314 DR. KNUD ANDERSEN ON BATS. [Apr. 7, 



General Remarhs. 



^ (a) Artificial and natural arrangement of the sjjecies. — The 

 arrangement of the species of Artibeus given in the foregoing 

 pages is based primarily on their number of molars : ^, |, or | ; it 

 has the practical advantage of facilitating the identification of the 

 species ; it is easier to count the teeth than to study their detailed 

 structure. But I have no doubt that it is thoroughly artificial. 

 Briefly epitomised the arrangement is this : — 



A. f molars. 



a. Cusp 7 of ml small : no living species. 



h. Cusp 7 of ml large : concolor, planirostris, hirsutus. 



B. f molars. 



a. Cusp 7 of m' small; glaueus, watsoni. 



b. Cusp 7 of ml large : jamaicensis. 



C. f molars. 



a. Cusp 7 of ml small ; chief eus, rosenhergi. 



h. Cusp 7 of ml large : qtiadrivittaUis, toltecus, phceotis, aztecus ; — furpis. 

 nanus. 



This and any other arrangement of the species based primarily 

 on the number of molars is, I believe, open to the following 

 objections : — Is it likely that the presence or absence of a rudi- 

 imentary tooth (m^ and nig) is moi'e important, in taxonomic respect, 

 than the general level of development at which the structure of 

 the upper molars (small or large cusp 7 in m^) has arrived ? It 

 has been pointed out in the foregoing pages that a certain small 

 percentage of individuals of the species with normally f molars 

 have |- only (see A . planirostris and hirsutus) ; that in the species 

 with normally ^ molars individuals occur which have | only 

 {see A. ja^naicensis); and that a few individuals of species with 

 normally | molars have ^ (see ^4 . rosenhergi and toltectos) : bearing 

 this in remembrance, is it then likely that a character which is 

 vacillating among individuals of the same species and geographical 

 race is of primary, and a character which is constant within the 

 species (the structure of the upper molars) of secondary import- 

 ance ? If the 7iuinber of molars were of fundamental importance 

 in this genus, is it then likely that we should find in the lowest 

 section (|^ molars) species which have the most advanced structure 

 of the upper molars (planirostris, hirsutus), and among the 

 species of the highest section (|^ molars) such as have retained a 

 low character (small cusp 7) in the upper molars ? Is it likely 

 that A. jamaicensis, because it has lost the rudimentary m^ 

 (although some of the races have as a rule retained the notch in 

 m'^ indicative of the former place of m^), is more closely related to 

 A. glauGUS and watsoni, from which it diifers in the structure of 

 the upper molars, than to A . planirostris, with which it accords in 

 this as in every other respect except the loss of a rudimentary 

 tooth ? These and similar considerations seem, with necessity, to 

 lead to the conclusion that the old and till now universally 

 accepted arrangement of the species according to their number of 



