While cutting our way inwards (Figs. 83 and 88), we found an angular piece of flint, which had evidently been chipped by human agency, and a water-worn fragment of a belemnite, which probably had been derived from the neighbouring marlstone rocks. Bones and teeth of the woolly rhinoceros, reindeer, stag, Irish elk, mammoth, hyæna, cave-bear, lion, wolf, fox, and horse rewarded our labours; and frogs’ remains, cemented together by stalagmite, were abundant at the mouth. The teeth preponderated greatly over the bones, and the great bulk were those of the horse. The hyæna-teeth also were very numerous, and in all stages of growth, from the young unworn to the old tooth worn down to the very gums. Those of the mammoth had belonged to a young animal, and one had not been used at all. The hollow bones were completely smashed and splintered, and scored with tooth-marks, while the solid carpal, tarsal, and sesamoid bones were uninjured, as in the Kirkdale Cave. The organic remains were in all stages of decay, some crumbling to dust at the touch, while others were perfectly preserved and had lost very little of their gelatine.

In 1860 we resumed our excavations; and, in addition to the above remains, found satisfactory evidence of the former presence of man in the cave. Our search was rewarded by one oval implement of white flint, of rude workmanship (Figs. 84, 85, 86, 87), one chert arrow-head, a roughly-chipped and a round flattened piece of chert, together with various splinters of flint, which had apparently been knocked off in the manufacture of some implement. Two rudely-fashioned bone arrow-heads were also found, which unfortunately were subsequently lost by the photographer to whom they were sent; they resembled in shape an equilateral triangle with the angles at the base bevelled off. All were found in and around the same spot, in contact with some hyæna-teeth, between the dark bands of manganese, at a depth of four feet from the roof, and at a distance of twelve feet from the present entrance (Fig. 83, a).

That there might be no mistake about the accuracy of the observations, I examined every shovelful of débris as it was thrown out by the workman; while the exact spot where they were excavating was watched by my colleague. The figured implement was picked out of the undisturbed matrix by him; the rest were found by me in the earth thrown out from the same place.

The lines of peroxide of manganese must have been accumulated on the old floors of the cave, because they were associated with numerous splinters and gnawed animal remains; and there can be no doubt that the latter were introduced by the hyænas. Those animals have a peculiar habit, as Dr. Buckland proved by experiment, of gnawing similar bones in precisely the same way; and a comparison of the relics of the meals of the hyænas in the Zoological Gardens with those in the cave, shows that the latter have passed between the jaws of a like animal that once inhabited Somersetshire. Coprolites of the same animal were very abundant, and in some places formed a greyish-white layer of phosphate of lime. There were also other equally unmistakeable traces of the animal in fragments of bone, polished by their tread, as in the Kirkdale cave. It is, therefore, only reasonable to suppose that these remains of animals were brought into the cave from time to time by hyænas, and left on the floors. That they were not introduced by water is proved by the preservation of the delicate processes and points of bone, which would certainly have been broken in transitu. Since, then, the implements, which, beyond doubt, had been fashioned by man, were underneath one of these old floors, it was certain that man was contemporary in the district with the hyæna and the animals on which it preyed, and the fact that they were found only on one spot implies that they were deposited by the hand of man. To suppose that a savage would take the trouble to excavate a trench twenty-four feet long—for twelve feet of the former mouth of the cave had been cut away—with miserable implements, and consequently with great labour, and having excavated it again to fill it up to the very roof, is little less than absurd. Nor could such an operation take place in such a deposit, without the stratification of the layers being destroyed. The absence of pottery and human bones precludes the idea of the cave ever having been a place of sepulture, such as Aurignac or Bruniquel. This discovery, therefore, of itself stamps the contemporaneity of man with the extinct mammalia, and following close on the similar discoveries in Brixham cave, to be mentioned presently, puts the question beyond all doubt.

