498 bulletin: museum of comparative zoology. 



more bowed than that of the type of Pachycyon, and in his opinion 

 the Peruvian Dog corresponded closely to a European Turnspit or 

 Dachshund, whence he calls it Canis ingae vertagus. The figures of 

 the skull of the same specimen likewise show an apparent similarity 

 in outline and proportions to that of the Arizona mummy. 



There seems thus to be no doubt that Pachi/ci/on robustiis is after 

 all only a breed of dog cultivated by the Indians of the southern 

 parts of North America and of Peru. It is therefore no longer to be 

 thought of as a problematical mammal of the Pleistocene. 



Among the dog-bones obtained by the University of California's 

 in^■estigations of the Indian shell-moimds on San Nicolas Island, off 

 the coast of southern California, are two crania nearly identical in 

 measurements with the Marsh Pass specimen that appear to repre- 

 sent this same small, short-nosed dog. They are characterized by 

 their broad brain-cases, spreading zygomata, wide palates, shortened 

 rostra, and small teeth. In profile the dorsal outline of the brain- 

 case is gently rounded, not fiat. The shortness of the rostrum does 

 not amount to real deformity however, for the lower jaw closes nor- 

 mally into its place and the premolars are not markedly crowded, 

 though p^ is turned at an angle of nearly 50° from the axis of the skull 

 to adapt its position to the sudden narrowing of the skull at this point. 

 Premolars 1 and 2 are normal in position, and there is a short diastema 

 between p^ and the canine. The ossification seems particularly 

 heavy, yet though old, neither skull has de\"eloped a sagittal crest 

 except at the interparietal region. In the dried mummy from Marsh 

 Pass, the shortened nose and elevated forehead give a characteristic 

 appearance to the head which is evident in these crania as well. No 

 liml)-bones that can be assigned to this dog, ha\"e appeared among the 

 Calif ornian collections. In both crania the opening of the posterior 

 nares is narrow, and a transverse line drawn at right angles to the 

 cranial axis at the posterior end of the palate falls behind the last 

 molar, iufiicating deviation from the normal condition. 



The following skull-measurements show close agreement. One of 

 the Calif ornian crania (r^is^) lacks any trace of the alveoli of ?»- 

 which are partly broken and partly resorbed. The first premolar is 

 wanting also. The proportions of the maxilla are, however, practi- 

 cally the same in both specimens. The Ancon specimen is figured 

 by Nehring (lS84b) of natural size and the measurements are taken 

 from this figure. It too lacks the first upper premolar, and in every 

 respect confoniis to the appearance of the other crania. 



