1905.] OF THE GENUS RHINOLOPHUS. 107 



essential points similar to Rh. affinis. It agrees with the now 

 existing affinis in the pandurate sella and the prolongation of 

 III.-. But it is considerabljj highei'-developed, chiefly in the 

 following respects : (1) the dentition ; (2) the wing-stvuctnre ; 

 (3) the length of the tail ; (4) the beginning, or complete, reduction 

 of the lateral mental grooves ; (5) the general size. 



The peculiar prolongation of the second phalanx of the third 

 finger, described above under A%. afftnis, is preserved in Rli.ferrum- 

 equmum : III.^ is more than (or, extremely rarely, at least equal 

 to) 1| the length of III.\ Also IV." is- lengthened, i. e. more than 

 1| of lY.^ ; it is an interesting fact that, in this particular point, 

 Rh.ferrimi-eqimium (all races) agrees with Rh. affinis himalayanus, 

 but not toith any of the other races of affinis. Besides these 

 two characters, which are simply inherited from an afflnis-\\kB 

 ancestor, there is an important modification in another part of 

 the wing, to which we have no parallel in any of the foregoing 

 forms*, viz. a change in the lyroportionale length of the third, 

 fourth, and fifth metacarpals, as shown in the subjoined table : — 



Forearm. 

 All the foregoing species 



(94 examples) 1000 



IZIi.ferrHm-equimim 



(all races ; 121 examples) 1000 



This table shows: — (1) In all the foregoing 21 fonias of this 

 group the fourth metacarpal is but very little longer than the 

 third (24 mm., for a supposed length of forearm of 1000 mm.), 

 and the fifth metacarpal is practically of the same length as the 

 fourth t. (2) In ferrum-equiniim a considerable shortening of the 

 third metacarpal has taken place ; at the same time a much smaller 

 reduction of the fourth metacarpal has occurred, so as to make 

 the fifth metacai-pal, slightly but decidedly, the longest of all. 



The tail is proportionately longer than in the foregoing species, 

 being, on an average, in the eastern races of ferrum-equinmn 

 {nippon, tragalus, regulus) exactly l^, in the typical form \h, 

 the length of the lower leg, whereas proximus, in this point (as 

 well as geographically), is intermediate between the eastern and 

 western races J. 



In all the foregoing forms, without exception, there are three 



* But thei-e is an exact parallel in an Ethiopian species, of the affinis t^'pe, viz. 

 nil. darliiiffi (see the "General Remarks," below, p. 118). 



t It would only have made the table more complicated if I had gi\'en separate 

 ciphers for all the foregoing species. The onlj- diiference (and an exceedingly small 

 one) is that in simplex, viegaphyllus, truncatus, namis, celehensis, horneensis, virgo, 

 and malayanus the fourth metacai'pal is, almost alwaj's, a mere trifle longer than the 

 fifth; in nereis, stheiio, roitxi, tJtomasi, and affinis a mere trifle sliorter than the 

 fifth. However small this difference is, it is evidently the first faint trace of the 

 modification definitely carried out \\\ ferrum-equhmm -. the fourth metacarpal always 

 shorter than the fifth. . i 



X It is hardly necessary to say that a short tail cannot be a primitive character in 

 the order Chiroptera, taken as a whole. But, for some reason or other, we find in the 

 most primitive species of the genus BJiinolophns a very short tail ; in the higher 

 forms of the 2}i'ese7it groiip we see, again, a lengthening of the tail. 