In April 1861 we resumed our excavations; and, as we made our way inwards, found that the cave began to narrow, and ultimately to bifurcate, one branch extending vertically upwards, while the other appeared to extend almost horizontally to the right hand. As we reached the middle constricted passage, the teeth became fewer, while the stones were of larger size than any that we had hitherto discovered. The great majority of the gnawed antlers of deer were found at this part, also the posterior half of a cervine skull, the right upper jaw of wolf, and, what is more remarkable, a stone with one of its surfaces coated with a deposit apparently of stalagmite: this, however, was much lighter than stalagmite, and not so good a conductor of heat; and, on analysis, I found that it consisted of phosphate of lime, with a little carbonate, and a very small portion of peroxide of manganese. Doubtless the surface of the stone, covered with phosphate of lime, formed part of the ancient floor of the cave, and hence was coated with album græcum; while the lower part, being imbedded in the earth on the floor, was not so coated. This deposit may, perhaps, explain the absence of round balls of coprolite, which, assuming that the cave at the time was more damp than that at Kirkdale, would be trodden down on the floor by the hyænas, instead of presenting a rounded form. The stone also itself exhibits tooth-marks underneath the coating of album græcum, and probably was gnawed by the hyænas, like the antlers, for amusement. This discovery proves that violent watery action had but small share, if any, in filling the cave; for in that case the soft covering would have been removed from the stone. Similar evidence is offered by the wonderful preservation of some of the more delicate fragments of bone, such as the palatine process of the maxilla of the wolf.

The section made in cutting this passage presented irregular layers of peroxide of manganese, full of bony splinters, and each more or less covered by a layer of bones in various stages of decay. These layers were absent from the upper portion of the passage. There were masses of prisms of calc-spar scattered confusedly through the matrix. After excavating the vertical branch as far as we dared (for the large stones in it made the task dangerous), we were compelled to leave off, having penetrated altogether only thirty-four feet from the entrance. No flint implements rewarded our search this year. Teeth were far more numerous than bones, probably because they are more durable as well as because of their rejection by the hyænas. One jaw was bitten in two, and the fragments found about a foot apart in the undisturbed matrix, just as they had been dropped from the mouth of the hyæna.

In the spring of 1862 Mr. Parker, Mr. Willett, and myself resolved to verify the association of articles of man’s handiwork along with the extinct mammalia, by cleaning out the cave, which was courteously placed at our disposal by the owner, Mr. Hodgekinson.

Our first task was to clear the contents out of the portion of the cave nearest the mouth, or the antrum (Fig. 83, A), and as we excavated onwards many traces of the presence of man were met with. A wide area on the left-hand side (b), where the roof and floor of the cave gradually met together, furnished innumerable fragments of charcoal, and many flint implements associated with the remains of the horse, rhinoceros, and hyæna. One fragment of bone in particular, belonging to the rhinoceros, had been calcined, and its carbonized condition bore unmistakeable testimony that it had been burnt while the animal juices were present. There were many other bones also burnt, which indicated the place where fires had been kindled, and food cooked. As we dug our way forward we met with a third area (c), that furnished flint and chert implements under the same conditions of deposit as that which tempted us to carry on our excavations. Its relation to the old floors of hyæna-occupation is shown by the dark lines over the area c in Fig. 88. At last the large open chamber (A) was cleared; it measured about thirty feet wide by six feet high, and it extended forty feet inwards. On the left there was a small upward-turning passage, very nearly blocked up with a mass of stalagmite; at the farther end a vertical fissure extended upwards (F), to the surface. This fissure has subsequently been proved to extend downwards to the right, and will doubtless furnish large quantities of animal remains to future explorers.

