308 BOTANICAL GAZETTE [ Novemper 
an absolute limit to the production of filaments was not found. 
A comparison of the limit found here with that found in the 
case of palmella material will be made later. Filaments floating 
on a strong solution change to the palmella form much more 
slowly than those at the bottom of the dish. 
The morphological details of the response just discussed are 
as follows. If a filament be placed in a strong solution the 
oldest cells are the first to be visibly affected. They continue 
rounding up beyond the barrel-shaped stage and become nearly 
or quite spherical. This process is accompanied necessarily by 
a splitting of the common walls from their margins inward, until, 
when the cells have attained the spherical form, they are practi- 
cally free from one another (figs. 4,5). They may not break 
entirely apart, however, but may remain clinging together in 
loose strands. The limitation upon the direction of new walls 
(cf. discussion of the responses of palmella form, /. 300) is nie 
apparently removed, and the cells proceed to divide in all direc- 
tions. The new cells may separate entirely (jig. ro), may 
remain loosely attached to one another (jig. 34), or may form 
an irregular parenchyma-like colony (jigs. 6, rr). Inthe latter 
case the rounding is of course incomplete, but such cells are 
always nearly isodiametric. Cell division goes on much more 
slowly in the palmella form than in the filamentous. y 
: ; : e hee. 
3. Response in reproduction.—The response is the sam a 
as in the other forms and is equally marked. The 
somewhat above 964 NV x 10%. The limit to their 
into long filaments is below 482 V X 10%. 
duced here in the way already described. 
filament, even in a weak solution, is limited by 
oldest cells are continually becoming zoosporang!@ - 
losing their contents (fig. 16). 
d_ thus 
General considerations. - 
ertaken to ansW je 
The question which this research was und 
has been answered very clearly as far as the a 
