296 NEW SPECIES OF BAT ALLEN. 



In Carollia these quantities are not fixed. The height of the auricle is 

 variable, but the length of the muzzle is constant, i^o estimate of re- 

 latioDS of measurements between them can be undertaken. 



The tragus varies in height from 5 to 6"^'". It presents different degrees 

 of thickness along the median border. As a rule very thick, this border 

 may be thin and membranous. The processes on the outer border 

 may be two or five, those toward the apex of the tragus tending to 

 merge in one another. This tendency appears to be most marked in 

 males. 



The size of the nose-leaf is constant, being 10™"^ high and G™"" broad. 

 The lower border shows striking peculiarities in some specimens. 



Three of the males exhibited warts arranged in one or two rows 

 across the upper lip on the line occupied in Artlbeus, PhyUostoma, etc., 

 with a well-defined ridge or border. This variation is one of generic 

 rather than specific value. At least it does not indicate any disposi- 

 tion to reversion to G. castanea, since in this species no warts are seen, 

 the intervals between the margins of membrane at the side of the base 

 of the nose-leaf simply being smaller than usual, and giving to the eye 

 the appearance of extending directly across the lip. I know of no 

 genus in which this variation of the nose-leaf of Carollia brevicauda is 

 a constant character. 



In C. brevicauda, the warts in the second row on the mentum are elon- 

 gate in all the nine examples, except one in which they are rounded 

 and do not differ from those of the first row. This arrangement resem- 

 bles that seen in C. castanea. 



It is probable that the two outermost rows of warts in C. brevicauda 

 coalesce to form the elongate wart, which, as a rule, exists. 



In reviewing the measurements of C. castanea, when placed in tabular 

 form with those of C. brevicauda, it is seen that in the species first 

 named that many of the measurements are the same ; that is to say, in 

 some one of the examples of G. brevicauda the measurement of a given 

 part will be found to be the same as in C. castanea. Thus the arm is 

 of the same length in three specimens of G. brevicauda. The length of 

 the bones of the digits find their complements in G. brevicauda, except- 

 ing the metacarpals of the fourth and fifth digits, which are shorter than 

 in any example of that species. The metacarpal of the first digit is of 

 the same length in one specimen of G. brevicauda, while the first pha- 

 lanx is shorter than in any. The head is shorter while the ear is longer. 

 The tragus remains the same in the two species. The thigh and the leg 

 are both shorter in G. castanea, while the tail is absolutely longer by 



The i^roportion of the widths of the second, third, and fourth inter- 

 digital spaces is shown to be subject to variation. Specimens numbered 

 3129, 3231, 312S, 3230, 3221), and 4192 are of those in the best condition ; 

 and it is seen that the differences are less than in the remaining speci- 

 mens. But after all possible sources of error are eliminated, it will be 