The large chamber now turned abruptly to the left, and we gradually worked our way into a small horizontal passage about four feet high. Here there was an interval of from three to four inches between the roof and contents, traversed by stalactites, which in some places formed a smooth undulating drapery with stony tassels, and in others tiny pillars extending down to the débris, and, as it were, propping up the roof. These pedestals (see Fig. 15) gradually expanded into round plates of stalagmite, which sometimes met and formed a continuous crust. In some places an infiltration of carbonate of lime had cemented organic remains, stones, and earth into a hard mass, which had to be broken up with gunpowder before it could be removed out of the cave. The excitement of extracting from these blocks their treasures was of the very keenest, for we could not tell what a stroke of the hammer would reveal. Sometimes an elephant’s tooth suddenly came to light, at others a hyæna’s jaw, or a rhinoceros’ tooth, or the antler of a reindeer, or the canine of a bear. The bones were so numerous that they scarcely attracted attention. In one fragment of this breccia, now in the Brighton Museum, are a tusk and carpal of mammoth, the right ulna of the woolly rhinoceros, and an antler of reindeer. In a second, two shoulder-blades and two haunch bones of the woolly rhinoceros, with a coprolite and lower jaw of cave hyæna. As the men removed the large blocks they were brought to the mouth of the cave to be broken up by our smaller instruments. Presently the passage narrowed to about six feet, and presented the following section (Fig. 89). On the floor of the cave there was a layer of red earth two feet in thickness, and, as usual, containing a few organic remains and many stones (Fig. 89, 1). Upon this rested a most remarkable accumulation of bones, and teeth, matted and compacted together, from three to four inches thick, and extending horizontally from one side of the passage to the other (Fig. 89, 2). Next came a layer of dark red earth (Fig. 89, 3), loose and friable, three to four inches thick, supporting in its surface a few rounded stalagmites, and a few stalactitic pillars, that spanned the interval of from three to four inches between it and the roof. This bone-bed was about seven feet wide and fourteen feet long, affording, therefore, a square area of ninety-eight feet (see dotted area B Fig. 83, and in Fig. 90). The enormous quantity of the remains of animals present cannot fairly be estimated even by the large number preserved, because most of the bones were as soft as wet mortar. The five hundred and fifty specimens obtained must be looked upon merely as a small fraction of the whole.

We presently passed beyond the bone-bed, and found that the passage bifurcated (Fig. 83, C and D), the smaller branch going straight forwards and gently upwards (Fig. 90), while the larger stretched at right angles from it and passed gently downwards. In the former there was a second bone-bed similar in every respect to that already described, which continued undiminished in thickness until it rested directly on the floor. It afforded a square area of about fifteen feet. The passage was about sixteen inches high and three feet wide, and gradually narrowed until at a distance of twelve feet from the bifurcation a stalactite six inches long reached the floor and formed a vertical bar, as if to forbid another ingress. When this had been explored as far as we could crawl, the larger branch (Fig. 83, D, and Fig. 91) engaged our attention, and we soon discovered a third layer of bones of the same character as the others, and in the same position, excepting that in some places it was in immediate contact with the roof. In width it was six, in length fourteen, and in square area eighty-four feet. From its further end to the termination of the passage there was not the slightest vestige of bones or teeth, and a stiff grey clay rested on a horizontal layer of sand on the floor. Here the passage suddenly turned upwards until it became so small and barren that it was not worth our while to pursue it farther. It doubtless rises to the surface, like the large fissure opposite the entrance of the cave shown in Fig. 88.196

The exploration was resumed the following year by Mr. Ayshford Sanford and myself, and yielded vast quantities of fossil remains. We cleared out the space marked 1863 in the plan, and discovered a flint implement at the point marked d, in Fig. 83. My friend the late Mr. Wickham Flower has also worked the cave, more particularly at the right-hand side of the entrance chamber.

