176 F. H. EDGEWORTH. 



branchialis, and posteriorly into the fifth levator and levator 

 scapulas. The musculature of the hyoid and mandibular 

 arches cannot be regarded as derivations either of a somite 

 or of a lateral plate ; it never takes on lateral-plate 

 characteristics. 



A difficulty in accepting- this theory is that it does not seem 

 to be of general applicability in Vertebrates. Thus, as stated 

 above, in the rabbit the lateral plate of the first branchial 

 arch is in front of the first trunk myotome, and those of the 

 second and third arches lie below the first trunk myotome; 

 and there is a gap between the dorsal edges of the lateral 

 plates of the second and third arches and the ventral edge of 

 the first myotome. A difference in the shape of the cell- 

 nuclei and in the rate of absorption of yolk-granules does not 

 appear to be of sufficient importance to justify a separation 

 into upper and lower portions of a structure which in other 

 animals has been called lateral plate. As far as I have been 

 able to observe, the ventral end of the lateral plate, taken in 

 its usual sense, does not degenerate into connective tissue, 

 but becomes converted into muscles. 



These difficulties lead to the following theory of the seg- 

 mentation of the head. It is probable that it originally con- 

 sisted of five segments only — the premandibular, mandibular, 

 hyoid, first branchial, second branchial — each having a njyo- 

 tonie, which, in the case of the four latter, contains a slit-like, 

 epithelium-lined cavity continuous with the cephalic ccelom 

 below. To each myotome passed a nerve — the Ilird, A"th, 

 Yllth, IXth, and Xtli.^ The gill-clefts are intersegmental. 

 New segments were added, one by one, behind the second 

 branchial, the iiead extended back into the body-region, and 



' According to Neaf ('98) only one enceplialomere corresponds to the 

 vagus — a fact which, if the enceplialomeres have segmental, or rather 

 inter-segmental, wortli, agrees with the tlieory that the vagus is not a 

 " zusammengesetzte Nervencomplex," but primarily, as regards its 

 motor brandies, of one segment only — the second branchial — and that 

 additional branches were developed as the number of gill-clefts and 

 l)ranchial sei^ments was added to. 



