30 BULLETIN 1452, U. S. DEPARTMENT OF AGRICULTURE 
The velocity of migration of the particles in distilled water appar- 
ently does not distinguish the different colloids as well as the quan- 
tity of methylene blue required to render them isoelectric. Several 
different colloids have the same migration velocities in pure water 
and the range in velocities is less wide than the range in dye values. 
However, when there is a difference in the migration velocities of 
the colloids the variations are in the same direction as variations in 
(he dye values, except in the case of the Susquehenna colloid, so 
that there is some correspondence between the two series of values. 
The milliequivalents of methylene blue which render the colloids 
isoelectric agree fairly well, except in the case of the Xorfolk and 
Susquehanna colloids, with the milliequivalents of Ca, Mg, K, and 
Xa exchanged by the colloid with a normal ammonium chloride solu- 
tion. In the case of the Xorfolk and Susquehanna colloids, the 
Ca, Mg. K, and Xa do not constitute the total quantity of exchange- 
able bases, since some H or Fe and Al is removed by the XHX1 
solution, as mentioned on page 24. A comparison of total exchange- 
able bases and methylene blue required for electrical neutralization 
shows a close agreement in the case of all these colloids, as pointed 
out in a previous publication (37-). 
The correspondence between the methylene blue values and the ex- 
changeable bases, as measured by the ammonium, chloride solution, 
is not surprising, since tests showed that the adsorption of methylene 
blue by the colloid is also an exchange adsorption. When a 10-gram 
sample of a Xorfolk colloid was treated with methylene blue. 1.32 
milliequivalents of the dye were absorbed and 1.36 milliequivalents 
of bases were displaced. A 2-gram sample of a Sharkey colloid ad- 
sorbed 1.01 milliequivalents of the dye. and 1.2 S milliequivalents of 
bases were brought into solution. 
The adsorption of methylene blue thus appears to consist in the 
replacement of equivalent quantities of other bases (Ca. Mg. K, 
and Xa with pos. ibiy some H or Fe and Al). Since the methylene 
blue required to render the colloids isoelectric was completely ad- 
sorbed and since this quantity corresponds with the exchangeable 
base content of the colloid, it is apparent that the colloids were 
rendered isoelectric when their exchangeable bases were replaced 
by the methylene blue ion. It is somewhat strange that the subtitu- 
tion of the methylyene blue ion for the exchangeable ions of the 
colloid should render the colloid electrically neutral : but according 
to the double-layer theory, as recently discussed by Michael is (39), 
this might be explained on the ground that the combination between 
colloid and methylene blue has no tendency to dissociate and form a 
double layer. This is in accord with the fact that methylene blue 
appeared to be completely adsorbed, up to the isoelectric point. 
The fact that the colloids had negative charges in water varying 
between 32. -1 and 13.7 millivolts 21 when they contained exchangeable 
Ca, Mg, etc.. but had no charge when these bases were substituted 
by the methylene blue radicle, shows that the charge of the particles 
is largely influenced by the character of the exchangeable bases. 
Further evidence that the charge on the particle is largely influenced 
by the nature of the exchangeable base was obtained by replacing 
the exchangeable cations in the Susquehanna colloid with Xa and 
a The charge was calculated from the migration velocities as shown on p. 29. 
