68 



SCIENCE 



[N. S. Vol. XLI. No. 1045 



wliieli he is able to comprehend without ma- 

 king extended explorations, under the personal 

 escort of one of the initiated. Until then, such 

 expressions as " Upper Sonoran," " Transi- 

 tion " and the like will be to him mere empty 

 names, or at best, they will recall to his mind 

 certain colored areas, on a map of North 

 America, the boundaries of which seem to have 

 been chosen quite arbitrarily. 



The second type of " distributional be- 

 havior " recognized by Grinnell is that which 

 he terms "faunal." The various life-zones 

 are each subdivided into a number of " faunal 

 areas " (or, more simply, " faunas "), " on the 

 causative basis of relative uniformity in humid- 

 ity." We must, at the outset, question the 

 wisdom of appropriating the word " fauna " for 

 use in such a restricted and technical sense, 

 particularly since quite a variety of meanings 

 have already been attached to it by previous 

 writers. 



Grinnell regards it as "probable that every 

 species is affected by both orders of geographic 

 control" (i. e., temperature and humidity), 

 though believing the influence of temperature 

 to be the greater of the two. While this belief 

 in the role played by humidity does not appear 

 to be based, in any single case, upon exact 

 observational data, it would surely be unrea- 

 sonable to throw it out of consideration on 

 that account. One does not require an accu- 

 rate hygrometer to sense the difference in 

 humidity between the atmosphere of the red- 

 wood district of northwestern California and 

 that of the Mojave Desert. We can not help 

 wondering, however, whether sufficient care 

 has been taken to disentangle the effects of 

 atmospheric humidity from those of rainfall 

 and soil humidity. Regions of high atmos- 

 pheric humidity may be regions of high rain- 

 fall as well, but the reverse is not infrequently 

 true, as witness the coast of southern Cali- 

 fornia. It is a matter of common knowledge 

 that vegetation is far more affected by the 

 though the latter is also an important factor.^ 

 moisture of the soil than by that of the air, 



2 Transeau {American Naturalist, December, 

 1905) contends that the controlling influence for 

 plants is the ratio of rainfall to evaporation. 



Since the distribution of animals is so largely 

 conditioned by that of plants, the indirect 

 effects of rainfall and soil humidity upon the 

 fauna of a region are beyond doubt. With 

 rodents and other burrowing animals it seems 

 not unlikely that the effects are much more 

 direct. 



The third " category of distributional con- 

 trol " recognized by Grinnell is that which he 

 terms " associational." By " associations " he 

 means " tracts of relatively uniform environ- 

 mental condition, including their inanimate 

 as well as living elements." They are, of 

 course, subdivisions of a "fauna," just as 

 " faunas " are subdivisions of a " life-zone." 

 The " association " proper to a species repre- 

 sents its habitat, in the narrower sense, as 

 distinguished from its geographical range. It 

 is here that we find the most conspicuous 

 correlation of the " so-called adaptive struc- 

 tures of animals . . . with certain mechanical 

 or physical features of their environment." 



The associations considered in the present 

 report are all (except one) named for some 

 characteristic plant, and, in fact, the term 

 itself is borrowed from the botanists, by whom 

 this conception was first developed. Ten of 

 these associations are distinguished in the 

 lower Colorado Valley traversed by the expe- 

 dition under consideration. This whole region, 

 however, belongs to the " Colorado Desert 

 Fauna " and to the " Lower Sonoran Life- 

 Zone." 



The report contains a considerable fund of 

 valuable ecological detail, palpably based upon 

 careful observation, and in a large degree co- 

 ordinated, so as to lead to conclusions, or at 

 least to definitely formulated problems. In 

 this last respect it stands in gratifying con- 

 trast to the recent output of some of the pro- 

 fessed exponents of the science of ecology. A 

 highly interesting special instance is Grinnell's 

 discussion of associational restriction, as illus- 

 trated by the various species of pocket-mice 

 (Perognathus). Here a truly quantitative 

 mode of treatment has been resorted to, and 

 very instructive results reached, despite the 

 comparatively small number of individuals. 



