102 DR. ST. G. MiVART ON [Mar. 18, 



to suppose that he can be mistaken as to obvious facts concerning the 

 many individuals which have, doubtless, passed through his hands. 



But a keen appreciation of facts does not guarantee a sound 

 drawing of inferences. I cannot persuade myself that he has not 

 been too apt to draw hasty inferences from insufficient data. Thus in 

 1856 (Fauna Brasil. p. 24) he separated C. cancrivorus and C. vetulus 

 generically from C. azarce and C griseus on the ground that in the 

 first two the sagittal ridge is present though weak, and that it is absent 

 in the last two ; that the upper fourth premolar is a little shorter than 

 the two upper molars in the former pair of species and much sliorter 

 in the latter, and, finally, that the pupil becomes elliptical in the one 

 pair and remains round in the other. 



I have, however, found the sagittal ridge to be very differently 

 developed in different adult skulls of undoubtedly the same species. 

 As to whether the fourth premolar is much shorter or only a little 

 shorter than the two molars behind it, I consider that a very useful 

 specific character but not a valid generic one — at least in the Canidce. 

 I regard the contraction of the pupil as a most unsatisfactory 

 distinction '. 



Twenty years later Dr. Burmeister seems to have come more 

 distinctly to recognize the variability of these Dogs. He says of the 

 South- American forms (Archiv f. Naturgesch. 18/6, p. 117) that 

 " In der Farburg sind nicht bloss alle diese Arten einander sehr 

 iihnlich sondern sie variiren auch etwas nach der Jahreszeit." He 

 also speaks, on the same page, of variation in the skull. 



Without any disrespect, then, to Dr. Burmeister, we must not deny 

 ourselves the right to criticise freely his various representations. 



With respect to the form he describes and figures as C. griseus, 

 he tells us that it is slenderer than C. azara. Now by C, azaros 

 he always intends that form which was described as G. azarce by 

 Mr. Waterhouse. But Dr. Burmeister seems more than once to 

 have changed his mind as to the identity of his and Mr. AVaterhouse's 

 G. azarce with the C. azarce of Wied. Thus in his ' Uebersicht 

 Thiere Brasiliens,' 1854, p. 99, he says of the form he describes as 

 the G. azarce of Waterhouse : " Ich habe auch den Canis azarce, 

 Pr. Max. wieder zu dieser Art gezogen ; " while in his ' Fauna 

 Brasiliens ' (1856, p. 37) he identifies the Prince's G. azarce with the 

 G. vetulus of Lund. In his ' Reise durch La Plata,' 1861, p. 405, 

 and in his description of the Argentine State, vol. iii. p. 147, he 

 leaves out all reference to Lund's C. vetulus amongst the synonyms 

 he there gives of his G. azarce. He distinguishes his own (and 

 Waterhouse's) G. azarm from his new species C, griseus as follows 

 (Fauna Brasil. p. 24) : — 



C. azarcB. G. griseus. 



"Forelimbsgrey to the carpus; "Fore limbs entirely reddish 



soles blackish brown." yellow; soles reddish brown.' 



He further tells us (p. 48), as we have said, that C. griseus is the 



1 For reasons before stated by me, see P. Z. S. 1882, p. 141. 



