1918] Barroics: Skeletal Variations in the Genus Peridinium 443 



P. islandicum Pauls, and P. tabidatum Ehrbg. These are the two 

 dorsal plate patterns most frequently found in this genus. 



Origin of the Asymmetrical Dorsal Plate Pattern. — Aside from 

 patterns caused by a reduced number of plates, there are, however, 

 two other patterns which are occasionally represented, and, curiously 

 enough, usually among forms taken from bays or sounds or regions 

 in which the regularity of the physical conditions of the open sea are 

 somewhat interfered with or altered. These two patterns are asym- 

 metrical and are caused apparently by the greater development of one 

 side, or of one lower or posterior corner of either plate 2a or 4". We 

 have assumed that the initiative for these readjustments lies in the 

 region of the posterior corners of plate 2a as the newest skeletal mem- 

 ber of a presumably plastic portion of the shell rather than in plate 4". 

 which forms part of a phylogenetically much older portion of the 

 shell, although the shape and to some extent the size of plate 4'' seem 

 to be more or less altered, conversely to plate 2a, to meet its changes. 



By the development of the posterior left corner of plate 2a the 

 pattern of fig. 7 is formed and this in the experience of the writer is 

 the more common of these two asjTiimetrical patterns. It will be seen 

 to be a combination of the patterns of figs. 5 and 6, and has been 

 found by Miss Bathgate in San Francisco Bay material as well as by 

 the writer (pi. 19, figs. 5 and 6). 



By the greater development of the posterior right hand corner of 

 plate 2a the pattern of fig. 8 is formed. This pattern has been figured 

 by Joergensen (1902, p. 7). 



There are, of course, other explanations which can be given for 

 the origin of these several dorsal patterns. The two skew or asym- 

 metrical patterns may have been formed by a sliding of the row of 

 accessory plates over the row of precingular plates because of some 

 force, probably internal, which may have pushed the whole row over. 

 This suggestion, however, seems rather improbable because it attempts 

 to account for the formation of the skew patterns on a hypothesis 

 separate and of a different order from any upon which the original 

 formation of these plates can be accounted for, and also because no 

 other adjustments of surrounding plates are to be noted, as would be 

 expected from a readjustment so great as the shifting of the position 

 of three plates in so prominent a position as that in which the acces- 

 sory plates are. 



Summary of Discussion of Dorsal Patterns. — It seems, then, that 

 the accessory plates appear first in the genus as a symmetrical pair 



