1871.] 323 [Wilder. 



dermal scales upon the belly of a serpent may vary widely in number in differ- 

 ent individuals of the same species. The rays of a small, sharply-outlined dor- 

 sal fin of a fish have no such variation in number as those composing a fin that 

 extends the'greater part of the length of the animal. The very numerous teeth 

 of a serpent cannot be rendered with the certitude that attaches to the dental 

 formula of a few-toothed mammal. In the lower families of birds possessing 

 more than twelve rectrices, the number is fallacious even as a specific charac- 

 ter, since it varies 'one or two pairs, at least, in different individuals of the same 

 species, whereas in birds with eight, ten, or twelve rectrices these numbers mark 

 whole familes, and the slightest variation is properly regarded as an anomaly. 

 The few digital phalanges of birds are so constant (much more constant than 

 their vertebras) that deviation from the ordinary number becomes a character 

 marking families. 



" But it is unnecessary to dwell upon this obvious point, the more so since it is 

 simply one part of the main proposition, that variation is greatest in organs 

 composed of the most similar parts— parts that are essentially either morphi- 

 cally or telically repetitive, and conversely, that the variation in numerical com- 

 position is least in the structures made up of more perfectly differentiated or 

 specialized parts. Any structure the essence of which admits of what is called 

 * vegetative repetition,' is susceptible of enlargement or curtailment by the de- 

 velopment of more or fewer segments or moieties, and variability is a necessary 

 result of such plasticity of organization. The examples adduced may be here 

 cited again in illustration Most of the caudal vertebra? of a long-tailed mam- 

 mal are precisely .similar in form and function — positive duplicates of each 

 other, and in such a mammal as the house-rat, the coccygeal formula can only 

 be given approximately, while the still more numerous dermal annuli of the 

 tail, though corresponding in a general way with the bones themselves, must be 

 enumerated simply in round numbers. The vertebras of a serpent, essentially 

 similar throughout the long series, represent no such fixed number as those of a 

 mammal where they are differentiated in several groups, each with its own char- 

 acter. And even surveying organs composed of few parts, we find striking dif- 

 ferences in variability. The presence, in an animal possessing five digits, of a 

 supernumerary one, is in frequency out of any calculable proportion to the ap- 

 pearance of two functional digits in an animal that, like the horse, has normally 

 but one — perhaps the improbability of the latter is on a par with that of the ap- 

 pearance of ten digits in a man. I am not informed as to the individual varia- 

 bility in the number of phalanges of cetaceans, and probably too few of these 

 animals have been dissected for correct estimation, but there is every reason to 

 suppose that the liability to variation here is as much greater than it is an or- 

 dinary mammal, as the increase in the number of phalanges. 



" The abrupt and marked increase in the number of phalanges of cetaceans 

 as compared with ordinary mammals, and the imperfect discrimination of pha- 

 langes, metacarpals and carpals in these mammals, seem to be explicable upon 

 the same principles that account for the great number, small size and mutual 

 resemblance of the vertebras of prehensile tailed mammals, and those that use a 

 long flexible tail as a balance. There is the same teleology in either case — it is 

 the production of perfect pliability ; and in both, the increase seems to be sim- 



