166 



The condition of this recurrent canal, it may be men- 

 tioned, varies considerably. On the eyeless side the rela- 

 tions of the snpratemporal canal to the ossicles was essen- 

 tially the same. The sections revealed a curious absence 

 of sense organs in the anterior portion of the supra- 

 temporal canal of the ocular side (fig. 28), there being only 

 4 sense organs as against 1-3 pores, but the canal on both 

 sides of the body was not completely developed, like the 

 infraorbital. On the eyeless side in the sections there 

 was a break in the middle of the supratemporal canal 

 occupied by three naked sense organs, only two of which, 

 however, seemed to be canal organs. This break is of 

 course converted into a canal later on in the ontogeny. 

 The recurrent canal was also absent, and there were only 

 eight pores as against 13 on the ocular side, although 

 there were at least 7 sense organs as against 4, not count- 

 ing the three naked sense organs above. The relative 

 position of the sense organs was also different. In the 

 lateral line itself the only difference of importance 

 between the two sides was that the third or last otic sense 

 organ of the ocular side was here innervated from the H. 

 supratemporalis vagi, and hence belonged to the lateral 

 canal, as in the Cod. 



Infraorbital Canal {inf. c). — After leaving the last 

 supratemporal ossicle, in which this canal on the ocular 

 side anastomoses with the lateral, it at once enters the 

 pterotic, and immediately receives the hyomandibular 

 canal, which comes up from below. There is no separate 

 dermal pterotic as occasionally occurs in the Cod. The 

 canal passing straight forwards, traverses the pterotic, 

 sphenotic and right frontal. Arrived at the space between 

 the first and second tuberosities (fig. 1) it anastomoses 

 with the right supraorbital canal and turns sharply down- 

 wards almost at right angles, afterwards curving forwards 



