392 ON THE ANTARCTIC WOLF, 



of the longisli hairs pulled at random from the back measured 

 only 1| inches (37 mm.) in length, the dark apical tip being 

 I inch (6 mm.) and the pale band below it ^ inch (3 mm.) in 

 length. The prevailing colour of the body is bi'own, relieved by 

 the fine speckling due to the narrow pale band on the individual 

 hairs just described. The lower side is white only on the 

 posterior portion of the belly and on the upper end of the throat, 

 thechin and lower jaw being white stained Avith a fuscous tint. 

 Apart from the areas described the ventral surface is brownish. 

 There is, moreover, as Mivart said, a fuscous patch above the 

 hock of the hind leg and the tail is pai'ticoloured, its basal 

 portion being like the back, its tip white, and the intermediate 

 area blackish, the colour of this area gradually blending proximally 

 with the brownish basal portion, but being quite black distally 

 and sharply defined from the white terminal area. 



Perhaps it may rightly be claimed that these differences in 

 the length of the ears and in colour do not count for much in 

 themselves. That may be so, Nevertheless, if G. atitarcticus and 

 O. latrans were only knoAvn from their skins, it is quite certain 

 that the latter would be placed in the same category with such 

 species as C. pallipes and C. lupaster, and that C, antarcticus 

 would be excluded therefrom. The latter would be difficult to 

 classify ; but there is one significant colour-feature connected 

 with the species. This is the presence of the dark patch above 

 the hocks ; and the interest of this lies in the circumstance that 

 it is a very common feature in vai'ious species of the smaller 

 South -American dogs and occurs in some of the species of Vulpes, 

 like V. chama *. 



There is one other little point that may be referi'ed to. 

 Darwin says he was informed that the cries of C antarcticus 

 resembled those of the South- American species C. azarce. I have 

 never heard C. azarm bai-k or howl, but the keeper in the Gardens 

 informs me that examples of wild dogs from Mar del Plata and 

 Cordova, which are closely allied to and perhaps only racially 

 distinct from C. azarce, bark after the manner of foxes. On the 

 whole, however, they are silent dogs in captivity, and, like the 

 foxes, never succum.b to the temptation of joining in the howling 

 concerts in which the dingos, jackals, prairie wolves, and large 

 wolves in the Gardens indulge, and which they seem unable to 

 resist contributing to. Personally I believe that voice in mammals 

 is often a good guide to affinity; and, in the present case, the 

 voices of C. antarcticus and C. latrans bear out njy opinion of 

 the relationship of these species to others, shown by structural 

 characters. 



Finally, if the conclusions above put forwaid are correct, 

 Huxley's classification, expressed in par. 5 (p. 384), must be 

 emended by transferring C antarcticus to the lowest section of 



* The presence of this patch in soino of tlie primitive Caiiidaa is well worth more 

 attention than it has received. J do not l<n(/w what it may mean, any more than 1 

 know what the pale area heliind the shoulder, observable in many Canida?, both wild 

 and domesticated, may mean. 



