82 MEMOIRS OF THE UNH'ERSITY OF CALIFORNIA 



Ivofoid (1907), namely, that Poli/krikos is a simple colonial gymnoclinimn-like 

 organism with the zooids remaining united in a permanent somatella after 

 binary fission, and the mmiber of nuclei equal to the number of zooids, or to 

 half that number. Temporary colonial organization of the linear or chain type 

 is often found in the dinoflagellates and permanent colonies and less or more 

 permanent somatellas are common in other groups of the Protozoa, which have 

 no suggestions of ciliate affinities attached to them. Variation in size of the 

 nuclei, resulting from successive mitoses, is also a common phenomenon and is 

 no more striking in Poli/krikos than in other protozoan forms. 



Cavers further suggests the possibility of the genus Erijthropsis affording 

 a transition stage between Poli/krikos and the CUiata. A comparison of the 

 structures of these two widely different organisms (figs. F, RR) will at once 

 demonstrate the futility and moii)hological folly of looking for the origin of 

 the Cilia ta near these highly specialized flagellates, the stiiicture of which differs 

 profoundly from that of present day ciliates. The types of specializations 

 found in the Diuofiagellata, represented on one line by the thecate forms, on 

 another by Enjtkropsis, and on still another by the extremes of torsion in 

 CochJodinium, are not intermediate types which lead on to other groups of 

 Protozoa, but are rather the culmination of their particular lines of development. 



Derivatiox. — The derivation of the Diuofiagellata from some primitive 

 flagellate allied rather closely to existing Cryptomonadina, with two anterior 

 flagella and xanthophyll chi'omatophores, seems the most probable of all pos- 

 sible lines of origin. The discussions of Pascher (1911), Klebs (1912), and 

 West (1916) serve to establish this, although their utilization of fresh- water 

 genera such as Wysotzkia and Protochrysis as primitive connecting links seems 

 less tenable. When the marine cr^i^tomonads are better known it may be that 

 as good or better morj^hological links will be discovered between cryptomouads 

 and dinoflagellates. 



The differentiation of one anterior flagellum to form the spirally wound, 

 ribbon-like collar of the craspedomonads, to form the trailing flagellum in the 

 Bodonidae, and to develop into the attached, posteriorly directed, undulating 

 membrane of Trichomonas, is suggestive of a widespread and deep-seated 

 tendency for the two or several flagella of primitive types to become differen- 

 tiated. This differentiation of the two flagella into a trailing thread on the 

 one hand and a closely sinuous rililDon on the other is the basis of diuoflagellate 

 evolution, and establishes the fundamental connections between the Adinif eridea 

 aud the Diuiferidea. It is to be expected that the primitive cr}-ptomonad with 

 dinoflagellate affinities will show this differentiation in structure or position, 

 or both, of these flagella. These conditions are fulfilled to some extent by 

 Haplodininm and Protochrysis, the first allied to the Adinif eridea and the 

 second to the Diuiferidea, but both are fresh-water forms. AVe may therefore 

 regard these genera as at least allied closely to the stock of cryptomonads from 

 which the primitive dinoflagellates took their origin, l)ut it is to be hoped that 

 stiU closer allies will be found among marine flagellates. 



