I cannot help suspecting that many tusks and teeth of the mammoth, said to have been found in peat, may be as spurious as are the horns of the rhinoceros cited more than once in the 'Memoirs of the Wernerian Society,' as having been obtained from shell-marl in Forfarshire and other Scotch counties; yet, between the period when the mammoth was most abundant, and that when it died out, there must have elapsed a long interval of ages when it was growing more and more scarce; and we may expect to find occasional stragglers buried in deposits long subsequent in date to others, until at last we may succeed in tracing a passage from the post-pliocene to the recent fauna, by geological monuments, which will fill up the gap before alluded to (p. 144) as separating the era of the flint tools of Amiens and Abbeville from that of the peat of the Valley of the Somme.
How far the lacustrine strata of North America, above mentioned, may help to lessen this hiatus, and whether some individuals of the Mastodon giganteus may have come down to the confines of the historical period, is a question not so easily answered as might at first sight be supposed. A geologist might naturally imagine that the fluviatile formation of Goat Island, seen at the falls of Niagara, and at several points below the falls,[1] was very modern, seeing that the fossil shells contained in it are all of species now inhabiting the waters of the Niagara, and seeing also that the deposit is more modern than the glacial drift of the same locality. In fact, the old river bed, in which bones of the mastodon occur, holds the same position relatively to the boulder formation as the strata of shell-marl and boggy-earth, with bones of mastodon, so frequent in the State of New York, bear to the glacial drift, and all may be of contemporaneous date. But in the case of the valley of the Niagara, we happen to have a measure
- ↑ Travels in North America, by the Author, vol. i. ch. ii.; and vol. ii. ch. xix.
