on Plant Growth in Nutrient Solutions, Cfc. 373 
From the data of Table VI and from the distribution of the areas 
representing high yields on the diagrams of Figs. 4, 5, and 6 it is at once 
apparent that good growth of soy bean tops in the Tottingham series E and 
F is correlated with only narrow ranges in the proportions of potassium 
nitrate and calcium nitrate, but with relatively wide ranges in the propor¬ 
tions" of the other two salts. For the cultures producing the highest five 
yields of roots in the two series, relatively wide ranges are shown in 
the proportions of all the salts except potassium nitrate. However, the 
ranges in the salt proportions correlated with high yields of both tops and 
> 
roots are always less extensive than the corresponding total ranges em¬ 
ployed, and with only one exception do these ranges include the highest 
proportions of any salt used, this exception being indicated for magnesium 
sulphate in culture T 1 R 1 C 1 , which produced a high yield of roots in series F. 
On the other hand, the ranges in the salt proportions associated with high 
yields of tops in these series include the lowest proportions of potassium 
nitrate, mono-potassium phosphate, and magnesium sulphate, and likewise 
those correlated with high yields of roots include the lowest proportions of 
potassium nitrate, mono-potassium phosphate, and calcium nitrate. 
The modified Tottingham solutions of series G which produced the 
highest five yields of both tops and roots are limited to those solutions of 
the series which comprise the lowest proportions of ammonium sulphate 
only. These solutions are further characterized by a low range in the 
proportions of mono-potassium phosphate and high ranges in the proportions 
of calcium nitrate and magnesium sulphate, the range in the proportions of 
the latter for the cultures producing the highest five yields of tops being 
co-extensive with the corresponding total range for the entire series. 
Comparing now the absolute dry-weight yields of series E with those 
of series F as given in Table VI, it will be observed that the maximum 
yield of tops from the series in which iron was supplied to the cultures in the 
form of ferrous sulphate (series F) is slightly superior to the corresponding 
yield from series E, in which the source of iron was ferric phosphate. The 
difference in these values, however, is scarcely large enough to be significant, 
although the average of the highest five yields from series F shows 
a pronounced superiority over the corresponding average from series E. 
The average of the highest five yields of roots, as well as the maximum 
yield from series F, show a marked superiority over the corresponding yield 
from series E. It thus appears that ferrous sulphate when supplied in the 
proper concentrations is a better source of iron for soy bean plants grown in 
the Tottingham solutions than is ferric phosphate under similar conditions. 
The maximum yield and the average of the highest five yields of tops 
from series G are approximately the same in value as the corresponding 
yields from series F and slightly superior to those from series E, but 
the maximum and average yields of roots from series G are much superior 
