78 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA 



It is with this background, at once obscure, yet replete with the beginnings 

 of far-reaching lines of influence, that we must search for the early affinities 

 of the dinoflagellates. In lieu of a complete knowledge of the life histories of 

 these forms, which is an important desideratum in any scheme of classification, 

 the morphological features characteristic of the group must be used in tracing 

 out their relationships. Throughout the Protozoa generally the motor organ- 

 elles have been found to be the most satisfactory basic feature, correlated with 

 other cytological characters, for systems of classification, and it is in the 

 arrangement, number and insertion of these that relationships may he most 

 surely found. 



In the dinoflagellates these relations of flagella and places of insertion with 

 the modifications connected therewith are typical and uniqiie. The two main 

 divisions of the subclass Dinoflagellata are based upon the extent of the devel- 

 opment of these peculiar modifications connected with the motor organelles. 

 In the Adiniferidea the girdle and sulcus are entirely lacking or merely sug- 

 gested by the projecting fins and flagellar notch. In the Diniferidea they are 

 usually well developed at some stage of the life cycle. In both groups the 

 flagella have the t}T.)ical dinoflagellate structvire, consisting of a transverse, 

 ri])bon-like flagellum and a longer, threadlike, longitudinal, trailing one. At 

 least it is possible so to orient the Adiniferidea as to harmonize them with this 

 interpretation. 



Kfxatioxships. — The relationships of the Adiniferidea, represented by the 

 genera Prorocentrum, ExuviaeUa, and Cenchridium, have long been a pro])lem 

 concerning which protozoologists have held widely divergent views. The reason 

 for this lack of agreement appears to be the fact that the Adiniferidea, though 

 they are relatively sunple in structure, yet have a bivalve theca. The thecate 

 forms (Prorocentridae and Peridiniidae) are probably derivatives of the athe- 

 cate ones (Protodiniferidae and G^noinodinioidae) , the prunitive forms of the 

 latter being more like the other and simpler types of Flagellata. 



This possession by the Adiniferidea of an exoskeleton or theca of an entirely 

 different type from that of the Diniferidea, which skeleton indicates an advance 

 in evolvition on the one hand, while on the other a relatively simple type of 

 structure and absence of the girdle, has led protistologists in opposite directions 

 in estimating the relationships of the Adiniferidea. Those who have felt the 

 importance of the skeleton in the dinoflagellates as a whole have regarded the 

 Adiniferidea as primitive representatives and have placed them at the base of 

 the phAdogenetic tree, as did Bergh (1881&) and Biitschli (1885fl), who first 

 attempted the systematic organization of the group. 



When this primitive role is assigned to the Adiniferidea a new difficulty 

 arises with regard to the relationships of the G^iimodinioidae, which are devoid 

 of skeleton and contain some of the simplest types of dinoflagellates, as well 

 as some of the most diversified and complex, possessing, for example, eyespots 

 with lens and pigment body, and even nematocysts. Both Bergh (1881Z)) and 



