THE HEAD PROBLEM 625 



resemble one another neither in development nor in relations. The same 

 objection may be made against the attempt to compare one of the aortic 

 arches with one of the intercostal arteries. Balfour's failure to distinguish 

 between the mesoderm of the visceral arches and that of the somites has 

 misled many, and this error has been perpetuated by H. E. Ziegler ('15). 

 Van Wijhe's head somites develop from the epimere, while the mesoderm 

 of the visceral arches comes from the hypomere. The segmentation of 

 the somites has been shown to be independent of that of the visceral 

 arches. There is nothing in the trunk region comparable with a visceral 

 arch. Branchiomerism is confined to the head. If branchiomeric 

 segmentation were the only evidence of head metamerism, we should have 

 to conclude that the head is a region sui generis. Whether or not there 

 was, or is, a correspondence between mesomerism and branchiomerism 

 remains today an open question. Their topographic relations in amphi- 

 oxus and cyclostome embryos justify the assumption of a primitive 

 correspondence, but the tendency of visceral arches and pouches to be 

 reduced and modified, makes them less satisfactory criteria of the primi- 

 tive segmentation than the somites. 



A number of objections have been made against the use of cranial 

 nerves as criteria of head segmentation and against the attempt to com- 

 pare them with spinal nerves. The dissimilarities of the two have been 

 emphasized. It has been asserted that cranial nerves, such as V, VII, 

 IX and X, are mixed motor and sensory, while the dorsal spinal nerves 

 are purely sensory; that the cranial nerves mentioned receive cellular 

 contributions from epibranchial ganglia, while spinal nerves do not; 

 that the ganglia of cranial nerves lie lateral to the somites, while those of 

 spinal nerves are medial; that spinal nerves have a metameric distribution, 

 while cranial nerves have an extensive polymeric distribution. 



To meet these objections, it has been pointed out that the dorsal 

 nerves of amphioxus are mixed in function, and that in this respect 

 cranial nerves retain an ancestral trait which has been lost by the dorsal 

 roots of spinal nerves. As a matter of fact visceral motor fibers occur 

 in the dorsal roots of spinal nerves of fishes and amphibians. They 

 disappear from these roots in amniotes. The difference in the rela- 

 tions of dorsal nerves to the somites and to epibranchial ganglia may be 

 explained as a result of the reduction in size of cranial as compared with 

 trunk somites, and the limitation of gills, with which epibranchial ganglia 

 are associated, to the head region. The wide peripheral distribution of 

 cranial nerves is regarded as a coenogenetic adaptation. There is some 

 evidence of metamerism in cranial nerves, however, in the relations of the 

 visceral branches of cranial nerves to the series of gill-clefts. 



The relations of the somatic motor cranial nerves, the oculomotor, 

 trochlearis, and abducens, resemble those of spinal somatic motor nerves. 



