1918] Barrows: Skeletal Variations in the Genus Peridinium, 419 



other of the rows of the skeleton. Correspondingly, a genus in which 

 a partial row of accessory plates has been interpolated on the dorsum 

 of either the epitheca or of the hypotheca is higher in the phylo- 

 genetic series, other things being equal, than a genus in which there 

 is no such row of accessory plates. It would seem probable, then, that 

 a genus having, for example, seven plates in the precingular row might 

 be derived from a form perhaps still represented in a related genus, 

 having but six plates in this row by the division of one of these plates 

 into two parts. We shall have need to refer to this suggestion again. 



Possible Method for ^Iultiplication of Plates 



It is possible, of course, that plate multiplication takes place by 

 the splitting of a former plate. Without saying that this may not 

 occur in certain genera it appears from the progressive series of pre- 

 cingular plates, for example, which can be constructed for the genera 

 related to Peridinium, that the splitting of a former plate into two 

 equal or nearly equal plates does not occur, but rather that new 

 plates arise at first as small skeletal elements either split off from the 

 corner of a former plate as, perhaps, in the formation of plates 1" and 

 7" of Peridinium or that new plates arise to fill the gap left by a 

 bulging of the internal protoplasm of the organism which would 

 cause three plates to be pulled apart at their point of articulation. 

 That is, the strain which causes the formation of a new plate seems 

 to require the construction of a new skeletal element rather than a 

 remolding of a formerly existing element. These elements, the plates, 

 moreover, seem to be limited in size, and it is probably the impossi- 

 bility of a shell of one piece to contain the active and growing organ- 

 ism which has caused the production of a shell composed of a number 

 of plates. Physical factors of the skeletal material or of the shape of 

 the organism may determine a limit beyond which the skeletal material 

 cannot exist as a single plate, the actual area of each plate as well as 

 its shape depending upon its location. In the few cases where very 

 large plates are found, as on the dorsal side of the epitheca of 

 P. excentricum, such excessively large plates are in regions where 

 there is but little change in shape or contour. In such a region as 

 the apex, where the contour changes sharply, there are usually found 

 a number of plates. However, in those forms in which there is found 

 a protracted apical horn the apical plates are narrow transversely in 

 the direction of sharp change of contour, but elongated axially in the 