The ashes and implements were found in positions, near the mouth of the cave, where man himself may have placed them (see Figs. 83, 88), with the exception of the flint implement at d, and an ash of bone imbedded in the earthy matrix between the canine tooth and a coprolite of the hyæna, and cemented to a fragment of dolomitic conglomerate. This was found far in the cave, either at the entrance of the passage B, or in the middle of the passage D. The latter passage yielded the only rolled flint without traces of man’s handiwork. The materials out of which the implements were made were used pretty equally. All those, like Fig. 84, were of flint; all those chipped into a rounded form and flat-oval in section of chert from the Upper Greensand; while the flakes consisted of both used indifferently. Besides these three typical forms, which were most abundant, is a fourth, in form roughly pyramidal, with a smooth and flat base, and a cutting edge all round. Of these we found but two examples, both consisting of chert. In form they are exactly similar to several hundreds found in a British village at Stanlake, in Berkshire, and to those I discovered in a cemetery of the same age at Yarnton, near Oxford. They strongly resemble a cast I have of one found by M. Lartet in the cave of Aurignac. Were it not for this similarity, I should look upon them as cores from which flakes had been struck. The rest are mere splinters, irregular in form, and probably made in the manufacture of the various flint and chert implements. All the flint implements have been altered in colour and structure, either by heat or, as is more probable, by some chemical action. Without exception, the old surfaces present a waxy lustre (by the absence of which forgeries are easily detected), the colour is of a uniform milk-white, and the ordinary conchoidal fracture is replaced by that of porcelain. Some are not harder than chalk. I have met with weathered and calcined flints in Sussex in which similar changes are observable, and in which the difference in the results of chemical action and heat can hardly be detected. The chert implements, on the other hand, show no traces of any such changes, but are similar in colour and structure to the rocks from which they came—the Upper Greensand of the Blackdown Hills.

All the fragments of calcined bone, with the exception of one already mentioned, were found near the entrance (see Fig. 83, b), and in a place more suitable for a fire than any other in the cave. I can identify none of them as human. The coarse texture, the structure, and the thickness of one indicate a fragment of a long bone of the rhinoceros.197 All resemble many splinters strewn about in other parts of the cave, which are not calcined, but were evidently introduced by the hyænas. The calcination may therefore be due to the accident of their lying upon the surface at the time the fire was kindled.

The remains obtained in 1862–3 from three to four thousand in number, afford a vivid picture of the animal life of the time in Somerset. They belong to the following animals, the numbers representing the jaws and teeth only, and the implements:—

The remains of these animals were so intermingled that they must have been living together at the same time. They lie large with small, the more with the less dense, and are not in the least degree sorted by water. There is no evidence of the hyæna succeeding to the cave-bear, or the reindeer to the urus, or that the bears came here to die, as in some of the German caves, or that the herbivores fell, or were swept into open fissures, and left their remains, as in the caves of Hutton and Plymouth. On the contrary, the numerous jaws and teeth of hyæna, and the marks of those teeth upon nearly every one of the specimens, show that they alone introduced the remains that were found in such abundance. And they preyed not merely upon horses, uri, and other herbivores, but upon one another (Figs. 92, 93), and they even overcame the cave-bear and lion in their full prime. Some of the bones of the larger animals, and in particular a leg-bone of a gigantic urus, have been broken short across and not bitten through—a circumstance which points towards one of the causes of the vast accumulation of bones in so small a cave. It is well known that wolves and hyænas at the present day are in the habit of hunting in packs, and of forcing their prey over precipices. The Wookey ravine is admirably situated for this mode of hunting, and would not fail to destroy any animal forced into it from the hill-side. It is therefore very probable that the hyænas sometimes caught their prey in this manner. They would not have dared to attack the bears and lions unless these had been disabled.

