1918] Barrows: Skeletal Variations in the Genus Peridinium 439 



which did not obtain in the primary condition. This explanation, 

 however, seems improbable, since in all other related genera there is 

 no suggestion that this mid-dorsal plate undergoes a great develop- 

 ment in advance of its fellows in the precingular row. 



That the mid-dorsal precingular plate does vary somewhat more 

 than its neighboring precingular plates, however, appears from the 

 fact that in those specimens having a small middle accessory plate 

 with the plate pattern of fig. 5, the mid-dorsal precingular, 4", occu- 

 pies a greater arc of the equatorial circumference of the organism 

 than in those species in which the middle accessory plate is large with 

 the pattern of fig. 6. There seems, then, to reside in the mid-dorsal 

 precingular plate, 4", some capacity for variation in relative size 

 from species to species, though there are other considerations to sug- 

 gest that the major stimulus for a real change in plate pattern may 

 take effect in the middle accessory plate just anterior to it. More- 

 over, since it is probable that the accessory row of plates is of the 

 more recent introduction, phylogenetically, in the skeleton of Peridi- 

 nium, it seems probable also that these plates are in a more plastic 

 condition than the plates of previously existing rows. 



It seems, then, that we may consider only the probability that the 

 seat of alteration of dorsal plate patterns lies among these lately intro- 

 duced plates of the accessory row. If so, is the center of expansion 

 by which articulations are changed, seated in the middle accessory 

 plate or symmetrically in the lateral accessory plates or in all these 

 plates of the accessory row? Upon regarding the proportionate size 

 of each of these three plates to the adjacent plates of the skeleton, it 

 appears that among the species illustrating the changes in dorsal pat- 

 tern the middle accessory plate varies more widely than the lateral 

 accessory plates relatively to the size of the adjacent plates, while the 

 lateral accessory plates undergo but comparatively slight variation in 

 size. It seems probable, then, that in the part of this genus contain- 

 ing these accessory plates it is the middle one which is the more plastic 

 or the more variable. 



Origiii of the Accessory Plates. — Of peculiar significance just here 

 is the absence of any example of a single accessory plate in this genus. 

 The accessory plates which first appear, do so as a pair, apparently 

 in response to a stimulus applied symmetrically on the anterior dorsal 

 region, and it is in connection with the symmetrical effect of this stim- 

 ulus for the interpolation of an accessory row of plates that the sig- 

 nificance of the absence of any form displaying a single accessory 



