8 BULLETIN 558, U. S. DEPARTMENT OF AGRICULTURE. 
ing a test weight varies widely at different elevators. Usually, how- 
ever, it is made with a hand scale (see fig. 1), which has a kettle hold- 
ing one or two quarts, the beam being graduated to read direct in 
pounds per bushel. The weight can be varied considerably by the 
manner in which the kettle is filled. Often it is merely dipped into 
the grain. Again it is partly buried and the grain scooped in over 
the sides with the hands, or the grain may be poured into the kettle 
from varying elevations, either a handful at a time or in a constant 
stream. In removing surplus grain (in order to level the kettle) an 
instrument is sometimes used that tends to pack it into the kettle. - 
Frequently the kettle is jarred to settle the grain before the surplus 
is struck off, while in many places the beam of the scale is used for 
striking off the surplus, and this eventually wears the edges to such 
an extent that inaccurate weights are obtained. Thus, unless a uni- 
form method of making the test prevails, there is opportunity for 
considerable variation in the test weight of the same sample. 
The department has now adopted, in connection with the enforce- 
ment of the United States grain standards Act, a standard method 
and apparatus for determining the test weight of grain, which is 
being used generally.t 
VARIATION IN TEST WEIGHT. 
During the summer cf 1915, 154 samples of wheat, oats, and rye 
were taken at various elevators in the grain belt, and the elevator 
operators were requested to supply the test weights as ascertained 
by them. The samples were then placed in air-tight containers, 
shipped to the Office of Markets and Rural Organization, and the 
test weight at once determined with a scale similar to the one used 
in the standard method adopted by the Department of Agriculture. 
The number of samples taken was not sufficient to justify exact 
conclusions, but the results obtained indicate that there is consider- 
able variation. 
Only 10 of the 154 samples showed a higher test weight here than 
at the country elevator, the maximum difference being 4.5 pounds. 
Sixteen samples showed no difference, while 128 tested lower here 
than at the elevator, the greatest variation being 6.5 pounds. The 
average of the test weights obtained at this office was 1.54 pounds 
lower than the average of the test at country elevators. 
The country dealers frequently allow the test weight that is shown 
when the poise is moved far enough on the beam to bear it down (as 
in fig. 1), while the inspectors at terminal markets use the weight 
shown when the beam balances, or at the nearest half-pound frac- 
tion before that point. In this way the farmer frequently secures the 
1 Boerner, E. G. Improved apparatts for determining the test weight of grain with 
a standard method of making the test. United States Department of Agriculture, 
Bulletin 472, 1916. 
dca aang SN oe EAE NO NRE en “fea 
