80 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA 



Cavers (1913) gives au extensive review of the interrelations of the flagel- 

 lates and lower Algae with a predominant utilization of the fresh-water forms, 

 which are perhaps better known than marine ones. In his phylogeuetie group- 

 ing he derives the branches leading to the GjTnnodiniaceae and the Prorocen- 

 traeeae from two different groups of the crypt omonads, Protochrysis and 

 W ysotzkia, with a questioned derivation of the diatoms from the Proroeen- 

 traceae. 



No connecting link between these two groups of dinoflagellates has hereto- 

 fore been found which would present a reasonable basis for claiming a closer 

 relationship between them. This gap we feel has been bridged by the genus 

 Protodinifcr, described in this paper, taken together with the genus Haplodi- 

 nium, recently described by Ivlebs (1912). The former lacks the theca of 

 ProroceMtrum , but has a smiilar orientation, with its tentacle extending from 

 the anterior end as the spine does on the theca of Prorocentrum. 



Thus Protodinifcr has the appearance of a Prorocentrum which has escaped 

 from its theca. A more profound difference separates them, however, in that 

 the former shows the beginnings of a girdle which is poorly developed and 

 apparently a temporary depression only, 0.5 transdiameter or less in length, 

 with a feebly developed sulcus. Haplodinium also lacks a theca of discrete 

 plates and tentacle as well as the girdle and sulcus of Protodinifcr, but its 

 orientation and arrangement of flagella are smiilar to those of Prorocentrum 

 (fig. R, 7). 



With these two genera as a starting point two possible lines of development 

 are open. The first of these starts near Protodinifcr and leads to the G}Timodi- 

 nioidae by way of Oxi/rrhis (figs. R, 2-4). The latter gemis still shows a poorly 

 developed but posteriorly located girdle. The second line of development goes 

 from Protodinifcr or other non-theeate forms to Haplodinium (fig. R, 5), with 

 its membranaceous, cellulose-like covering of the body, and to the theeate forms 

 EjcuviacUa (fig. R, 6) and Prorocentrum (fig. R, 7). Of these genera the 

 athecate Protodinifcr is obviously to be regarded as the more primitive. 



The development of the theeate from the athecate forms is a natural process 

 of evolution which is repeated for each individual in the growth of a new theca 

 or part of a theca at every recurring division. 



In thus estal^lishing the relationships of the Adiniferidea and the Dino- 

 feridea a monox^hyletic origin for the Dinoflagellata becomes a logical conclu- 

 sion. This, as has been suggested by Bergh (1881&) and more fully by Cavers 

 (1913), would probably be found in the cryptomonads, though not as the latter 

 has postiUated, as two distinct series of independent origin. Of the two genera 

 which Cavers has suggested as the starting point for his two series leading up 

 to the two separate groups, Protocliri/sis to the G}^nnocliniaceae and Wysotzkin 

 to the Prorocentraceae, W ysotzkia (fig. R, 1) seems more nearly related to the 

 Adiniferidea than probably any other genera of the cr\"ptomonads. With a 

 form nearly allied to these as a starting point the progressive evolution of the 



