174 Fretp Museum or Natura History — Zoéxoey, Vou. X. 
and relatively heavier dentition with the last upper molar particularly 
large and the whole toothrow in consequence more crowded. 
Measurements of the skull and teeth are as follows: Occipito-nasal 
length 137.5; basal length 144.8; basilar length 136; basion to spheno- 
ideum 37.4 (44) ;*sphenoideum to gnathion 102 (112);zygomatic breadth 
82.3 (91); breadth across postorbital processes 37; interorbital con- 
striction 26.7; breadth of braincase 49.5 (51); length of nasals 54 (62); 
length of palate 77.3 (82); length of Pm.! 4.6 (4); Pm.? 9 (8); Pm.® 10.7 
(10); length and breadth Pm.‘ 17.7 x 8.3 (15 x 7); M.1 10.6 x 14.5 (10x 
14); M.26.5 x ro.r (5.5 x 6.5); length of Pm.; 4.4 (4); Pm. 9. 3 (8); Pm.z 
to (10); Pm. 10.9 (11); length and breadth M., 17x 7 (26 x7); Mes 
8x 5.7 (8x 4); M.s 3.9 x 3.3 (3 x3). 
For the present, I have retained this species and elnsay allied forms 
in the genus Canis, feeling that this course is desirable until such time 
as an exhaustive phylogenetic study of all the South American canids 
is made. An able contribution to this subject has been made recently 
by Pocock} who points out certain cranial and dental characters dis- 
tinguishing Cants antarcticus from Canis latrans.t He also demonstrates 
a degree of similarity between Canis antarcticus and C. thous. The 
thous group, however, shows similarity to Urocyon, and speaking gen- 
erally it might be said that true Canis and Urocyon are at opposite ends 
of a series in which species like C. antarcticus and C. thous form the 
intermediate steps. In this series C. antarcticus unquestionably falls 
nearer to true Canis than to Urocyon. Just where generic and sub- 
generic divisions should be made, therefore, depends upon more thor- 
ough investigation than has yet been made.§ 
* Measurements in parentheses are those of the type of C. magellanicus published 
by Mivart (Monogr. Canide, pp. 55-56, 1890). 
t Proc. Zool. Soc. Lond., pp. 382-393, 1913. 
t Much larger series of the /atrans group than were available to Pocock bear out 
his general observations, although occasional skulls show a slightly developed 
parietal shield quite equal to the condition in certain species of ‘‘ Cerdocyon.”’ 
§ While the above is in press, the current (March) number of the Annals and 
Magazine of Natural History comes to hand with its important discussion of the 
names of South American Canide by Thomas. It is regrettable from the standpoint 
of those less fortunately situated, that one so eminently competent and so favored 
with material should not have characterized the various groups otherwise than 
nomenclaturally, for at least part of them have not been defined zodlogically, 
The Peruvian specimen mentioned above is doubtless referable to Pseudalopex 
culpaus andina of Thomas, described as slightly different in color from P. ¢. reisst 
of Ecuador and probably also of that part of Peru in which our work was done. 
P. c. magellanicus is recognized as a third subspecies and P. lycoides of Tierra del 
Fuego is given specific rank, 
