THE SEGMENTATION OF THE HEAD. 203 



Balfour introduced another element, the mesodermal somites, 

 into the discussion ; and his method, developed by Marshall, 1 and 

 still farther by van Wijhe, is that which has given results most 

 often quoted in connection with this subject. Van Wijhe con- 

 sidered mesodermal somites, gill clefts, and nerves, and tried to 

 utilize the purely motor nerves (III, IV, VI,) as ventral roots 

 of the preauditory nerves. He recognized nine mesodermal 

 segments, and the relations of these to the segmental nerves is 

 given here in tabular form. 



DORSAL NERVE ROOT. VENTRAL NERVE ROOT. 



Ophthalmicus profundus, Oculomotor. 



Trigeminal less op. prof., Trochlearis. 



Acustico-facialis, ( Abducens. 



( None. 

 Glossopharyngeal, None. 



Not recognizable. 



Va US JHypoglossal. 



Since van Wijhe wrote, others have tried to add to his 

 structure, and some have claimed to recognize eighteen or nine- 

 teen of these head somites. It is pretty certain that there is at 

 least one somite in front of the first recognized by him (Figs. 

 121 and 122 a). Others have taken the sense organs as their 

 b.isis, including in this not only ear and nose, but sensory struc- 

 tures developed in connection with the gills, and considering the 

 f icialis as compound, have figured out eleven head segments 

 (Heard). Again, the early condition of the neural tube has 

 shown the existence of nervous segments (neuromeres, p. 49), 

 eleven in number in the head region (Orr, Maclure, Hoffmann) 

 back to and including the vagus. Locy has claimed that the 

 same number of segments can be recognized in the medullary 

 plate of elasmobranchs, amphibia, and birds before it is infolded 

 to form the neural tube, but his conclusions are in dispute. 



The questions asked at the beginning of this section cannot 

 as yet be fully and finally answered. That the head is truly 



1 Balfour recognized eight or nine somites; Marshall nine (eleven in notidanid 

 sharks). 



