settle in, the front portion would be shut off. This condition would hold until the stream found its new outlet and the branch cave had become entirely filled; and when these processes were completed the floor of the cave would be practically at its present level. Under the circumstances exploration would probably, almost certainly, be fruitless. The company which owns the cave would also wish it restored to something like its present state.

ELLIS CAVE.—On the estate of Dr. Ellis, 19 miles north of Fort Payne and 3 miles from Sulphur Springs, are two caves known locally as Big-mouth and Little-mouth. The smaller is closed by a locked gate. The larger has a rather imposing appearance from the outside. From a ledge of rock, in place, in front of it, one looks down a steep slope in which rocks up to 40 or 50 tons weight are imbedded. At a vertical depth of 30 feet is a level space not more than 8 or 10 square yards in area. From this a narrow crevice goes to the right. Within a few yards it reaches a hole which can be descended only by means of a rope or ladder. Persons have, however, gone several hundred yards in it.

On the left of the level space and bounded on each side by solid rock walls is a pit 10 feet deep, caused by inflowing storm waters which have created this depression in seeking a small outlet, also toward the left. The height from the bottom of this sink to the roof of the cave is nearly 50 feet.

Crossing this pit on a foot log, which rests on loose rock and earth at its farther end, a crevice varying from 6 to 10 feet wide goes inward for 50 feet. Earth covers the loose rock at the level of the foot log almost at once, and this earth has a steep ascent toward the rear. The crevice widens beyond the distance mentioned, though irregularly, being in some places 25 feet from side to side. So far as progress is concerned, the cave terminates 150 feet from the doorway in a blank wall. It may be that if the earth were out of the way further progress would be possible.

Considerable digging has been done for saltpeter, but except near the front it has been only superficial.

The top of the earth at the extreme rear of the cave is almost or quite as high as the roof at the front, which means that, if the bottom should be level, the thickness of this accumulated deposit is not less than 35 feet. As the dip is toward the rear and quite sharp, about 10 or 12 degrees, the earth here may well be much thicker than indicated.

Excavation would be tedious and costly, as it would be impossible to dispose of the dirt except by blasting a deep trench through the rock in front to make room for wheeling it out.

KILLIAN CAVES.—There are two of these, both on the west slope of Lookout Mountain. One is near Brandon, 6 miles south of Fort Payne. The entrance is a large sink hole on the side of the mountain, descent into which is difficult owing to the steepness and large rocks. At the bottom the water which flows in over the muddy floor from the slope above—several acres in extent—rushes into a hole choked with loose stones and disappears.

The second cave is about 3 miles northeast of Collinsville. Debris from the mountain has formed a wall across the entrance, which is naturally wide and high and opening out on a little flat in front. Some digging has been done for saltpeter at the front part of the cave, reaching about 30 feet back from the inner foot of the accumulation. In the pit thus formed water stands after every rain until it soaks away. Where it ends the "face" is about 5 feet high. On top, farther in, there is much travertine or stalagmite; in some places it extends entirely across the floor. In other places the floor is bare. There is constant drip, and in one room there is a little gully, where surface water in wet weather, entering from a small branch cave on one side, has cut an exit through the earth at the foot of the wall on the other side. The hole in which it disappears extends beyond the rays of a lamp, and a stone thrown in goes down a slope several feet in length. Very little working is needed to reduce any of the earth to soft, slippery mud, hence no excavation was possible.

MARSHALL COUNTY

FEARIN CAVE.—This is in a bluff on the right bank of the Tennessee River, 10 miles below Guntersville. It has three divisions. Shortly after passing the spacious entrance a branch turns to the right. In a few feet a wall is reached which can be scaled only with a ladder. Climbing this, a large chamber is reached, totally dark, and the home of innumerable bats whose "guano" covers the floor and fills the air with a stifling odor. This branch comes to light again more than a mile away on the side of the mountain.

