16, 6 Trelease: The Growth of Rice 619 
(the photosynthesizing surface). The process thus resembles 
autocatalysis, where one of the products of the reaction acts 
as a catalyzer and hastens the reaction; but the increase does 
not go on indefinitely, because some other condition checks it. 
He further states that the first addition of nitrate causes a 
marked rise in the weight of the grain and the proportion of 
grain to the total product, but that successive additions show 
no further rise, and that an excess of nitrogenous fertilizer 
causes the proportion of grain to decrease, cereal crops produc- 
ing a very high proportion of straw if the nitrogen supply is 
excessive. As will become evident from a discussion of the 
comparative plant measurements, in the next following section 
of this paper, that feature was not observed with the concen- 
trations of ammonium sulphate employed in the study upon rice 
here reported. 
The middle graph, B, of fig, 4 should indicate whether or 
not a correlation exists between the proportion of potassium 
_ sulphate and the yield of grain. The arrangement of this graph 
is similar to that of graph A, except that the cultures in the 
middle graph are arranged as abscissas in the order of the 
magnitude of their potassium sulphate content. This graph is 
irregular and fluctuates most widely in its latter half, where 
the yield values with each amount of potassium sulphate vary 
from the highest to approximately the lowest value of grain 
yield. The character of this graph indicates that, with the 
soil, the plants, and the general conditions of the present experi- 
ment, there was little or no general relation between the potas- 
sium sulphate content of the fertilizer mixture and the yield of 
grain. The only suggestion of a general trend in this graph 
is an upward slope in passing from left to right. This might 
suggest an inverse relationship between the potassium sulphate 
content of the fertilizer and the yield. But, if the conclusion 
reached regarding the rough proportionality between yield and 
ammonium (nitrogen) supply be true, then this indicated up- 
ward slope, in passing from left to right, may be explained by 
the distribution of the salt proportions tested. It will be ob- 
served that, with 0.8 of the salt mixture due to potassium sul- 
phate, only 0.1 was due to ammonium sulphate; while with lower 
proportions of potassium sulphate much higher proportions of 
the ammonium (nitrogen) salt were used. Thus, the average 
content of the ammonium salt increases in passing to the right 
from one group of cultures to the next. If the yield is roughly 
proportional to the ammonium (nitrogen) content of the mix- 
171585——_6 
