258 
AMERICAN JOURNAL OF BOTANY 
[Vol. 8 
is, Knop's solution — modified by the addition of o.i gram of sodium chloride 
per liter of solution, and diluted to one tenth the concentration usually 
listed (table i); and of five further modified culture solutions in each of 
which one of the elements magnesium, calcium, potassium, phosphorus, and 
nitrogen was reduced to one tenth of the quantity present in the normal 
solution. 
Table i. The Composition of the Modified Knop's Solution Used as the Normal Culture 
Solution 
Concentration of Salts Used in " Normal " Solution 
Salts Grams per Liter of Solution Percentage Composition 
Ca(N03)2 0.08 0.0533 
KNO3 0.02 0.0133 
KH2PO4 0.02 0.0133 
MgS04 0.02 0.0133 
NaCl O.OI 0.0067 
Total salt 0.15 o. 1 000 
Although these deficient elements were reduced to such an extent that 
they seriously hindered the development of the plant, yet they did not in 
any case prevent the production of grain. In other words, the various 
nutritive elements were reduced to about the lowest concentration that 
would still support the production of grain by the plants growing therein. 
It was necessary, of course, to base the concentration of the element reduced 
in each solution on the constituents in greatest demand by the plant, namely, 
phosphorus and nitrogen, and then to prepare all of the solutions in which 
one of the essential nutritive elements is deficient with the same concen- 
tration of the deficient element. By reducing the amount of the various 
nutritive elements below that required for normal development, it was 
hoped to bring out more strikingly the relation of the elements to the 
composition of the plant. 
The various cultural solutions were calculated and prepared upon the 
basis of equal osmotic concentration by varying to a slight degree the amount 
of sodium chloride added to the respective solutions. To state it differently, 
the required nutritive salts were added to the solution and their respective 
osmotic activity was calculated from the dissociation of each constituent 
salt, then sufficient sodium chloride was added — never enough to exert a 
toxic influence — to make the total osmotic concentration equal to that of 
the ''normal" solution. It was not found practicable to take into account 
the effect of one constituent salt upon the dissociation of another in the 
culture solution, but the degree of dissociation of the various solutions was 
checked by the soluble salts present, by measurements of electrical conduc- 
tivity, and by freezing-point determinations. 
The quantity of sodium chloride used in balancing the modified nutritive 
solutions was in no case greater than that added to the normal solution, 
that is, 0.0067 percent of the total salts added (table i). Therefore, the 
presence of this salt should have no appreciable difference in effect upon the 
