596 The Philippine Journal of Science 1920 
especially with reference to root yields, by the addition to the 
solution of a proper amount of potassium chloride. 
Since the wheat plant is not very sensitive to changes in 
concentration of potassium chloride, the osmotic value of the 
solution must be changed when potassium chloride is added 
in large enough amounts to cause a physiological response. If 
the nutrient solution is to remain the same in total osmotic value, 
the concentrations of the other three salts must be decreased 
when potassium chloride is added to give high partial concen- 
trations. It is impossible to decide whether such plant responses 
as are thus obtained are due to the high partial concentration 
of the chloride or to the low partial concentrations of the three 
other salts. This dilemma was always encountered in seeking to 
obtain evidence on the points just mentioned, and this whole 
general problem is too complicated to treat experimentally. 
It is also difficult to interpret the results in such a way as 
to determine the action of the chlorine ion (Cl), as such. 
The problem is complicated by the fact that additions of the 
chlorine ion to the nutrient solution were always accompanied 
by equal additions of the potassium ion. It is possible that such 
additions of the potassium ion may have modified any effects 
of the chlorine ion that might have been observed under other 
conditions. 
Data on root development are never so easily interpreted as 
are those on the development of tops in such plants as wheat, 
a difficulty experienced also in the present study. But effects 
upon the root system are of especial physiological importance, 
because this is the part of the plant that is in immediate contact 
with the culture solution, and it is through the roots that all the 
water and salts absorbed by the plant must enter. It appears 
to have been true, in a general way at least, that those salt pro- 
portions of the four-salt solution with chloride that gave high top 
yields also gave high root yields. Conversely, salt proportions 
giving low top yields generally gave low root yields also. This 
statement is the opposite of the one given by Shive, based upon 
his studies with the three-salt solutions. It is possible that the 
addition of the chlorine ion to the ones already present in the 
three-salt solution, or some special influence of potassium chlo- 
ride, may account for this difference. The general indefiniteness, 
however, of the root data makes superfluous a further discussion 
of this point without further experimentation. 
Certain of the sets of salt proportions with chloride produced 
a characteristic form of leaf injury, which was observed in 