Returning to the lower chamber and going back about 100 feet from the main entrance, a wall similar to the first is reached, above which is another large cave. Bats never inhabit this, and the floor is of loose dry earth. But no ray of daylight penetrates it, and as a great amount of saltpeter was made here during the War of 1812 scarcely any of the earth retains its original position. During the Civil War the floor of the lower or main cave was also dug up for making saltpeter and much of the leached earth piled in front of the cave. This acts as a dam against encroachment of the river except in the highest floods. There seems, however, to be a passage between the cavern and a spring under the river bank, for water appears on the floor as soon as it reaches the same height outside and the two surfaces maintain a constant level until the freshet subsides. On account of these facts no excavations were made.

HARDIN'S CAVE.—Nine miles below Guntersville, on the right bank of the Tennessee, is a ferry known as Honey Landing. It is at the lower end of a steep bluff which forms the river front of a high hill or mountain, as such elevations are called here. A few feet above high-water mark a narrow ledge or shelf projects, which can be reached only from a point on the side of the hill just above the ferry. About 100 yards from here the ledge reaches a cave, which has a high and wide entrance, with ample space for several families to live on a fairly level, well lighted floor. If the cave were dry, it would be an ideal primitive home. But water continually seeps down the hill above and falls over the roof at the entrance, while a gully through the cave and several minor washes, as well as the mud spread over the floor, show that a large amount of water flows through the cave in wet seasons and covers all the floor except an area some 15 feet in diameter. This is dry on top, but would be muddy at a depth of 3 or 4 feet, the level of the bottom of the gully, so no exploration was attempted.

WELBURN'S CAVE.—Six miles northeast of Guntersville is a cave in which many human bones have been found. It is only a burial place and could never have been used as a dwelling. The entrance, barely large enough to crawl into, is at one side of the bottom of a large sink hole due to the falling in of a cave roof. It receives all the rainfall of more than an acre and is nearly choked with mud and driftwood. It may have been somewhat larger at one time, as there is a tradition that a deer was chased through the cave, coming out at Bailey's Cave, a mile away. Within a few rods the water sinks into the earth, and the floor of the cave, rising beyond this point, is dry. It was on this dry earth, not in it, that the skeletons were found. The floor is uneven, at some places permitting a man to stand, and at others rising to within 3 feet of the roof. Explorations can not be made, as there is no method of disposing of the removed earth.

BAILEY'S CAVE.—This cave is 7 miles northeast of Guntersville. The entrance is high and wide and there is a large, well-lighted area within; but the cave is flooded every time Town Creek gets out of its banks. Bailey's Cave is the other end of Welburn's Cave, as persons have gone through the hill from one to the other.

BARNARD CAVE.—This cave, which is also called Alford's and is still more commonly known as Saltpeter Cave, is on the left bank of the Tennessee 10 miles below Guntersville and opposite the Fearin property. The entrance is at the foot of a bluff overlooking a strip of bottom land a fourth of a mile wide, but the opening is above any flood that has occurred since the country was settled. At the foot of the slope is a bayou filled with Tupelo gums. Between this and the river the ground can be cultivated.

The cave is so straight and the walls so smooth as to look like an artificial tunnel. The entrance is in plain view from a point 380 feet back, and the change of direction, even at that distance, is very slight. The saltpeter miners started at the entrance and removed all the earth lying from 3 to 6 feet higher than the present floor, which is nearly level. They carried their work along the surface of a stratum of gravel, sand, and clay, which is so compact as to be difficult to remove with a pick, and seems to belong to the stream which carved out the cavern. The "face" where they quit work is 5 feet high, and the earth is quite dry, breaking down in angular fragments and separating from the walls so freely as to leave no residue on them. Its original depth at any point, however, may be very easily ascertained by noting the different tints or shading of the wall rock, the lower part, which was protected by earth, being distinctly lighter in color than that above, which was exposed to atmospheric weathering and, for a time, to the smoky torches and candles of the workmen.

The distinct lamination of the saltpeter earth, as shown in the "face," proves it to have been laid down slowly and intermittently in still water. It could not be determined whether this was due to the river in flood periods, or to a gentle stream from the interior whose volume varied in accordance with weather conditions. There is also a small channel along the top of the earth, filled with gravel and sand, as if the overflow of a stream far back in the mountain had been diverted in this direction after the laminated deposits had become dry and settled.

