669 



able talk of an original hexa- or heptadactylism. No support to 

 this view is given by the Stegocephali ; here we never have more than 

 five digits, very often only four ; and entirely limless forms are found 

 even in the carboniferous. The fact is that after the pentadactyle 

 limb was once established, the terrestrial forms did not add any di- 

 gits (the praepoUex and praehallux are of secondary origin). Re- 

 duction however took place very often, for instance in the horse-series 

 from the pentadactyle to the monodactyle condition; and in hundreds 

 of other cases. It is of importance to observe that the limb of the 

 tailed Amphibia develops in a different way from the limb of the 

 Amniota. In the tailed Amphibia it is a regular process of budding. 

 The first two digits appear first and than in succession the third, 

 fourth and fifth. In the Amniota all digits make their appearance 

 simultaneously, and very often rudiments of one or more digits are 

 present in the embryo, which disappear later; in the birds for in- 

 stance a rudiment of the fifth toe is present in the embryo, but never 

 in the adult. Such a condition is never seen in the tailed Amphibia; 

 one digit appears after the other, and there are never more digits in 

 the larvae than in the adult. The development of the limbs of Pro- 

 teus have been studied; but there are never found more than three 

 digits in the hand, and never more than two digits in the foot in the 

 course of the evolution. The same is true for Necturus, where not 

 more than four digits are found in the earliest stages, in the hind foot ; 

 a 5th tarsal appears occasionally in the later stages. But if the chei- 

 ropterygium developed from a fish-fin with many rays, we ought to 

 find traces of these supernumerary rays in the embryos; in the same 

 way as we find traces of former digits in the embryos of such Am- 

 niota, the ancestors of which possessed more digits than at present. 

 We see therefore, that the process of development of the limbs of the 

 tailed Amphibia is different from that of the Amniota. Another im- 

 portant fact is, that in none of the Stegocephalia do we find more 

 than four phalanges in the digits. From all these considerations it 

 seems very probable that the ichthyic ancestors of the Stegocephali 

 must have had a limb with few digits and few phalanges. The first 

 segment of the limbs must have consisted of one, the second of two 

 elements. 



We know that the fins of fishes are characterized by finrays; 

 but that finrays are never found in the Amphibia. The finrays must 

 have been reduced before the batrachian limb could be developed. 

 The reduction of finrays is very well seen in the different members of 

 the living Dipnoans. 



