272 HARRIS, GORTNER, AND LAWRENCE: 
increase were an adjustment on the part of the cells to hydrostatic 
head and to increased resistance of conducting tracts. 
Freezing point lowering furnishes a measure of the concentra- 
tion in moles and ions of all the solutes, non-electrolytes and non- 
dissociated and dissociated electrolytes. Specific electrical con- 
ductivity furnishes a measure of the concentration of dissociated 
electrolytes only. 
It would be highly important to differentiate the electrolytic 
and the non-electrolytic elements in the solutes to which the whole 
of the concentration in moles and ions is due. Such differentiation 
presents, however, in the case of a mixed solution of non-electro- 
lytes and of electrolytes composed of inorganic salts and of weakly 
dissociable organic salts and acids, a problem of great difficulty. 
We have therefore contented ourselves with the calculation of the 
ratio «/A, which is useful although obviously only a substitute 
for more refined analysis. 
Before proceeding to the analysis of the results, a word should 
be said concerning the trustworthiness of the constants. 
There are two sources of error: the probable errors of random 
sampling in the collection of the leaves and the technical errors 
in the extraction of sap and in the determination of the constants. 
The determination of the actual magnitude of these errors presents 
a problem of considerable difficulty, and we are not in a position 
to say exactly how large they may be. It is quite clear from the 
consistency of our results that they are not so great as to obscure 
the general law, the existence or the non-existence of which we 
have been seeking to demonstrate. 
We do not, however, lay emphasis upon any single constant 
or difference, and we do not attempt to explain individual incon- 
sistencies. In the rather onerous task of collecting material for 
and carrying out the determination of 72 freezing points and 66 
conductivities, it is highly improbable that we should have been 
able to avoid entirely probable errors of random sampling in the 
collection of the tissues and errors of laboratory technique at least 
as great and sometimes greater than the actual differences of the 
kind which we have been seeking to demonstrate, if existent. 
Thus, even if osmotic concentration increases with height of 
insertion, one should expect to find a negative relationship indi- 