The walls are 10 feet apart near the entrance, but are not more than 8 feet elsewhere and in some places the width narrows to less than 3 feet. They also have an inward slope at the bottom, so the cave is either shallow or else so narrow at no great depth as to be uninhabitable. This fact, and the character of the material deposited by the ancient drainage stream, make it hopeless to expect result from exploration.

MCDERMENT'S CAVES.—There are two caves 100 yards apart, in Brown's Valley, 11 miles southwest from Guntersville. The larger has a descent of 21 feet from the front to the general level of the first floor. All this part is well lighted. The drainage from several acres of the mountain side above pours over the roof at the entrance and runs down the inner slope. It has worn a gully, and the first level it reaches is quite muddy. Leaves and trash 3 or 4 inches deep are piled on and against the loose stones toward the side where the water seeks an outlet. It has worn a crooked channel along this side of the chamber, and falls into a hole which at a depth of 10 or 11 feet below the floor makes a turn and passes from sight. So it is certain that soft wet clay extends more than 30 feet below the level of the entrance. The drier deposits of this room have been extensively worked for saltpeter, and a much greater quantity of earth would have been removed but for the fact that masses of stalagmite, too thick to break off with a sledge hammer, and scores of columns, some of them 6 or 8 feet in diameter and many tons in weight, cover a considerable part of it. The first room is succeeded by several others, all of which are dry and of large size, but in total darkness, and the floors in all have been more or less disturbed in the search for niter. The general direction of the bottom is downward. The last floor is probably 50 or 60 feet lower than the entrance, and is reached by a slope on which it is difficult to retain a footing. In nearly every part the earth is covered by stalagmite, much of it so heavy that the miners could not remove it, but were compelled to dig under it as far as they could reach; and in no place is a rock floor to be seen.

The thickness of stalagmite on the floor, and the great size of the columns, is proof of their antiquity, while the depth of earth beneath must have been thousands of years in accumulating before the deposits began to cover them.

Excavations here, while quite desirable, would be very expensive. Much stalagmite would have to be blasted; upward of a thousand yards of earth moved, and all of it taken out of the cave, because there is no room for it inside. As a man can not push a wheelbarrow up such an incline, a trench must be cut through to the exterior slope; and as solid rock lies not more than 5 feet below the surface at any point, blasting would be necessary the rest of the way. The task is equal to opening a stone quarry.

The second cave on McDerment's place has a good opening. A trench 4 feet wide and 6 feet deep where the rock is thickest has been blasted out to make a level approach to the entrance. Masses of stalagmite on each side, sloping like solid rock from the walls, leave barely room for a man to walk for the first 30 feet. Here the walls recede somewhat, and a pit nearly 15 feet deep yawns before the explorer. After continuing for some distance with this depth, there is another drop of 10 feet which holds until the end of the cave is reached. This entire depression is due to the removal of earth for making saltpeter. It is evident that a vast amount of material has been carried out.

As in the first cave, excavation would be very difficult and expensive. All rock and earth would have to be carried up a steep grade, or a deep cut made to wheel it out. As the light is very dim at the first widening of the walls, it is not probable the space farther back would be occupied unless as a refuge.

Both caves were eroded by water running into the hill, and the end of each is abrupt, the roof being higher and the walls farther apart than at any point nearer the entrance. The original outlets are now filled with earth, and apparently have been so for ages.

FORT DEPOSIT CAVE.—Six miles below Guntersville the highway to Huntsville crosses the Tennessee River at Fort Deposit Ferry and passes out through a narrow valley between two bluffs. Less than 100 yards above the landing, on the north, or right, bank, is a large cave from which the spot takes its name; there being a tradition that it was used by General Jackson as a storage room for supplies during the Creek Indian war. On either side the bluff is vertical to the water's edge, making the cave now inaccessible except by boat. In front of the entrance the rock is worn in ledges which can be easily ascended.

