478 



Mr. G. Bidder. 



[May 12, 



Fiq. 4. — Basal spherules of collar cells, Sycon raplianus. At A the typical plinth 

 and column is shown. At B (from the same preparation) the fission is less 

 typical, but a large ovum or sperm cell is feeding on the spherules. There 

 are no collars, but many cells show sharply flat, or even concave tops, suggest- 

 ing that the collars are being formed anew. Osmic acid and hematoxylin. 



My conception of the metamorphoses of the collar cells is as fol- 

 lows : — 



Tn Heteroccela (probably it is similar in Silicea) the collars of 

 the collar cells are at first mere fringes, which help to retain the 

 food and filter the water as it passes from the base of the cell to 

 the moving tip of the flagellum. When the cell is satiated the 

 flagellnm ceases to move, and degenerates ; the collar unites with 



