26 
BULLETIN 1452. U. S. DEPARTMENT OF AGRICULTURE 
Gedroiz (lo. 18. 19) has shown that the dispersability of the col 
loid and the swelling of a soil are markedly affected by altering the 
kind of exchangeable bases present by treatment with neutral salts. 
The introduction of sodium into the colloid, for instance, gives a 
far greater dispersion than the introduction of calcium. Thus it 
would appear also from Gedroiz 's work that the physical properties 
of the colloid should be determined by the relative proportions of 
the bases. The correlation between properties and relative propor- 
tions of bases would probably be more apparent in the case of col- 
loids containing approximately the same quantity of total exchange- 
able bases, than in the case of these six colloids, which contain widely 
different quantities. 
It has already been shown in Table 12 that there is a fair degree 
of parallelism between the total exchangeable bases and the adsorp- 
tion of ammonia vapor, which in turn follows fairly well other prop- 
erties of the colloid. Inasmuch as there is some uniformity among 
the different colloids regarding the part of the bases which is present 
in exchangeable form, it follows that there should also be some 
parallelism between properties and total content of nonexchange- 
able monovalent and divalent bases. 
Table 15 shows the quantities of the bases which are present in 
exchangeable and nonexchangeable forms in the six-colloids. 
Table 15. — Exchangeable and nonexchangeable monovalent and divalent bases 
in soil colloids • 
Ca 
Mg 
K 
Na 
Total 
Total 
Kind of colloid 
Ex- 
change- 
able 
Nonex- 
change- 
aole 
Ex- 
change- 
able 
Nonex- 
change- 
able 
Ex- 
change- 
able 
Nonex- 
change- 
able 
Ex- j Nonex- 
change-, change- 
able j able 
change- 
able 
bases 
change- 
able 
ba^es 
Fallon 
MiUi- 
equiva- 
lents 
0.760 
.474 
.357 
.139 
.061 
.046 
MW.i- 
equiv cl- 
ients 
0.082 
.054 
.042 
None. 
.039 
.018 
Milli- 
equiva- 
lents 
0.168 
.179 
.144 
.0 4 
.030 
.030 
Milli- 
equiv cl- 
ients 
2.470 
1. 0S1 
.823 
.615 
.198 
.163 
Milli- 
equiva- 
lents 
0.072 
.021 
.027 
.008 
Trace. 
.004 
MiUi- 
equir cl- 
ients 
0.403 
.374 
.412 
.212 
.117 
.191 
Milli- 
equir cl- 
ients 
0. OSO 
MiUi- 
equ ira- 
te nts 
0.094 
Milli- 
eqvira- 
lenti 
1.080 
.687 
.528 
.227 
.091 
.087 
Milli- 
equiva- 
lents 
3.049 
Sharkey 
ftlai shall 
. 013 . 035 
1.574 
1.277 
Sassafras 
.016 .084 
Trace. ' - 100 
.911 
.454 
Susquehanna 
.007 
.087 
.459 
It will be seen that the different colloids stand in the same order 
when arranged according to the content of either exchangeable or 
nonexchangeable bases. Properties of the colloid, such as heat of 
wetting, adsorption of vapors, etc.. correlate slightly better with 
the total exchangeable monovalent and divalent bases than with 
the content of nonexchangeable bases, but the proportionality is 
almost as good in one case as in another. 
There appears to be more or less of a balance between the total 
content of exchangeable and nonexchangeable bases. In the case 
of the Fallon, Marshall, and Sharkey colloids, which are neutral or 
slightly alkaline, the Ca-f-Mg+K+Xa exchanged fairly represent 
the total exchangeable bases ; and here the total of the nonexchange- 
able bases is about 2.5 times the total of the exchangeable. In the 
case of the other three colloids, which are somewhat acid, the 
Ca-j-Mg-f-K+Xa exchanged do not quite represent the total ex- 
changeable bases: and here the total of the nonexchangeable bases 
