416 University of California Publications in Zoology [Vol. 18 



out decades of study the number and arrangement of these plates 

 have been found to be sufficiently constant to be relied upon as bases 

 for the descriptions of genera ; and the system of rows of plates upon 

 which the whole matter of plate arrangement seems to rest, extensible 

 to all the genera of this large family, are other considerations which 

 still further suggest that certain group relationships are to be revealed 

 in a comparison of the plate formulae when taken into account 

 together with other skeletal features which may in one part of the 

 genus or another be of phylogenetic importance. There is indeed no 

 other character which reveals more, not only of the detailed but also 

 of the general relationships of the Peridinidae, than this character of 

 the development and arrangement of the plates of the shell. 



Progressive Plate Development in Genera of Peridinidae 



We find ourselves confronted in viewing the family Peridinidae by 

 a considerable assemblage of genera which can be arranged in a series 

 according to the number of plates composing the skeleton. All of 

 these genera have at least four circumferential rows of plates and 

 several genera have one or more plates of a partially complete acces- 

 sory row. At one end of this series stand such genera as the following : 



CeraiMm -4'-0''-5"4s-5"'-0P-2"" 

 Proioee?-atmm,-2'-0''-6"( ?)-68( ?)"'-0p-3"" 

 Goniodoma-l (to 3)'-0''-7"-(2)s-5"'-0p-2"" 

 Heterocapsa,-4: (or 5)'-0''-.5 (or 6)"-?«-5"'-0p-2"" 

 Pyrophacus-A (to 6)'-0°-0 (to 12)'?b_9 (to 12)"'-0p-(3 to 5)'" 



The last genus mentioned is especially noteworthy because of the 

 large number of plates contained in each row. 



In another group of genera an incomplete row of plates has been 



introduced on the dorsal side of the epitheca or of the hypotheea or 



of both, such as the following : 



Eeterodinmm,-3'-l''-6"l^-7'"-0'^-3"" 

 GonyauUx-1 (to6)'-0 (to 4)''-6"-6s-6"'-1p-1"" 

 Centrodinium-2 (to 4)'-0«-6"-?s-6"'-1p-4"" 

 Spiraulax,-A'-l''-6"-6s-6'"-V~l"" 

 Peridinium,-4:'-0 (to 3y-7" -3^-5" -0^-2"" 

 Perindinella,-4:'-0 (to 3)"-7"-3e-5"'-0P-2"" 



It is not possible to arrange all of these and the other genera of 

 this family in a strictly linear series because of the two ways, at 

 least, in which the number of plates of the shell seem to differ, i.e., in 

 the number of rows and in the number of plates in each row. More- 



