6i6 



COMPARATIVE ANATOMY 



prolongation of the premandibular cavity. Later, Julia Piatt found these 

 segments in squalus embryos, and observed that they develop anterior 

 to the premandibular cavities and independent of them. Finding 

 rudimentary muscle fibers in the walls of these cavities, she concludes 

 that they are serially homologous with mesodermal somites, and named 

 them anterior cavities (somites). If her interpretation of the anterior 

 cavities as somites is correct, this adds another segment to the series 

 discovered by van Wijhe. 



The existence of the anterior somites has been repeatedly confirmed 

 by subsequent investigators, but opinion is divided as to their significance. 

 Van Wijhe, Dohrn, and Goodrich doubt their segmental value and assume 



SOMITE 3 



SOMITE 2 

 1ST GILL-POUCH 



SOMITE 4 



HYPOMERE 

 SOMITE I 



a CROSS SECTION X. 

 C. CROSS SECTION Y. 



SOMITE 2 



A. PARASAGITTAL SECTION. 



1ST GILL-POUCH- 



CYCLOSTOME EMBRYO. 



Fig. 509. — Diagrams of sections of an eight-day Petromyzon (cyclostome) embryo, 

 showing the segmentation of the mesoderm in the head region. The "anterior" somite 

 of Miss Piatt, which is present in elasmobranch embryos, is wanting in cyclostomes. 



that they are derivatives of the premandibular somites, Goodrich urges 

 that in any scheme of cephalic metamerism the anterior cavities should be 

 ignored, since "they soon disappear and form no permanent structure, 

 are scarcely or not at all developed in other elasmobranchs, have not 

 been found at all in other gnathostomes, and are absent in petromyzon." 

 None of these objections, however, appears to invalidate Miss Piatt's 

 interpretation. In making them, Goodrich seems to have overstrained 

 his point. If, for example, every embryonic feature should be ignored if 

 it "soon disappears and forms no permanent structure," we should have to 

 ignore also van Wijhe's fifth and sixth somites which, like the anterior 

 somite, break up into mesenchyma. If we are to ignore the anterior 

 somites because they are " scarcely or not at all developed in other elasmo- 

 branchs," to be consistent, we should have to ignore also the tubular 



