560 The Philippine Journal of Science 1920 
The chemical composition of each of the forty solutions in 
series II is shown in Table 5, which corresponds in arrange- 
ment to Table 1. The calculation for this series followed the 
general method employed in series I and resulted in the partial 
molecular .concentrations given in Table 6, which were used 
directly in the preparation of Table 5. 
TABLE 6.—Partial concentration of each of the four salts required to pro- 
duce from 0.1 to 0.7 of the total osmotic concentration of 1.60 atmos- 
pheres; for series II. - 
Fractional Salt. 
parts of 
1.60 atmos- 
pheres. KCl. KHePO.. | Ca(NOs)z. | MgSO,, 
M. M. M. M. 
0.1 0, 0033 0. 0083 0. 0023 0. 0088 
UND Ee hecoer area 0. 0045 0.0081 0. 0052 
Geddy. 0. 0056 0. 0039 0. 0067 
0.2 0. 0067 0. 0068 0.0047 0, 0081 
ese 0. 0080 0. 0055 0. 0096 
O2E | 0.0091 0. 0064 0.0111 
0.3 0.0101 0. 0108 0. 0072 0.0126 
WOR co 0. 0121 0. 0085 0, 0149 
Cee ee 0.0126 0.0089 0.0156 
0.4 0.0135 0.0138 0. 0098 0.0171 
Va! Bae ae 0.0161 0.0115 0.0201 
0.5 0.0170 0.0173 0.0124 0, 0216 
0.6 0. 0205 0.0208 0. 0150 0. 0261 
0.7 0.0240 0. 0248 0.0177 0. 0309 
RESULTS OF SERIES H 
Appearance of plants.—Root and top development took place 
in a way very similar to that described for the preceding series. 
The various differences in the appearance of the plants described 
as occurring in series I were observed in the present series to 
about the same degree with reference to the range of salt pro- 
portions. The distribution of the two forms of magnesium in- 
jury in this series is shown by the diagrams of fig. 5, which 
is to be interpreted like fig. 2. 
Dry weights—The dry weight data for this series are shown 
in Table 7, in which the various items are arranged as in Table 
4, excepting that in the present case the actual values corre- 
sponding to the two duplicate cultures are given, followed by 
the relative average. Thus these relative average weights cor- 
respond to the relative weights given in Table 4. The letters 
H and L are used in the same way as in the preceding series, 
but it will of course be noted that the “high” and the “low” 
