NO. 2057. THE GRASSHOPPER MICE—HOLLISTER. 429 



which to diagnose the two species of Onychomys, for though the pe- 

 cuHarities they exhibit are less easily discerned than are the super- 

 ficial differences that have heretofore answered, they are, of course, 

 of far greater importance and phylogenetic meaning. 



The subspecies of each gi'oup are closely related and intergradation 

 is in almost every case shown by the material examined. The races 

 are, in the main, well marked over a considerable territory and usually 

 exhibit some conspicuous deviation in color accompanied by varia- 

 tions in dimensions or relative cranial characteristics. 



The marked diversity in color of skins of Onychomys from almost 

 any locality has given rise to a general impression that true dichro- 

 matism is, in a measure, responsible for the wide differences exhibited. 

 That such is not the case has been one of the general results of the 

 present study. Cases of supposed dichromatism represent various 

 stages in a definite sequence of pelages, shown by a comparison of 

 animals of various ages as determined by the relative wear of the 

 teeth. In torridus and its subspecies this sequence of pelages is 

 especially complex and, in both groups, abrasion and fading, combined 

 with a gradual moult and renewal, has made the study of the pelages 

 an interesting problem. In an account of the various subspecies it is 

 necessary in nearly every case to describe fully the color of several 

 stages of pelage, so widely do individuals of the same form differ 

 with age and season. 



MEASUREMENTS. 



In the tables of measurements appended only the dimensions of 

 fully adult animals have been considered. Measurements of selected 

 specimens showing considerable tooth wear and from as many 

 localities as possible have been used. The total length, tail, and 

 foot measurements are from collector's notes, made from fresh speci- 

 mens. The length of ear has been taken from the dry skin. The 

 cranial and dental measurements were made with sliding caU]iers 

 registering tenths of a millimeter. Condylobasal length was taken 

 from condylion to alveolar point, not to most anterior point of the 

 premaxillse. 



ACKNOWLEDGMENTS. 



An intelligent revision of the grasshopper mice would not have 

 been possible without assembling the collections from various 

 museums for study. Thanks are due the custodians of several 

 mammal collections for the loan of material, which has, in several 

 cases, included types and type series. To Dr. J. A. Allen, of the 

 American Museum of Natural History, New York; to Mr. Joseph 

 Grinnell, Museum of Vertebrate Zoology, University of California; 

 to Mr. Charles B. Cory and Mr. W. H. Osgood, Field Museum of 

 Natural History, Chicago; to Mr. Samuel Henshaw and Mr. Outram 



