u 



rSKRIFT FÖR LILLJEBORG 12 



on the central part we sec that the meshes are somewhat concentrically arranged 

 and compressed towards the direction of the concentricity. 



The cytoplasm situated in the upper part of the cell is by far the most in- 

 I gin with it may be stated that it stains less intensely than the cap, 

 - much as the cylinder which connects the cap and the floor. In cross-sec- 

 tion it has the appearance of two wings, just as figured by Horst; but with careful 

 fixing and staining I find that this general appearance is deceptive, and that the real 

 shape of this part of the cytoplasmic mass might be roughly compared to three 

 I'unn differenl one inserted in the other, and the two upper ones resting 



with their edge- against the cell wall. This might be easier seen in the figures 7, 8 

 and 9 than described by words. The lowest of these funnels is the shortest. Its 

 r part rests against the upper part of the cap, while the upper reflexed part 

 supports the middle funnel. On the edges of this lower funnel, if indeed so it might 

 be called, we find a number of thicker strands of cytoplasm, probably the upper mar- 

 gins of a r<>w of plates serving as a support for the cytoplasm above, as seen in 

 figures 7 and 8. These strands with their plates, however, run far up in the cyto- 

 plasm, but are nevertheless quite distinct from the structures above. The middle 

 funnel is large and thick, and fills the upper one-fourth of the cell. It seems especially 

 adapted to support the roof of the cell, the upper part of the cytotheca. More or 

 intimately connected with this most funnel-shaped part of the cytoplasm is the 

 smaller upper funnel, the sides of which bend inward, forming between them a 

 vacuole, situated at the very top of the cell, immediately below the cuticle. In 

 this large, almost globular, vacuole is seen a smaller or larger, nearly always globu- 

 lar, body, otosome, entirely isolated from any of the other cytoplasmic parts. The 

 nature of this globule is doubtful. It is very transparent, generally globular with 

 slightly roughened surface, and seemingly composed of slightly or irregularly layered 

 cytoplasm, the layers having the appearance of being somewhat concentric. The 

 ball is not strongly refractive, and is not calcareous, like the concretions of the Cala- 

 is diverticula. It docs not take any of the stains tried on it and enumerated 

 It varies in .size, to some extent, but is always seen in cells not injured 

 ■ning. In one cell the ball consisted of two joined balls (fig. 17). For this 

 body 1 propose the name of otosome. 



Nerve endings or ganglia. On each side of the cylinder between the 



iid the floor I find always in properly preserved cells what I take to be two 



all ganglia or probably only nerve endings or plates. They appear in my prepara- 



ä dark spots from which line fibres are spreading out (fig. 7, n. e.) towards the 



ind along the cytoplasmic floor. It seems to me that these supposed 



