438 University of California Publications in Zoology [Vol. is 



degree necessary to meet plates 2" or 6" without the growth, whieli 

 is unknown, of these plates to meet it half wa3^ On the other hand, 

 it is easy to see how plates 1" and 7", beginning, as it were, as small 

 plates, budded off from plates 2" and 6" when these latter were the 

 end plates of the precingular row and may by increasing in size have 

 inserted themselves between plates 1' and 2" on the left hand for 

 plate 1" and between plates 1' and 6" on the right hand for plate 7" 

 until this primary articulation is completely destroyed and plate 1" 

 has come to meet broadly plate 2', and plate 7" to meet plate 4', We 

 have thus before us the ventral plate patterns of the known stages 

 for the three subdivisions of this genus proposed by Joergensen. 

 First is that of Paraperidinium represented by our fig. 1 in which 

 there is only a small precingular plate at each end of the precingular 

 row adjacent to the rhomboid plate. A stage preliminary to this may 

 be found in Gonyaulax, for example, in which but one such small 

 plate, that of the right hand, has as yet split off. It is significant to 

 note also that this plate first seems to split off on the right hand end 

 at the beginning of this series of development and that later in the 

 series it is found that the right side proceeds more rapidly in devel- 

 opment than the left side, which seems to lag, while one stage of 

 development on the left side seems in present faunas not to be repre- 

 sented at all. Secondly, the plate pattern of Metaperidinium is that 

 in which the precingular end plate on the right side has increased in 

 size until it meets plate 4', completelj- separating plates V and 6". 

 The corresponding asymmetrical stage in which the precingular plate 

 on the left end has increased rather than that on the right end seems 

 not to be known. Finally, we find the pattern represented in Ortho- 

 peridinium in which both the end plates of the precingular row have 

 increased uniformly, or if one pleases to so regard it, in which plate 1". 

 on the left, has caught up with the maximum development of plate 7'' 

 which can be permitted by the pressure of surrounding plates or by 

 its own capacity to hold together. 



Turning now to the dorsal surface of the epitheca, we find that in 

 that portion of the genus having three accessory plates the variations 

 in plate pattern occur in a pair of regions at or near the anterior 

 corners of the mid-dorsal precingular jolate, 4". Among possible 

 explanations for the variation of the suture pattern at these points, 

 we may note that these variations may be due to the increasing size of 

 the mid-dorsal precingular plate, 4", so as to crowd the middle acces- 

 sory plate and effect an articulation with the lateral accessory plates 



