26 
BULLETIN 122, U. S. DEPARTMENT OF AGRICULTURE. 
The method applied by them is as follows: Ten grams of the sample 
were separated into sands of five different grades and into silts and 
clays. The determinations were made on subsamples of very fine 
sand (1-0.05 mm. diameter) and on the coarser silts (0.05-0.005 mm. 
diameter). Six or eight counts were made on each sample, and from 
the number of potash feldspars and micas and the total number of 
particles present the percentages of the forenamed minerals have 
been determined. 1 
From this data and from the amount of each separate in the soil 
the minimum percentages of potash feldspars and muscovite have 
been determined. The minerals in question in the coarser sands and 
clays are not included, for these separates are difficult to examine. 
The percentage of potash-bearing minerals in the clay is very often 
higher than in any other separate, as shown by Failyer, Smith, and 
Wade. 2 
Table V shows the relation between the total amount of potash in 
the soil as shown by fusion analysis and that estimated from the 
amount of potash minerals determined in the sands and coarse silts. 
The theoretical 3 percentage of potash in orthoclase was taken as 
the basis for calculation and 8 per cent was used for muscovite. 
Table V. — Relation between the actual amount of potash present and the amount of 
potash in the minerals estimated. 
Number 
of soil 
sample. 
Potash. 
Actual. 
From min- 
eralogical 
data. 
1 
16 
19 
22 
Per cent. 
0.10 
.61 
3.96 
4.07 
Per cent. 
0.22 
.32 
2.33 
2.77 
Excepting sample No. 1, where the actual amount is small — and 
for this reason the exception is not important — there is much more 
potash present than is accounted for mineralogically. These figures 
simply show that the estimates based on an examination of the min- 
erals are certainly not too high. 
In all the soil types examined there were present either potash 
feldspars or potash mica in large amounts. In one case only is 
potash feldspar lacking and, likewise, muscovite was not found in one 
iln this calculation it is assumed that in each mechanical subdivision the average weight of the particles 
of each mineral in question is equal to the average weight of the particles of any other mineral. While this 
is not strictly true, it is sufficiently accurate for the purpose. The error is probably greatest in the mica 
determinations. 
* Bui. 54, Bureau of Soils, U. S. Dept. Agriculture. 
8 Orthoclase as ordinarily found does not contain the calculated amount of potash. 
