OENNIS, ON FOSSIL LIAS. 261

OENNIS, ON FOSSIL LIAS. 261 The existence of MAMMTFERS anterior to the deposition of the LIAS, demonstrated from the MICROSCOPIC STRUCTURE of a BONE from the RIVER-BED DEPOSIT, LYME REGIS. By the Rev. J. B. P. DENNIS, Bury St. Edmunds. (Plate XVI.) THROUGH the kindness of the President of the Geological Society, a Paper of mine was read before the Society on the 19th of March last, in which I brought under their notice certain facts which seemed to me strongly to indicate the existence of Mammifers at a period when the Lias had not been deposited, and in a deposit well known by the name of the Bristol bone-bed. Since then, through the kindness of Professor Owen, Dr. J. E. Gray, and Professor Huxley, I have had great opportunities of carrying on my investigations, and I trust that, though errors may be found in my inductions, the result of my inquiries will be found to have added some fresh truth to the treasures of science. Being desirous of giving publicity to some of the results of my labours, the Editors of the Microscopic Journal have courteously offered me an opportunity of so doing; and as my investigations have been for the most part microscopic, that journal seems the most fitting medium for the introduction of my views. The microscope, like the telescope in another field, has already revealed its wonders and unlocked many of the oncehidden mysteries of Nature. No small authority has said, " By the microscope the supposed monarch of the Saurian tribes, the so-called Basilosaurus, has been deposed and removed from the head of the reptilian to the bottom of the mammiferous class. The microscope has degraded the Saurocephalus from the class of Reptiles to that of Fishes." And Mr. Quekett, adverting to these brilliant results of Professor Owen, justly inquires, "Why should not the minute fragments of the other parts of the skeletons of extinct animals afford us, by the same method of manipulation, some indications of the particular class to which such fragments belong?" Such reasoning is irresistible: and I may mention as a corroborative circumstance in favour of the use of the microscope, that a geologist gave me what he considered, and what I believed to be the bone of a pterodactyle, but which the microscope proved to be crustacean. In my examination of the microscopic structure of bone, I have observed certain facts that have induced me to suspect a law which, if I am right in the discovery, will be of great importance to science. I have noticed, for instance, in ani-

262 DENNIS, ON FOSSIL LIAS. mals that have the power of springing, a preponderance of pointed, oval lacunae, and it is curious in this respect, to compare the microscopic structure of the tiger's femur with that of the kangaroo, or the frog's tibia with that of the newt. The toad agrees very nearly with the frog, only the lacuna; are longer, a character I have observed in animals that climb. Those of the newt are quite dissimilar, and the structure of the tiger and the kangaroo is so very similar that it is difficult at first to discriminate between them. The same oval lacunae are present in birds, and 1 cannot but think they indicate a power possessed by the animal, of springing. The ulna of the lesser flying opossum is a very beautiful illustration, the bone, in the shape of its lacunae, most resembling that of birds, though still retaining sufficient evidence of the mammal in its character. The pterodactyle, that singular flying lizard, has the same pointed, oval lacunse. The bat, also a flying quadruped, has the same 5 and it is curious to observe in Mr. Quekett's book that the only bird that has not similar lacunse is the parrot, a bird that never springs from its perch, but climbs by its bill and claws. The force and rapidity with which some birds rise from the ground, as the partridge does, is perfectly surprising, and is quite as wonderful as the spring of the tiger, the bound of the gazelle, or the flight of the opossum: but the strain upon the bones of these animals must be very great, and may well account for a particular and suitable structure. In the tarsus of a small Australian parrot, I find very few of the pointed lacunae, but numerous long ones ; and in the ulna of the same bird, the pointed ovals are much more numerous. The Arctic fox, in its leg-bone, beautifully exhibits the pointed oval lacunse; so does the red Indian squirrel The same may be seen in the dog, cat, common fox, mouse, &c.; and it would appear that they are present in all bones that are subject to great or violent muscular action, as a bird in its flight, or a mammal and a reptile in its bound. It is curious also to compare the leg-bone of the ornithorynchus with that of the turtle ; for there is a very great similarity between them, both in the haversian canals and the shape of the lacunse ; and certainly that strange mammal does approach in its habits the chelonian. If a very thin vertical section is taken from the same part of the humerus of a kangaroo and an otter, you will observe in the former numerous narrow-pointed lacunse, similar to those in the tibia of the frog or in the bones of birds; in the latter you will not see one, the ovals seeming almost round. Compare the otter with the beaver, another aquatic animal, and

