28 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA 



to the posterior junction of girdle and sulcus (figs. LL-PP). A few exceptions 

 to this may ])e found. In Ne))iatodininm armatiim (figs. NN, 1, 2) it is located 

 sliglitly posterior to the distal junction of the girdle and sulcus, as also in 

 N. partitum (fig. NN, 4). In Pouchetia purpurescens (fig. 00, 7) the ocellus 

 is located on the right side of the intercingular sulcus instead of the left when 

 viewed from the ventral face. This species and Protopsis neapolitana (fig. 

 LL, 2) are the only ones in whi<'h this organelle is thus located in the morpho- 

 logical right side of the body, insiiead of the left as in the other species. Owing 

 to the torsion of the body and the varying positions from which it may be 

 viewed, as well as its great transparency, this relation of ocellus and sulcus is 

 often hard to determine. It is easily the most conspicuous part of the body 

 and tends to obscure other structures lying near it. When the same organism 

 is viewed at several angles, however, it is quite evident that the ocellus is 

 situated on the morphologically left side of the body. 



The orientation of the ocellus is also fairly constant throughout the group, 

 with the lens anteriorly placed with respect to the melanosome. A few ex- 

 ceptions to this are found. In Protopsis neapolitana (fig. LL, 2) the mela- 

 nosome is anterior, with the axis of the ocellus as a w^hole lying at an angle of 

 about 45°with the main axis of the Iwdy. This is the only instance of this 

 apparent anteroposterior reversal of the usual relations of the lens and mela- 

 nosome to be found in the group. 



A position in which the axis of the ocellus forms a right angle with the main 

 axis of the body is more frequently met, as in the species of Ncmatodinium 

 (fig. NN), Pouchetia purpurescens, and P. maxima (figs. 00, 7, 2). Nearly 

 every gradation from this horizontal position of the axis to a vertical one is to 

 be found among the species of Pouchetia (figs. 00, PP). 



In the genus Erijthropsis fewer variations are found. The ocellus is in all 

 cases premedian and protuberant. One species only, E. hispida (fig. SS, 2), 

 lacks the anteroposterior orientation of the ocellus to any marked degree. Here 

 the two structures lie side by side, Avith the axis passing through both parts, 

 thus taking a horizontal position. In the remaining species the ocellus is 

 directed anteriorly. In E. extruders (fig. SS, 11) the amoeboid melanosome 

 at times nearly covers the lens, thus ol)scuring these relations. In the con- 

 tracted state, however, the axis showed only a small change from a vertical 

 position (fig. TT, 3). 



Of the finer structure and constitntion of the ocellus very little indeed is 

 known. Paure-Premiet (1911) found that the lens was not acted upon by 

 solvents for oil, such as alcohol and acetone. In alcoholic solution of iodine 

 it became yellow. In our own material we foimd that it is the last part of the 

 body to disappear in cytolysis, twenty minutes being required in one case for 

 its dissolution in sea water. 



The melanosome, at least the part outside of the core, is evidently similar 

 in its constitution to the pigment that is frequently met with in other parts of 

 the body. This is most clearly seen in the diffuse type of melanosome where 



