88 MEMOIES OF THE UNIVERSITY OF CALIFORNIA 



HH. 5), C. helix (fig. HH, 8), and reaches its greatest development in the genus 

 in C. cavatum (fig. T, 1). In Pouchcfia this same mobility occurs, as seen in 

 P. macidata (fig. T, 2). A further development is found in Protcrijtliropsis 

 crassicaudata (fig. T, 3). This line of specialization reaches its climax in the 

 genus Enjthropsis (fig. T, 4), with its highly developed prod, provided with 

 protractile and retractile fibrillae on the prod, which has developed from the 

 mobile region of the sulcus. 



Another form of specialization is found in the development of a tentacle. 

 In the genus Protodinifer this is anteriorly located (fig. R, 2). In Pavillardia 

 (fig. JJ) it is posteriorly placed, is more mobile, and may reach a greater 

 length than in Protodinifer. In Noctilnca the greatly modified form of the 

 body, which here becomes a huge sphere, makes it difficult to give to the tentacle 

 an exact location that will link it with that in the other genera. It is closely 

 connected with the remains of a furrow and Avith the flagellum, hence it prob- 

 ably has its origin in a part corresponding to the sulcal area of other dinofla- 

 gellates. There is little evidence if any that these three types of tentacles have 

 any derivative connection other than origin in common territory. 



Another line of development that closely links the two groups may be 

 found in the melanosome of the highly specialized ocellus of Ponchetia and 

 Eri/thropsis. In the G^innodiniidae, where the lens fonnation has not yet taken 

 place, the pigment is scattered through the body, often massing at the apices or 

 near the girdle. The pigment occurring in the family Pouehetiidae has become 

 massed around the lens in the typical melanosome near the girdle. There is 

 no ocellus elsewhere in the Dinofiagellata, or indeed in the Protozoa as a whole, 

 comparable with this integrated organ composed of a laminated, oriented lens 

 eml^edded in a melanosome enclosing a highly pigmented, presumably sensory 

 core. The stigma of fi'esh-water dinoflagellates is a single mass, ellipsoidal, or 

 broadly horseshoe-shaped, lying in or near the sulcus, but not consistently at 

 its left. It does not appear to be similar in structure to or homologous with 

 the ocellus of the Pouchetiidae, but the two are rather independent specializa- 

 tions of pigment-producing plasma. 



The production of pigments is widespread among the more highly differ- 

 entiated species of the related genera Cochlodinium, Gyrodinium, and even in 

 the more generalized species of Gymnodinium. Melanin, however, is produced 

 in these lower genera only in two species, Cochlodinium atromaculatum (pi. 7, 

 fig. 71) and Gyrodinium spumantia (pi. 7, fig. 72). The occurrence of such 

 pigments and of refractive spherviles within the cytoplasm of these lower genei'a 

 has doubtless been the starting point from which the evolution of the ocellate 

 species of this family has proceeded. Various stages of the integration of the 

 ocellus are foimd within the group. 



The evolution of the ocellus within the Pouchetiidae presents several ortho- 

 genetic series proceeding along different lines independently of one another but 

 parallel in the different genera, in each case starting with the diffuse, scattered, 



