390 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA 



girdle and sulcus, compels the esta1)lislinieut of a separate genus to include this 

 aberrant species. The first step in this divergence is to he found in the aber- 

 rant Gymnodinium gJaucum, which forms a connecting link between the two 

 genera. 



The nucleus is conformable to the body in x^roportions and shape, being, in 

 T. teredo (fig. II, 4), eight times as long as it is wide. It is exceptionally trans- 

 parent and difficult to detect because of the delicacy of its membrane and trans- 

 parency of the excei^tionally coarse and ropelike, longitudinal, spirally twisted 

 chromatin threads. 



The marked asjTmnetry of the body is exhibited not only in the as^inmet- 

 rical contour of the short hAiJocone but also in the grouping of the rhal)dosomes 

 which are linear in form and massed entirely on the left side of the body in 

 four parallel rows and in a star near the apex. It is possible that these struc- 

 tures, which we designate as rhabdosomes, are in reality chi'omatophores. 



Comparisons 



The structure of this genus is so unique and so far removed from other 

 genera that its relations are in no case immediate. It is clear, however, that 

 in the torsion of the sulcus it is nearer to Cochlodinhiin than to GijnniodiuiioH. 

 In its posterior location of the girdle it is unique and at the opposite extreme 

 from Erythropsis. in which the girdle is far anterior with a compensating re- 

 duction of the epicone. The extreme reduction of the liw[)ocone in Torodinium 

 is ^Aithout jiarallel among free-living G^Tiiuodinioidae, except as foreshadowed 

 in Giimnodiuium glaucum. 



It is also worthy of note that the cytoplasmic differentiations of pigmented 

 rhabdosomes (or chromatophores?) are sinistral as is the ocellus of Erythropsis, 

 but they are in the epicone instead of the h;\'7)ocone as in the latter genus. 



Historical Discrssiox 



The first species now included in this genus was described by Pouchet 

 (1885a) as Gynmodinmm teredo. Schiitt (1895) found both species, but did 

 not diffei'entiate them, referring them lioth to G. teredo. We have found only 

 the stouter of these two species, Torodinium robustum, in which we include the 

 first two of Schiitt 's figures (1895, pi. 23, figs. Tli-j) of G. teredo, referring the 

 remainder to the species to which he assigned them. 



Key to Species of Torodinium 



Length less than 3.5 greatest transdiameters, sulcus with reversed terminal apical loop 



- robustum sp. nov. 



Length more than 4 transdiameters. no loop in terminal part of the sulcus teredo (Pouchet) 



