12 CIRCULAR 558, U. S. DEPARTMENT OF AGRICULTURE 
The values for loss in weight were 6.75, 7, 9, and 10 percent. For 
three lots of Siberian wild-rye, losses in weight were 9, 10, and 14 
percent. Sufficient allowance for variation between lots should there- 
fore be made when the results of any test are put to practical use. It 
is recommended that preliminary trials with no less than 30 pounds of 
seed and 3 minutes of full operation be made with each lot of seed 
when a high degree of efficiency or accuracy is desired. Although 
these values are for loss in weight, values for other criteria of the re- 
sults of processing would also show similar variation between lots. 
LIVE PURE SEED 
Processing increased the percentage of live pure seed in five of the 
seven grasses and decreased 1t shghtly in Siberian wild-rye and bulbous 
barley (table 3). The decrease in Siberian wild-rye was within the 
limits of experimental error. The values for percentage purity and 
germination are the components of the values representing changes in 
quality as expressed by the percentages of live seed. The large in- 
creases in live pure seed obtained by processing bluebunch wheat- 
grass, tall oatgrass, and squirreltail, the original samples of which 
contained considerable inert matter, are due chiefly to the increase in 
the purity of the seed, made possible by the fact that milling facili- 
tated cleaning. The germination of bluebunch wheatgrass was in- 
creased, because milling removed nonviable shriveled seed. Reduc- 
tion of germination in bulbous barley and squirreltail is correlated 
with injury, which could not be avoided; nevertheless, the quality 
was sufficiently increased to compensate for the loss from injury. 
TEST WEIGHT 
Another important advantage of processing is the increase in test 
weight per bushel. In table 3 the values are expressed in terms of a 
factor that indicates the numerical increase. The reciprocal of this 
factor indicates the volume of the final product in relation to the origi- 
nal. For example, the volume of processed Canada wild-rye is 
approximately one-fifth that of the untreated seed. The advantages 
of this reduction to storage and handling are immediately apparent 
and are of commercial value. 
CostTs 
Costs for processing grass seed ranged from 67 cents to $1.50 per 
hundred pounds, the lower figure being for blue wild-rye and the higher 
for bulbous barley. The cost of milling alone was 20 and 80 cents per 
hundred pounds, respectively, for the same species. Since cleaning 
is essential, even if the seed is not deawned, this part of the operation 
might have been deducted, but it was included in the costs shown in 
table 3 because the cleaning operation is influenced by milling. Costs 
for milling and cleaning are not shown separately in table 3. It 
should also be pointed out that since the costs are based on relatively 
small lots they may be somewhat high. 
