We have evidence that this change took place at a very remote epoch. The Swiss anatomist, Dr. J. Kollmann, has published a critical investigation of the most ancient skulls discovered in America, as the one I have already referred to from Calaveras county, California, one from Rock Bluff, Illinois, one from Pontimelo, Buenos Ayres, and others from the caverns of Lagoa Santa, Brazil, and from the loess of the Pampas. All these are credited with an antiquity going back nearly to the close of the last glacial period, and are the oldest yet found on the continent. They prove to be strictly analogous to those of the Indians of the present day. They reveal the same discrepancy in form which we now encounter in the crania of all American tribes. The Calaveras skull and that from Pontimelo are brachycephalic; those from Lagoa Santa dolichocephalic; but both possess the wide malar arches, the low orbital indices, the medium nasal apertures and the general broad faces of the present population. Dr. Kollmann, therefore, reaches the conclusion that “the variety of man in America at the close of the glacial period had the same facial form as the Indian of to-day, and the racial traits which distinguish him now, did also at that time.” The marked diversity in cranial forms here indicated is recognizable in all parts of the continent. It has frustrated every attempt to classify the existing tribes, or to trace former lines of migration, by grouping together similar head-measurements. This was fully acknowledged by the late Dr. James Aitken Meigs, of Philadelphia, who, taking the same collection of skulls, showed how erroneous were the previous statements of Dr. Morton in his _Crania Americana_. The recent studies of Virchow on American crania have attained the same conclusion.[34] We must dismiss as wholly untenable the contrary arguments of the French and other craniologists, and still more peremptorily those attempted identifications of American skulls with “Mongolian” or “Mongoloid” types. Such comparisons are based on local peculiarities which have no racial value. Yet it must not be supposed from this that carefully conducted cranial comparisons between tribes and families are valueless; on the contrary, the shape and size of the skull, the proportion of the face, and many other measurements, are in the average highly distinctive family traits, and I shall frequently call attention to them. The lowest cephalic index which I have seen reported from an American skull is 56, which is that of a perforated skull from Devil river, Michigan, now in the medical museum at Ann Arbor university;[35] the highest is 97, from a Peruvian skull, though probably this was the result of an artificial deformity.