Such evidence as there is seems rather to suggest that South Africa was the cradle of the placental mammals. We shall find that many of our mammals originated in Africa; there, too, is found to-day the most primitive representative of the Tertiary mammals, the hyrax; and there we find in especial abundance the remains of the mammal-like reptiles (Theromorphs) which are regarded as their progenitors. Further search in the unexplored geological treasures and dense forests of Africa is needed. We may provisionally conceive the placental mammals as a group of the South African early mammals which developed a fortunate variation in womb-structure during the severe conditions of the early Mesozoic. In this new structure they would have no preponderant advantage as long as the genial Jurassic age favoured the great reptiles, and they may have remained as small and insignificant as the Marsupials. But with the fresh upheaval and climatic disturbance at the end of the Jurassic, and during the Cretaceous, they spread northward, and replaced the dying reptiles, as the Angiosperms replaced the dying cycads. When they met the spread of the Angiosperm vegetation they would receive another great stimulus to development.
They appear in Europe and North America in the earliest Cretaceous. The rise of the land had connected many hitherto isolated regions, and they seem to have poured over every bridge into all parts of the four continents. The obscurity of their origin is richly compensated by their intense evolutionary interest from the moment they enter the geological record. We have seen this in the case of every important group of plants and animals, and can easily understand it. The ancestral group was small and local; the descendants are widely spread. While, therefore, we discover remains of the later phases of development in our casual cuttings and quarries, the ancestral tomb may remain for ages in some unexplored province of the geological world. If this region is, as we suspect, in Africa, our failure to discover it as yet is all the more intelligible.
But these mammals of the early Tertiary are still of such a patriarchal or ancestral character that the student of evolution can dispense with their earlier phase. They combine in their primitive frames, in an elementary way, the features which we now find distributed in widely removed groups of their descendants. Most of them fall into two large orders: the Condylarthra, the ancestral herbivores from which we shall find our horses, oxen, deer, elephants, and hogs gradually issuing, and the Creodonta, the patriarchal carnivores, which will give birth to our lions and tigers, wolves and foxes, and their various cousins. As yet even the two general types of herbivore and carnivore are so imperfectly separated that it is not always possible to distinguish between them. Nearly all of them have the five-toed foot of the reptile ancestor; and the flat nails on their toes are the common material out of which the hoof of the ungulate and the claw of the carnivore will be presently fashioned. Nearly all have forty-four simply constructed teeth, from which will be evolved the grinders and tusks of the elephant or the canines of the tiger. They answer in every respect to the theory that some primitive local group was the common source of all our great mammals. With them are ancestral forms of Edentates (sloths, etc.) and Insectivores (moles, etc.), side-branches developing according to their special habits; and before the end of the Eocene we find primitive Rodents (squirrels, etc.) and Cheiroptera (bats).
From the description of the Tertiary world which we have seen in the last chapter we understand the rapid evolution of the herbivorous Condylarthra. The rich vegetation which spreads over the northern continents, to which they have penetrated, gives them an enormous vitality and fecundity, and they break into groups, as they increase in number, adapted to the different conditions of forest, marsh, or grass-covered plain. Some of them, swelling lazily on the abundant food, and secure for a time in their strength, become the Deinosaurs of their age, mere feeding and breeding machines. They are massive, sluggish, small-brained animals, their strong stumpy limbs terminating in broad five-toed feet. Coryphodon, sometimes as large as an ox, is a typical representative. It is a type fitted only for prosperous days, and these Amblypoda, as they are called, will disappear as soon as the great carnivores are developed.
Another doomed race, or abortive experiment of early mammal life, were the remarkable Deinocerata ("terrible-horned" mammals). They sometimes measured thirteen feet in length, but had little use for brain in the conditions in which they were developed. The brain of the Deinoceras was only one-eighth the size of the brain of a rhinoceros of the same bulk; and the rhinoceros is a poor-brained representative of the modern mammals. To meet the growing perils of their race they seem to have developed three pairs of horns on their long, flat skulls, as we find on them three pairs of protuberances. A late specimen of the group, Tinoceras, had a head four feet in length, armed with these six horns, and its canine teeth were developed into tusks sometimes seven or eight inches in length. They suggest a race of powerful but clumsy and grotesque monsters, making a last stand, and developing such means of protection as their inelastic nature permitted. But the horns seem to have proved a futile protection against the advancing carnivores, and the race was extinguished. The horns may, of course, have been mainly developed by, or for, the mutual butting of the males.
