338 bulletin: museum of comparative zoology. 



The polar protoplasmic continuity of schizonts exemplified else- 

 where in Polykrikos is here even more extended in areas and persists 

 moreover through more generations and individuals. To this is 

 added the formation de novo of an entire new skeleton such as is seen 

 in some other species of Gonyaulax after ecdysis with or without free 

 cell division. 



The interpretation of the condition of the anterior member of the 

 chain from which the epitheca is missing is difficult. The protoplas- 

 mic contents are intact and show no conclusive evidence of breakage 

 yet the adjacent hypotheca and girdle plates are fairly well formed. 

 It is possible that the epitheca has been accidently torn away or shed 

 as the result of some physiological condition of the cell body. On the 

 other hand it is obvious that another division of the distal member 

 remains possible without thecal fission, ecdysis, or rupture so long as 

 the theca is not formed about the end of the protoplasmic cell body. 

 The posterior member of the chain is entirely enclosed in a delicate 

 pellicle except where it joins its neighbor. Should the whole of this 

 metamorphose into thecal plates the further progress of fission in that 

 region without skeletal fission, ecdysis, or rupture of the hypotheca 

 and budding will be definitely stopped. 



The difference between this chain of G. series and that of G. catenata 

 or of a typical Ceratium chain will be best appreciated if one compares 

 Levander's figures (Plate 2, fig. 6) together with one of a bit of Cera- 

 tium chain (Plate 2, fig. 5) with those of G. series (Plate 1, figs. 1, 2; 

 Plate 2, fig. 5). It will be seen that in G. catenata and in Cera- 

 tium chains generally, the schizonts are of equal size and similar 

 proportions (except in the antapical horns) and present similar mark- 

 ings. It is only immediately after fission or rarely among members 

 of a chain that the old and newer parts of the theca which has 

 recently undergone fission can be readily distinguished. The schi- 

 zonts of G. series on the other hand diminish in size progressively 

 toward the ends of the chain. The posterior theca is wholly in the 

 earliest stages of formation, and the anterior ones also have their 

 suture lines less developed, and the epitheca is not formed (or has 

 been removed) on the most anterior member. There is evidence 

 neither of skeletal fission nor skeletal transmission. 



Another striking difference between the chain of G. series and the 

 others is in the method of attachment of the adjacent members. In 

 Ceratium (Plate 2, fig. 6) the attachment is by a slender protoplasmic 

 strand emerging from the apical pore of the posterior schizont and pass- 

 ing into the cell body of the anterior member at the ventral edge of 



