230 DR. KNUD ANDERSEN ON BATS. [Api\ 7, 



the first to make an attempt to subdivide Artibeus into three 

 genera characterised by their number of molars, as follows : — 



(1) Artibceios, molars 4; species three, viz. '■'■ A. jamaicensis," 

 according to the figure of the teeth clearly not A.jamccicensis Jje'Ach, 

 but A. ])lanirost7ns Spix ; the localities given by Gervais, " de la 

 Jamai'que, de la Guadeloupe et de Cuba " are undoubtedly wrong ; 

 further, "^4. lineatus " (^= Vanvpyrops lineatus), and " A. undatus " 

 ( = Stenoderma rufum) ; 



(2) Pteroderma, molars f ; species one, " Pt. 2jersjncillatu7n," 

 "repandue au Perou, au Bresil, et a la Guayane " ; this is, as 

 shown by the figures of the teeth, A. jmnaicensis Leach (sensu 

 lato) ; 



(3) Der')nanura, molars |- ; species one, " D. cinerea" i. e. 

 A. cinereus of the present paper. 



In July 1865*, Peters divided Artibeus into two "groups" 

 (subgenera), viz. : — 



(1) Artibeus Leach, molars |^ or t; species four: " ^. per- 

 spicillatus Geofli"." (i. e. A. jamcdcensis lituratus of this paper), 

 " A. jamaicensis Leach" (i. e. A. jamaicensis jamaicensis), A.fallax 

 (i. e. A. planirostris fallax), and^. concolor ; 



(2) Der'tnanura Gervais, molars ^ ; species three : D. cinerea, 

 ? D. tolteca, and D. quadrivittata. 



Peters placed the species with |^ and f molars together in one 

 group, " weil dieser kleine Zahnstumpf allein kein Grund sein 

 kann, Arten, die sonst im Schadel- und Zahnbau, so wie in jeder 

 anderen Beziehung ganz mit einander iibereinstimmen, generisch 

 von einander zu trennen." But he was not quite consistent ; if 

 "dieser kleine Zahnstumpf" (m'^) is not suflicient reason to 

 separate, as difi"erent groups, species with | and f molars, it is 

 difficult to see why the other, equally rudimentary tooth (nig) 

 furnishes a valid reason to separate, as a distinct section, the 

 species with | molars from those with | molars. From Peters's 

 standpoint there would seem to be two alternatives only, either 

 not to subdivide the genus, acknowledging that the presence or 

 absence of a perfectly rudimentary tooth is a chai'actei' of specific, 

 but not of subgeneric or generic importance, or to subdivide it 

 into three groups. — Peters himself has probably felt the incon- 

 sistency of his classification. At all events, in spite of his own 

 argument, that the species with -| and -| molars " sonst im 

 Schadel- und Zahnbau, so wie in jeder anderen Beziehung ganz mit 

 einander iibereinstimmen," he, only a few months later f, proposed 

 a new subgeneric name, Uroderma, for the species with |^ molars. 

 And, finally, in June 1866 +, he evidently regarded the sections no 

 more as siibgenera, but as genera. — Thus Peters had now, in 1866, 

 practically adopted Gervais's view, that Ai'tibeus is to be divided 

 into three genera, according to the number of molars, viz. : — 



(1) Ui'oderma Peters 1865 (synonym: Artibceus Gervais 1856, 



* Peters, MB. Akad. Berlin, 13 July 1865, p. 356, footnote, 

 t Peters, MB. Akad. Berlin, 13 Nov. 1865, p. 688, footuote. 

 X Peters, MB. Akad. Berlin, 25 June 1866, p. 394. 



