266 
JOURNAL OF MAMMALOGY 
form is the first treated. The accounts include the principal synonymy, a 
“diagnosis,’’ discussion of relationships, and full paragraphs under several other 
headings — the type locality and distribution, together with localities and numbers 
of specimens examined, being placed at the end. Measurements, including “ear 
from crown,” when available, are given under each species, and in many instances 
the weights in grams, a valuable feature. 
In regard to methods of measuring the author states that total length given “is 
the distance (with body and tail straightened out) from the tip of the nose to the 
tip of the last caudal vertebra, taken usually after skinning. If this measure- 
ment is taken in kangaroo rats before skinning, there is a chance of error because 
the skin sometimes slips backward some millimeters free from the actual tip of 
the vertebral series. Tail vertebrae is length of tail alone, from a point on upper 
side at base where tail can be bent at right angles to back, to tip (as just desig- 
nated).” It seems to the reviewer that such a method of measuring specimens 
is an attempt to attain a degree of accuracy impracticable in general work. The 
most valuable field measurements, on the whole, are those taken by as nearly 
standardized methods as possible adopted by preparators in general, because 
these will be fairly comparable, while variations from the general standard may 
be misleading in application without the constant repetition of explanations. 
Measurements taken by different collectors, whether before or after skinning, 
vary with the tension exerted or with other individual peculiarities of method, 
and they vary, moreover, with the degree of relaxation of the body of the animal. 
Tables of measurements usually reveal considerable range of individual variation 
in total length and length of tail in animals of comparable age, and the average 
is obviously governed by the relative number of large or small examples chosen. 
These measurements must therefore usually be regarded as approximations only, 
which for most practical purposes should be taken before skinning, the body and 
tail being carefully straightened or extended to the natural limit, but not 
stretched. 
The taxonomic treatment of a group, especially the number of forms recognized, 
their status as species or subspecies, and the distribution area assigned to each 
is likely to vary always with the varying standards set up by different workers. 
Perhaps no two would arrive at exactly the same conclusions. Local variants 
are of great interest to the close student of speciation and geographic distribution, 
and the current tendency is toward greater refinement in the process of “split- 
ting,” in such classes as mammals and birds at least, until the forms, if accepted, 
become so numerous and their ranges so restricted that the determination of new 
accessions of specimens is accomplished only with the greatest difficulty. The 
tendency to set up an excessive number of local forms is apt to be marked in faunal 
papers, in which the irregular, minor geographic variations presented by many 
widely ranging species are appraised at more than their true value. The reviser 
of the genus has the advantage of the wider viewpoint of the group as a whole. 
In so diversified a region as California the desirability and practical utility of 
recognizing a considerable number of well-marked geographic races of kangaroo 
rats, some of which have very restricted ranges, is freely granted. But in the 
opinion of the reviewer the author in some of his conclusions accords undue impor- 
tance to characters ascribed to certain forms, examples of some of which are cited. 
‘^Dipodomys levipes,” orginall3?’ described as a subspecies of “Perodipus microps,” 
