568 
S. C. BROOKS 
peduncle, to make it ordinarily impossible to construct a composite 
curve which should fairly represent the characteristic effect of the 
salt. Therefore experiments have been selected which represent as 
nearly as possible the mean of all the experiments with the same salt, 
and the curves representing the progressive changes in rate of recovery 
plotted in Figure 2. The original data of these experiments are given 
in Table i. 
The difference in the behavior of the three groups of kations em- 
ployed is very striking. The curves for sodium, potassium and am- 
monium salts lie everywhere above those for the mixed solution; the 
rate of recovery, and therefore the rate of penetration of the salt, is 
and remains greater than that normal for the protoplasm. The rate 
of penetration of calcium and magnesium salts is considerably, and 
that of salts of the trivalent kations (cerium and aluminium) very 
much below the normal.^ It is to be noticed that sodium, potassium, 
and ammonium salts cause a marked increase in the rate of recovery 
between the twentieth and thirtieth minutes. 
The secondary decrease, in the light of the experiments on the 
influence of salts on exosmosis, is probably to be attributed to the fact 
that the increase of permeability leads to a considerable rate of exos- 
mosis, thus retarding the increase of the intracellular osmotic pressure, 
and hence decreasing the rate of recovery. It will be seen that sodium 
and ammonium chlorides, the most toxic among the three salts of 
univalent kations, cause the most rapid secondary fall of the curve, 
while potassium nitrate, the least toxic, causes only a slight fall. This 
fact also favors the supposition that the cause of secondary retardation 
in the rate of recovery is to be sought in an increase rather than a 
decrease of permeability; but this explanation must remain purely 
hypothetical pending the accumulation of further evidence with respect 
to the phenomenon. These experiments are thus found to be in 
essential agreement with those of Osterhout in which Laminaria was 
the plant used, although the fluctuations in permeability induced by 
pure salts were greater and more rapid than those in Laminaria. The 
3 The first recovery times are, however, usually in the same sequence in a series 
of salts as are the later recoveries. 
^ These curves were probably not disturbed by differences in the acidity of the 
solutions, except possibly in the case of cerium and aluminium chlorides, whose 
hydrogen ion concentrations, as determined by the use of a hydrogen electrode, were 
3 X io~W and 4 X IQ~^M at the concentrations used. 