DENNIS, ON FOSSIL LIAS. 263 again you see none of the structure of the kangaroo, but a similarity in the shape of the lacunae and the branching appearance of the canaliculi of the beaver to those of the otter; and compare both with the same bone of the newt, and though the reptilian character of the latter is apparent, there is a striking similarity in the appearance of the lacunae. Examine a leg-bone of the ring-tailed monkey, and you will find long lacunas more abundant than the oval; the same will be seen in the bear. These long lacunae are very remarkable in the radius of the chimpanzee, and would seem to be connected with suspensive or pulling movements. It is a step gained if it is found, upon comparison with the bones of different animals who possess in common some faculty, that the structure of their bones indicate it. The next step will be to discover their points of difference as well as their points of agreement with other animals. It may then be possible not only to determine, for instance, whether the animal could spring, but also whether it obtained its prey by a spring, or by bounding escaped from the destroyer. All this implies a thorough knowledge of the microscopic structure of the bones of animals from observations made in different parts, and exact comparisons made with the same bones in different animals; and until this has been done, no satisfactory conclusion, in this respect, can be arrived at. If, however, there is any appearance of truth in this opinion, it is for the mathematician to show what effect the difference of shape in the lacunae may have upon the strength and uses of any particular bone. It may be only the sportive fancy of Nature, as she has delighted to besport herself in the varied structure of the foliage of plants; yet if that sportiveness has only method and arrangement, it may prove of admirable use in distinguishing animals by the microscopic structure of their bones, as plants already have been by that of their leaves. In the fragment of a bone, what clue can we have to the character and habits of the animal to which it once belonged, unless the microscopic structure indicates it. As far as I have been enabled at present to carry my investigations, everything has tended to show that there is a singular correspondence in the microscope structure of animals of similar movements. This matter, indeed, is well worthy of investigation, since if any indication of habits may be deduced from the formal arrangement of the lacunae, we shall be able to reconstruct, in some degree, the history of a primaeval animal, of which only a fragment of its bone remains. The presence or absence of haversian canals are no proof for or against a bone being mammalian, as the radius of one of T 2

264 DENNIS, ON FOSSIL LIAS. our common bats does not exhibit* them, not to mention other instances In the thick portion of fig. 2, PI. XVI., there are apparent traces of haversian canals which very probably have been ground away in the more transparent portions of the bone, a circumstance of frequent occurrence in grinding vertical sections of fossil bone. Figs. 2 and 2 a show admirable lacunae that are in connection with an haversian system, and very nearly compare with, that of the walrus, fig. 8. Having made these preliminary observations, we will turn - at once to the consideration of the fossil bone, of which fig. 1 is a representation, and attempt to determine its true relations. If by the microscope alone it may be shown to be possible to assign it its due position in the scale of animated beings, tlie achievement will be a brilliant one; if failure ensues, at least it should be pardonable. Figs. 2 and 3 represent small vertical sections taken from the lower part, and magnified 100 diameters. Figs. 2 a, 2 b, portions of fig. 2, magnified 400 diameters in fig. 2 b, the lacunae are irregular, and agree, in this respect, with some of the three-toed sloth. The upper end of the bone is composed of fine cancellated structure : the appearance externally is very fine in the grain like ivory. The bone came from the bone-bed deposit west of Lyme Regis, which rests upon the upper beds of the new red sandstone, and is identiral with that deposit which has been called from, I believe, the circumstance of its first having been observed at Bristol the Bristol bone-bed. It is principally composed of minute portions of bones and fishes' teeth, and was supposed chiefly to contain the remains of fishes. For convenience sake, I shall refer at first to Mr. Quekett's very excellent histological work ; and we have to deal therein with Fishes, Reptiles, and Mammals, for Birds may well be put out of the question. None of the plates on Fishes demand our notice, until we arrive at Plate V., and this requires attention. At once we may dismiss the Lepidosiren, as the size and shape of its lacunae, not to mention other differences, forbid us to compare our fossil with it; no more can the Megaltcthys Hibberti, figs. 1, 2, 3, 4, and 5, be compared with it. We may also dismiss the figures of the Burdie house fossil, for the lacunse of that sauroid fish more nearly agree with those of the pterodactyle. We have now only the supposed Rhizodus to deal with, and a very slight examination of figs. 16 and 18, the transverse and vertical sections of the cranium with the plate, will show at once that no agreement exists * At least at present I have not detected auy. They are hardly apparent in the rib of the rat, and wanting in the scapula, &c.

