394 HooKER: MOVEMENT IN DROSERA ROTUNDIFOLIA 
They contract at each end so that they have the appearance of 
dumb-bells, and finally separate into several disk-shaped parts. 
In extreme cases they separate into a large number of droplets 
which exhibit Brownian movement. No marked difference was 
observed in the size of the rabdoids on opposite sides of the tentacle 
in the bending region, and no alteration was visible during move- 
ment. The function of the rabdoids is not apparent. It seems 
probable, however, that they are connected in some way with the 
process of secretion, rather than with the mechanics of movement. 
Gardiner (’85) suggested that they might be reserve material or 
some substance used up during secretion. This seems to be borne 
out by the fact that they are of protein composition according to 
Tunmann (’13, p. 481). 
3. UNBENDING:*BY PLASMOLYSIS 
Although a two per cent. solution of potassium nitrate usually 
plasmolyzes the cells on the convex side of a bending tentacle, this 
does not cause the tentacle to become straight (see De Vries, ’86, 
p. 5). More concentrated solutions produce unbending. For a 
short time after the tentacle is fully bent complete plasmolysis 
still causes unbending, but when the tentacle has been bent a 
considerable period and during unbending, plasmolysis does not 
alter its shape. 
The xylene experiment which W. H. Brown (12, ’16) made on 
Dionaea and Mimosa was tried on Drosera tentacles, but no positive 
results were obtained. Bent and bending tentacles were killed 
in boiling water, treated with 95 per cent. alcohol, absolute alcohol 
and xylene. No unbending ensued. This may indicate a dif- 
ference between the mechanics of movement in Drosera and in 
such plants as Dionaea and Mimosa. However, the failure of the 
experiment may have been owing to faulty technique, although 
the description of the process by Brown (’16, p. 78) was followed 
in detail. 
III. DISCUSSION 
I. INCREASED TURGIDITY AS THE MEANS OF MOVEMENT 
The straightening of bent tentacles by plasmolysis shows the 
cell elongation, which is the immediate cause of bending, to be a 
