AQUATIC MAMMALS 



the other three. It can accordingly be deduced that if the flipper of an 

 aquatic mammal be employed as an oar the digit upon its anterior border 

 will likely be more robust than the others, while if it be used purely in 

 equilibration there will probably be no digit markedly heavier than the 

 others. 



As in the sea-lion and sirenians the subcutaneous tissue between the 

 cetacean digits has been built up so that the flipper surfaces are plane. 

 This interdigital tissue is very tough, and at least in adults of some 

 sorts (Neomeris) is strengthened by a network of fibers which have a 

 tendency to converge distad from the bases of the digits, suggesting the 

 manner in which the heads of the femur and humerus are ordinarily 

 strengthened osteologically. 



As there is such variation in the width of the cetacean manus it is 

 only to be expected that there would occur a corresponding stimulus for 

 digital differentiation, and there might conceivably be a tendency for 

 digital reduction in one group of whales and the opposite tendency in 

 another. Consequently, in spite of the fact that there are but four 

 digits in the rorquals, it was not inconsistent with the possibilities for 

 Kiikenthal to claim that at least the broad-handed beluga (Delphinap- 

 terus) exhibits a trend toward the polydactylous condition of the ichthyo- 

 saurs (in which there occur as many as nine digits) and fish, because 

 he discovered in the white whale that the two phalangeal elements (but 

 not the metacarpal) of the pollex were double, one pair being situated 

 beside the other. This is a possibility that can not be lightly discarded, 

 but as the only beluga in the National Museum with articulated manus 

 has a pollex entirely normal, and no one else besides Kiikenthal has 

 ever encountered such a doubling in this or any other cetacean, 1 must 

 presume that his material was pathologic. 



The same investigator (1893) argued with his customary vigor that 

 it is not the pollex which is absent in the rorquals but the third digit. 

 Attempting to carry this argument still further he made the remarkable 

 statement that in Balaena mysttcetus the metarcarpus of digit 1 is not 

 really a part of the pollex but represents a prepollex, in spite of the fact 

 that in Eubalaena the same detail is provided with two phalanges. His 

 reason for this course of action was that he had "occasionally" discovered 

 in Sibbaldus musculus a structure which he interpreted as constituting 

 atrophied phalangeal segments lying between digits three and four, and 

 entirely disassociated from the carpus. Such a situation would be quite 

 astonishing. In the first place it would be extremely unlikely indeed that 

 the more distal part of a digit could persist as a remnant in the connective 



[258] 



