RESEARCH· ·1970·

..;
ppm. Then U = 0.7 and s.d. = ·o.15 + (0.063) (0.7)
= 0.19 ppm, or the expected standard deviation
would be about 0.2 ppm, and the expected coefficient
of variation would be 27 percent after rounding.
TABLE 2.-The arithmetic mean and standard deviation for duplicate
uranimn determinations grouped by ranges regardless of
plant species
Number of pnlrs of dotcrmlnntlons
43 _____________________ _
106 ____________________ _
78----------------------
51 _____________________ _
17 _____________________ _
20 _____________________ _
11 _____________________ _
z 3.0
0
::i
..J
:::E 2.5
0:
w
Q..
~ 2.0
0:
~
~ 1.5
z
0
i=
=:; 1.0
> w
0
~ 0.5
c(
0
z
c(
1-
rn
0 5 10 15
Rnngc (ppm)
0. 0- 0. 8
. 8- 1. 6
1. 6- 3. 2
3. 2- 6. 4
6. 4-12. 8
12. 8-25. 6
25. 6-51. 2
Arithmetic
menn
(ppm)
0. 54
1. 15
2. 31
4. 45
9.46
18. 22
34. 77
Regression line
Standard deviation in parts per
million= 0.15 + 0.063 X uranium
concentration (For range 0.4 to
35 ppm)
HUFFMAN AND RILEY
Standard
deviation
(ppm)
0. 14
. 21
. 31
. 39
. 73
1. 48
2. 26
20 25 30 35 40
URANIUM CONCENTRATION IN PLANT ASH,
IN PARTS PER MILLION
FIGURE 1.-Arlthmetlc means and standard deviations from
table 2 and their 1 re~ression line.
B183
2. Observed concentration of uranium in ash = 10.0
ppm. Then s.d. = 0.15 + ( 0.063) ( 10.0) = 0.78 ppm,
or the expected standard deviation would be about
0.8 ppm, and the expected coefficient of variation
would be about 8 percent after rounding.
For many purposes the equation could be rounded to
standard deviation = 0.2 + 0.06 U. If ·m!Mly solutions
are needed, the use of a graphical method similar to
that shown in figure 1 may be helpful.
Two precautions in the use of this equation should
be noted : ( 1) the precision measured is for the analysis
of uranium in a given sample of plant ash; the equation
does not give the precision to be expected from two
samples from the same tree; and (2) the equation
should not be extra pol a ted beyond the range of the
uranium content of the samples used in its derivation
(0.4-35 ppm U in ash). ..
A study of the accuracy of the fluorimetric method
for determining uranium in plants, as contrasted with
its precision, has not been attempted. However, the use
of blanks and of known solutions as standards, as
described in the procedure, gives considerable assurance
that there is no appreciable bias in the m~thod.
REFERENCES
Cannon, H. L., 1954, Botanical methods of prospecting for
uranium [Colorado Plateau]: Mining Eng., v. 6, no. 2,
p. 217-220.
Cannon, H. L., and Starrett, W. H., 1956, Botanical prospecting
for uranium on r. .. a Ventana Mesa, Sandoval County,
N. Mex. : U.S. Geol. Survey Bull. 1009-M, p. 391-407.
Grimaldi, F. S., May, Irving, }~letcher, M. H., and Titcomb,
Jane, compilers, 1954, Collected papers on methods of
analysis for uranium and thorium: U.S. Geol. Survey Bull.
1006, 184 p.
Kinser, C. A., 1954, The Model VI transmission fluorimeter for
the determination of uranium: U.S. Geol. Survey Circ. 330,
9 p.
· ..