232 DR. KNUD ANDERSEN ON BATS. [Apv. 7, 



individuals of their species ; others, in which the absence of nig 

 is normal, show, individually, a tendency to reversion to the 

 more j^rimitive stage in which this small tooth was present. — A 

 character which is not only in itself very insignificant, but not 

 even individually perfectly constant is evidently unsuitable for 

 the separation of groups of generic or subgeneiic rank. 



(3) By subdividing Artiheus according to the number of molars, 

 the " genei-a ■ ' or "subgenera" would be these three : — molars |-, 

 A. concolor, 2^^<^f'^^'^'>'0^t^'^^j Mrsutus ; molars |-, A, jamaicensis, 

 glaucus, loatsoni ; molars |^, A . cinereths, rosenbet^gi, toltecits, qua- 

 drivittatus, phceotis, aztecus, turpis, nanus. — But A. planirostris is 

 rauch more closely related to A . jamaicensis, which is placed in a 

 difierent genus or subgenus, than to A. concolor, with which it is 

 associated in one group. A. glaiocus and watsoni ai-e put together 

 with A. jamaicensis, solely because they like this latter have a 

 rudimentary m^, but in all other respects they are much more 

 closely related to A. ciner&us, which has permanently lost mg. 

 The eight species with |- molars constitute a strangely hetero- 

 geneous section ; A . cinereus is nearer to A . glaucus and watsoni 

 than to any of the forms with ^ molars with which it is put 

 together; A. rosenbergi is unique in the genus in the strong 

 reduction of m"; A. toUecus, quadrivittatios, and aztectis are, 

 probably, rather more closely allied to A . jamaicensis than to any 

 species with % molars ; and, finally, A . turpis and nanus form 

 a small natural group characterised by the unusually strongly 

 depressed and slightly upwardly dii'ected cranial rostrum. — From 

 this it will be evident that a subdivision of the genus based on the 

 presence or absence of m^ or m^ would give only a very distorted 

 view of the mutual aflinities of the species. 



(4) A study of the species of Artiheus has led the writer of this 

 paper to the conclusion that they, probably, fall into two natui-al 

 groups, which have nothing to do with the hitherto proposed sub- 

 divisions of the genus, viz., those species in which cusp 7 of 

 m^ is relatively sinall {A. glaucus, tvatsoni, cinereus, and rosen- 

 bergi), and those in which it is relatively lai-ge (all the other 

 species). This point, which has more theoretical than practical 

 interest, will be discussed in the last section of the present paper, 

 pp. 314-316. 



Artibeus concolor Pet. 



1865. Artiheus concolor Peters, MB. Akad. Berlin, p. 357. — Paramaribo (Surinam). 

 1878. Artibeus planirostris (not Spix), var. a, Dobson, Cat. Chir. Brit. Mus. p. 518. — 



Upper Amazons. 

 1892. Artibeus concolor Pet., Thomas, Ann. & Mag. N. H. (6) x. pp. 409-410, 



footnote (Nov. 1892). — Some cranial measurements of the type. 

 1901. Artiheus concolor Pet., Thomas, Ann. & Mag. N. H. (7) viii. p. 191 (Sept. 



1901).— Para. 



Diagnosis. — An Artibeus with |^ molars, the maxillary tooth- 

 row measuring about 7 "2, the forearm about 50 mm. 



Teeth. — The teeth of A. concolor are proportionately very much 

 smaller than in the two other species with |^ molars, A . plani- 



