KOFOID AND SWEZY: UNARMORED DINOFLAGELLATA 79 



Biitselili (1885rO solved this diffir-ulty by putting Giimnodininm and allied 

 genei-a at the toi) instead of the base of their phylogenetic tree, thus giving rise 

 to the conception that the athecate condition is secondarily derived and not a 

 primitive one. 



With the increase in our knowledge of the structure, abundance and 

 speciation of these naked dinoflagellates there has come a reversal of opinion 

 in respect to their being derivative instead of primitive t^^es. This is shown 

 in the system proposed by Schiitt (1896) in his treatment of the Dinoflagellata 

 (Peridiniales) in Engler and Prantl's Pfanzenfamilien. He placed the 

 G}^nnodiniaeeae as the first or most primitive family, the Prorocentraceae next 

 and the Peridiuiaceae as the terminal family. This linear arrangement brings 

 the Prorocentraceae between the Gymnodiuiaceae and the Peridiuiaceae and 

 thus parts near allies with an entirely divergent family intervening. Klebs 

 (1912) held to a similar view, placing the G}^nnodiniaceae in the primitive 

 position and deriving the Prorocentraceae from them liy way of Ainphidiniam. 

 Poche (1913) grasps the other horn of the dilemma and places the Prorocen- 

 trinea in the primitive position with the Peridiniinea next, within which he 

 brings the two related tribes, Gjannodinioidae and Peridinioidae in the sequence 

 of their evident relations. This arrangement places the thecate Prorocentrinea, 

 however, as the most primitive group. Doflein (1911) gives the primitive 

 position to the same group in his Adinida with the Dinif era containing G>inno- 

 dinidae, Peridinidae and Dinophysidae. Cavers (1913), and later West (1916), 

 have suggested still another solution to these difficulties by giving to the Dino- 

 flagellata a diphyletic origin, with the Prorocentraceae having a separate origin 

 from the Gpnnodiniaceae. 



The dilemma which has thus arisen is plainly one resulting from the fact 

 that we are here dealing with the divergent and largely terminal branches of 

 an evolving group, of the more primitive forms of which we have had too little 

 knowledge to elaborate any satisfactory systematic arrangement which ade- 

 quately represents the true relationships of these main divisions of the 

 Dinoflagellata. 



The search for more primitive representatives of the Dinoflagellata which 

 will show connections with other divisions of the Flagellata and, at the same 

 time, afford a common starting point for the Prorocentridae and the Peridi- 

 nioidae is well illustrated in the contributions of Klebs (1912), Cavers (1913), 

 and West (1916). Their search leads them among the Chrysomonadina and 

 CrjT3tomonadina, which include among others biflagellate, asjanmetrical flagel- 

 lates with yellow chromatophores. Klebs adds a new genus, Haplodiniiim , with 

 characteristic dinoflagellate nucleus, two characteristically differentiated fla- 

 gella, and a continuous, cellulose-like mem])rane covering the body without sub- 

 divisions or pores. He includes this in the Prorocentraceae and modifies the 

 definition of the family to provide for the inclusion of this new, brackish-water 

 genus. 