The opening or mouth of the cave is oval in form, about 18 feet high and 15 feet wide. The sides are uneven, there being a projecting shelf on each side near the floor. At 40 feet from the opening these disappear, owing to the narrowing of the cavern. There is a gradual ascent of the floor toward the rear, the rise being about 2 feet in the first 60 and more rapid from that point onward. A thin deposit of dried mud on each side, where it escapes the feet of visitors, shows that the river enters the cave at times, but not to a depth that carries it back more than 25 feet. The present ferryman says the flood of 1867 is the only one which has reached so far within that period.

After clearing away the earth, roots, and rocks at the front, a straight vertical face at a distance of 18 feet from the entrance measured 91/2 feet at top and 5 feet at the bottom between the solid rock wall on each side, and was 4 feet 4 inches high. The floor was not of solid rock entirely across, there being a crevice less than 4 feet wide which was not cleaned out, because no one could have lived in it. About the middle of this bank (vertically) streaks of red earth, burned elsewhere, extended 31/2 feet out from the right wall; there was very little ashes and no charcoal mixed with it. Above this red the earth was dark like garden soil and contained a few shells and fragments of pottery, with a little charcoal and ashes; it had all been disturbed and apparently resulted from scraping the debris away from camp fires. Below this, the line of demarcation being very distinct, the earth was yellow and sandy, like river bottom land, and contained no foreign matter except roots of trees growing outside. Figure 23 shows a section on this line; the crevice is omitted from this and the subsequent illustrations.

At 20 feet in, a foot below the top of the dark earth, was some charred corn. The yellow earth became irregular, thinner, and higher against the side walls than at the center. (See fig. 24.)

At 22 feet the yellow earth had nearly run out, there being only a small amount against either wall, while the darker earth reached down into the crevice that opened in the narrow strip of rock floor. In the lower portion were mingled a few shells, pebbles, and specks of charcoal, as if it had been thrown there. Across the upper portion of the deposit extended fire beds, burned earth, ashes, shells, broken pottery, and occasionally a fragment of bone. (See fig. 25.)



At 24 feet it was found that what had been taken for a solid floor in the last section represented was only a large flat rock which had fallen into the crevice and wedged tightly. When this was passed the yellow earth reappeared, at a slightly lower level.





At 26 feet the yellow earth became mixed with red. It was excavated to a depth of 5 feet in the endeavor to discover the reason for this. As there was not the slightest trace of ashes or charcoal, the red admixture must be a natural result of staining by iron in some form and not due to heat. Above the yellow was the usual stratum of dark earth, containing culinary debris. In the central portion of this was a mass, sufficient to fill a wheelbarrow, of angular, unburnt fragments of limestone from 3 to 15 pounds in weight. On the surface of the dark earth were some ten or twelve fire beds, reaching from wall to wall, the edges overlapping and interlacing in so confusing a manner that the exact number could not be made out. (See fig. 26.) At this stage it appeared that the crevice, or at least its upper part, had been filled by river floods and a slight ridge of sand thrown across the mouth of the cave. The Indians, it seems, occupied both this ridge and the lower area behind it, throwing debris to the rear to fill up the depression instead of carrying it all to the outside. It is equally possible, however, that this waste was brought from points farther back and thrown here to fill and level the floor. These heavy fire beds came to an end at about 28 feet on the right and 29 feet on the left. A section at 28 feet is given in figure 27. At their inner margin, among the ordinary refuse characteristic of such deposits, were many fragments of human bones, including ulnas of two individuals, one much larger than the other. They plainly indicated cannibalism, as they were broken when thrown here. Besides the ulnas, there are pieces of ribs, scapula, tibia, and feet.



At 29 feet the underlying yellow earth became comparatively level across its upper surface, again closely resembling a river deposit. The darker earth above it contained a greater amount than heretofore of ashes, bones in small pieces, potsherds, mussel, snail, and periwinkle shells, and the like. More charred corn was found along here.



