24 
BULLETIN 1193, U. S. DEPARTMENT OF AGRICULTURE. 
cases the unextraetable colloid is more nearly equal to the extractable 
colloid in adsorptive capacity for ammonia than in adsorptive capac- 
ity for water or dye. This is apparent also on averaging the results 
of the ten samples. The adsorptive capacity of the unextracted 
colloid averages 64 per cent that of the extracted colloid for dye, 58 
per cent for water and 83 per cent for ammonia. 
Table 6. — Relative adsorptive capacities of extractable and unextraetable colloidal material, 
as indicated by the relative percentages of colloid shown by microscopical and by cor- 
rected adsorption ratio determinations. 
Soil fraction in which colloid present. 
Colloid 
present 
by micro- 
scopical 
count. 
Colloid present by corrected— 
Adsorptive capacity of 
unextraetable colloid 
for dye, H2O, and 
NH 3 , expressed rela- 
tive to adsorptive ca- 
pacity of the extract- 
able colloid taken as 
100 in each case. 
Dye ad- 
sorption 
ratio. 
Water 
adsorp- 
tion ratio. 
Ammonia 
adsorp- 
tion ratio. 
Dye. 
H 2 0. 
NHa. 
Cecil clay loam soil: 
Fine fraction 
Per cent. \ Per cent. 
74 22 
Per cent. 
18 
Per cent. 
36 
28 
2 
63 
68 
45 
43 
44 
68 
101 
136 
23 
25 
38 
80 
49 
Fine fraction (duplicate) 
55 | 1 j 14 
48 i 30 18 
25 17 j 20 
3S 17 19 
42 18 26 
Huntington loam: 
Soil, fine fraction 
26 
21 j 
22 
31 
35 
24 
101 
23 
54 
Subsoil, fine fraction. 
84 
Sassafras silt loam: 
50. j 58 
Sharkey clay soil: 
Fine fraction 
62 71 
Separation by supercentrifuge— 
Fine fraction 
52 23 
25 17 
97 98 
32 
21 
85 
62 
84 
88 
67 
96 
Vega Baja clay loam, soil (separation 
by supercentrifuge): 
Fine fraction 
104 
Coarse fraction 
14 
64 164 
Conclusions regarding representativeness of a sample of colloidal 
material. — The data obtained on the adsorptive capacities of suc- 
cessive lots of colloidal material isolated from the same soil and on 
the relative adsorptive capacities of the extractable and unextraet- 
able colloid show that a small sample of colloidal material is not 
exactly representative of all the colloid present in many soils. Part, 
therefore, of the disagreements between the quantities of colloid 
indicated by the dye, water, and ammonia adsorption ratios in Table 
2, may be ascribed to failure to obtain an exactly representative 
sample of the colloidal material. Sampling of the colloidal material, 
however, does not appear to be the only cause of disagreement in the 
adsorption ratios. 
A small sample of the extractable colloidal material which is 
dispersable into particles less than 0.3 micron in size appears to be 
about 90 per cent representative of all this class of material. The 
chief error in sampling appears to be due to the unextraetable 
collodial material in many, but not all, soils having a different 
adsorptive capacity from that of the extractable colloidal material. 
A small sample of the colloidal material seems to represent more 
nearly all the colloidal material in its adsorptive capacity for ammonia 
than in its adsorptive capacity for water or dye. 
