618 Philippine Journal of Science 1920 
comparison. This graph shows a strong tendency to fall rather 
uniformly in passing from left to right. 
A generalized graph representing the downward slope of this 
graph might be shown by a straight line drawn from the value 
of 100, for the culture having 0.8 of its fertilizer addition due 
to ammonium sulphate, to a value of about 24, for the group 
of cultures at the right having 0.1 due to ammonium sulphate. 
It will be noted that the control culture, having no fertilizer, 
has a still lower value (18). Thus, with decreasing proportions 
of ammonium sulphate in the fertilizer mixture there was a 
general tendency for the yield of grain to decrease, which 
strongly suggests that the yield of grain, under the conditions 
of this experiment, was approximately proportional to the am- 
monium (nitrogen) content of the fertilizer mixture applied to 
the soil. This graph exhibits irregularities or fluctuations, 
which are in part due, no doubt, to unknown or uncontrolled 
conditions (such as initial internal differences in the plants 
used), and also in part due to the fact that the suggested 
proportionality between yield and ammonium sulphate content 
of the mixture does not hold rigidly—the proportions of the other 
salts also influencing the yield. It must be added that still 
higher yields may be expected when tests are made with salt 
mixtures containing still greater amounts of ammonium sul- 
phate than any used in the present series. The results of a 
test with a salt mixture having approximately the proportions 
of culture R1C8 have indicated that increase in yield accom- 
panies an increase in the concentration of total salt added, until 
the latter has an osmotic value slightly higher than 2 atmos- 
pheres. Thus, increase.in yield may be expected to follow an 
increase in the ammonium (nitrogen) content of the salt mix- 
ture up to more than four times the amount used in the present 
tests, 
The general proportionality between yields and the ammo- 
nium (nitrogen) content of the fertilizer is of course not sur- 
prising, since it has long been known that the amount of plant 
growth may usually be greatly increased by means of proper 
nitrogenous fertilizers. Russell,(12) page 32, in discussing the 
effects of nitrogen on cereal crops, states that the increasing 
effects produced, up to a certain point, by successive increments 
of suitable nitrogenous compounds may be due to the circum- 
stance that the additional nitrogen not only increases the con- 
centration of nitrogenous compounds in the soil, but also in- 
creases the amount of root (the absorbing surface) and of leaf 
