THE COLLAR-CELLS OF HETEROCGLA. 13 
The protoplasm is in life greenish, and in normal condition 
of ground-glass appearance. Lach cell contains from four or 
five to a dozen spherical granules, up to 1 yw, or rarely 2m in 
diameter, rather more refracting than the surrounding proto- 
plasm. I have called such granules “ basal spherules ” (18, 
p. 476) from their strong tendency to segregation in the base 
of the cell. 
The cells have nearly the form and relation to each other 
of full corn-sacks standing side by side in a granary (v. figs. 
1, 3, 9a, 15, and 19); in the normal condition they are dis- 
tinctly but not widely separated, appearing to be actually in 
contact only at their bases. The generally barrel-shaped 
lateral surface of the cell always shows a clear smooth line 
in optical section; the circle marking upon it the base of the 
collar is also a smooth and sharply defined line. On the other 
hand, the convex or irregular area inside the collar (intra- 
choanal area) has nearly always a fainter outline, as though it 
were less refractive; it is often irregular, often finely punc- 
tated, often strongly granular. This was observed also in S. 
raphanus. 
CoLuar. 
The living healthy collar is from 2u to 7y in height, in- 
variably an almost perfect cylinder, very little constricted at 
its base ; ending sharply above without either rim or expan- 
sion (figs. 1, 2,3,and 19). It has no vertical cleft, thereby 
differing from the spathiform collar of Choanoflagellata as 
described by Franze (19). From observations in life the 
thickness of collar or flagellum appeared to be 4 to fu. 
Once in 8. compressum, and once in S. raphanus, I 
observed in a fresh preparation the free edge of a collar, looked 
at from above, to present a “ milling ” or beaded appearance, 
as in fig. 7; in each case the cell had been some time under 
the cover-slip. Since the accompanying plate was engraved I 
have re-examined all my permanent preparations with a Zeiss 
nucleoid bodies about ‘007 to (008 mm. in diameter.” The identification of 
the chambers in my living specimen was unmistakable. 
