292 University of California Puhlications in Zoology. [Vol. 8 



The surface is marked by faint salient lines and a delicate tracery 

 which rarely attains the condition of a meshwork. Several prominent 

 pores are found along the girdle and several on the apical plates. No 

 ventral pore could be with certainty distinguished. A delicate fin 

 guards the left margin of the ventral area and two tapering antapical 

 spines rise nearly 0.5 transdiameter apart from the antapical plate. They 

 are 0.5 to 0.75 girdle width in length. The left one is sometimes double 

 (fig. 5). Two or three accessory antapicals arise from the margins of 

 the ventral area. Fins are scarcely developed upon any of the spines. 



The individuals both isolated and in chain exhibit very clearly the 

 attachment pores through which the protoplasmic continuity of sister 

 cells in chain is maintained. The anterior pore {ant. att. po., pi. 18, 

 fig. 2) lies just dorsal to the apex in the upper edge of apical 3', not in 

 the apex itself. The posterior pore (post. att. po., pi. 18, fig. 1) is on the 

 right side of the posterior plate of the ventral area. The apex and an- 

 terior attachment pore of Ceratium (see Kofoid, 1909) coincide. This 

 fact raises the question as to the homology of the apex of Ceratium and 

 that of Gonyaulax. 



The plasma in the two individuals in chain (pi. 18, fig. 3) contained 

 constricted nuclei (pi. 18, fig. 3); and all in the collection were crowded 

 with globular, ring-shaped and occasional comma-shaped starch granules. 



Chain formation was exceedingly common and chains of even ten 

 individuals were noted. 



Dimensions : Length, 30/x ; transdiameter, 33/x ; dorso-ventral 

 diameter, 16-22/^,; width of girdle, 3-5/i.; length of antapical 

 spine, 2-Afx. 



Synonomy : Cleve (1901a) is probably correct in regarding 

 Glenodiniiim cinctum of Pouchet (1883) as Levander's (1894) 

 Peridinium catenatnm. Mennier (1910) includes this species in 

 his new genus Amylax, a genus based on the presence of starch 

 grains. On the basis of skeletal morphology this species is, how- 

 ever, a Gonyaula.r, and I therefore reject his assignment. He 

 has not analyzed the skeletal morphology of any species of his 

 newly created genus. 



Comparisons : Resembles G. series Kofoid in the fact that it 

 forms chains but not in the method of their formation. In the 

 presence of numerous antapical spines, wide ventral area, and 

 especially wide posterior plate, this species is nearest Gonyaulax 

 triacantha. It also approaches this species in having four an- 

 terior intercalaries in the right dorsal region. G. triacantha has 

 two, while none, or but one, in the right ventral region is the 

 usual arrangement. This species is perhaps worthy of subgeneric 

 distinction. 



