1915.] Allen, Review of the South American Sciuride. 1Di 
published in May, 1899 (Proc. Washington Acad. Sci., I, pp. 15-110, pila 
and ii), based on 795 specimens.! 
From a study of this material Mr. Nelson recognized 41 forms (excluding 
3 from north of Mexico),— 28 species and 13 subspecies, of which 6 were 
described as new, in contrast with my 9 in 1877, and Alston’s 7 in 1898. 
According to Mr. Nelson’s synonymies, my 9 Mexican forms included 13 of 
his, 28? of his 41 forms being based on new material not available in 1877. 
But the quality of the material at these two periods was as different as 
was the amount. In contrast with the roughly prepared and frequently 
distorted, half-filled skins, giving little clue to the size, proportions, or 
external appearance of the animal, and without measurements taken from 
the specimen while in the flesh, or definite data as to sex, locality, and date 
of capture, and the skulls often unavailable for examination, of the earlier 
period, the bulk of the new material consisted of smoothly-made skins, 
giving the correct proportions of the animal, each with its well prepared 
skull, accompanied by labels giving the fullest field data, and the external 
measurements made before skinning. In Mr. Nelson’s case he had the 
further immense advantage of an intimate personal knowledge of the geo- 
graphic conditions of the greater part of the area to which his revision of the 
squirrels related. 
The excessive ‘lumping’ of the earlier period, in respect to mammals, 
was due in large part to the limited amount and the poor quality of the 
material available for study, and in part to the mistaken assumption that 
it was adequate for permanent revisionary work; also to the belief, of at 
least some of the prominent investigators of that day, that the mammal 
fauna of North America, and of some other parts of the world, was already 
pretty well known. Indeed, the announcement of the discovery of new 
forms, in certain groups at least, was looked upon with suspicion. Yet the 
prosecution of extensive field work under new methods of collecting and 
of preparation disclosed new genera as well as new species in supposedly 
previously well-worked fields.? With this increase of material and the 
improvement in its quality has come the possibility, as well as the inclina- 
tion, to recognize finer distinctions than formerly, so that many forms given 
nomenclatural status at present would not have been thus honored had these 
1This may be contrasted with my total of 79 specimens for the same region, which 
included three of Sciurus arizonensis (the only ones then known), referred to S. collie: of 
Mexico. Alston must have had access to much more extensive material, as well as to the 
types of most of the previously described species. 
2 Of these 4 were described by me between 1889 and 1895 — S. nayaritensis (1889), 
cervicalis (1890), apache (1893), alfari (1895). 
3 Miller in 1912, in his ‘List of North American Land Mammals’ (Bull. 79, U. S, Nat. 
Mus.), listed 2138 forms (species and subspecies), as against 363 given by True in 1885, and 
1405 given by Miller and Rehn in 1900. 
