l!V (i. I. PLAYKAIR. 795 



velopment for its solution. In the following notes, if the figures 

 (dorsal view) of p. emulation var. guild f'ordense (Text-fig. 5, «) 

 P. geminum type (Text-fig. 9. a) and P. umbonatum var. cen- 

 tenniale (Text-fig. 14, a) be compared, it will at once be seen 

 that not only are they all of a size and of quite remarkable 

 similarity in characteristics, but that P. geminum holds the cen- 

 tral position between the other two. In that species four plates 

 meet at a point a little below the apex. Let this point become 

 either a vertical line by longitudinal development, or a hori- 

 zontal line by transverse growth and either P. caudatum var. 

 guildfordense or P. umbonatum var. centenniale is the result. 

 From their minute size it is certain that all three are the out- 

 growth of resting-cells such as I have already figured (op. cit., 

 PI. lvii., f. 4) . Is it not possible, though they are representa- 

 tives of distinct types (types which are still more distinct in 

 their larger and more mature forms) that they are the results of 

 three diverging lines of growth from the same resting-cell ? 



General remarks. — There is another point that calls for re- 

 mark. Such a large proportion of the cells have the antapical 

 plates set obliquely; very often also the apex of the epivalve 

 is eccentric. This is neither a specific nor varietal characteristic, 

 but is common to almost all cells in which the hypovalve is not 

 regularly domed over and the terminus of the longitudinal fur- 

 row definitely fixed. Along with this peculiarity goes another 

 viz. that the transverse furrow is more or less spiral, being 

 dragged downwards always on the left side in ventral view. In 

 such cells, the fact is, there are two longitudinal axes, a dorsal 

 which is straight and vertical, and a ventral which is curved (cf. 

 Text-figs. 14, 15; PI. xli., fig. 3 b, b') . 



The convex side of the curve is always on the left in ventral 

 view, and the plane of the antapical plates is at right angles to 

 the curved axis, hence its obliquity with respect to the other. 

 The apical pit in the epivalve, which always tends to be ventral 

 rather than dorsal, is very often drawn aside (but to a less ex- 

 tent) conformably to the same influence. As I consider the 

 shape and relative proportions of the antapical plates to be of 

 more importance than their position on the cell I have always 

 drawn the diagram of the hypovalves on the oblique plane. The 

 cause of this asymmetry seems to be a longer period of growth 

 (or a stronger growih) on one side, this perhaps being due to 

 the greater action there of the tip of the transverse flagellum. 



