I'LASTIC CHAKACTEKS. xli 



16. Tail. 

 Varies ;is follows : — 



(1) Tailequul in length to forearm: — the oligocene ArchcFopterOjxis 

 (8 free caudal vertebnc) ; one genus of living Megachiroptera, of the 

 subfamily Macroglosshue, viz. Notopteris (" 10 vertebne,"' statement 

 taken from literature, not verified by the writer). 



(2) Tail considerably reduced, subequal to or somewhat shorter 

 than tibia, but longer than hind foot with claws : — one genus of the 

 Cynopterus section, Xycthnene (7 free caudal vertebrae). Though 

 the tail of Nyciimene is distinctly longer than in llonsettus, the 

 number of vertebrae is the same as in Itousettus amplexiauidatus. 



(3) Tail further reduced in length, from about one-half to the 

 full length of the foot with claws : — (i^owsc^^HS section:) Eidolon 

 (4 free vertebrae), liouseltus (5-7), Boneia, Bobsonia; — {Cynoptenis 

 section :) Myonycteris, Cynopterus (4), Ptenochinis, Dyacopterux, 

 Fenthetor ; — {Macroylossimv :) Eonycferis (7). 



(4) Tail rudimentary, as a rule reduced to a small knob ex- 

 ternally, more easily traceable by touch than by eye : — {Epomophorns 

 section :) PJerotes (tail absent?), Epomops (2 free vertebrae), Epo- 

 mophorus (2-3), Mkropteropus (3), Kanonycteris, Scotonycteris, 

 Casinycteris; — (Cynopterus section:) Thoopteriis; — (Macroglosshia:) 

 Megaloghssus (5 vcrtebrao, terminal two or three rudimentary), 

 Macroglossus (3 or 2), Syconycteris. 



(5) External tail absent : — (Bouseti us section :) Pteropus (no free 

 vertebrae), Acerodon, PteraJopex, SiyJoctenitim, Ilarpyionycteris ; — 

 {Epomoplioriis section:) Hypsigmtthus (no free vertebrae); {Cyno- 

 pterus section :) Megcerops, Balionycteris (no vertebrae), Chironax 

 (no vertebrae), Spluerkis ; — {Macroglossinai :) Melonycteris, Xeso- 

 nycteris. 



In the first, second, and third of the stages recorded above the 

 basal portion of the tail is included in (connected by its dorsal 

 integument with) the intcrfemoral, the tip freely projecting (whether 

 or not this was the case also in Archceopteropits is unknown) ; in the 

 fourth stage the tail rudiment is usually unconnected with the 

 intcrfemoral. 



Two facts are evident from the above, first, that the tail is con- 

 siderably reduced in length in all living Megachiroptera, with the 

 single exception of JS'otopteris, second, that the degree to which it 

 is reduced varies, as a rule even very conspicuously, within each of 

 the primary sections of the suborder. The Bousettus section falls 

 into three natural subsections, the Rousettus, Pteropus, and Dob- 

 sonia subsections ; in the first (Eidolon, Bousettus, Boneia) the tail 

 is on the third stage of reduction, in the second (Pteropus, Acerodon, 

 Pteralopex, Styloctenium) entirely absent, while in the third it is 

 either on the third stage (Dobsonia) or absent ( Harpy ionycteris). 

 In the whole of the Eponiophorus section the tail is rudimentary 

 (fourth stage), except in HypsignatJius (and Plerotes'!), which has 

 no tail. Within the Cynopterxis section, it is on the second stage 

 in Nyetimene only, on (he third in Myonycteris. Cynopterus, Pleno- 

 chirus, Dyacopterus, and Penthetor, on the fourth in Thoopleri'.!. 



