RACCOONS OF NORTH AND MIDDLE AMERICA 53 



interorbital and postorbital areas usually broader; maxillary tooth 

 row shorter; posterior upper premolar and carnassial smaller. 



Measurements. — Adult male from Lochiel, Santa Cruz River, Ariz.: Total 

 length, 895 mm.; tail vertebrae, 365; hind foot, 121. Adult female from Deming, 

 N. Mex. : 840; 305; 125. Adult male and female from Fort Lowell (near Tucson), 

 Ariz., respectively: 890, 820; 325, 305; 131, 125. Skull: Adult male from 

 Lochiel, Ariz., and adult female from Deming, N. Mex.: Greatest length, 121.1, 

 120.6; condylobasal length, 114.6, 115.4; zygomatic breadth, 77, 78; interorbital 

 breadth, 24.2, 24.2; least width of palatal shelf, 55.7, 55.9; maxillary tooth row 

 (alveoli), 42.5, 42.7; upper carnassial, crown length, 8.7, 8.6, crown width, 9.7, 9.7. 

 Adult male and female from Fort Lowell, Ariz.: Greatest length, 123.3, 116.2; 

 condylobasal length, 115.6, 110.9; zygomatic breadth, 83.5, 78.8; interorbital 

 breadth, 28.8, 25.3: least width of palatal shelf, 17.6, 16.4; maxillary tooth row, 

 42.7, 43.6; upper carnassial, crown length, 8.5, 8.8, crown width, 9.5, 9.6. 



Remarks. — P. I. mexicanus shares extremely pale coloration with 

 P. I. yallidus of the Colorado River Valley, typical examples of the 

 two being externally indistinguishable. They are evidently very 

 closely allied, some specimens being practically identical, but the 

 combination of cranial characters pointed out usually serves to 

 separate them as geographic races. The specimen described by 

 Baird and regarded as the type of tnexicanus is a fully adult, but 

 inidersized individiuil, probably a female (greatest length of skull, 

 113) which does not properly reflect the true characters of the sub- 

 species as shown by other specimens from the type region. The 

 type locality was originally given as Espia, Sonora. Examination 

 of Boundary Survey reports and statements by C. B. R. Kennerly, 

 the collector, show that it was taken at Espia, shown on modern 

 maps in extreme northwestern Chihuahua. The error was doubtless 

 due to lack of information in regard to the exact location of the 

 Sonora-Chihuahua boiuidary. To this subspecies are referred speci- 

 mens from Alazatlan, Sinaloa. A specimen from Mazatlan formed 

 the basis of the detailed description by Geofi'roy-Saint Hilaire (1855, 

 p. 125) of a Mexican raccoon, "variete mexicaine," which, however, 

 he did not name. A specimen from northern Xayarit is referred to 

 mexicanus, but those from localities farther south seem more properly 

 assignable to P. I. heniandezii. Specimens from the upper part of 

 the Gila River Valley grade toward and might be referred to paUidus. 

 The series of 47 specimens from Escuinapa, Sinaloa, affords an unusual 

 opportunity for the study of individual variation. 



Specimens examined. — Total nimiber, 87, as follows: 



Arizona: Fort Huachuca, 1; Fort Lowell, 2; Lochiel, 1; San Bernardino Ranch, 

 Cochise County, 1; Santa Catalina Mountains, 1 (skull only); Santa Rita 

 Mountains (McCleary's Ranch), 1. 



Chihuahua: Casas Grandes, 1; Colonia Diaz, 1; Espia, 1 (type, skull only); San 

 Luis Moiintains, 1. 



