286 HARRIS, GORTNER, AND LAWRENCE: 
which should equal unity within the limits of experimental error 
aif increase in osmotic concentration were an adjustment to hydrostatic 
head alone, varies widely, with the average value of 1.508. This 
excess might be expected from the resistance due to the conducting 
tracts. The closeness of agreement of theory and observation 
must depend upon the value found for R, a value upon which 
experimenters are not at all in accord. 3 
We must, however, point out that an agreement between the 
observed increments in osmotic concentration and the theoretical © 
values calculated from the hydrostatic head and the resistance 
of the conducting elements cannot be taken as proof of an adjust- 
ment on the part of the cells to the back pull due to these factors. 
Such agreement might be purely accidental. Until the possible 
external factors to which increase in osmotic concentration at 
higher levels might be due are eliminated by experimental studies, 
any such agreement should be regarded as a coincidence. Meas- 
urements of the extent of differentiation in evaporating power of 
the air and in the intensity of illumination to which leaves growing 
at various heights are subjected are urgently needed. 
In emphasizing the fact that the results of this study neither 
substantiate nor disprove the assumption that the back pull due 
to hydrostatic head and to the resistance of the tracts stand ina 
causal relation to the increase in osmotic concentration, we must 
point out that if osmotic concentration of the leaf sap be a factor 
of importance in the rise of the transpiration stream, the increase 
in the osmotic concentration with height of insertion may bear 
a very significant causal relation to the ascent of sap in trees. 
The fact that the relative concentration of electrolytes de- 
creases from lower to higher levels would indicate that the dif- 
ferences are due to increased photosynthesis in the upper regions 
of the tree rather than to the concentration of salts from the soil 
solution by increased transpiration. 
In conclusion, it is perhaps hardly necessary to suggest the 
desirability of correlating differences already known to be asso- 
ciated with height in trees—for example, heavier fruiting or the 
development of the so-called shade leaves—with sap properties. 
STATION FOR EXPERIMENTAL EVOLUTION, 
CoLp SprinG Harsor; New YorxK 