DENNIS, ON FOSSIL LIAS. 265 between them, either in size, shape, or number of the canaliculae. Nor do we meet with better success in viewing the plates on saurians: most of them much exceed the fossil in the size of their lacunae, and those that do not, differ in form. We are therefore obliged, so far as Mr. Quekett's book is concerned, to turn to his plates on Mammals, when at once we cannot help noticing a striking similarity, especially when we come to the sloth, and even more so when we reach Plate XI., and find ourselves amongst the cetaceans. The mammalian characters of the lacunae and canaliculi of the fossil in question, are certainly unquestionable; and besides tliis, there is a very striking similarity in them to those of the edentata and cetacea. I shall now attempt the proof from my own practical observations ; and let me here observe, that if any person wishes to make himself acquainted with the structure of bone, the best way is for him to do as I have done grind for himself every bone he can lay his hands on. It is a matter more of time than of difficulty, except in the manipulation of fossil bone, which requires all the care imaginable. Above all, let him not put recent bone near Canada balsam, as all the minute structure of bone is destroyed by it, unless polarised; and even the canaliculi and lacunae are filled with the balsam, and often obliterated. Old or burnt Canada balsam should be used when any is required, as is the case often in fossil bone, when the canaliculi require to be better exhibited; and as the small passages of fossil bone are generally filled up, the balsam, of course, cannot then enter, while it has the advantage of making other parts of the bone more transparent.* Some of the fossil fishes with which our bone was a contemporary were remarkable, some for the massive character of the scales, and others the armature of their jaws. Through the kindness of Mr. Adams,! a gentleman residing at Buriton, Petersfield, and who has in his cabinet some interesting Stonesfield fossils, I have been enabled to examine the structure of the jaws of the Lepidotus and Pycnodus : the structure of the former tubular, something like the fistularia; the latter I have figured, as its lacunae and canaliculi well illustrate the icthic characters when they are present in fish, which are as follows: lacunae of irregular shape, often stellate ; canali- * I have obtained the best sections by the aid of marine glue. t This gentleman has also enabled me to figure a fossil saurian from Stonesfield, fig. 5; it compares almost exactly with the crocodile (vide D and E of Plate), and is well worthy of notice from its showing that nature has not deviated from her primeval plan.

