564 The Philippine Journal of Science 1920 
groups here primarily include ten cultures each, instead of 
twenty-one as in the other case. The relative average yields 
simultaneously obtained with Shive’s and with Tottingham’s 
best solutions are again given at the bottom of the table. The 
data of Table 7 are shown diagrammatically in fig. 4 (tops) 
and fig. 6 (roots), in which the method of plotting is the same 
as that for series I (figs. 1 and 3.) 
Water absorption.—The data of the total water absorption 
for each of the cultures in this series are shown in Table 8, 
where arrangement and notation are the same as heretofore 
followed, 
Water requirement.—The relative average values for the 
amounts of water absorbed per unit of dry weight of tops and 
of roots are also presented in Table 8. 
DISCUSSION OF SERIES II 
Appearance of plants.—The appearance of the plants and the 
occurrence of magnesium injury were practically the same as 
in series I, and these features need not be discussed in detail. 
Dry weights.—The distribution of the high and low values of 
top yields on the diagram for this series (fig. 4) agrees in 
general with the distribution of these values for series I (fig. 
1). The highest yields of tops for series II was 0.804 gram for 
culture T2R4$C1%, which had virtually the same salt propor- 
tions as the one giving the highest yield in series I (T2R4C2). 
Culture TIR1C1 gave the lowest top yield in series I and in 
both the duplicates of series II. The lowest yields for the two 
duplicates of series II were 0.573 and 0.529 gram (an average 
of 69 per cent of the highest), while the corresponding lowest 
yield for series I (carried out earlier in the winter) was 0.457 
gram. In a similar way, the two duplicates giving the highest 
yield in series II (culture T2R42C12) were 0.804 and 0.803 
gram, while the highest yield from series I (culture T2R4C2) 
was 0.761 gram. This suggests that the aérial conditions for 
the two series were such as to give a somewhat higher yield 
of tops in the second series. The average top yield for four 
parallel cultures of Shive’s best solution was 0.801 gram, prac- 
tically the same as the best yield from the series with chloride. 
This value was also markedly higher than the corresponding 
value (0.684 gram) for the Shive solution in series I. A sim- 
ilar, though less pronounced, difference between the average top 
yields for Tottingham’s solution in series I (0.714 gram) and 
in series II (0.751 gram) may be noticed. As far as these 
