Scharff — On Dohrn' s Theories on the Origin of Vertebrates. 23 



firmed by their early appearance, their primitive position between the 

 distal and proximal portions of the visceral muscles, their insertion, 

 and the appearance of secondary cartilages. The latter are homo- 

 logous with the cartilaginous rays, but on account of their appearing 

 very late they have not yet been described. 



If we compare the elasmobranch arch with that of Petromyzon, 

 the main difference is exhibited in the separation of the former 

 into four pieces. But supposing, says Dohrn, this reduction did 

 not take place — in fact, if the arch remained as a single piece — 

 and if we further supposed that the constrictor and inter-branchial 

 muscles which are parts of the distal and the adductors, inter- 



V. 



Cartilaginous arch of an Elasmobranch, showing the principal muscles. 



c. a. = cartilaginous arch. m.i. = musculi interbranchialis. 



a. m. = adductor muscle. c. m. = constrictor muscle. 



arcuales and coracobranchiales which are parts of the proximal 

 portion, remained all in connexion with one another, we should 

 have exactly what we find in Ammocoetes, the young of Petro- 

 myzon. 



No one will doubt that the separation into four parts of the 

 gill-arches is one of the later modifications, and that originally 

 there was only one cartilaginous rod. Hence we may reasonably 

 suppose that both the elasmobranchs and the cyclostomata are 

 derived from such an ancestor whose hypophysis discharged the 

 duties of an independent cleft. 



In his previous pamphlet Dohrn drew attention to the fact, 

 that in almost all animals the gills were protected by special 

 contrivances. This necessity of protecting the gills must have 

 arisen when the present habits of cyclostomata (burrowing in mud 