266 DENNIS, ON FOSSIL LIAS. cumfow in number, rather thick at their bases, tapering off, branched, straggling, reticulate. I have examined other fossil fish, as the Dapedius and others, but could find no appearance whatever of lacunae or canaliculi in their structure. One, a very powerful-boned fish, that I found myself at Lyme Regis, and which I believe is at present undescribed, is entirely destitute. In the skull of the Pagrus, a recent sparoid fish, whose jaws are armed with a most formidable array of canines and molars, I have not been able to discover a vestige, either of a lacuna or a canal icul us; and this I should presume applies to all the class. The sturgeon only shows them in its fin-bones and dermo-skeleton, where the lacunas are very thick together. Other large fish exhibited none. Our common roach has them in the ribs. When, however, they are present, as in the conger-eel, sturgeon, &c, they present precisely the same characters, especially the very small number of canaliculi, and their straggling, spider-like character. There is nothing whatever to show that the bone under our consideration, as the slightest examination of the figs. 21, 22, tells us, ever belonged to a fish, for the lacunae and canaliculi have no icthic characters. The contest, therefore, runs between saurians and mammals; but I think I can satisfactorily show that there are certain marked characteristics between the lacunae and canaliculi of mammals and reptiles, and that in these the fossil agrees with mammifers, though certain higher saurians do approach the lower mammifers in apparent similarity, as might be expected ; and much more irregularity of structure is observable in the bone of the walrus or dugong than we observe in the cat or giraffe. There are certain well-marked and distinctive differences in the lacunas and canaliculi of Reptiles and Mammals; and which, when made apparent, render the task easy in determining the question about a fossil bone, if its structure is sufficiently preserved so as to show some of the lacunae with their canaliculi pretty perfect and well-defined. The lacunae in Reptiles are more generally irregular in their shapes; and by this I do not mean that irregularity which is observable in some mammifers, where the lacunae are some long, others oval, while some are nearly round; for a regularity is observed in one sense, as far as the individual shape of a particular lacuna is concerned; but in the reptile the lacuna is more often irregular in its shape. We must be careful, however, not to confound a transverse or tangential section of an haversian canal with a lacuna. The crocodile, which of all reptiles that I have observed, ap-

DENNIS, ON FOSSIL LIAS. 267 proaches nearer in the general appearance of its lacuna: to mammals, having its lacunae much more regular in form, is still not free from the reptilian character. Great caution, however, is required, as the walrus, and other lower mammals, are inclined more or less to an irregularity in this respect. The canaliculi of lizards are in general thicker than those of mammals, and do not branch, but run in wavy courses from the lacunse, where a number interlace. They have something of the appearance of trellis-work. In the boa-constrictor the saurian characters of the canaliculi are well displayed, which more or less are apparent in all saurian bone that I have examined. The canaliculi also are much less numerous, though their number varies in different Saurians. The definition, therefore, of the general character of the lacunae and canaliculi of Saurians will be : lacunae more generally irregular-shaped generally larger than mammals; the canaliculi generally larger, in a similar manner extending from all parts, not so branched, long, thick and wavy. There are apparent exceptions, of course, to this general definition; but in the main it is truthful and correct. The irregularity of the lacunae in Reptiles sometimes gives the appearance of branching; but then a thinner part of the lacuna is mistaken for a canaliculus. I have observed this in the boa-constrictor. In examining mammal bone, the lacunee are observed generally to be (whether they are round, oval, or long and fusiform) of a pretty regular and uniform appearance, especially in the higher mammifers, and are generally of an oval character. They are also smaller than the generality of Saurians. The canaliculi are much more numerous than those of Saurians, and often fork out at their base ; and besides this, give out branches. One end of a lacuna, when viewed with its long axis, generally looks something like the head of a stag's-horn beetle, the canaliculi being bifid, and branched like antlers. In most of the lacunee of the crocodile, I have observed t the greatest apparent approach to this character; but it is more an apparent one {vide lacuna D). In the mammal there is a considerable open space often left where the extreme.canaliculi bifurcate; whereas in the crocodile they simply radiate all round, with little or no bifurcation, or run out into a long point. The mammal definition will then be: lacunae generally of regular form, most frequently oval, bifurcate, with forked, branching, fine and numerous canaliculi. These remarks and definitions refer to vertical sections, and the lacunae are defined as they appear when seen with their long axis in the field. Transverse or tangential sections give the lacunae always a more or less irregular form, and the la-

