i8 
Walter Stiles 
even necessary that an insoluble compound should be formed so long 
as the new combination of substances in the vacuole does not exert 
any higher osmotic pressure than previously. That such possibilities 
are by no means to be ignored is made^yery clear from the work of 
Pfeffer on the penetration of dyes described earlier in this chapter. 
Similarly, Stiles and Jprgensen (1917 b) have shown that storage 
tissues actually plasmolyse in solutions of common salt and remain 
so for 20 hours or more, while Stiles and Kidd (1919 a, h) have shown 
that sodium chloride is actually absorbed with some rapidity by such 
tissues. 
But if such complication does not enter into the case, the pene¬ 
tration of a substance into the cell will raise the osmotic concentration 
of the cell sap, and hence the external solution will no longer be able 
to effect plasmolysis. In order to bring this about a higher concen¬ 
tration of the solution will be necessary, and even then, if the entrance 
of solute continues the cell will recover from plasmolysis. 
If then the osmotic concentration of a substance required to 
produce plasmolysis is higher than that required in the case of a 
substance which is known not to penetrate the cell, it may be con¬ 
cluded that the former substance penetrates the cell. 
Sucrose is generally chosen as a substance which penetrates the 
cell extremely slowly or not at all, and it is fortunate that the values 
of the osmotic pressure of this substance are more firmly established 
than those of other substances. The concentrations of other substances 
isotonic with sucrose solutions have been determined chiefly by in¬ 
direct methods, such as calculations from the freezing point lowering, 
or from determinations of the electrical conductivity, as very few 
direct determinations of osmotic pressure of substances other than 
sucrose have been made (cf. Chapter VI). If the isotonic coefficient 
found for a substance by the plasmolytic method is lower than that 
found by direct measurement or by calculation from other physico¬ 
chemical data, it is concluded that the cells used for the determination 
of the coefficient are permeable to the substance examined. 
In this way de Vries (1888 a, b,c) showed that to produce plasmo¬ 
lysis of the cells of Spirogyra nitida a solution of glycerol was necessary 
of considerably higher concentration than that isotonic with a solu¬ 
tion of sucrose which would produce plasmolysis. This method was 
later extended to a large number of different cells, and it was found 
that glycerol is very generally permeable to plant cells, although 
there are exceptions. Thus de Vries showed that the cells of the bud 
scales of Begonia manicata are impermeable to glycerol and also to 
