8 BULLETIN 1193, U. S. DEPARTMENT OF AGRICULTURE. 
It was expected that some evidence concerning the reliability of 
the method would be afforded by the closeness with which the ratios, 
adsorption per gram of soil 
adsorption per gram of colloid 
obtained by the adsorptions of dye, water, and ammonia agreed with 
each other. 
EXPERIMENTAL PROCEDURE. 
SOILS USED. 
The soils selected for a trial of the proposed method of colloid 
determination were some of the most important agricultural types. 
They varied in texture from sands to such heavy clays as the Sharkey 
clay and the Stockton clay adobe soils. In most cases the surface 
soil represented the first 8 to 12 inches of material and the subsoil 
samples were taken to a depth of 2 to 3 feet. The samples were 
received in a fairly moist condition and were kept moist until ready 
for use. since in previous work there had been some evidence that 
thorough drying was likely to make extraction of the colloid more 
difficult. M. H. Lapham, J. E. Lapham, and W. Edward Hearn of 
The Soil Survey collected the samples. 
ISOLATION OF A SAMPLE OF THE COLLOIDAL MATERIAL. 
No attempt was made in the first part of the work described here 
to extract all the colloidal matter in the soils. The idea was to isolate 
from each soil enough of the colloidal material for chemical analysis 
and for determination of the adsorptive capacity for malachite green, 
water vapor, and ammonia gas. The samples of colloid for the 
determination reported in Table 1 were extracted virtually as de- 
scribed in the preliminary paper by Moore, Fry, and Middleton (22), 
although in the subsequent work this procedure was altered some- 
what. 
One hundred pounds of each soil were agitated in barrel churns 
with 500 pounds of distilled water for two hours. After standing 
for 18 hours the turbid liquid was siphoned off and the soil was again 
agitated with water to obtain a further yield of suspended matter. 
Frequently the second treatment of the soil gave a heavier suspension 
than the first, although a few soils failed to give any appreciable 
quantity of suspended matter on three such treatments. In such 
cases a fourth treatment with water containing approximately 1 part 
of ammonia to 3,000 parts of water was Very effective in producing 
a suspension. The turbid extract from the soils was run through a 
high power centrifuge (known in the trade as a supercentrifuge) driven 
at the rate of 17,000 revolutions per minute, which threw out a large 
part of the suspended matter. In this process the particles were 
exposed to a force of approximately 17,000 gravity for about 3 
minutes. 
The colloidal suspension which passed through the centrifuge was 
first concentrated by sucking the water off through batteries of 
Chamberland-Pasteur filters (bogie F), the colloidal matter collecting 
on the outside of the filters. 5 When the colloidal matter had thus 
6 The slimy film of colloid which collected on the outside of the filters was readily removed from time to 
time by releasing the suction and blowing air into the tubes. 