268 DENNIS, ON FOSSIL LIAS. cunae generally are then presented with their short axis in view. The mammal that shows most irregularity in its lacunae, as far as I have observed, is the walrus, and perhaps the dugong. The definitions given appertain only to those lacunae that appear in the haversian interspaces. Those that form part of nn haversian system have uniform and nearly straight canaliculi proceeding from them, which in the mammal generally are more numerous and finer, and less wavy than in the Saurian. The lacuna, fig. 4, is most remarkable for the great number of canaliculi proceeding from it; and from a photograph that T have of it, they appear so numerous, that they are not distinguishable one from the other. Figs. 2,2 a, beautifully exhibit some of the lacunae in the fossil connected with an haversian system, and very clearly show their mammal character.* They bear a close comparison with the walrus ; and since that comparison is made with the recent bone of a mammal, it is easy for any person who knows something about the microscopic structure of bone, to form an opinion as to whether I am right or wrong. For my own part, I cannot entertain a doubt upon the question as to the bone belonging to some lower mammal. But, at any rate, a new field is opened for the microscope ; for the question I have raised must have a fair and scientific solution. For if we are not able to discover the bones of mammals and birds (and why we should not discover them I know not, especially as the footmarks of the latter have been already noticed), we shall at least be able to add new genera and species to our existing list of reptiles and fishes. The Bristol bone-bed is a mine of animal matter. Let only the microscope, with its splendid powers, be scientifically used, and it must bring to light new treasures for science. In investigating these matters, one thing ought ever to be borne in mind, and which I have already adverted to the connection or relation that the bone bears to the animal to which it belongs. If this is not considered, I think our inquiry is almost useless. We want not only to say that such a fragment of bone belongs to a mammal or a fish, but we also desire to be able to discover something of the general character of the animal to which it belonged. In forming our judgment, then, concerning this or any other fossil bone by its microscopic structure, we must do pretty much what an anatomist would do under similar cir- * I had intended that fig. 2 a andfig.8 should have been magnified the same number of diameters; the correspondence then would have been more striking: but the reader must bear in mind thatfig.2 a is magnified twice as many times asfig.8.

DENNIS, ON FOSSIL LIAS. 269 cumstances, in determining about a particular bone or tooth j that is, try to discover the class and order of animals it most resembles in its general characters. As I have adverted to before, great variety is observed in form and position of the lacunae in the bones of animals ; the canaliculi also vary. Having first of all determined the class, by a strict comparison with known forms, then an attempt should be made to form a judgment of the kind of animal, based upon inductive reasoning, the result of numerous and accumulated facts; and unless we study recent bone with this end in view, we shall only half do our work. It is a matter of difficulty, no doubt, and the subject is quite a novel one; but it affords most interesting material for microscopic inquiry. Too great praise cannot be given to Mr. Tuffen West for the extraordinary accuracy of his engraving. He says, and truly says, that each lacuna, with each of its canaliculi, is a study from nature, and though I see my name at the foot of the Plate, yet I beg to say that all I have had to do with it was in some measure to design its form, and to give Mr. West the sections of bone to engrave from. I left them in his hand to be produced in their truth and perfectness, and most admirably he has accomplished the work. I ought to mention that, for the sake of ready comparison, some fossil and recent sections of bone have been introduced into the Plate. They are as follows: A, single lacuna, with its canaliculi, human; B, tiger's; C, boa constrictor's; D, crocodile's; E, fossil saurian's, Stonesfield ; F, turtle's, G, conger-eel's: these I term typical. Fig. 19, crocodile, transverse section. Fig. 5, Stonesfield, fossil saurian. Fig. 9, fossil vertebra of a whale. Fig. 7, fossil mammal, probably a palaeotherium; tertiary strata, Touraine. Fig. 20, toad. Fig. 21, fin of sturgeon. Fig. 22, fossil fish, pyonodus, Stonesfield. Fig. 10, fossil mammoth, Till. Fig. 11, anteater. Fig. 12, three-toed sloth. Fig. 15, ant-eater, vertical section. Fig. 16, sloth, ditto. Fig. 13, dolphin, ditto. Fig. 14, dugong. Fig. 17, dolphin, vertical section. Fig. 18, dugong, ditto. Figs. 1, 2, 2 a, 2 b, 3, 4, and 6, representing the fossil in question.