KOFOID AND RIGDEN: SCHIZOGONY IN GONYAULAX. 339 



the distal end of the girdle at the point where the girdle abuts against 

 the ventral area. It is in a precisely homologous position that the 

 connection is made in Gonyaulax catenata (Plate 2, fig. 5). Upon de- 

 tachment a small scar or attachment area (att. area) is visible on G. 

 catenata in this region. In G. series on the other hand the adjacent 

 schizonts are attached by very broad polar regions of cell plasma which 

 preclude the formation of nearly the whole apical series and the 

 anterior intercalary (if any exists) plates on the epitheca and much 

 of the antapical plate and the posterior and possibly several interme- 

 diate plates of the ventral area. The lower part of the posterior inter- 

 calary is also suppressed. We find no evidence that these plates are 

 forming inside of the line of junction of adjacent thecae. The theca 

 wherever formed is wholly exposed. The region in G. series homolo- 

 gous to the attachment pore of G. catenata is apparently fully exposed, 

 especially in anterior members of the chain and no evidence of a 

 specialized "attachment area" appears elsewhere. It is possible 

 that the attachment region in G. series will be formed in a more pos- 

 terior location or that these areas are not homologous in different 

 species of the same genus, but it seems from the evidence in hand 

 more probable that in general these regions are homologous but that 

 in G. series, in the absence of skeletal fission, the de novo origin of the 

 theca while the schizonts are still in chain but subsequent to nuclear 

 division and some protoplasmic constriction, profoundly modifies 

 the relations of adjacent thecae in this matter of the attachment area. 

 The later history of the chain must be known and the completion of 

 the constriction between adjacent schizonts be observed, in case it 

 occurs, before the final skeletal relations of the attachment regions 

 can be determined. It is quite possible that they are of a wholly 

 different type in G. series from that acquired in schizogony when this 

 is accomplished by skeletal fission as in Ceratium and in Gonyaulax 

 catenata. 



The striking feature in which this chain of G. scries throws new light 

 on the process of reproduction in the Dinoflagellata and at the same 

 time adds to its complexity, is the formation, de novo, of the theca, 

 following schizogony in which the members take on a linear arrange- 

 ment and at the same time retain nearly the maximum possible pro- 

 toplasmic continuity. 



In Pyrophacus, for example, the two, four, eight, or sixteen daughter 

 cells or spores escape from the mother theca and synchronously each 

 converts its thin hyaline pellicle into a complete sutured theca. 

 The association of the daughter cells within the mother theca 



