56 COMPARATIVE ANATOMY. 



for the head a inetameric origin of the same nature as that of 

 the body. This, however, is not admissible, inasmuch as the above- 

 described segmentation of the visceral section of the skull by 

 the formation of gill-slits does not correspond to a segmentation 

 of the same nature as that seen in the body, but arises quite 

 independently. To express it briefly Metamerism does not 

 correspond to branchiomerism. It follows that a direct 

 parallelism of the branchial nerves to the intercostal nerves which 

 correspond with trunk-metameres does not exist, and the attempt 

 to solve the problem of the Vertebrate skull by indirect methods, 

 i.e. those of Comparative Anatomy, must lead to crude theories and 

 false conclusions. 



The result of the above considerations may be shortly expressed 

 as follows : 



1. The Vertebrate skull is not a structure sui generis, but has 

 been derived by a metamorphosis of the most anterior section of 

 the skeleton of the body. 



2. The proof of this lies in the common origin of both the 

 cranial and vertebral skeleton out of the protovertebra? (somites, 

 metameres). 



3. The skull is divided into two main sections, a dorsal and a 

 ventral. The former encloses the brain, and is spoken of as the 

 cranium, while the latter lies in the region of the fore-part of the 

 alimentary tube, has primitively to do with branchial respiration, 

 and is called the visceral skeleton. 



4. The cranial section alone is to be looked upon as made up of 

 a series of mesoblastic somites : the segmentation of the visceral 

 skeleton must be regarded as a secondary acquisition, for the gill- 

 arches are developed as secondary supports for the hypoblastic 

 gill-clefts. 



5. The attempt to explain the adult skull as a series ot 

 vertebrae 1 fails completely; it is a question of protovertebra3 

 (somites) only, and thus is one that can only be solved along the 

 lines of Embryology, and not those of Comparative Anatomy. 



6. The number of mesoblastic somites concerned in the 

 formation of the skull may be fixed at nine, 2 according to researches 

 up to the present time on Cyclostomes, Elasmobranchs, and Am- 

 phibians. In no case are there fewer, in many instances possibly 

 more. 



1 Rosenberg has, however, shown that in Carcharias glaucus, but apparently 

 not in other Selachians, the portion of the cranium lying between the exit of the 

 vagus and the vertebral column is clearly composed of three vertebrae, which gra- 

 dually fuse with, and constitute a part of, the occipital region of the skull (Gadow 

 finds four vertebrae in embryos of Carcharias which thus become modified). It follows 

 that the cartilaginous cranium is not completely homologous throughout the Verte- 

 brata ; the skull of Carcharias con-responds with that of Scyllium, for instance, plus 

 certain of the anterior vertebrae. 



Sagemehl has found a somewhat similar modification in Ganoids. 



* Beard, in a recent paper, increases the number of segments in the head in 

 Sharks to eleven. 



