THE RARER ELEMENTS IN SOILS AND PLANTS. 5 
The <> 21 weights of the samples were not taken. On receipt at 
the laboratory the samples were air dried at a slightly elevated tem- 
perature, ground, and stored in stoppered bottles. The proper 
grinding of the samples presented many difficulties. An Excelsior 
coffee mill was found the most efficient machine for the purpose, but 
the plant fiber in many cases was so tough and leathery that there 
was always a contamination by the material of the grinder. 
For determinations like sulphur, chlorine, and the alkalies, where 
a small sample was taken for analysis, the material was ground finer 
in an agate or porcelain mortar. 
There was some foreign material in the samples, mostly soil grains 
adhering to the majority of plants. Nearly all of this material could 
be removed by brushing and picking, though some of it had to be 
separated chemically, as described later. The ash of most plants is 
low in silica, consequently an admixture of even a small quantity 
of silica from the soil is to be avoided in this case, whereas it would 
make little difference in ground-rock samples. 
DESCRIPTION OF THE SAMPLES. 
LEGUMES. 
1. Alfalfa—oOne-half mile west of Penn Square, Norristown, Pa. 
Soil, Penn silt loam. This sample was 2 feet high and in bloom 
when cut. 
2. Alfalfa—One mile north of Conshohocken, Pa. Soil, Hagers- 
town loam. This sample had not bloomed when cut. 
3. Alfalfa—Stein farm, Carollton, Mo. Soil type not determined. 
This sample was mature. It had some fine sand adhering to it so 
persistently that it was impossible to brush or pick it all off. 
4, Beans.—String variety. Two and one-half miles northwest of 
Charlotte, N. C. Soil, Cecil clay. The plants were about 2 feet 
high and contained blossoms and small pods. There was a heavy 
rain in the locality just before the plants were gathered, and it was 
impossible to remove all the soil that had been splashed on the lower 
leaves. 
5. Beans.—Seeds of the string variety. Three miles east of Marl- 
boro, N. H. Soil, Gloucester stony loam. 
1The iron particles ground off of the machine were very efficiently removed by sliding 
the sample through a soft iron trough which was in contact with a powerful magnet. 
As a check on the efficiency of this treatment, the following test was made: A sample 
of beans (seeds) was coarsely ground, then sampled and analyzed for iron. A _ sub- 
sample was passed several times through the mill, then run over the iron trough and 
analyzed for iron. A contamination of 5 milligrams FeO; to 100 grams of bean meal 
was noticed. Reckoned on the dry basis, this amounts to 0.005 per cent. The grinding 
of this particular sample was more severe than with ordinary samples, and it can be 
safely assumed that the contamination of iron in the samples reported was well under 
0.005 per cent. In the figures given in the tables no attempt is made to correct for this 
contamination, 
