14 BULLETIN 353, U. S. DEPARTMENT OF AGRICULTURE. 
RELIABILITY OF AIR-DRIED SAMPLES. 
The reliability of air-dried samples may be determined in three 
ways: (1) By a comparison of the percentages of moisture loss in 
the samples with that in the 100-pound and 500-pound quanti- 
ties, which, on account of their bulk, approximate field methods; 
(2) by a careful comparison of the relation between the moisture 
lost in air drying and the total moisture content as revealed by oven 
drying; and (3) by noting the variation in the percentage of moisture 
remaining in the air-dried material. A comparison of the moisture 
loss in air-dried samples with that in bulk lots of the same material 
is given in Table II. 
TaBLE IJ.—Comparison of the loss of moisture in green and field-cured forage when air 
dried in small samples and in large bulk. 
. . . Moisture in field-cured ~ 
Moisture in green material. aintora 
Place. Crop. 
Loss in | Loss in Loss in | Loss in 
Total. samples.| bulk. Total. samples.} bulk. 
Per cent. | Per cent. | Per cent. | Per cent. | Per cent. | Per cent. 
Chico Calecs see Minin ss2ck soe ae 76.9 74.5 73.0 22.3 : ink 
Arlington Farm, Va.-.-.| Tall oat-grass and 72.0 66.3 64.3 29.0 13.4 13.5 
orchard grass. 
New London, Ohio...} Timothy...-....... 58.0 50.5 49.2 20.3 7.2 10.1 
Amarillo, Tex...-...- Sorghum.......... (lez 54.2 58.2 43.2 20.5 16.8 
PAV Se KanSerneeeen eeeee GO n see eeeeas|aee ccc eee 65.8 60:9) |oS2 oss 26.0 Pa | 
It will be seen that the losses in the small samples of green material, 
except for those of sorghum at Amarillo, Tex., which were not well 
cured, averaged from 1.3 to 4.9 per cent greater than it did in the 
bulk lots. This was to be expected, since the small sample naturally 
dries out more completely than the bulk. The difference, however, 
is shght, and the loss of moisture in the small samples seems to be 
fairly consistent with the loss which was found in the bulk lots. 
The comparison of small samples with bulk lots of field-cured 
material is not so favorable to the use of the sample method as in 
the case of the green material. Table II also shows that the mois- 
ture loss in the samples, when compared with the total moisture con- 
tent, is not quite so ree as the percentage of moisture loss in 
the bulk lots. 
A better way to Hetennne the reliability of the sample method is 
by a study of the percentages themselves, especially in the column 
devoted to percentage of moisture in the air-dry material. The uni- 
formity of these percentages throughout one crop means that the air 
drying of samples can be depended upon to bring samples to a nearly 
uniform moisture content, and this method therefore serves the pur- 
pose of correcting field weights almost as well as to oven dry the 
samples. The moisture content of the air-dry samples is not en- 