But if all the remains of the animals were introduced by the hyænas, they certainly in some cases do not occupy the exact position in which they were left by those animals. One of the bone layers (Fig. 91) for instance, actually touched the roof. This, indeed, has been used as an argument in favour of their having been introduced by water, from some unknown repository. But if this hypothesis be admitted, we are landed in the following dilemma: either the introducing current of water must have passed down the vertical passages, or upwards through the horizontal mouth of the cave. In the former case the three bone layers would not have been found in the narrow passages, but would have been swept out into the wide chamber, where the force of the hypothetical current must have abated. In the latter case the great bulk of the remains would have been found in the chamber, and not in the smaller passages. Moreover, the absence of marks of transport by water, and especially of that sorting action which water as a conveying agent always manifests, renders the view of their being so introduced untenable. On the other hand, the horizontality of the layers of bone, and the presence of sand and of red earth, imply that water was an agent in re-arranging the bones and in introducing some of the contents of the cave. The only solution of the difficulty that I can hazard is the occurrence of floods from time to time, during the occupation of the hyænas, similar to those which now take place in the caverns of the neighbourhood. A few years ago, the outlet of the Axe in the great cave was partially blocked up, and the water rose to a height of upwards of sixteen feet, leaving a horizontal deposit of red earth of the same nature as that in the hyæna-den. Now if we suppose that similar floods were caused by an obstruction in the ravine below the hyæna-den, it may have been flooded, just as the upper galleries of the great cave, and the water laden with sediment might have elevated the layers of matted bone, and some of the scattered remains on the surface, while the current was insufficient to disturb the stones, or to affect to any extent the deposits of former floods. The buoyancy of the organic remains is not required to be greater, on this hypothesis, than in that of their having been introduced by a current through the vertical passages. Some of the wet bones taken straight from the cave were sufficiently light to be carried down by the current of the Axe.

All these facts taken together enable us to form a clear idea of the condition of things at the time the hyæna-den was inhabited. The hyænas were the normal occupants of the cave, and thither they brought their prey. We can realize those animals pursuing elephants and rhinoceroses along the slopes of the Mendip, till they scared them into the precipitous ravine, or watching until the strength of a disabled bear or lion ebbed away sufficiently to allow of its being overcome by their cowardly strength. Man appeared from time to time on the scene, a miserable savage armed with bow and spear, unacquainted with metals, but defended from the cold by coats of skin.198 Sometimes he took possession of the den and drove out the hyænas; for it is impossible for both to have lived in the same cave at the same time. He kindled his fires at the entrance, to cook his food, and to keep away the wild animals; then he went away, and the hyænas came to their old abode. While all this was taking place there were floods from time to time until eventually the cave was completely blocked up with their deposits.

The winter cold at the time must have been very severe to admit of the presence of the reindeer and lemming.

The district of the Mendip Hills at a higher level than now.

When we reflect on the vast quantities of the remains of the animals buried in the caves of so limited an area as the Mendip Hills, it is evident that there must have been abundance of food to have enabled them to live in the district. The great marsh now extending from Wells to the sea, and cutting off the Mendips from the fertile region to the south, was probably a rich valley at a higher level than at present, joining the westward plains now submerged under the Bristol Channel. An elevation of from 100 to 300 feet would produce the physical conditions necessary for the sustenance of the herbivora found in the caves both in South Wales and Somersetshire.

The characters of a Hyæna-den.

The remains of the animals which have been eaten by the cave-hyæna, may be recognized by the following characters. All are more or less scored by teeth, and the only perfect bones are those which are solid, or of very dense texture. The skulls are represented merely by the harder portions. That of the woolly rhinoceros, for example, by the hard pedestal which supports the anterior horn (see Fig. 30). Several of these pedestals occurred in the Wookey hyæna-den. The lower jaws also have lost their angle and coronoid process, and are gnawed to the pattern of the shaded portion of Fig. 92, the less succulent part bearing the teeth being rejected. This holds good of the jaws of all the animals so persistently, that out of more than two hundred from Wookey there was only one exception. The jaw of the glutton (Fig. 82), from Plas Heaton, is also gnawed to the same shape, and one of those of the cave-bear from the cavern of Lherm, considered by M. Garrigou to have been fashioned by the hand of man into an implement, seems to me, after a careful comparison in company with Dr. Falconer, referable solely to the gnawing of the hyæna. In Fig. 92, the lower jaw of an adult hyæna is represented, and in Fig. 93 (1) the upper and lower jaws of a hyæna-whelp. In the latter the teeth marks a and b are remarkably distinct.199

