N THE early part of the present century collections of mammals were from widely scattered localities throughout North America, and nothing was known about the vast intervening areas. It was natural, therefore, to assign full specific rank to nearly every kind of mammal that was dignified with a latinized name. As collections increased, and as geographic gaps narrowed, it became apparent that some of the species had more or less unbroken ranges that extended to those of other named kinds. As a consequence, the number of species decreased and the number of subspecies increased. At the time of Osgood's revision of the genus Peromyscus (1909) most of the named kinds of North American mammals were given full specific rank. Accordingly, there were some 28 species of Peromyscus that were destined to become subspecies or synonyms of what Osgood considered to be one species, maniculatus. Thus, it was Osgood's classical work that focused thinking in terms of the trinomial rather than the binomial, as had been common practice among his predecessors as well as many of his contemporaries. To cite two more examples, Miller listed 24 species of weasels (Mustela) in his 1911 checklist (1912) and 25 species in his 1923 checklist (1924); Hall (1951) reduced the number of species of weasels occurring over the same geographic area to three, with 60 subspecies. In the coyotes of the genus Canis, Miller (op. cit.) listed 14 species; these were recently grouped under one species by Jackson (1951), with 19 subspecies. This does not indicate that evolution has been going on in reverse. It does indicate that our interpretations now are quite different from those of students of 50 years ago. Along with the change in evaluation of the lower categories, there grew up a school, of which I was one, of subspecies describers. I fear that the objective of many in this group was to discern small differences in samples of populations from different localities rather than to study the similarities of the samples, or the variability within a population. Our most horrible example of what can result from this indiscriminate naming of local populations is to be seen in the present literature on pocket gophers of the genus Thomomys. In one species, T. bottae, we now have listed more than 150 subspecies, and there are more to come. If this continues, eventually every colony of Thomomys bottae will bear a formal name recognized by the International Commission on Zoological Nomenclature. It is now necessary to have a large-scale map just to indicate the type localities of all the subspecies of Thomomys bottae. What does all this show from the biological point of view? It shows that Thomomys bottae is highly variable throughout its range. But is it necessary to have 150 or more names in the literature to show this? I think not. I sincerely believe that the variations in most of the morphological characters used as bases for naming the subspecies could be shown with symbols on a map, as some of the botanists have been doing with plants and as Blair (1953) attempted to do with Peromyscus. Blair's symbols
Read full abstract