16, 5 Espino: Salt Requirements of Young Rice Plants 487 
There was generally good agreement in the appearance of the 
six plants of any culture. Occasionally a seedling died, but this 
was rare. For such cases the quantitative data have been com- 
puted to represent six plants, although actually based on only 
five. 
As to the different sets of salt proportions, the appearance of 
the plants became progressively poorer (especially with refer- 
ence to chlorosis and size) with higher relative partial concen- 
trations of monopotassium phosphate; that is, as the top apex 
of the tetrahedral diagram is approached, the plants are shown 
to have become more and more stunted and chlorotic. The best- 
appearing plants represent sets of salt proportions given in the 
lowest or the lowest two triangles of the diagram and the poorest 
plants represent the sets at or near the top apex. 
Dry yield of tops.—The dry yield of tops from series 14 to 
20 is shown in Table 5, all values being expressed in terms of the 
corresponding value for culture T1R1S1 of the same series. 
The actual value (fraction of a gram) for the last-named culture 
is given in parentheses beneath the value 1.00, in each case. 
Values marked H belong to the highest seven values obtained 
in the given series, those marked L belong to the lowest seven. 
These two groups of seven values (there are generally more than 
Seven cultures included, on account of like values from two or 
more of them) will be called the high and the low groups. The 
date of beginning of each series is given beneath the series 
number. 
Inspection of Table 5 brings out the fact that, where repeti- 
tions were made there are frequent discrepancies, a given solu- 
tion giving a high relative value for one trial and a medium 
or low value for another trial. These discrepancies are attrib- 
utable to unknown conditions (perhaps largely to the non- 
Solution conditions called climatic). It will be remembered that 
these different trials were made at different seasons of the year, 
as indicated by the dates. It has been emphasized by other 
writers that the relative value of any solution must be expected 
to be different for different sets of non-solution conditions, pro- 
Viding, of course, that the latter differences are large enough 
to be effective. Whatever may have been the conditional dif- 
ferences that were at work in these cases of discrepancy, it is 
clear that comparisons between the physiological properties of 
the solutions here considered (as to whether they gave high, 
medium, or low yields of tops) must be made only in a broad 
and general way. It is to be supposed that, had series 14 and 
T7114 
