1918] Barrows: Skeletal Variations in the Genus Periclinium 407 



cant in supporting the suggestion that the origin of the plates of the 

 dinoflagellate skeleton is b,y the fragmentation of an originally homo- 

 geneous shell circumferentially rather than meridionally, and that 

 other plates though derived from the plates of this or secondarily 

 formed rows still preserve this arrangement. 



Kofoid's system has certain points in common with that of Stein 

 and of Biitschli in recognizing rows of plates and in enumerating 

 these in a left-to-right direction. Broch (1912, p. 39, fig. 8) in his 

 figure recognized the arrangement of the plates in rows, and so does 

 Faure-Fremiet (1912, figs. 3-14), though the figures of Faure-Fremiet 

 suggest that he conceived of the apical plates as lying in three meri- 

 dional rows in addition to the rhomboid plate. The implication from 

 the nomenclature of these figures is that Faure-Fremiet regarded the 

 accessory plates as derived by splitting off from the apicals, while the 

 implication from the nomenclature of the accessory plates in Broch 's 

 (1912, fig. 8) figure is that these accessory plates may have been 

 derived from the precingular plates. There are indications of shape 

 in many of the plates, however, which make the manner of origin of 

 the accessory plates doubtful. In view of this lack of more definite 

 knowledge upon the origin of the accessory plates, Kofoid's system 

 committed by implication to neither view would seem to be the more 

 conservative. The rhomboid plate in Kofoid's nomenclature is re- 

 garded as an apical which has not divided but which has become 

 extended to meet the girdle. 



The plates of the ventral area seem never to have received a defi- 

 nite nomenclature, probably because of the great difficulty in deter- 

 mining them. On account of the considerable variability of form 

 and extent of the ventral area in different species it is to be expected 

 also that the plates of which it is formed will display much variation, 

 and this has been demonstrated for shape as well as number in several 

 forms already worked out. A system of nomenclature for the major 

 plates of the theca in order to permit expression of the relationships 

 of the various groups of the dinofiagellates need not include, there- 

 fore, the ventral area in its consideration, for the main relationships 

 of the dinofiagellates seem to be sufficiently if not mainly expressed 

 in variation in the arrangement of the major plates of the theca ex- 

 clusive of those of the ventral area. 



