Ma)itmals J'r<»ii Jnjuj. 11)5 



10. Dasyprocta varietjata bolivun, Tlios. 

 Immature skull, c^. Villa Ciiroliua. 500 m. 



17. Galea comes, TIios. 



c? . 635, eeC^, 670, 691, 708 ; $ . 643, 657, 661, 064:, 696, 

 700, 701. Villa Carolina. 1258 in. 



18. St/Jvi/agiis brasiUensis gihsoni, Thos. 



S. 690. Villa Carolina. 500 m. 



Not fully adult. Nape-patch less rufous than in type. 



19. Marmosa^ hudini, sp. n. 



c?. 714. Altura de Yuto, Rio ISan Francisco. Alt. 

 500 m. 



"Caught in an upland wood.'' — E. B. 



A medium-sized species, grey above and bufFy yellowish 

 below. 



* I am quite uuable to accept the nomenclatmal rcsultsof Dr. .Matscliio's 

 recent paper ou the Didelphiidse (SB. Ges. Nat. Fr. Berl. 1916, p. 25!t). 

 because, as in other cases, he bases his wliole work ou the obsolete and 

 now generally discarded principle of elimination, instead of using mode: n 

 methiids for the identitication and selection of genotypes. Some of his 

 conclusions in the present case would he specially unacceptable to workers 

 in general, such as his entire ignoring of my selection in 1888 of brachi/- 

 tirus as the type of Feramys, Less., and his long and complicated argu- 

 ments that because the other species of the original Feramys — bracfiyurus, 

 tristriald, aud pusillu — fall into other genera, the fourth species men- 

 tioned — crassicnudata — must be taken as the genotype. Such a definite 

 selection of the genotype of Perafiiys (hvachyitra) as tliat in the ' Catalogue 

 of Marsupials ' is in accordance with modern usage and cannot be ignored. 



With regard, however, to Monodelphi^, Biirm., altlioiigh Dr. ^[atschie's 

 selection of its genotype is obtained, as I consider, in the wrong wav,yet 

 he has made a selection, and, in the absence of an earlier one, that woiild 

 bt) valid, and I would therefore accept *' braclujura " as its genntvpe. In 

 consequence Monndrlphis would antedate and supersede Peramys for the 

 genus containing- the common short-tailed opossum. 



All Dr. Matschie's recent nomenclature work is similarly based on this 

 unsound principle of elimination, so that his e.vceptional literary know- 

 ledge is rendered nugatory so far as the utilization of his results is 

 concerned. 



Incidentally I may note that the group called Micoureut by Matschie, 

 who quotes its type as I), luniyer, Desm., appears to need a new nanu\ as 

 M Huunns, Le.^s.)n, with type by sub.-.equent selection I), citierea (Thomas, 

 1888), properly goes to quite a different group. I would suggest the 

 name Slftllodflphya iox the former, with I), laniyer i^« its genotype. It 

 .should, I think, rank as a subgenus of the gtiius I'/iilunder, whoso geno- 

 type, by tautonymy, is rhilander philundcr, L. 