The marrow-containing bones are also universally splintered away, until either the articular ends alone are left, as in Fig. 80, or in some cases, as in that of the femur of woolly rhinoceros (Fig. 94), the dense central portion bearing the third trochanter is preserved. This fragment is extremely abundant in nearly all the hyæna-caves in this country. From the invariable habit of the hyæna leaving the bones of its prey in fragments of this kind, their dens are characterized by the absence of perfect long-bones and skulls, and consequently, when these occur in a cave it is certain proof that it was not occupied by these animals. In a great many caves, however, the gnawed fragments are associated with the perfect bones, as, for example, at Banwell, a circumstance that may be accounted for by the untouched carcases and the gnawed fragments being swept in from the surface by a stream falling into a swallow-hole. In all hyæna-dens also are large quantities of album græcum, as well as fragments of bone more or less polished by the friction of the hyæna’s feet.

The Caves of Devonshire.

The ossiferous caves on the south coast of Devonshire, explored during the last fifty years, are by far the most important in this country, since they were the first which were scientifically examined, and the first which established the co-existence of man with the extinct mammalia.

We owe the full details of their history to the labours of the distinguished cave-hunter Mr. Pengelly, F.R.S.,200 whose writings are freely used in the following account.

The Oreston Caves.

The first intimation of the presence of fossil bones in the district was furnished by Mr. Whidbey, the engineer in charge of the construction of the Plymouth breakwater, who discovered numerous bones and teeth, imbedded in clayey loam, in some cavernous fissures at Oreston, which were brought before the Royal Society by Sir Everard Home in 1817. Thus Dr. Buckland’s researches in Kirkdale were anticipated by four years. From time to time, since that date, several other fissures and caves close by have furnished remains of rhinoceros, mammoth, hyæna, lion, and other animals. Among the bones and teeth originally sent up by Mr. Whidbey are several which were identified by Prof. Busk,201 as belonging to the Rhinoceros megarhinus, a species that is vastly abundant in the pleiocene strata of northern Italy and is also represented in the early pleistocene forest-bed of Norfolk and Suffolk, and in the lower brickearths of the valley of the Thames at Grays and Crayford. This is the only case on record of the discovery of the animal in a cavern deposit.

The cavernous fissures in the neighbourhood of Yealmpton,202 about seven miles east-south-east from Plymouth, explored by Mr. Bellamy and Colonel Mudge, R.A., F.R.S. in 1835–6, contained the remains of the hyæna and rhinoceros, and the other animals more usually associated with them. They were probably filled, as in the case of Oreston, mainly by the streams which introduced the pebbles. They may, however, from time to time have been inhabited by the hyænas, although the presence of three skulls of that animal forbids the supposition that they dragged in all the fossil bones.

The Caves at Brixham.

The series of fissures accidentally discovered in 1858, in quarrying the rock which overlooks the little fishing town of Brixham, known as the Windmill cave, was selected by the late Dr. Falconer,203 as a spot in which thorough investigation would be likely to decide the then doubtful question of the co-existence of man with the extinct mammalia. Kent’s Hole had been disturbed by repeated diggings, and the results might be viewed with suspicion. He, therefore, urged the importance of a systematic examination of this virgin cave with such effect, that it was undertaken by the Royal and Geological Societies, and a committee was appointed, comprising, amongst others, Dr. Falconer, Prof. Ramsay, Mr. Prestwich, Sir Charles Lyell, Prof. Owen, Mr. Godwin-Austen, and Mr. Pengelly. To the superintendence of the last is mainly due the minute care with which the exploration was conducted. The remains have been identified by Dr. Falconer and Prof. Busk. The work was commenced in July 1858, and completed in the summer of 1859.204

The cave consists of three principal galleries, with diverging passages, running in the direction of the joints from north to south, and from east to west, communicating with the surface at four points. The following is the general section (Fig. 95) of the deposits in descending order.

