592 The Philippine Journal of Science 1920 
very different in form from those just considered. This graph 
exhibits a marked maximum at 4.50 atmospheres, instead of at 
1.60 or 2.50 atmospheres. The fourth graph, that for the 
T1R1C1 cultures, having 0.7 of their total concentration due to 
magnesium sulphate and 0.1 due to each of the other three salts, 
resembles the first two graphs much more closely than it does 
the third. With these solutions having very high proportions 
of magnesium sulphate, the optimum concentration was 1.60 
atmospheres, and the yields decreased rather slowly with increas- 
ing concentrations above that value. Very high total concentra- 
tions retarded root production much less when the solution had 
high proportions of potassium chloride than when it had high 
proportions of magnesium sulphate. 
This series furnishes very conclusive proof of the point already 
brought out by Shive and Tottingham ; namely, that the optimum 
total concentration may be expected to vary with the sets of 
salt proportions used, and that for any total concentration an 
optimal set of salt proportions may be found. There appears 
to be no such thing as an optimal total concentration without 
reference to the salts and salt proportions used, and there is 
no optimal set of salt proportions except for some specific range 
of total concentrations. 
Water absorption . — The most striking feature of the graphs 
shown in fig. 11 is the nearly linear decrease in total water ab- 
sorption with an increase in concentration between the limits 
1.60 and 7.00 atmospheres. It will be observed that the graph 
for the T2R4C2 cultures tends to parallel and lie slightly above 
that for the R5C24 cultures. The graph for the T7R1C1 cul- 
tures (which have 0.7 of their concentration due to potassium 
chloride) lies considerably below these two and tends to parallel 
but remain above that for the T1R1C1 cultures (having 0.7 of 
their concentration due to magnesium sulphate). 
The graphs for the T2R4C2 and the R5C2^ cultures are ir- 
regular below 1.60 atmospheres. The R5C24 cultures, it will be 
remembered, have no potassium chloride but otherwise are ap- 
proximately the same in composition as the T2R4C2 ones. It is 
impossible to determine from these graphs what concentration 
may be expected to give maximum water absorption, because the 
values for 0.50, 1.00, and 1.60 atmospheres are so nearly alike. 
Eut a markedly lower absorption is evident for concentrations 
higher than 1.60 atmospheres. Both graphs show an approxi- 
mately linear decrease in water absorption with an increase in 
