believe that it was the first of these periods that produced the glacial epoch. Guided by the rate and amount of post-glacial erosion, however, Dr. Croll concludes that it was the more recent period which corresponded with the Quaternary glacial epoch; but he suggests that the earlier period may have coincided with the Miocene glacier. He thinks that the glacial period of the Quaternary lasted from shortly before the last great maximum until about 80,000 years ago (when the eccentricity was 0·0398, corresponding to a difference of twenty-two days in the length of the seasons), or for about 160,000 years—including, of course, the alternating interglacial periods.
Aside from the strong inherent evidence of the approximate correctness of this determination of the date of the glacial epoch (any uncertainty being due to the imperfection of Leverrier's formulæ), there is an abundance of independent testimony leading to substantially the same conclusion. Many eminent geologists have calculated the duration of post-glacial time from various data—generally the rate and amount of erosion or deposition in stated localities—with results usually ranging from 100,000 to 300,000 years. The mean of several of the most reliable is a trifle less than 200,000 years. Now, while each of these results, when viewed singly, may properly be regarded, in the words of Sir John Lubbock, "not as a proof, but as a measure of antiquity," they may, when viewed collectively, justly be considered reliable within wide limits, and to prove that no less a period than 40,000 or 50,000 years can have elapsed since the retreat of the ice sheet from temperate latitudes; and the time has now come when he who endeavors to fix a later date for that event, without showing why these estimates should be rejected, need not be astonished if his efforts only bring him into contempt. So great is the weight of this independent testimony, indeed, as to warrant the suggestion that the glacial epoch of the Quaternary did not extend down to the period of high eccentricity 80,000 years ago, as Dr. Croll intimates, but closed 160,000 or 170,000 years ago. This last maximum might, then, be represented by the Reindeer epoch of Europe, and possibly (if it may be permitted in this strongly reactionary age to suggest the bare possibility that the pioneers in the field of American archæology, At water, Morton, Squier, and their compeers, did not err most egregiously in their estimates of the antiquity of the earlier works of our prehistoric races) by the migration into Mexico and Central America of the mound-builders of the Mississippi Valley—a migration which might thus furnish a parallel with that southerly migration of the Pliocene mammalian fauna at the inception of the glacial period, to which the Indian geologists attribute the richness and variety of the Siwalik and related fossil fauna of the Orient. It might further be urged that, if the duration of the ice age were so great as Dr. Croll suggests, we would be likely to find more unequivocal evidence of the fact in structural variations in those species which survived the cataclysm.
