16, 5 Espino: Salt Requirements of Young Rice Plants 493 
the two low-medium series and between the three high-medium 
series, also between the low-medium and the high-medium con- 
centrations, as indicated by the groups (high, medium, and low) 
in which the various sets of salt proportions fall. 
These agreements and disagreements are set forth in Table 8, 
which also includes the corresponding data for root yields. In 
this table the letters H, M, and L are employed to show in which 
group any given set of salt proportions fell for each trial. In 
the case of the lower of these two medium total concentrations 
there were two trials, and disagreement between the two is 
shown by the use of both letters (thus, HM denotes that one 
trial showed the solution in question as in the high group, while 
the other trial showed it as in the medium group). In the case 
of the higher total concentration there were three trials. When 
all were different in their indications all three letters are given 
in the table; when two of them agreed but the third was dif- 
ferent, only two letters appear, but one is lower-case, to denote 
that only one trial is thus referred to. Thus, if the three trials 
call for L, L, and M respectively, the designation is Lm, etc. 
Table 8 shows that solutions T1R1S1, T1R1S2, and T1R2S1 
agree in giving high top yields, while solutions T2R2S3, T3R183, 
and T4R1S2 agree in giving low top yields, for both total con- 
centrations. The other solutions show medium top yields or are 
more or less questionable, as far as the data go. It seems safe 
to conclude that the first three solutions just named may be 
expected to produce excellent top growth for this sort of plants, 
when the total concentration is between 0.0016 and 0.0038 gram- 
molecule per liter (0.08-0.2 atmosphere), while the last three 
mentioned may be taken as representing the poorly balanced sets 
of salt proportions of these series. The distribution of these 
and the other solutions on the tetrahedral diagram, together with 
the chemical interpretation of this, will be deferred till after 
the root yields have been presented. 
Dry yields of roots.—The relative dry yields of roots from 
Series 14 to 20 are presented in Table 9, in a manner quite like 
that followed for the top yields in Table 5. 
From Table 9 it appears that disagreements are fully as fre- 
quent among the root yields as among those of tops, and the 
remarks made concerning the discrepancies in the other case 
will apply here also. 
It appears that solution T1R1S1 and T2R1S1 agreed in giving 
high root yields for all series, and that no one of the thirty-five 
sets of salt proportions gave low root yields for all series. 
