PLASTIC CHARACTERS. XXXIU 



Supermimerary teeth tliat arc certainly, or at least probably, merely accidental 

 outgruwthi : — 



"13." — In one Pteropus vampynis skull (55.12.26.90) a tooth is present, on 



both sides, behind i^ and leaning against the inner side of the canine. 



The possibility of this tooth being really an i^ (lost in all Mega- 



chiroptera, but preserved in some Microchiroptera) cannot of course 



be denied, but from the position and general shape of the tooth it 



appears more likely that it is an accidental outgrowth. 



i' reduplicated on both sides. — In one Maoroglossus mininuis sohrinus (3.2.6.17). 



i- reduplicated on one side. — In one Cynoptenishrachyotisangnlatus (6.11.6.42). 



" P2," /. e. a tooth occupying the position of a ]i^. — Present on one side, in three 



eases, viz. one Pteropus scapidatus (86.11.1.1), one Epomophorus gamhi- 



aniis (Berlin Museum, 10171), and one Eonycteris spelcsa (1.3.9.1), 



Present on both sides, in one EpomopJiorus yamhianus (99.6.15.3). 



It would be interesting if this supernumerary premolar really repre- 

 sented P2, a tooth lost in all Chiroptera, and, as in the case of 

 "ia" (abovf), the possibility cannot be altogether denied. But it 

 should be remembered that Pteropus scajndatus and Eonycteris are 

 forms with somewhat degenerated teeth, therefore liable to accidental 

 anomalies in the dentition, and the teeth of Epomophorus gamhianus^, 

 though not exactly .degenerated, are remarkably small for the size of 

 the bat ; further, that the diastema between p, and p^ is in all three 

 forms considerably wider than elsewhere in the lower jaw, so that, if 

 there is any latent tendency in the jaw to a development of a super- 

 numerary tooth, this wide diastema so to speak " invites" it to crop u]) 

 there ; and finally, that in two of the species, viz. Pt. fcopiilatus and 

 E. gainliiairiis, cases are known (see below) of the appearance of a 

 supernumerary premolar between p^ and p^ (compare a Eousettus 

 angolensis with an abnormal tooth squeezed in between m' and m^). 

 Tliis, to say the least, is a warning against too rash homologizations. 

 Supernumerary tooth between p3 and p^. — On one side (tooth small) in one 

 Epamophorus gamhianus (99.6.15.2). 



On both sides (tooth well-developed), in one Pteropus scapulatus (8.8.8.3). 

 Supernumerary tooth between m' and m'. — On one side (tooth extremely 



narrow, abnormal in shape), in one Rousettiis angolensis (6.12.4.5). 

 " inj." — On one side, in four cases, viz. one Pteropus gigantens (]06.d), one 

 Mctcror/lossus laqochilus nouns (10.3.3.3), two Macroglossiis I. lagochilus 

 (10.3i 23 ; L'.S. National Museum, 125316). 



On both sides, in two cases, viz. one Macroglossus mininuiS minimus 

 (.10.4.7.4), one Macroglossus lagochilus nanus (10.3.2.2). 



In no other geuus are anomalies in the cheek-tooth series so frequent 

 as in Macroglossus (see p. 754). 



13. Palate-ridges. 



Eight ridges forming regular curves from side to side, the inter- 

 dental ridges almost equidistant, the postdental a little more 

 narrowly spaced, and some of the latter slightly notched at the 

 middle, as if tending to split into a right and left half, such is the 

 simplest, and therefore presumably the most primitive, type of 

 palate-ridges known in Megachiroptera (see fig. 28, p. 484, the 

 Epomophorine geuus Pleroten). A similar, more or less slightly, 

 but never profoundly, modified arrangement of the ridges is seen in 

 the large majority of genera of the Rousettine, Cyuopterine. and 

 Macroglossine sections ; the number of the ridges may be a little 

 increased, and, if so, the increase generally takes place chiefly on 

 tlic postdental portion of the palate, the spacing of the ridges may 



