Evolution of Animals 463 



over Cope says (152: S'2): "The feet are first simple at the 

 extremity, but soon become bifurcate, or with two digits. 

 This was first observed by Rusconi in 1821 in the European 

 Hemisalamandra cristata. Professor Baird first observed this 

 in the Amhlystoma punctatum in larvae of 1.4 mm. in length. 

 The same has since been observed by myself in 1869, and 

 later by Gotte in his essay on the development and regenera- 

 tion of the foot-skeleton in the triton; also by Strasser and 

 by Hoy. In Amhlystoma punctatum the number of digits 

 succeed each other as follows: First stage, 2 before, hind limb 

 none; second, 3 before, hind limb none; third, 3 before, hind 

 limb 2; fourth, 3 before, hind limb 3. Sometimes the anterior 

 digits are completed in number before the posterior foot ap- 

 pears, and we have combinations of numbers from 4-0 to 

 4-3, and 4-4 to the full number, 4-5, which is found in all 

 specimens of 2.5 mm. and upwards." 



Now in connection with the above numerical advance it 

 may be well to inquire whether, in Sirenidse and other groups 

 of Urodela, a corresponding advance is not shown in the car- 

 tilaginous and bony internal parts. Here again we are met 

 by the degeneracy theory, which must be balanced with all 

 the facts of the case. 



In Proteus, which has feeble fore and hind limbs, the three 

 digits of the former are united to the radius and ulna by three 

 carpals, of which two are in line with the ulna. In the later 

 formed hind limb there are two digits similarly placed in line 

 with the tibia and fibula. In Siren (152: Plate XLVI) there 

 are six carpal cartilage bones joining the limbs. In Amphiuma 

 the number of carpals and tarsals varies from three, placed as 

 in the last, to 4, 5, 6, or 7. In higher urodeles the number may 

 be nine to ten. Does such then represent a reducing series 

 along a degeneration line, or an increasing series along a pro- 

 gressive line.f* Later additional facts, drawn from other sys- 

 tems, may enable us to decide. 



Having accepted it, as strongly confirmed by the evidence 

 in last chapter, that the notochord represents the now modified 

 proboscis sheath of nemerteans, we find it running almost 

 to the anterior end of the head in Amphioxus, thereby strongly 

 suggesting that this genus is derived from one of the nemer- 

 teans in which the proboscis and its sheath opened at the front 

 of the head, whereas in land nemerteans generally it opens 

 within the oral cavity as in Plate 15b-d. This is seen also in 



