whelming idea does this give us of the destruction of whole piles of rock, miles in thickness and covering areas comparable with those of continents; and how great must have been the loss of the innumerable fossil forms which those rocks contained! In view of such destruction we are forced to conclude that our palæontological collections, rich though they may appear, are really but small and random samples, giving no adequate idea of the mighty series of organism which have lived upon the earth.1

Admitting, however, the extreme imperfection of the geological record as a whole, it may be urged that certain limited portions of it are fairly complete―as, for example, the various Miocene deposits of India, Europe, and North America,— and that in these we ought to find many examples of species and genera linked together by intermediate forms. It may be replied that in several cases this really occurs; and the reason why it does not occur more often is, that the theory of evolution requires that distinct genera should be linked together, not by a direct passage, but by the descent of both from a common ancestor, which may have lived in some much earlier age the record of which is either wanting or very incomplete. An illustration given by Mr. Darwin will make this more clear to those who have not studied the subject. The fantail and pouter pigeons are two very distinct and unlike breeds, which we yet know to have been both derived from the common wild rock-pigeon. Now, if we had every variety of living pigeon before us, or even all those which have lived during the present century, we should find no intermediate types between these two-none combining in any degree the characters of the pouter with that of the fantail. Neither should we ever find such an intermediate form, even had there been preserved a specimen of every breed of pigeon since the ancestral rock-pigeon was first tamed by man—a period of probably several thousand years. We thus see that a complete passage from one very distinct species to another could not be expected even had we a complete record of the life of any one period. What we require is a complete

1 The reader who desires to understand this subject more fully, should study chap. x. of the Origin of Species, and chap. xiv. of Sir Charles Lyell's Principles of Geology.

record of all the species that have existed since the two forms began to diverge from their common ancestor, and this the known imperfection of the record renders it almost impossible that we should ever attain. All that we have a right to expect is, that, as we multiply the fossil forms in any group, the gaps that at first existed in that group shall become less wide and less numerous; and also that, in some cases, a tolerably direct series shall be found, by which the more specialised forms of the present day shall be connected with more generalised ancestral types. We might also expect that when a country is now characterised by special groups of animals, the fossil forms that immediately preceded them shall, for the most part, belong to the same groups; and further, that, comparing the more ancient with the more modern types, we should find indications of progression, the earlier forms being, on the whole, lower in organisation, and less specialised in structure than the later. Now evidence of evolution of these varied kinds is what we do find, and almost every fresh discovery adds to their number and cogency. In order, therefore, to show that the testimony given by geology is entirely in favour of the theory of descent with modification, some of the more striking of the facts will now be given.

Geological Evidences of Evolution.

In an article in Nature (vol. xiv. p. 275), Professor Judd calls attention to some recent discoveries in the Hungarian plains, of fossil lacustrine shells, and their careful study by Dr. Neumayr and M. Paul of the Austrian Geological Survey. The beds in which they occur have accumulated to the thickness of 2000 feet, containing throughout abundance of fossils, and divisible into eight zones, each of which exhibits a wellmarked and characteristic fauna. Professor Judd then describes the bearing of these discoveries as follows—

"The group of shells which affords the most interesting evidence of the origin of new forms through descent with modification is that of the genus Vivipara or Paludina, which occurs in prodigious abundance throughout the whole series of freshwater strata. We shall not, of course, attempt in this place to enter into any details concerning the forty distinct forms of this genus (Dr. Neumayr very properly hesitates to call them all

species), which are named and described in this monograph, and between which, as the authors show, so many connecting links, clearly illustrating the derivation of the newer from the older types, have been detected. On the minds of those who carefully examine the admirably engraved figures given in the plates accompanying this valuable memoir, or still better, the very large series of specimens from among which the subjects of these figures are selected, and which are now in the museum of the Reichsanstalt of Vienna, but little doubt will, we suspect, remain that the authors have fully made out their case, and have demonstrated that, beyond all controversy, the series with highly complicated ornamentation were variously derived by descent-the lines of which are in most cases perfectly clear and obvious-from the simple and unornamented Vivipara achatinoides of the Congerien-Schichten (the lower division of the series of strata). It is interesting to notice that a large portion of these unquestionably derived forms depart so widely from the type of the genus Vivipara, that they have been separated on so high an authority as that of Sandberger, as a new genus, under the name of Tulotoma. And hence we are led to the conclusion that a vast number of forms, certainly exhibiting specific distinctions, and according to some naturalists, differences even entitled to be regarded of generic value, have all a common ancestry."

