THE CHEMICAL COMPOSITION OF SOIL COLLOIDS 25 
cording to the theory of soil development, it might be expected that 
the composition of the colloidal matter of more mature soils would 
show a closer relation to climate than younger soil colloids. There 
is some evidence of this here. The Houston and Sharkey soils are 
young in the localities in which they are now found. The colloids 
from these soils have considerably higher silica and lime plus soda 
ratios than would be expected from the present rainfall. 
The effect of parent material on the chemical composition is dim- 
cult to evaluate. In certain cases the effect is no longer evident. 
For instance, the colloids from the Hagerstown, Sassafras, and Manor 
soils are much alike in composition, yet the parent materials are 
respectively limestone, outwash material from the Piedmont Plateau, 
and mica schist. On the other hand, the Houston colloid plainly 
shows the effect of the parent material in its high lime content. Thus 
It can be seen that colloids derived from calcareous formations may 
be high or low in lime, depending probably on the extent of decom- 
position of the parent material. 
It is apparent from Table that the general composition of the 
colloidal matter can in many cases be predicted from the color. 19 All 
greyish or black colloids have a silica ratio above 2.10 and the red- 
dish or yellowish colloids generally have a lower ratio. Since there 
is a pretty good correlation between the lime plus soda ratio and the 
silica ratio, we would expect the lime plus soda ratio to be usually 
low in red and yellow colloids and high in blackish or gray colloids. 
From the data given in Table 6 it is apparent that, in general, 
the colloidal matter developed in dry regions will be higher in silica 
and lime plus soda than colloids developed in humid regions. The 
composition of any particular colloid, however, can not be predicted 
with much cer{aintv from the present rainfall. The similaritv in 
general composition of the colloids Nos. 21, 7, 22, 28, 33, 17, 18, 34, 
15, and 14 (see also Table 3), which were extracted from presumably 
old soils in parts of Maryland and North Carolina, where the rain- 
fall and temperature are much alike, suggest that the colloidal mat- 
ter in mature soils may be fairly constant in composition in a given 
climatic region. Sufficient information, however, has not been ac- 
cumulated to substantiate this generalization. 
COMPOUNDS PRESENT IN COLLOIDAL MATERIAL 
FRAGMENTS OF SOIL-FORMING MINERALS 
The ultimate chemical composition of the colloidal material has 
been discussed, but no evidence as to the manner in which the various 
consitutents are combined has been given. In this connection a 
common idea that the colloidal matter in soils is made up of the 
same minerals as the larger particles of the soil, but in a finer state 
of subdivision, will be considered. 
It is known that the soil contains considerable quantities of such 
minerals as the feldspars, micas, and hornblende in particles 5 to 50 
microns in diameter (silt size), and it would seem probable that 
such minerals would be present in fragments less than 0.3 micron 
in diameter. Hall (19, p. 41), however, from reasons advanced by 
10 It is interesting- to note that the soil and corresponding- colloid have the samp color : 
the color of the colloid, however, is more intense. Hence there is some relation between 
the color of the soil and the chemical composition of the colloid. 
