368 JOHNS HOPKINS UNIVERSITY MORPHOLOGICAL MONOGRAPHS. 



between the two posterior limbs of the eye (e') to innervate the ventrally 

 placed thin-walled ends of the rod cells of the second, the anterior por- 

 tion of the eye (e"). The rod cells, then, in all regions of the eye of the 

 chain Cyclosalpa pinnata are innervated at their thin -walled ends, the 

 ends distal from the pigment layer. Such a fundamental difference in 

 the innervation of the different regions of the eye as Goppert describes 

 would be utterly inexplicable. It is but fair to say that in the adult eye, 

 which alone Goppert studied, the fibers of the optic nerve are more 

 closely pressed to the rod cells than in the immature individuals. It is 

 therefore a matter of great difficulty to determine the innervation from 

 the study of serial sections of the adult. The minute fibers innervating 

 the rod cells of the anterior portion of the eye cannot be traced in sec- 

 tion. The optic nerve appears to stop at the junction of the posterior 

 limbs of the eye with the anterior curved portion. That this appearance 

 is really deceptive is shown by my sections of the immature individuals. 

 But even in the absence of the actual proof I have brought forward we 

 would be justified in believing that the innervation in all regions is 

 uniform, unless there were much more definite observations to the 

 contrary than any Dr. Goppert has made. 



In the immature chain Cyclosalpa pinnata it is plainly seen that the 

 rod cells in all regions are innervated from their thin-walled ends. In 

 the adult chain Cyclosalpa pinnata the innervation of the anterior region 

 of the eye cannot be made out. In other chain-salpae of the pinnata 

 group the innervation of the anterior region of the dorsal eye can no 

 more be made out in the adult than in the adult Cyclosalpa pinnata. 

 The innervation of the posterior region is plainly seen to be the same as 

 in the latter species. The anterior portion of the dorsal eye is clearly 

 homologous in all these species. In consideration of these facts we must 

 believe that the whole innervation in all the species is as in Cyclosalpa 

 pinnata. In the chain Salpa democratica-mucronata, which I have shown 

 belongs to the pinnata group since the dorsal eye is here reversed, Gop- 

 pert himself says the rod cells of all regions are innervated from their 

 thin-walled ends. In the more primitive group, including Salpa scuti- 

 gera-confederata and Salpa bicaudata, my figures show as clearly as can 

 be shown from sections of the adult eye that the manner of innervation 

 is the same. The innervation of the anterior horns of the horseshoe- 

 shaped eye of the solitary form is also difficult to make out, but since 

 the rod cells of the posterior curved portion are plainly seen to receive 

 the fibers into their thin : walled ends, there is no room for doubt that the 



