JSS9.] ^ [Allen. 



With the exception of the foot, which is constantly 11 mm. long, all the 

 measurements are subject to variation — indeed, no two specimens in all 

 respects are alike. This statement is made while making due allowance 

 for the difficulty in taking some of the measurements, as for example those 

 of the thigh and of the membranous expansions. Specimens which had 

 been macerated in weak alcohol were rejected. But among those which 

 were accepted it was not always possible to determine (owing to the con- 

 traction of the tissues), the exact extent to which the parts should be ex- 

 tended, so as to represent as far as possible the position of the wings in 

 flight. One of the most interesting measurements is that of the width of 

 the third digital interspace. This space, so small in Pteropidae, Molossi, 

 and in Noctilio, is wide in Phyllostomidse, excepting Phyllostoma. An- 

 other interesting feature is the extent of the incision on the free margin of 

 the interfemoral membrane. In well-preserved specimens of G. brevicaudu 

 the incision is conspicuous, while in the type of G. caslanea, which is also in 

 good condition, the incision is absent. Yet in slightly macerated speci- 

 mens of 0. brevicauda the incision disappears, showing that it is a charac- 

 ter which is dependent upon tonicity and not on any distinctive structural 

 peculiarities, and cannnot, therefore, have much value. One of the marked 

 ranges of measurements is seen in the length of the tail. The shortest tail 

 is 5 mm. long and the longest 7 mm. The tip of the tail answered in three 

 specimens to the middle of the femur, in four to the junction of the middle 

 with the lower third, and in two lack one-fifth only in being as long as 

 the femur. In none, therefore, was the tail as long as in the single exam- 

 ple of G. castanea. 



The length of the thigh varies from 12 mm. to 15 mm. Hence the rela 

 tive lengths of these quantities will be also variable, especially so since even 

 in the same individual the length of the-tail does not tautogenize* with 

 the length of the femur. The length of the tibia — a character of value in 

 Cheiroptera — varies from 14 mm. to 18 mm. 



The length of the forearm, perhaps the most important single measure- 

 ment which can be taken, varies from 35 mm. to 40 mm. 



The following includes the variations of the manus and their range : 



mm. mm. mm. 



First metacarpal , from 4 to 6 Range 2 



Second " " 24 " 35 " 9 



Third " " 33 " 37 " 4 



Fourth " " 32 " 36 " 4 



Fifth " " 35 " 37Jr " 2£ 



First phalanx first digit 4 " G " 2 



fFirst " second" " 3 " G " 3 



First " third " " 13 " 17 " 4 



* Tautogeneily — a word introduced by Prof. Rolleston as a more correct term in this 

 connection than corelatiou. 



fAn apparent anomaly exists on the left side of specimen No. 3993. There are two 

 phalanges to the second digit. 