(A.) On the floor was a layer of stalagmite, varying from a few inches to upwards of a foot in thickness, and containing only twenty-five bones, among which were the humerus of a bear, and the antler of a reindeer.

(B.) Reddish cave-earth with fragments and blocks of limestone, and of stalagmite, generally averaging from two to four feet. In it 1,102 bones were discovered irregularly scattered through its mass, and belonging to mammoth, woolly rhinoceros, lion, cave, grizzly, and brown bears, reindeer, and others. They varied in state of preservation, and some were scored and marked by teeth. Associated with these, thirty-six rude flint implements were met with, of indisputable human workmanship, and of the same general order as those figured by the Rev. J. MacEnery from Kent’s Hole. Among them was one lanceolate implement with rounded point and unworked butt end, considered by Mr. John Evans, F.R.S., of the type of those usually found in the valley gravels.205 There was, therefore, the most conclusive evidence that man inhabited the neighbourhood, either before or during the time of the accumulation of B, and before those physical changes took place by which the red silt ceased to be deposited, or the stalagmite above began to be formed.

(C.) At the bottom of the cave-earth was a deposit of gravel, principally of rounded pebbles and devoid of fossils.

The early history of the cave, as shown by these deposits, is given by Mr. Prestwich, in the report presented to the Royal Society, as follows:—

“Looking at all the phenomena of Brixham cave, the conclusion your reporter has arrived at is, that the formation of the cave commenced and was carried on simultaneously with the excavation of the valley; that the small streams flowing down the upper tributary branches of the valley entered the western openings of the cave and, traversing the fissures in the limestone, escaped by lower openings in the chief valley, just as the Grotto d’Arcy was formed by an overflow from the cave taking a short cut through the limestone hills, round which the river winds. These tributary streams brought in the shingle bed (Fig. 95, C), which fills the bottom of the fissure. It was only during occasional droughts, when the streams were dry, that the cave seems to have been frequented by animals, their remains being very scarce in that bed, while indications of man are comparatively numerous. As the excavation of the valley proceeded, the level of the stream was lowered and became more restricted to the valley-channel. The cave consequently became drier, and was more resorted to by predatory animals, who carried in their prey to devour, and was less frequented by man. At the same time with the periodical floods, which there is every reason to believe, from other investigations, were so great during the quaternary period, the cave would long continue to be subject to inundations, the muddy waters of which deposited the silt forming the cave-earth, burying progressively the bones left from season to season by succeeding generations of beasts of prey. By the repetition at distant intervals of these inundations, and by the accumulation during the intervening periods of fresh crops of bones, the bone-bearing cave-earth, B, was gradually formed. During this time the occasional visits of man are indicated by the rare occurrence of a flint implement, lost, probably, as he groped his way through the dark passages of the cave. As the valley became deeper, and as with the change of climate at the close of the (pleistocene) quaternary period the floods became less, so did the cave become drier and more resorted to by animals. At last it seems to have become a place for permanent resort for bears; their remains in all stages of growth, including even sucking cubs, were met with in the upper part of the cave-earth, in greater numbers than were the bones of any other animals. These animals resorted especially to the darker and more secluded flint-knife gallery, where 221 out of 366 of their determinable bones were found, whereas only twenty-six were met with in the reindeer gallery.