At 30 feet the yellow earth began to rise, and at 32 feet it was very little more than 3 feet lower than the top of the highest ashes. A section at this point is shown in figure 28. At 35 feet the strata became quite regular and uniform from wall to wall. The dark earth, next above the yellow, measured 3 feet in thickness at the center, and while showing by its admixture of ashes, etc., that it had been thrown here, had evidently formed the floor for a considerable time. The upper foot was burned red or dark from long-continued fires, the ashes above it being from 6 to 8 inches thick, and forming the present floor of the cave at this place. The dark earth contained much less of refuse than nearer the entrance; such shells and ashes as appeared were promiscuously distributed and not in little piles or masses as before. A section at 351/2 feet appears in figure 29. It may be remarked here that this is the only sketch in which the upper line coincides with the surface of the deposits. In the others a thin covering, less than 6 inches at any point, of disintegrated material from walls and roof covers the ashes left by aboriginal fires. This is omitted from the drawings.



At 38 feet the yellow earth had risen until it was within 3 feet of the top of the entire overlying deposit. The latter contained little of the dark earth, being mostly composed of ashes and burned earth, some of which resulted from fires made on the spot, but the greater part being thrown from other points. The rise of the yellow earth, consequently, is more rapid than the rise of the material covering it.



At 40 feet there was a dip in the yellow earth, extending for 4 or 5 feet and descending 2 feet at the deepest point. This may be due to drainage at a lower level.

At 471/2 feet a pocket of the dark earth extended a few inches into the underlying yellow earth. A hole seems to have been dug into the latter. There was no more of foreign material in this hole than elsewhere in the dark earth above and around it. It is shown in figure 30.



The amount of shells, pottery, etc., had been decreasing for several feet before this point was reached; indeed, from 40 feet onward there was very little of it—enough, however, to show that all the dark earth had been disturbed and thoroughly mixed. The fire beds, too, while holding their depth of about a foot, contained more earth between the successive layers of ashes, showing as great age, probably, as those nearer the entrance, but less continuous occupation. This condition prevailed to about 60 feet from the entrance, at which point the yellow earth, now mixed with sand and gravel, was only 3 feet below the surface of the floor. The appearance of this line is sketched in figure 31.



At 62 feet there was a dip in the yellow earth, extending to 67 feet and 2 feet deep at its lowest point; it then rose to the usual level.

At 70 feet ashes appeared in greater quantities; at 73 feet the dark earth was only a foot thick, the ashes and burned earth being 2 feet thick and apparently all dumped, as there was no definite arrangement of the various parts. (See fig. 32.) A small perforated disk and a double-pointed bone needle were found here.

The fire beds now began to thin out rapidly, the dark earth also diminishing in quantity, until at 80 feet, from which point the entrance was no longer visible owing to curvature of the walls, there was only 5 or 6 inches of them in all, resting directly on the yellow earth, which contained much more clay than farther toward the front. The walls began to diverge here, forming a room whose greatest width was 11 feet 6 inches at 95 feet. At 100 feet a reverse curve brought the cavern on a course parallel to that which it had held up to 60 feet.



At 90 feet there was evidence of fire at one side, the ashes and burned earth being 5 inches thick at the wall, and thinning out to a feather edge within 4 feet. This was the last fireplace discovered which may not with certainty be attributed to white men. The yellow earth, presenting no evidence of having been disturbed since originally deposited, reached from the superficial layer of loose dry earth to the bottom of the trench, a depth of 4 feet 8 inches. Below this point the walls were less than 4 feet apart, and the space filled with gravel, as shown in figure 33. This gravel had exactly the appearance of that in gullies on the hills outside, and plainly dates back to the period at which the cave was formed. The stream which aided in the erosion, or which flowed through from some sink hole or other outside opening, carried this gravel into the crevice. Consequently, even if the space between the walls had been ample for dwelling purposes, an attempt to live here when the gravel was being carried in would result in the intending settler having his effects washed out into the river.



