M. M. METCALF ON THE EYES AND SUBNEURAL GLAND OF SALPA. 325 



gated ovoid, with the more pointed end anterior. Its long axis is nearly 

 horizontal, and is directed forward and about twenty degrees to the right 

 or left. It lies wholly in front of the ganglion ; its posterior end abutting 

 on the antero-dorsal surface of the ganglion. Fig. 7, Plate LIII, shows a 

 vertical section of the eye and ganglion, nearly in the direction of the 

 long axis of the eye. Fig. 8, Plate LIII, represents a cross-section of 

 the basal portion of the eye. The relative position of the histological 

 elements is much the same as in Salpa cylindrica. In the posterior 

 portion of the eye the rod cells are dorsal and the pigment cells ventral. 

 In the anterior portion this arrangement is reversed, the pigment being 

 dorsal and the rod cells ventral. The innervation is as in Salpa cylin- 

 drica. The optic nerve arises from the antero-dorsal part of the gang- 

 lion, running over the dorsal face of the posterior part of the eye, inner- 

 vating the rod cells of this region, and then bores through the eye to 

 innervate the thin-walled ends of the ventrally directed rod cells of the 

 anterior portion. The structure of the rod cells is the same as in Cyclo- 

 salpa pinnata or Salpa cylindrica. The pigment cells are fewer in number 

 than in either of the latter species and are not massed together. They 

 are branched, having an appearance as if amosboid. They are large, 

 with large nuclei and finely granular protoplasm in which is found a 

 great quantity of pigment granules of different sizes. The intermediate 

 layer of the retina is present and the cells are remarkably distinct. Fig. 

 8, Plate LIII, shows their appearance. They are small cells with small 

 nuclei. Their protoplasm is clear and almost homogeneous, like that 

 of the nucleated ends of the rod cells. The most interesting feature is 

 that they each send up one (or more?) processes toward the rod cells. 

 Although I have been unable to make successful macerations and can- 

 not speak with absolute certainty, still I am convinced, after careful 

 study of serial sections, that each of these processes connects with the 

 protoplasmic core of the thick-walled end of one of the rod cells. Fig. 8 

 shows that the delicate membrane that surrounds the eye does not 

 include the pigment cells, but intervenes between them and the inter- 

 mediate cells. The pigment cells are attached, some to this membrane, 

 and some to the basement membrane of the ectoderm, while others are 

 attached to both. A similar, though less distinct membrane is found 

 between the intermediate cells and the pigment cells in certain other 

 species : e. g., Salpa cylindrica, solitary form ; Salpa hexagona, solitary 

 and chain forms, and Salpa cordiformis-zonaria, solitary form. Also in 

 several species, where the intermediate cells are not present, the rod cells 

 are separated from the pigment by a similar membrane. 



