516 
Philippine Journal of Science 
1920 
mental data are so limited that detailed discussion of the dif- 
ferences is not warranted. It is clear, from the yield data, that 
the three sets of salt proportions selected as most generally prom- 
ising are shown as most promising by the water-absorption 
data, as well as by the data for green weight. Also the two sets 
of salt proportions selected to represent badly balanced cultures 
are shown as belonging to the low group in fig. 9. 
FOUR-SALT TYPE B (MONOPOTASSIUM PHOSPHATE, MONOCALCIUM PHOSPHATE, 
MAGNESIUM SULPHATE, AMMONIUM SULPHATE) 
As has been mentioned, the tests with the four-salt solution 
type B were carried out to determine whether the nitrate ion in 
the solutions of type A might be unnecessary, or even harmful. 
The lowest section of Table 3 shows the six experimental series 
that were carried out with this type of solution, three total con- 
centrations (0.0008, 0.0016, and 0.0038 gram-molecule per liter, 
or 0.04, 0.08, and 0.2 atmosphere) being simultaneously employed 
for the period beginning August 3 (series 23, 25, and 27), and 
again for that beginning September 9 (series 24, 26, and 28) . 
The better cultures of each series showed very satisfactory 
growth, excepting that the plants were always markedly affected 
in the latter part of the three-week period by a characteristic 
drying of the leaves backward from the tips. This was so pro- 
nounced and so characteristic of these type B solutions that it 
must be regarded as a serious form of physiological injury. The 
dry yields of tops and of roots gave values not markedly lower 
than the corresponding ones obtained with solution type A, but 
the presence of the injury just mentioned made it impossible 
to regard even the best sets of salt proportions of type B as in 
any way approaching physiological equality with the best ones 
of type A. For this reason the data of series 23 to 28 will not 
be presented in the present paper. It may be noted further, 
however, that the highest total concentration here used (corre- 
sponding to the high-medium of solution type A) proved to be 
too high for the best growth, and that the lowest one here used 
seemed to be somewhat too low. The most promising solutions 
of this type had what was termed in the preceding discussions 
the low-medium total concentration (0.0016 gram-molecule per 
liter, or 0.08 atmosphere) . It will be recalled that this total con- 
centration seemed to be somewhat more promising than the 
others for type A. For type B, however, emphasis appears to 
be placed on the lowest and the low-medium total concentration 
rather than on the low-medium and the high-medium. 
On account of the pronounced leaf injury present even in the 