At 93 feet the side walls confining the yellow clay narrowed to a little less than 5 feet apart. The upper portion of the one to the left has been eroded into a recess or cavity, forming the chamber above mentioned. The earth on the rock floor in this recess is nowhere more than a foot deep. A section is presented in figure 34.

At 100 feet the room came to an end. The space between the walls was 71/2 feet at the floor level and 4 feet at a depth of 4 feet. At 105 feet the nearly vertical walls were only 5 feet apart on the floor; at 112 feet the space increased to 7 feet. A section showed about a foot of loose earth mixed with ashes; 3 feet of yellow clayey earth, rather compact; then gravel and sand. The latter was dug into for a foot, at which level the walls were converging and it was useless to go any deeper. Enough was done, however, to verify the supposition that this stratum was due to the action of running water seeking its outlet at the mouth of the cave.

At 103 feet, at the bottom of the yellow clay and on top of the gravel, was a chalcedony pebble about 21/2 inches in diameter. The material is foreign to this locality. It had plainly been used as a hammer stone, and is the only object of human origin found anywhere below the dark earth. There was not the slightest evidence of any disturbance of the clay in which it rested.



At 120 feet the side walls were only 5 feet apart. At 125 feet they again diverged slightly, and a recess on the left forms a chamber 12 feet across. At 150 feet they had drawn in to 8 feet at the widest interval. A section showed loose dry earth, some of it cemented by drip from the roof until about as hard as lump chalk; then compact clayey earth, also with travertine in small lumps; below this the gravel and sand. The latter, at this point, seems to have been deposited in the last stages of the formation of the cave. Occasionally, along here, a small patch appeared that seemed to be ashes; but none of it was more than 6 inches below the top of the ground, and the substance may not have been ashes at all, but the fine white limestone dust that wears off from the stone. There was nothing in the trench, at any depth, after the chalcedony pebble, that could possibly be due to human intervention, except these small patches of ashes, if ashes they are.

At 165 feet from the entrance the cave made its fourth turn and expanded into a chamber about 15 feet wide. Along the sides of this and in the various crevices opening from it were great quantities of clean ashes, plainly enough thrown there from fires made in the central part. The gravel came to within 3 to 5 feet of the top, being quite irregular. On the gravel was dry clay, seamed and fissured in all directions so that it fell out under the pick in clods like angular pebbles from an inch to 3 or 4 inches across. This was clearly the result of muddy water settling in a hole and thoroughly evaporating. There was also some travertine in small lumps here and there through the clay, and above it was a mass fully 2 feet thick at one side of the trench but running out before it reached the other side. It was porous, almost spongy, and seemed to be the lime dust from the roof and sides cemented by dripping water. Above all this, so far as the trench extended toward the sides of the cave, was an inch to 4 inches of loose, dry, dark earth, which on the left dipped down to the clay, thus replacing the travertine.





At 175 feet the gravel had leveled down and was more or less mixed with clay and sand. Above this was another "mudhole deposit" of clay which had thoroughly dried out and become checked and cracked in all directions. On the right this was covered with travertine slightly mixed with earth and clay; on the left, above it and also at one place within it, was a coarse gritty earth fallen from the roof but not converted into a compact travertine. The section appears in figure 35. At 180 feet the trench was carried to a depth of 6 feet. This exposed a fine clay and sand, or silt, like that deposited in the eddies of streams. Above this was another deposit of "mudhole" material which had thoroughly dried out, checked and cracked in all directions so that it formed angular masses of various sizes, and had then become wet again so that it was now soft and sticky. To the left of this, on the silt also, was a small amount of the gravel. It had the appearance common to a bank of such material on the side of a little stream which has undermined and carried away part of it. Clearly, these three formations were of an age that witnessed the erosion of the cave. Next above them was a stratum of loose dark earth similar to that noticed in the front part of the cavern; but here were found no traces whatever of man's presence. Into the right side of this stratum projected the wedge-like edge of a mass of travertine, which was not traced to a termination. Over all lay a deposit 3 or 4 inches thick of dark, nearly black earth, mixed with ashes. This is quite modern. The section appears in figure 36.