It is, as Professor Judd remarks, owing to the exceptionally favourable circumstances of a long-continued and unbroken series of deposits being formed under physical conditions either identical or very slowly changing, that we owe so complete a record of the process of organic change. Usually, some disturbing elements, such as a sudden change of physical conditions, or the immigration of new sets of forms from other areas and the consequent retreat or partial extinction of the older fauna, interferes with the continuity of organic development, and produces those puzzling discordances so generally met with in geological formations of marine origin. While a case of the kind now described affords evidence of the origin of species complete and conclusive, though on a necessarily very limited scale, the very rarity of the conditions which are essential to such completeness serves to explain why it is that in most cases the direct evidence of evolution is not to be obtained.

Another illustration of the filling up of gaps between existing groups is afforded by Professor Huxley's researches on fossil crocodiles. The gap between the existing crocodiles and the lizards is very wide, but as we go back in geological time we meet with fossil forms which are to some extent intermediate and form a connected series. The three living genera-Crocodilus, Alligator, and Gavialis-are found in the Eocene formation, and allied forms of another genus, Holops, in the Chalk. From the Chalk backward to the Lias another group of genera occurs, having anatomical characteristics intermediate between the living crocodiles and the most ancient forms. These, forming two genera Belodon and Stagonolepis, are found in a still older formation, the Trias. They have characters resembling some lizards, especially the remarkable Hatteria of New Zealand, and have also some resemblances to the Dinosaurians-reptiles which in some respects approach birds. Considering how comparatively few are the remains of this group of animals, the evidence which it affords of progressive development is remarkably clear.1

Among the higher animals the rhinoceros, the horse, and the deer afford good evidence of advance in organisation and of the filling up of the gaps which separate the living forms from their nearest allies. The earliest ancestral forms of the rhinoceroses occur in the Middle Eocene of the United States, and were to some extent intermediate between the rhinoceros and tapir families, having like the latter four toes to the front feet, and three to those behind. These are followed in the Upper Eocene by the genus Amynodon, in which the skull assumes more distinctly the rhinocerotic type. Following this in the Lower Miocene we have the Aceratherium, like the last in its feet, but still more decidedly a rhinoceros in its general structure. From this there are two diverging lines-one in the Old World, the other in the New. In the former, to which the Aceratherium is supposed to have migrated in early Miocene times, when a mild climate and luxuriant vegetation prevailed far within the arctic circle, it gave rise to the Ceratorhinus and the various horned rhinoceroses of late Tertiary times and of those now living. In America a

On "Stagonolepis Robertsoni and on the Evolution of the Crocodilia," in Q. J. of Geological Society, 1875; and abstract in Nature, vol. xii. p. 38.

number of large hornless rhinoceroses were developed — they are found in the Upper Miocene, Pliocene, and PostPliocene formations—and then became extinct. The true rhinoceroses have three toes on all the feet.1

The Pedigree of the Horse Tribe.

Yet more remarkable is the evidence afforded by the ancestral forms of the horse tribe which have been discovered in the American tertiaries. The family Equidæ, comprising the living horse, asses, and zebras, differ widely from all other mammals in the peculiar structure of the feet, all of which terminate in a single large toe forming the hoof. They have forty teeth, the molars being formed of hard and soft material in crescentic folds, so as to be a powerful agent in grinding up hard grasses and other vegetable food. The former peculiarities depend upon modifications of the skeleton, which have been thus described by Professor Huxley :

"Let us turn in the first place to the fore-limb. In most quadrupeds, as in ourselves, the fore-arm contains distinct bones, called the radius and the ulna. The corresponding

region in the horse seems at first to possess but one bone. Careful observation, however, enables us to distinguish in this bone a part which clearly answers to the upper end of the ulna. This is closely united with the chief mass of the bone which represents the radius, and runs out into a slender shaft, which may be traced for some distance downwards upon the back of the radius, and then in most cases thins out and vanishes. It takes still more trouble to make sure of what is nevertheless the fact, that a small part of the lower end of the bone of a horse's fore-arm, which is only distinct in a very young foal, is really the lower extremity of the ulna.

"What is commonly called the knee of a horse is its wrist. The C cannon bone' answers to the middle bone of the five metacarpal bones which support the palm of the hand in ourselves. The pastern, coronary, and coffin bones of veterinarians answer to the joints of our middle fingers, while the hoof is simply a greatly enlarged and thickened nail. But if

1 From a paper by Messrs. Scott and Osborne, "On the Origin and Development of the Rhinoceros Group," read before the British Association in 1883.