PLASTIC CnAKACTERS. 



fingers. If hi^i identiiicalions M^ere correct, the indices of the third, 

 fourth, and fifth fingers would be, respectively, 1486, 1450, and 

 1046, i.e. the third and fourth lingers would l>e very short and 

 practically subequal, the fifth enormously lengthened, a condition 

 unthinkable except on the supposition that the form of the wing of 

 Archcpopteropus was fundamentally different from that of any other 

 bat, living or extinct. It is obvious that what Meschinelli takes to 

 be the fifth finger is the third, and what he considers tlie third is 

 the fifth. With this correction the indices of the metacarpals and 

 phalanges of the three long fingers (calculated from Meschinelli's 

 measurements) are as indicated in the bottom line of the following 

 table : — 



And the indices of the total lengths of all fingers are these : 



A comparative analysis of the hand gives the following 

 results : — 



The first finger diilerod in no respect from that of living 

 ilegachiroptera. 



The second finger was somewhat less reduced in length (843, as 

 compared with 608-786); as its first phalanx (inde.x 126) was 

 very nearly similar to the average of recent forms (in which the 

 index varies from 89 to 149), and the metacarpal (541), though 

 unusually long, is equalled by that of the longest-winged living 

 Fruit-bats (total variation 406-543), the greater length of the 

 whole finger was due chiefly to the rather less reduced second 

 and third phalanges (together 176, as against 64-162 in recent 

 Fruit-bats, but the latter maximum index includes the claw, 

 whereas in the fossil form only the claw iihalanx is included). 



The third metacarpal was conspicuous]}' shorter than the fourth, 

 this again somewhat shorter than the fifth. Although in some 

 living Fruit-bats the fifth distinctly tends to be the longest, so 

 great a disproportion in the lengths of the three long metacarpals 



