152 PATAGONIAN EXPEDITIONS: ZOOLOGY. 



Canis ttrostictiis Mivart, as Dr. Matthew informs me, but as first published 

 it did not include this species, it covering only three previously published 

 fossil species from the Miocene of Oregon, which were later only provis- 

 ionally referred to this genus, and which prove to be, as would be ex- 

 pected, not congeneric with Canis nrostictiis of Brazil. In allusion to 

 this embroilment I propose for this group, should it be deemed worthy 

 of recognition, the name Eunothocyon, with Canis sladcni Thomas as the 

 type.^ Here also belongs Canis parvidens Mivart,- which, like C tirostic- 

 tus, was described as from " Brazil," without definite locality. 



The " Maned Wolf" of southern Brazil, Paraguay and northern Argen- 

 tina is the type and only known representative of Hamilton Smith's "sec- 

 tion" Chrysocyon [Chrysocyon jnbatns= Canis jiibatus Desm.). 



beds of Oregon were referred to it. The genus was first described nearly three months later 

 (June 21,1 899) in a paper by Wortman and Matthew (op. cit., p. 1 24) as " Nothocyon, gen. nov.," 

 with Canis nrostictiis Mivart, as (inferentially from the context) the type, to which genus were also 

 referred Canis pat~videns Mivart and also "provisionally the three John Day species \_Galecymis'\ 

 latidens Cope, [6^] lemur Cope, and Canis geisntarianus Cope." Hay, however, in his " Bibli- 

 ography and Catalogue of the Fossil Vertebrata of North America" (1902, p. 771), in entering 

 this genus, says : "No type was designated, but Canis geisniariamis vm.y h^ \.?kzx\." Palmer, 

 two years later, in his " Index Generum Mammalium" (1904, p. 462), cites Hay, and says 

 " type fixed " (/. e., by Hay, as above). This, according to current rules of nomenclature and cur- 

 rent usage, was a perfectly correct proceeding, although when the genus Nothocyon was defined 

 it was founded on a skull of a living species from southern Brazil (Am. Mus. Osteol. Coll. Mamm., 

 No. 391, from Chapada, Matto Grosso, Brazil = Canis sladeni Thomas, April, 1904), identified 

 and figured as Mivart's Canis iirostictus. Thus the real intention of the authors was inadvert- 

 ently defeated by the original publication of the name without any reference being made to the 

 intended type. The point is covered by Canon XXII of the A. O. U. Code, which is : " In no 

 case should the name of a genus be transferred to a group containing none of the species origi- 

 nally included in the genus." There is also a similar provision in most other modern codes of 

 nomenclature. 



' As stated in the preceding footnote, the real basis of the diagnosis of Nothocyon was a skull 

 from Chapada, Matto Grosso, which for this purpose was described in detail and figured, and 

 identified as Ca?iis nrostictiis Mivart. Three years later Thomas described (P. Z. S., 1903, II, p. 

 235, pi. xxvii, April I, 1904) a small species, based on Chapada specimens, as Canis sladetii, 

 of which the skull used as the basis of Nothocyon is unquestionabiy'a topotype. The type of 

 Eunothocyon is therefore Canis sladcni Thomas. The group will apparently include, besides E. 

 sladcni (Thomas), E. nrostictiis (Mivart), E. parvidens (Mivart), and perhaps other small forms 

 described and figured by Burmeister under apparently preoccupied names, as noted in the next 

 footnote. 



^ Mivart considers, and probably correctly, that the Caiiis vetnliis of Burmeister is not the 

 Canis vetnlus of Lund, leaving Burmeister's C vetiilns without a name, but which he provision- 

 ally refers to his parvidens, as he does also Canis fulvicaudus Burmeister (not Lund). 



