NYcriMErTE. 693 



entirely suppressed. The constant action of the lower canines 

 iie;ain8t the single pair of upper incisors has moulded these into 

 their jteciiliar obliquely triangular shape ; on the other hand, the 

 iipper incisors acting upon the posterior surface of the extremity of 

 the lower canines have ground the cutting-edge of these as sharp as 

 that of the lower incisors of a Rodent, whereas the outer edge of 

 the lower canines, which moves against the inner edge of the upper 

 canines, is worn smooth and rounded. The secondary cusp of the 

 upper canines, if present, is produced by the action of the lower p^; 

 individuals in which p, barely touches or only slightly bites against 

 the external edge of the upper canine show no secondary cusp or 

 have this latter indicated only by a more or less distinct notch in 

 tlio edge. Tliat the posterior " heel " of P3 and p^ is rather more 

 well-marked tlian in other Cynopterine bats is due to the action of 

 the high external cusps of p'' and p* ; similarly, the high cusps of 

 p. and p^ have produced, respectively, a well-developed cingulum in 

 the up[)or canines and well-marked posterior heel iu p^ ; in some 

 individuals p' and the heel of p^ may effect a more or less distinct 

 depression in (" bilobed " shape of) the cutting-edge of the outer 

 cusp of, respectively, Pg and p^. 



(3) Other modifications.— According to Robin, the cardiac 

 portion of the stomach (distinctly differentiated in other Frait^bats) 

 is wanting or, perhaps more correctly, represented only by a conical 

 enlargement of the oesophagus not differing from the latter in tlie 

 structure of its mucous membrane; the fact that this is the type of 

 stomach commonly found in Microchiroptera seems to lend support 

 to the statement that Nyciinune is (at least to some extent) insec- 

 tivorous. The three circumvallate papillae are in Vynopterus and 

 other Fruit-bats arranged in the figure of a triangle, two in front 

 and one larger behind ; the latter is in Nydimene replaced by two 

 smaller (fig. 05 A, p. 725) ; whetiier this difference has anything to 

 do witli the supposed insectivorous habits of the present genus is 

 unknown (the fact that all Microchiroptera, so far as kno\A'n, 

 possess two papilla?, except the bloodsucking Desmodus which has 

 none, would seem to indicate that a difference in diet may liave 

 an influence on the development of the circumvallate papilla3 ; in 

 Insectivora the usual number appears to be three or two). The 

 characters of the wing (yellow spots and proportions of metacarpals 

 and phalanges, as in Bcdionycteris), the colour of the fur, and the 

 sexual colour-difference have been discussed above under the 

 headings " External characters," " Colour," and " Sexual difieren- 

 tiation." 



All characters considered, it is impossible to deny that jS''yctimeni', 

 though aberrant in some respects, is phylogenetieally very inti- 

 mately connected with the other genera of tlie Cynopterine group. 

 To separate this genus as a distinct subfamily, IJarpyiincB or 

 Kyctiineniufv [an proposed by llobin, 1. s. c, 18S1, on the basis of 

 some anatomical peculiarities, the more important of which have 

 been referred to above, and again, probably independently of Eobin, 



