PAIRED FINS WITH THEIR NERVES, IN LEPIDOSIREN AND PROTOPTERUS. 629 



myotomes participating in its muscularisation. It retains this position relative to the 

 myotomes throughout life. 



The foregoing account applies to Lepidosiren. In Protopterus the scarcer material 

 did not allow the development to be followed in such detail. It shows, however, that 

 the ventral processes of MM 2, 3, 4, and 5 slope forward towards the fin rudiment, the 

 myotomes behind 5 having ventral processes running straight downwards or sloping 

 backwards (text figure 7). The pronephros has the same relations to the fin rudiment 

 as in Lepidosiren. The attachment of the limb is approximately in the same position 

 (at the front end of M2). 



The indication of M5 taking part in the muscularisation of the fin is in agreement 

 with the composition of the brachial plexus in this genus (p. 623). 



After giving off the myoblasts to the fin rudiment, the ventral processes 

 concerned again become compact, definitely outlined masses (but more than two- 

 layered), and take their part in the formation of the obliquus and rectus muscles, 

 similarly to the ventral processes in the hinder part of the body. It is worthy of 

 note that whereas in an earlier stage the ventral processes of Ml and M2 are very 

 widely separated, the gap between them is obliterated by forward growth of M2 

 and M3 (cf. text figure 6), so that in the adult the coraco-hyoid (MM y, z, 1) forms 

 a direct continuation of the rectus muscle (MM 2, 3, 4, etc.). The pectoral girdle is 

 situated in the myoseptum M1-M2. 



As was stated on p. 623 the nerve 1 contributes both to the cervical and the brachial 

 plexus. It is probable, however, that the nerve fibres from Nl do not enter the limb. 

 As we have seen, the ventral process of Ml seems to be absolutely cut off from par- 

 ticipation in the fin mesoderm by the presence of the pronephros, and it seems probable 

 that the fibres from Nl leave the brachial plexus by some other of the numerous 

 branches given off by it to the composite muscle inserted into the occipital rib (the 

 costal element of the neural arch in the myoseptum Mz-Ml). It is, of course, just 

 possible, though unlikely, that there might be some interchange of substance between 

 the proximal portions of the ventral processes of Ml and M2, which are in contact (text 

 figure 7). 



Furbringer emphasises the fact that from Dipnoi onwards the cervical and brachial 

 plexuses tend to become distinct, and we see that it probably is so in Lepidosiren and 

 Protopterus* (The separation is sometimes complete in Amphibia, and probably 

 always in the Amniota.) In all other fishes the plexuses overlap, some nerve or nerves 

 sending fibres to both hypobranchial and pectoral girdle muscle. The overlapping is 

 greatest in Elasmobranchs. t May we not seek the cause of this separation in a con- 

 dition of the pronephros, together with a flattening out of the ventral processes over a 



* Note, however, that I make the separation between Nl and N2 instead of between Nz and Nl, as Furbringer 

 makes it in Ceratodus and Protopterus. 



t It must, however, be noted that Furbringer considers an extensive overlapping to be secondary, as this condition 

 is found more highly developed in the less primitive Elasmobranchs. 



