various South- American Mammals. 351 



always been involved in extreme confusion, partly, no doubt, 

 because of the doubtful standing of the groups tiiemselves, 

 but more because of the constantly incorrect determination of 

 the species on whicii the names are founded. 



The chief of these causes of error lies at the door of 

 " Cariis azanv, Wied/^ a name which, though made in 

 honour of Azara, with mention of the latter's Agouaradiay, 

 was clearly based on a member of the Crabeater group. It 

 has, however, been commonly used for the Agouarachay of 

 Paraguay, and from this error much of the confusion has 

 arisen. 



Dr. J. A. Allen, in his ' Mammalia of Soutiiern Pata- 

 gonia^*, has made a valiant effort to clear up the confusion, 

 but, owing to his not appreciating the various misdetermina- 

 tioiis of species that liave taken place, his results do not seem 

 to be completely acceptable. I propose, however, to accept 

 as far as possible his selection of the genotypes wherever 

 these are doubtful. 



The names JSpeot/ios (syn. Icticyon) for venaticus and 

 Clinjsocyon for juhatus are clearly settled, and do not need 

 further reference. 



Dusicyon, Ham. Smith (1839), contained four species, of 

 which, following Allen, we may accept Cams antarcticus as 

 the genotype. On this basis it forms a group distinct from 

 anything on the S.-American continent, and peculiar to the 

 Falkhmd Islands. Two species, botii now extinct, from the 

 AVest and East Falklands respectively, are contained iu it, 

 the second one being described below. 



The next name is Cerdocyon, Ham. Smith. From the 

 four species included, Dr. Allen selects " Cants azarce, Wied, 

 and Vulpes magellanicus, Gray," as being its basis and being 

 congeneric. Put this is not the case, for Cants azarce, Wied, 

 is a Crabeater, while niayellanicus is one of the Agouarachay 

 group. I shall return to this name below. 



The next is Lycalopex, Burm. (1^51), and for this 

 Dr. Allen selects vetulus as genotype, but, on the ground that 

 vetulus, Burmeister, is not the same as vetulus, Lund, he 

 renauK'S this group Eunothocyon, Matthew's Nothocyon having 

 been accidentally rendered unavailable for it, lor reasons 

 whicli he explains. 



Put whether or not Purraeister and Lund's vetulus are 

 sptLiJically difFeient from each other (which, after seeing 



« P. 151 et sojrj. (1005). 



