35G Mr. 0. Thomas on 



few that I hesitate for tlie ])resent to use tliis name for the 

 Brazilian animal, as better series may brini^ out some constant 

 distinguishing character. C. thons ranges westwards into 

 Venezuela, and it is possible that the Santa Marta C. aquilus 

 should be united with it. 



At Chapada, Matto Gro?!SO, Mr. Robert obtained three 

 dogs of this group, two of them quite like brasiliens/s, but 

 the third standing out by its long skull and larger teeth from 

 all the rest ; and I see no alternative but to consider this as 

 a special form, like as it is externally to the others. Curiously 

 enough, the same thing happened with Lyadopex, for he 

 obtained at Chapada the type of L. sladeni as well as examples 

 of L. vefulus. Probably the locality being at the edge of the 

 Brazilian plateau, animals from both highlands and lowlands 

 may be collected there. 



The skull of C. mimax in some respects approaches that 

 of the Amazonian C. microtis'^, but the external differences 

 between the two are (juite marked. 



Cerdocyon thous lunaris, subsp. n. 



Allied to C. t. savannarum, but with even smaller skull 

 and teeth. 



External appearance very much as in savannarum, the 

 general colour a little darker and less suffused with buffv. 

 Under surface whitish, suffused with buff on the belly ; chin 

 bhick. Hands and feet grizzled grey above, black on palms 

 and soles. 



8kull markedly smaller than that o^ savannarum, itself the 

 smallest Cerdocyon hitherto described. Brain-case propor- 

 tionally large, its breadth practically equal to that of savan- 

 narum. Muzzle narrow and delicate. Teeth conspicuously 

 .snialler througiiout (see measurements), the difference in the 

 size of ni^ particularly noticeable. Canines short and slender. 



Dimensions of the type : — 



Hind foot 113 mm. 



Skull; condylo-basal length 11G*.5; zygomatic breadth G5"5; 

 nasals on middle line 38; breadth of muz/.le outside p^ 17'.5 ; 

 interorbital breadth 23 ; across postorbital processes 35 ; 

 intertemporal constriction 30 ; l)rain-case, breadth 43 ; j)ost- 

 orl)ital j)rocess to back of deltoid ridge 52 ; palatal length 59. 



Teeth : length ot'^r G'4, ji"^ G"4, p* on outeiedge 11 ; in^ on 

 outer edge 8'8, greatest diameter 1 r4 ; length of lower 

 carnassial 13. 



* I am suHicieiitly a *' ono-l<'ttorist " to see no reason to alter the name 

 microtis because it is antedated bv inicrotm. 



