still showing us in life to-day the more or less modified representations of forms which elsewhere have long since passed away from among us, leaving but rare and scattered fragments—relics "sealed within the iron hills."
No member of the Australian families of the kangaroo's order has left its relics in European strata more recent than the secondary rocks. But the American family, Didelphidæ, is represented in the earliest Tertiary period by the remains of an American form (a true opossum) having been found by Cuvier in the quarries of Montmartre. He first discovered a lower jaw, and, from its inflected angle, concluded that it belonged to a marsupial animal, and that therefore marsupial bones were hidden in the matrix. Accordingly he predicted that such bones would be found; and, proceeding to remove the enveloping deposit with the greatest care, he laid bare before the admiring eyes of the bystanders the proof of the correctness of his prediction. It is noteworthy, however, that, had this fossil been that of an animal like the Tasmanian wolf, he would have been disappointed, as, though marsupial, it has, as has been already said, not marsupial bones, but cartilages.
But relics of creatures more closely allied to the kangaroo existed in times ancient historically, though, geologically speaking, very recent. Just as in the recent deposits of South America we find the bones of huge beasts, first cousins to the sloths and armadilloeswhich live there now, so in Australia there lived beasts having the more essential structural characters of the kangaroo, yet of the bulk of the rhinoceros. Their bones and teeth have been found in the tertiary deposits of Australia, They have been described by Prof. Owen, and are now to be seen preserved in the British Museum and that of the Royal College of Surgeons. It may be that other fossil forms of the middle mesozoic or even of triassic times may, so some believe, have belonged to creatures of the kangaroo's family; but at least there is no doubt that such existed in times of post-tertiary date.
As to our third point—the geological relations of the kangaroo—we may say, then, that "the kangaroo is one of an order of animals which ranged over the Northern Hemisphere in triassic and oolitic times, one exceptional family lingering in Europe to the Eocene period, and in America to the present day. That the kangaroo itself is a form certainly become fossil in its own region, where, in times geologically recent, creatures allied to it, but of vastly greater bulk, frequented the Australian plains."
We may now, then, proceed to answer finally the question, "What is a kangaroo?" We may do so because the meaning of the technical terms in which the answer must necessarily be expressed (if not of undue length) has been now explained, as far as space has allowed.
We may say, then, that "the kangaroo is a didelphous (or marsupial) mammal, of the family Macropodidæ; an inhabitant of the
