DEVELOPMENT OF A CHIM.EKOID. 285 



cally. Thus in comparison with other piscine groups, they are 

 allied with one another much as earliest ganoids with the teleosts. 

 Taxonomically we have therefore to revert to Bonaparte's early 

 arrangement (of about 1 840) and regard elasmobranchs as a sub- 

 class, and selacha and holocephala as natural orders (or super- 

 orders). 



II. That in some regards, in comparison with selachians, the 

 chimseroid has retained the more primitive developmental fea- 

 tures, e. g. the total segmentation of the egg, and the less 

 modified early gastrula. That in other respects it has acquired 

 more highly specialized characters, e. g., restriction of the blas- 

 toderm to a smaller region of the egg, appropriation of yolk 

 via the external gills (and gut), extraordinary egg capsule 

 and its adaptation to the embryo. (The foregoing conditions, 

 wherein high specialization is found associated with archaic 

 developmental processes, are essentially in keeping with our 

 knowledge of the history of the chimaeroid group as derived 

 from anatomy and palaeontology. Descended from earliest 

 sharks, this group may well have retained some of their peculiar 

 developmental characters, c. g., in earlier stages; on the other 

 hand new and modified processes of growth doubtless arose in 

 connection with advances which were taking place in the special 

 direction of chimseroid structures.) 



III. Of more general significance, I believe, are : 

 (A} The early processes of gastrulation in Chimtzra, which 

 elucidates the corresponding developmental stage in sharks. In 

 these forms, long studied among vertebrates, gastrulation has 

 been subject to widely different interpretations - - indeed in the 

 latest time so careful an observer as Samassa has even denied the 

 presence in sharks of any process of gastrulation scnsu stricto. 

 Chimcsra, it now appears, indicates that the blastopore of sharks 

 is a secondary structure. 



(B) The accessory mode of nutrition of the late embryo. In 

 sharks the yolk is appropriated by means of a constant extension 

 of the blood-producing area and a progressive differentiation of 

 the vitelline circulation. In Ckim&ra, on the other hand, this 

 mode of nutrition of the embryo is less extensively established, 

 for it has been supplemented by the fragmentation of the yolk 



