258 University of Calif oruia Puhlicatiouft in Zoologjj. [Vol.8 



which separate the subgenera and groups which I have recognized. 

 In the discussion of the species, especially of those with the S'-O" 

 combination I have repeatedly called attention to an incipient 

 suture in apical 3', which if realized would give rise to a S'-l" 

 or a d'-O" arrangement. It may be noted in passing that in the 

 individualized plates of the theca of the Dinofiagellata we come 

 about as close to a realization of unit characters in these so-called 

 lower and supposedly simple organisms as we do anywhere in 

 the organic world. This exhibition in Gonijaulax of a tendency 

 on the part of plate 3' to divide into two elements and its appar- 

 ent realization in certain species (e.g. in G. diegensis) is strongly 

 suggestive ot a gradual rather than an abrupt transition from a 

 three-unit to a four-unit phase in this region of the skeleton. 



The differentiation in form and in position of ends of girdle 

 and in form of ventral area modify certain plates of the thecal 

 wall to an unusual degree. These plates are apical 1\ anterior 

 intercalary :/". precingular 6", the anterior plate of the ventral 

 area, postcingular 1'", and the posterior intercalary 1p. Three 

 significant relations of these plates appear to be correlated with 

 their relatively greater degree of diversification. 



In the first plact. they are, with two exceptions (apical 1' and 

 anterior intercalary J") immediately in contact with the region 

 of protoplasmic motion, the furrows in which the two flagella 

 are active, during the time of thecal formation ; in the second 

 place they are all in immediate contact with the line along which 

 the theca is parted in cell division, and thus in a region of con- 

 stantly repeated liberation along one margin, from restraining 

 contacts with other plates. 



It should, however, be noted in this connection that the girdle 

 plates 1-6 are absolutely constant in number and exhibit few 

 modifications in the genus, and that other plates along the fission 

 line (see text figs. A-D, p. 195) do not manifest an equal amount 

 of diversification within the genus. In a third particular, how- 

 ever, the plates named, together with the whole ventral area which 

 is quite variable, do occupy a unique relation among the elements 

 of the theca. They all lie in the midventral region adjacent to 

 the flagellar pore precisely in the region where the spiral move- 



