352 Mr. K. Andersen on 



" moderately long but heavy muzzle, flattened and scarcely 

 concave frontal region ; sagittal crest weakly developed, 

 almost obsolete." In Pt. rodricensis the muzzle is moderately 

 long but heavy ; the sagittal crest in young adults undeveloped 

 or almost obsolete, in aged individuals well developed but 

 low ; the frontal region (between postorbital processes) in 

 young adults flattened and scarcely concave, in aged specimens 

 distinctly concave. The type of Pt. mascarinus is no doubt 

 a young or young adult specimen. 



The canines in Pt. mascai-inus are " long and sharp, both 

 above and below, with unusually acute and prominent postero- 

 internal basal ledges, those of the maxillary in particular " ; 

 the same is the case in Pt. rodricensis. y? in Pt. mascarinus 

 is " very minute, but would have been still persistent in the 

 type and standing in the tooth-row " ; 'p^ i'^ Pf- rodricensis is 

 minute and deciduous; the presence of the alveoli of this 

 tooth in the type of Pt. mascarinus is additional evidence 

 that it is a young or young adult individual. 



It must be admitted that the above characters of the skull 

 and teeth in P^ m(7scamiM.9 are rather vague ; they are, in 

 fact, common to a large number of species of Pteropus, and 

 consequently do not give any conclusive evidence of the clo.se 

 affinity of Pt. mascarinus to Pt. rodricensis, far less of its 

 possible identity with the latter species. But when we 

 turn from the description to the measurements of the skull 

 and teeth we are on much safer ground ; nearly all tlio 

 measurements of Pt. mascarimis show the most striking 

 similarity to those of Pt. rodricensis. In the table below I 

 compare the princii)al measurements of the skull and teeth of 

 Pt. mascarinus, as given by Mr. Mason, with those of three 

 sl<ulls of adult specimens of Pt. rodricensis (one of them the 

 type of the species) in the collection of the British Museum. 

 It will be seen that the measurements of the two species are 

 practically identical, with the following two exceptions, both 

 of which are, most probably, apparent only, not real : — 

 (1) 'J'he approximate basilar length of the skull of Pt. masca- 

 rinus is stated to be 44 mm., the approximate condylo-basal 

 length 46 mm.; in Pt. rodricensis the measurements are 

 respectively 46 and 51"7 mm.; the basilar lengtli is practi- 

 cally the same in the two skulls (taking into consideration that 

 the measurement of mascarinus is approximate only, and the 

 type probably a young adult), whereas the condylo-basal 

 length would seem to be widely diff'erent ; but it must be 

 said that, provided the basilar length in mascarinus is 

 approxiuiately 44 mm., the condylo-basal length cannot 