The extinction of these races will remind many readers of a theory on which it is advisable to say a word. It will be remembered that the last of the Deinosaurs and the Ammonites also exhibited some remarkable developments in their last days. These facts have suggested to some writers the idea that expiring races pass through a death-agony, and seem to die a natural death of old age like individuals. The Trilobites are quoted as another instance; and some ingenious writers add the supposed eccentricities of the Roman Empire in its senile decay and a number of other equally unsubstantial illustrations.
There is not the least ground for this fantastic speculation. The destruction of these "doomed races" is as clearly traceable to external causes as is the destruction of the Roman Empire; nor, in fact, did the Roman Empire develop any such eccentricities as are imagined in this superficial theory. What seem to our eye the "eccentricities" and "convulsions" of the Ceratopsia and Deinocerata are much more likely to be defensive developments against a growing peril, but they were as futile against the new carnivores as were the assegais of the Zulus against the European. On the other hand, the eccentricities of many of the later Trilobites—the LATEST Trilobites, it may be noted, were chaste and sober specimens of their race, like the last Roman patricians—and of the Ammonites may very well have been caused by physical and chemical changes in the sea-water. We know from experiment that such changes have a disturbing influence, especially on the development of eggs and larvae; and we know from the geological record that such changes occurred in the periods when the Trilobites and Ammonites perished. In fine, the vast majority of extinct races passed through no "convulsions" whatever. We may conclude that races do not die; they are killed.
The extinction of these races of the early Condylarthra, and the survival of those races whose descendants share the earth with us to-day, are quite intelligible. The hand of natural selection lay heavy on the Tertiary herbivores. Apart from overpopulation, forcing groups to adapt themselves to different regions and diets, and apart from the geological disturbances and climatic changes which occurred in nearly every period, the shadow of the advancing carnivores was upon them. Primitive but formidable tigers, wolves, and hyenas were multiplying, and a great selective struggle set in. Some groups shrank from the battle by burrowing underground like the rabbit; some, like the squirrel or the ape, took refuge in the trees; some, like the whale and seal, returned to the water; some shrank into armour, like the armadillo, or behind fences of spines, like the hedgehog; some, like the bat, escaped into the air. Social life also was probably developed at this time, and the great herds had their sentinels and leaders. But the most useful qualities of the large vegetarians, which lived on grass and leaf, were acuteness of perception to see the danger, and speed of limb to escape it. In other words, increase of brain and sense-power and increase of speed were the primary requisites. The clumsy early Condylarthra failed to meet the tests, and perished; the other branches of the race were more plastic, and, under the pressure of a formidable enemy, were gradually moulded into the horse, the deer, the ox, the antelope, and the elephant.
We can follow the evolution of our mammals of this branch most easily by studying the modification of the feet and limbs. In a running attitude—the experiment may be tried—the weight of the body is shifted from the flat sole of the foot, and thrown upon the toes, especially the central toes. This indicates the line of development of the Ungulates (hoofed animals) in the struggle of the Tertiary Era. In the early Eocene we find the Condylarthra (such as Phenacodus) with flat five-toed feet, and such a mixed combination of characters that they "might serve very well for the ancestors of all the later Ungulata" (Woodward). We then presently find this generalised Ungulate branching into three types, one of which seems to be a patriarchal tapir, the second is regarded as a very remote ancestor of the horse, and the third foreshadows the rhinoceros. The feet have now only three or four toes; one or two of the side-toes have disappeared. This evolution, however, follows two distinct lines. In one group of these primitive Ungulates the main axis of the limb, or the stress of the weight, passes through the middle toe. This group becomes the Perissodactyla ("odd-toed" Ungulates) of the zoologist, throwing out side-branches in the tapir and the rhinoceros, and culminating in the one-toed horse. In the other line, the Artiodactyla (the "even-toed" or cloven-hoofed Ungulates), the main axis or stress passes between the third and fourth toes, and the group branches into our deer, oxen, sheep, pigs, camels, giraffes, and hippopotamuses. The elephant has developed along a separate and very distinctive line, as we shall see, and the hyrax is a primitive survivor of the ancestral group.
Thus the evolutionist is able to trace a very natural order in the immense variety of our Ungulates. He can follow them in theory as they slowly evolve from their primitive Eocene ancestor according to their various habits and environments; he has a very rich collection of fossil remains illustrating the stages of their development; and in the hyrax (or "coney") he has one more of those living fossils, or primitive survivors, which still fairly preserve the ancestral form. The hyrax has four toes on the front foot and three on the hind foot, and the feet are flat. Its front teeth resemble those of a rodent, and its molars those of the rhinoceros. In many respects it is a most primitive and generalised little animal, preserving the ancestral form more or less faithfully since Tertiary days in the shelter of the African Continent.