92 
Journal of Agricultural Research 
Vol. XXIV, No. r 
The second series was in most respects a repetition of the first, except 
that the lead arsenates were dry and used in slightly larger quantities 
per sample. The sodium stearate soap was also almost dry in this case, 
but the quantity of soaps used (dry basis) was the same as before. 
The samples, treatment, and soluble arsenic found are given in Table II. 
The grinding in the water or in the soap solution was perhaps more 
thorough than in the first series, but the shaking was continued for only 
four days. 
It will be noted that the pair of tests 5 and 6, in Table I, are not in 
agreement with other pairs of similar tests. The general tendency is, 
the greater the concentration of soap, the greater the quantity of soluble 
arsenic found, but in this one pair this tendency seems to be reversed. 
It is believed this indicates some mistake, possibly in the numbering of 
the bottles, or of the record. The corresponding pair in Table II, No. 
21 and 22, is in accord with the tendency noted. 
EFFECT OF THE LIQUID USED FOR GRINDING 
Comparison of the quantity of arsenic made soluble in pairs of tests 
otherwise identical, but in one of which the grinding was in the presence 
of pure water and the other in the presence of soap solution, shows that 
the quantity of arsenic made soluble is seldom the same. It frequently 
yaries by several per cent of the arsenic present, but does not always 
vary in the same direction. The cause of this lack of uniformity was 
not determined, but conceivably might be as follows: The shells of lead 
soap formed about the small solid particles of lead arsenate are some¬ 
what plastic and the real effect of the grinding might be a “churning'' 
one that would serve to stick some of the particles together and increase 
the protective action of the shells of lead soap. In other cases the grind¬ 
ing might remove the coating of lead soap, as was the intent when the 
work was done. 
SOLUBLE ARSENIC DUE TO ACTION OF FREE ALKALI DEDUCTED 
Since the free alkali in the sodium stearate soap would probably exert 
some solvent action, it is worth while to compute how much of the solu¬ 
ble arsenic found might be due to that cause. Starting with the reaction 
Na^COg + PbHAsO, = PbCOg + NagHAsO^, 
the highest ratio possible between sodium carbonate and soluble arsenic 
pentoxid is i to 1.085. On this basis Table III is computed from the 
data already given, showing the quantity of sodium carbonate present, 
the equivalent quantity of ASgOg that might be made soluble, the quan¬ 
tity of soluble arsenic as actually found, and, by difference, the quantity 
that must be made soluble by the neutral soap, the quantity of AsgOg 
present in the lead arsenate used, and the percentage made soluble by 
the action of the soap aside from the free alkali. 
Since the oleic soap contained no free alkah, no similar table is pre¬ 
sented for correction of the results on oleic soap in Tables I and II. 
