PHYSICAL PROPERTIES OF HAWAII SOILS 
33 
of the soil; (2) that the property employed in the estimation is not 
altered during extraction; and (3) that the property employed in 
the estimation is possessed by the colloidal fraction only, or at least 
to an overwhelming extent. In connection with this, it was assumed 
that all the organic matter exhibiting adsorptivity is obtained in the 
colloidal fraction. Soil particles less than 2 microns in diameter 
were considered as colloidal. 
Within recent years the power of the soil to adsorb moisture from 
a saturated atmosphere has been used frequently to determine the 
colloidal content. The method used consists in dispersing the soil in 
water and allowing the sand and silt fractions to settle according to 
Stokes' law. This gives the colloidal fraction in suspension whence 
the colloids are obtained by filtering or by supercentrifuging, and 
afterwards drying. The adsorption of water vapor is then deter- 
mined for both the soil and the extracted colloid by exposing them to 
a saturated atmosphere of water vapor over 3.3 per cent sulphuric 
acid at a certain temperature for the length of time required to assure 
complete adsorption. From the adsorption of the soil and that of 
the colloidal fraction the percentage of colloids is calculated. This 
method was tried on a number of common, and a few unusual types 
of Hawaii soils, the characteristics of which are given in Tables 1, 
2, 3, and 4, and in the Appendix. 
In order to obtain a representative part of the colloidal phase of 
the soils — which has the same power of adsorption it has in the soil — 
50-gram aliquots were repeatedly shaken with distilled water in an 
end-over-end shaker for four hours and allowed to settle according 
to Stokes' law for the length of time required to leave only particles 
smaller than 0.002 millimeter in suspension. The suspension was 
then siphoned off, the collected fractions were mixed, and the col- 
loids in suspension were passed through Pasteur-Chamberland filters. 
The separated colloids were dried at room temperature and passed 
through a 60-mesh screen. A fairly large amount of soil colloids was 
obtained, as is shown in Table 14. 
Table 14- 
-Percentage of total colloids extracted by three successive dispersions 
Soil No. 
Colloids 
extracted 
(a) 
Proportion 
T t t of total 
3ff> "Sffi 
w extracted 
(o/6) 
Soil No. 
Colloids 
extracted 
(a) 
Total 
colloids 
(6) 
Proportion 
of total 
colloids to 
colloids 
extracted 
(a/6) 
2 
Per cent 
40.82 
50.16 
8.12 
3.64 
9.36 
1.88 
9.60 
.96 
4.62 
Per cent Per cent 
51. 5 79. 2 
66. 6 75. 3 
27. 30. 1 
15. 24 2 
23. 9 39. 2 
15.5 12.1 
49. 3 19. 5 
17.0 5.6 
23. 1 20. 
42 
48 
49 
56 
57 
65 
66 
Per cent 
7.16 
10.04 
2.76 
22.76 
23.24 
15.88 
12. 14 
9.94 
2.58 
Per cent 
32.5 
30.1 
21.3 
48.0 
49.1 
35.1 
30.6 
25.8 
28.1 
Per cent 
22.0 
6 
33.4 
9 
13.0 
12 
47.4 
21 
47.3 
23 
45.2 
31 
39.7 
33 
76 
77 
38.5 
37 
9.2 
This method of procedure permitted presumably little change in 
the adsorptivity of the soil colloids during treatment. 
Although all the soils, with the exception of soil No. 69, contained 
only a small amount of soluble salts, the samples for determination 
were washed with water to prevent the salts from influencing the 
