Wilder.] 324 [June 21, 



ply a matter of repetition. It is probably as impossible to homologize individ- 

 ual bones of a cetacean manus with those of an ordinary mammal, as it is to 

 homologize the immense number of caudal vertebras of the genus Mus, for in- 

 stance, with the few of a neighboring genus, Arvicola. In all such cases as 

 these, where variability is at a maximum, the importance of numerical composi- 

 tion, either as a taxonomic or as a morphical character, is obviously at a mini- 

 mum. If the Cetacea agreed with ordinary mammals in other respects, the 

 composition of the manus would afford no better grounds for these wide separa- 

 tion than the number of caudal vertebras in certain other families. 



" If we take the other extreme, of a solidungulate animal, we find such strong 

 differentiation of the osseous elements of the manus, that every single one of the 

 few bones has its own shape and size, and each of the distal segments, at least, 

 performs a perceptibly distinct function ; even a sesamoid is elevated, function- 

 ally, almost to the rank of a phalanx. Here the variability is virtually nil ; if 

 it occur at all, it would be entirely abnormal ; and the slightest normal differ- 

 ence in numerical composition, either in number of digits or of their phalanges, 

 has a generic, if not a higher, value. 



" The value of numerical composition of the pollex and primus as a mor- 

 phological character, has been estimated by different anatomists at its two possi- 

 ble extremes — some considering it an insuperable objection to the antitropic 

 homology of pollex with quintus, and others finding it little or no obstacle to 

 such a view. Two considerations have had great weight with me, in reducing 

 my estimate of its value so low, that it presents itself as no valid objection^ 

 when taken in connection with the strong evidence derived from other sources. 

 In the first place, the question can only arise in respect to five-fingered mam- 

 mals, a part, at least, of the digits of which have three phalanges each ; and 

 since here we have the maximum known number of digits, and the next to the 

 maximum known number of phalanges (Cetacea alone having more) the suscep- 

 tibility of variation in numerical composition is nearly at a maximum, accord- 

 ing to the principles already laid down, and hence the value of numerical com- 

 position is nearly at a minimum so far as the manus is concerned. 



" Secondly, it is certain that pollex and primus are telically correspondent 

 (analogous), and no less so that the modification each has undergone in its com- 

 position is simply telical. Both have been strongly differentiated from the other 

 digits in the same way, and for the same purpose. It is presumed that no anat- 

 omist questions the homology of the whole manus of a bird, a reptile and a 

 mammal ; yet the homology cannot be pushed to the individual osseous ele- 

 ments without recognition of vastly more difference in numerical composition 

 than we are called upon to admit in the present case of pollex and quintus, and 

 hence without tacit depreciation of the morphical import of mere number. The 

 manus and the pes of a bird cannot be homologized with each other, according 

 to any one of the current modes of comparison, without greater allowance still 

 for telical modification in the matter of numerical composition. For myself, if 

 I attempt to recognize any homology between the manus of a man, for example, 

 and that of certain chelonians and of a cetacean, beyond a homology of the 

 members in their aggregate, I must consider that a medius digit, for example, 

 with three phalanges, corresponds to one with several more than three, and be- 



