48 THE BOTANICAL MAGAZINE. [Vo j. xxiv. No. 278. 



Ag. But different from the latter, the spore of the present 

 species has a short hy aline pedicel which serves to push out the 

 spore from the epitheca (PI. III). No special content could be 

 seen within the pedicellar room. The perizonium is complete 

 and entire from the beginning of appearance so that no com- 

 munication of plasm is allowed between the pedicel and spo- 

 rangium proper. And the walls of both parts have different 

 chemical properties as shall be stated below. The base of 

 pedicel fits to the inner side of epitheca, and gradually dimin- 

 ishes upwards in the diameter. The height is nearly as large 

 as the diameter of base. We were not able to ascertain how 

 the pedicel has been formed. It may be generated as an out- 

 growth of sporangium during the latter is increasing in its size 

 and afterward septated, or it may be formed as a special cell 

 while the sporangium was still confined within the frustule. 

 Absence of contents, however, suggests us to adopt the former 

 view as more probable case. 



As a pedicel lifts up sporangium, the lower pole of the latter 

 is naturally freed from epitheca earlier than from hypotheca 

 (Fig. 1). Just after the valves have detached from sporangium 

 the prints of the trabecular ridges are seen as radiating shallow 

 grooves on the polar regions (Fig. 2). 



In the next stage, the sporangia swell up by degrees, the 

 trabecular prints of course disappearing. Thos: which have 

 attained to the full size are compressed globular in shape (Fig. 

 5, 6). The plane of complanation is at right angles to the ad- 

 herent plane of the mother frustule. Masses of protoplasm 

 aggregate at the center of both convex surfaces and mark the 

 future positions of the pseudo-nodule (Fig. 5, 6). The chromo- 

 plasts are especially dense around the plasmic masses and are 

 disposed radially towards the margin of sporangium. 



While the process is going on, a new membrane is formed 

 lining the inner surface of the already existing perizonium. This 

 membrane begins to separate from one of the convex surfaces 

 of perizonium, gradually approaching to a plane though with 

 round edge (Fig. 7). The radial arrangement of chromoplasts 

 becomes more vivid than before and each radial line indicates a 



