1908.] DR. KNUD ANDERSEN ON BATS. 231 



not Artiheus Leach 1821), molars |; species U. bilohatum Pet., 

 ''''U. fallax Pet.," and TJ. concolor ; 



(2) Artiheus Leach 1821 (synonym: Pteroderma Gervais 1856), 

 molars |-; species "A. persjyicillatus Geoff.'" and '■'■A. jamaicensis 

 Leach " ; 



(3) Dermanura Gervais 1856, molars |; species as above. 



By this airangement Peters, as already said, had practically 

 gone back to Gervais's standpoint ; these words are true also in 

 the sense that his arrangement is in no respect an improvement 

 upon the older one ; both of them are typical examples of artificial 

 classification. Gervais selected as the only leading character for 

 his subdivisions the presence or absence of a vanishing tooth ; 

 Peters did precise!}^ the same. Gervais proved the fallacy of the 

 taxonomic character selected by him, in so far as he placed together 

 in one "genus" {Artibceus) a true Artiheus, a Vampyroj^s^ and a 

 Stenoderma, because they, though diflei'ent in many important 

 respects, happen to have -^ molars ; and, on the other hand, sepa- 

 rated into two genera [Artibceus and PteroderincC) two species so 

 closely related as to be sometimes extremely difficult to distin- 

 guish (^J. pfe;^/rOc<f^r«"s and ^'aHtajcevisis). Peters proved the same, 

 by putting together in one "genus" [Uroderma) two generically 

 widely different forms {U. hilobatum and ^' A. fallax"), because 

 the}^ both happen to have f molars, at the same time separating 

 into two "sections" or genera " ^. /«Z^«,;*; " m\A A. jamaicensis, 

 which differ in next to nothing but the presence or absence of a 

 rudimentary tooth. 



In tlie description of the genus Uroderma (above, p. 217) I 

 have given my reasons for keeping U. bilohatum and tJiomasi 

 generically separate from Artiheus. The next question, therefore, 

 is, if, having removed these two species from Artiheus, it might be 

 convenient to divide it into three subgenera or genera, according 

 to the number of molars. Also in this modified shape I am unable 

 to accept Peters's proposal, for the following reasons : — 



(1) The series of species here referred to the genus Artiheus 

 form one natui-al group the members of which are perfectly 

 similai' in all essential cranial, dental, and external characters. 



(2) A. jjlanirostris has ^ molars ; but of 73 skulls examined of 

 this species, two lack m' on one side, two on both sides, and one 

 of these latter also lacks mg on one side. A. hirsutus has f molars ; 

 but of 8 skulls, two lack m'' on one side. A. jamaicensis has f 

 molars; but of 182 skulls, two lack nig on one side, four on both 

 sides. A. rosenhergi has f molars, but of the only two individuals 

 known, the one has an mg on one side. A. toUecus has f molars, 

 but of 26 skulls, one has an mg on one side. — None of the in- 

 dividuals here referred to are aberrant on account of very young 

 or very high age. Some of them, it will be noticed, have lost the 

 rudimentary molar* (m^ or mg) which is normally present in 



* It is liardlj'- necessary to say that in all the aberrant individuals referred to 

 above m^ or ma (or both) are entirely lost, i. e. no trace of their alveoli has 

 been left. 



