80 
Walter Gardine«. 
seems lo me difficult to imagine hy Avliat means a drawiug-out of fresli 
substanee from the main protoplasmic mass is occasioned. My view is as 
follows. When the cell is acted upon by the salt solution, the contraction of 
the primordial utricle which takes place, is caused by the fact that there 
is a rapid diffusion of the less concentrated cell-sap of the vacuole into 
the more concentrated salt solution. Now the water passes from the vacuole 
into the salt solution much more quickly than does the salt solution into 
the vacuole. Finally the protoplasm becomes for the time abnormally 
shrunken in consecjuence of the rapid pragress of the dehydration and while 
in this state the Strands connecting it lo the cell-wall, will be at their 
maximum degree of tenseness, and will be draw n out to such a degree of 
lenuity as to be invisible. But after a time when the diffusion is beginning 
lo cease, and llie entire solution lends to assume a uniform specific gravity, 
a certain quantity of the salt solution will pass back into the shrunken proto¬ 
plasm to supply the place of the water which it had so violently abstracted: 
the vvhole body will swell, the Strands will become less tense and at the 
same time will stricken, and in so doing will gradually come into view. 
Finally owing to the further expansion on the part of the protoplasmic body 
they will become so slack as to admit even of lateral vibratory movement. 
ft might be supposed that these Strands of altachment, as wo may torm 
them, which are thus brought into view in plasmolysed cells are held in 
Position, in consequeuce of their being connected, in the mode described 
above, with similar Stands in neighbouring cells. This is doubtless true in 
some eases, but by no means in all: for the Strands of altachment in the 
case of any one pit are more numerous than the filaments actually perfora- 
ling the closing membrane, and in the case of an unpittcd cell-wall like that 
of Tamus they far exceed in number the filaments which actually traverse 
the wall. 
It is obvious that in Plasmolysis the perforating filaments in the cell- 
wall, or pit, will tend to be pulled out of their channels and there is no 
reason why the Strand coming from such a lilament should appear different 
from those coming from the general and apparently imperforate cell-wall. 
At any rate a careful examination of the plasmolysed cells of many endo- 
sperms, and other like tissues, gave no clue as to there being any discern- 
able difference between the threads. 
In the coarser Plasmolysis induced by the action of powerful reagents 
the protoplasm is soon coagulated, and killed, and hence the assumption by 
the contracting protoplasm, of the roundcd form, and the formation of Strands 
of altachment between protoplasm and cell-wall is prevented. The proto¬ 
plasm now remains in a passive condition and is mechanically held to the 
wall at those points where the most inlimate relation between the two 
exists.- Such shrinking as occurs is the expression of the rapid dehydration 
of the vacuole, but only those portions of the protoplasm contract which are 
