252 



Dll. KNUD ANDERSEN ON BATS. 



[Apr. 7, 



hand, the loss of m^ is, in the large majority of individuals (about 

 90 p. ct.), more or less compensated by a slight increase in the 

 postero-internal portion of m^ (by a filling up, partly or completely, 

 of the notch befcween cusps 5 and 7), and, consequently, mg has 

 almost as much to work against in the upper jaw as in those 

 species which possess an m^ ; the disappearance of mg is, probably 

 for this reason, of extreme rarity in the southern races. 



So far as m^ and the notch in the posterior margin of m" are 

 concerned, the various stages represented by A. planirostris and 

 jamaicensis may be tabulated as follows : — 



(A) m3 present on both sides : almost all A. planirostris (94 p. ct.). 



(B) ra^ present on one side, entirely wanting on the other : 3 p. ct. of A. plani- 



rostris. 



(C) m-* entirely wanting on both sides : 3 p. ct. of A. planirostris ; all A. jamai- 



censis. 



(1) Notch in hinder margin of m-, between cusps 5 and 7, perfectly preserved 



(i. e. as distinct as in any Artibeus which possesses an m^) : the large 

 majority (77 p. ct.) of the individuals of the northern races of A. Jamai- 

 censis {A. J. parvipes, yucatanicus, jamaicensis, cequatorialis) ; a small 

 minority (11 p. ct.) in the southern races {A. j. lituratus, palmarmn 

 praceps). 



(2) Notch in hinder margin of m^ decidedly reduced : a minority (19 p. ct.) in 



the northern races ; a large number (38 p. ct.) in the southern races of 

 A. jamaicensis. 



(3) Notch in hinder margin of m" almost or completely filled up : a vanishing 



minority (4 p. ct.) in the northern ; half the iiumber of individuals 

 (51 p. ct.) in the southern races oi A. jamaicensis. 



The subjoined table and diagram are intended to give a view 

 of the gradual reduction of the notch between cusps 5 and 7 of 

 m^ in the various races of A. jamaicensis (one race, A. j. prceceps, 

 as being too poorly represented in the collections examined, 

 is excluded from the table; it apparently agrees with its nearest 

 relatives, A. j. pahnarum and liti(,ratus) : — 



