24, GEORGE BIDDER. 
The degree of change differs in different specimens; but 
usually after two hours every chamber presents an appearance 
it may be convenient to call “ striated,” the lumen being 
greatly reduced, the elongated thin cells forming a herring- 
bone pattern down the chamber, and appearing (if it be not, 
indeed, the fact) as if many of them became free. 
Carter (2) draws two collar-cells from teased living prepara- 
tions of 8S. compressum, of which his fig. 1 corresponds 
exactly to my fig. 10, and his fig. 2 to my fig. lla. He de- 
scribes changes on the slide to amceboid forms; but he is 
treating entirely of cells “ scratched out from the body of the 
sponge,” whereas I have confined my observations to cells in 
situ. 
Fig. lla is a sketch made from S. raphanus (Naples, Aug. 
1892), with the note ‘‘ a tendency to elongation of cells as the 
preparation dies ;” while figs. 114 and lic, made at the same 
time, bear the note “flagella” [in other parts] “still in 
motion, certainly none on these cells.”” I have already quoted 
Schulze’s observation (3) as to occasional concrescence of 
collars in this species; the cells drawn by him as normal 
appear to have mostly entered on the phase of my figs. 6 and 
8, that is, to have been twenty or thirty minutes under the 
cover-slip; his Taf. 14, fig. 4, is practically in the stage of my 
figs. 11 and 12. 
The form of cell produced by this series of changes appears 
identical with that described by Dendy in his “x ”’ chambers 
(12, 20), and is certainly so with the ‘ column-and-plinth ” 
cells described by me (18, p. 477). Similar cells in permanent 
preparations show the nucleus in most cases at the extreme 
distal end of the cell, the granules are in the base. The upper 
surface of this base coincides—at least in most instances— 
with the upper surface of the intercellular jelly, and the con- 
tours of the uppermost enclosed granules lie in the same plane. 
With the ordinary immersion objective the appearance of a 
septum is in many cases convincing, so that even now it is 
only with the apochromatic lens that I find it possible to 
resolve it into its component optical elements. 
