368 TT. B. Scott — Variations and Mutations. 



possible, but they have not been found yet and the large num- 

 ber of phyla in which the law of reduction has been demon- 

 strated is evidence that this law represents the normal method 

 of procedure in the lines of descent of the terrestrial mammals. 



In the case of the horses we may trace the gradual reduc- 

 tion of live digits to one and throughout all the many species 

 which compose this great line there is no loss or even obscuring 

 of the individuality and homology of the parts involved.- Noth- 

 ing is ever added, and what is lost disappears gradually, by suc- 

 cessive well recognized steps. The phalanges first dwindle and 

 are then suppressed, accompanied by the distal portion of the 

 metapodials. The latter become splint-bones, which in some 

 cases shorten to nodules and then vanish. Last of all the change 

 affects the carpus and tarsus. So far as the horse is concerned, 

 the only podia! element which is actually suppressed is the trape- 

 zium, but in the tarsus the ankylosis of the ento- and mesocu- 

 neiforms brings about a similar diminution of the number of 

 separate elements, though there is no actual loss of parts. 

 While the lateral digits are thus gradually dwindling away, the 

 median or third digit is being gradually enlarged in the same 

 proportion, until it alone bears the entire weight of the body. 

 This change requires a gradual readjustment of the carpal and 

 tarsal bones to the altered character of the strains transmitted 

 by them and all of these transformations may be followed 

 step by step in the successive species. In the ruminants and 

 the camels analogous changes have been produced in the same 

 slow and steady manner, with a remarkable directness of devel- 

 opment, individual variations apparently not affecting the re- 

 sult. 



If the facts of digital variation, as Bateson has collected them, 

 possessed any real phylogenetic significance, we should expect 

 to find a close correspondence between these facts and those 

 exhibited by the extinct forms, but nothing of the kind is 

 observable. Indeed the utterly different character of the two 

 classes of facts is obvious on the most superficial comparison. 

 The cases of polydactylism which are so far from infrequent 

 among the recent horses, demonstrate their radical difference 

 from the processes at work in the phyletic history of that 

 species by the fact that the carpus and tarsus in such cases are 

 unlike those of any normal form whatever. For this reason 

 Gegenbaur very properly refused to admit such cases as 

 examples of " reversion." 



Examining Bateson's conclusions as to digital variation more 

 in detail, we reach the following results. (1) The frequency 

 of digital variation differs greatly in different animals ; the 

 same is true of digital reduction. (2) In each kind of animal 

 variation tends to produce a particular type of structure, which 