“Finally, as the cave became out of the reach of the flood waters, the drippings from the roof, which up to this period had, with the single exception before mentioned, been lost in the accumulating cave-earth, or deposited in thin calcareous incrustations on the exposed bones, now commenced that deposit of stalagmite which sealed up and preserved undisturbed the shingle and cave-earth deposited under former and different conditions. The cave, however, still continued to be the occasional resort of beasts of prey; for sparse remains of the reindeer, together with those of the bear and rhinoceros, were found in the stalagmite floor. After a time the falling in of the roof at places (and any earthquake movement may have detached blocks from it), and the external surface weathering, stopped up some parts of the cave, and closed its entrances with an accumulation of débris. From that time it ceased to be accessible, except to the smaller rodents and burrowing animals, and so remained unused and untrodden until its recent discovery and exploration.”206

Mr. Pengelly points out207 an episode in the history of the cave, between the formation and the filling up with its present contents, which is of considerable importance, viewed in relation to the deposits in Kent’s Hole. Over the empty space in D, of Fig. 95, is an ancient stalagmite floor, E, constituting the present ceiling, and shutting off D from the true roof above, E. At the time this was formed, the cave must have been filled up to that level with débris, fragments of which are set in the inferior portion of the calcareous sheet. Subsequently, and before the present contents, A and B, were introduced, the whole of this material has been swept away, probably by an unusual flood similar to that alluded to in the second chapter in the Clapham cave. The pieces of stalagmite in the cave-earth are, probably, some of the relics of the older floor. This filling up, re-excavating, and re-filling with its present contents, are phenomena which considerably complicate the problems offered not merely by Brixham cave, but also by those of Kent’s Hole.

Two other caverns in the neighbourhood of Brixham, the “Ash Hole” and “Bench,” have also yielded the remains of the reindeer, hyæna, and several other pleistocene species, and are fully described by Mr. Pengelly, in his essays contributed to the Devonshire Association.208

Kent’s Hole.

The celebrated cave of Kent’s Hole,209 known from time immemorial, was first found to contain fossil bones by Mr. Northmore, and Sir W. C. Trevelyan in 1824, and was subsequently explored by the Rev. J. MacEnery in the five following years, during which he met with flint implements in association with the extinct animals in the undisturbed strata, and obtained the teeth of the sabre-toothed feline, named by Prof. Owen Machairodus latidens, which has never before or since been discovered in any other cavern in Britain. His manuscripts unfortunately were not used until they passed into the hands of Mr. Vivian, of Torquay, who published an abstract in 1859. Subsequently they were published in full by Mr. Pengelly, in 1869. The discovery of the flint implements, verified by Mr. Godwin Austen in 1840, and six years later also by a committee of the Torquay Natural History Society, was received with incredulity by the scientific world, until the result of the exploration of the Brixham cave had placed the fact of the co-existence of man with the extinct mammalia beyond all doubt. In 1864 a committee210 was appointed by the British Association for the carrying on the investigation, which from that time to the present has been conducted under the careful supervision of Mr. Pengelly.

The cave consists of two parallel series of chambers and galleries, an eastern and a western, which penetrate the low cliff of Devonian limestone in the direction of the joints, with a northern and southern entrance, very nearly at the same level, “about fifty feet apart, from 180 to 190 feet above the level of mean tide, and about seventy feet above the bottom of the valley immediately adjacent.” The largest chamber of the eastern series is sixty-two feet from east to west, and fifty-three from north to south. The extent of the cave has not yet been ascertained.

The contents, examined with the minutest care (on Mr. Pengelly’s method, see Appendix I.), were found to be arranged in the following order.

(A.) The surface was composed of dark earth varying in thickness from a few inches to a foot, on which rested large blocks of limestone, fallen from the roof. It contained mediæval remains, Roman pottery, and combs fashioned out of bone, similar to those discovered in the Victoria and Dowkerbottom caves in Yorkshire, which prove that the cave was frequented during the historic period. A barbed iron spear-head, a bronze spear-head, other bronze articles, and polished stone celts, establish the fact that it was also used during the iron, bronze, and neolithic ages. This stratum contained the broken bones of the short-horn (Bos longifrons), goat, and horse, large quantities of charcoal, and was to a great extent a refuse-heap like that in the Victoria cave.