PROCESSING SEED OF GRASSES AND OTHER PLANTS ll 
first milling over a scalper and milling the awned seeds a second time. 
_ Both fractions were then combined and cleaned. 
Screens with /%-inch round perforations were satisfactory for mill- 
ing all grasses except tall oatgrass, which required a %¢-inch screen. 
In general, the correct size of hammer-mill screen was found to be one 
in which the diameter of the perforation was slightly greater than the 
length of the caryopsis. 
INJURY TO SEED 
The injury to grass seed caused by milling can be held to less than 3 
percent except with bulbous barley and squirreltail, for which it 
increases to approximately 15 and 20 percent, respectively. The values 
for injury in milling are obtained by subtracting the percentage injured 
in the check from the percentage of processed seed that was injured 
(table 3). Bulbous barley seed is large and brittle. It has a narrow, 
pointed embryo at the proximal end, and the lemma with its long, stiff 
awn is characteristically fused to the caryopsis. Some of these features 
are seen in figure 3, A. The seed of squirreltail is slender, fragile, and 
tapered at both the proximal and distal ends. The lemma i is fused to 
the caryopsis and terminates in a long tapered awn. It 1s obvious 
that milling,even when carefully done, w “ill cause damage to theseseeds. 
LOSS IN WEIGHT 
Loss in weight was chiefly a loss of inert matter except in bulbous 
barley and squirreltail, for which a small percentage of the loss can be 
attributed to ground seed. Part of the loss in weight of bluebunch 
wheatgrass, tall oatgrass, and squirreltail was in the chaff and straw 
that could not be removed by cleaning awned seed but was easily 
separated from milled seed. This is reflected in the percentage of 
purity. If the purity of any seed was increased 10 percent or more by 
processing, as it was for these three grasses, the untreated sample con- 
tained a considerable amount of inert matter. The amount of straw, 
chaff, and sterile glumes was especially high in squirretail, and purity 
was increased from 41.0 percent to 92.9 percent, or more than 50 
percent. The loss in the weight of bulbous barley, 55.0 percent, was 
to be expected because of the relatively high proportion of awns, 
sterile glumes, and persistent rachilla segments in the untreated sam- 
ple. The proportion of awns to seed was also high in Canada wild- 
rye and squirreltail. These relationships are reflected in the deter- 
minations of seeds per pound, high values for percentage loss in weight 
correlating with high values for percentage increase in seeds per 
pound, when proper deductions are made for changes in purity due to 
the removal of inert matter. Essentially, ther efore, a loss in weight 
incurred by processing grass seed is indicative of a loss of inert matter, 
the removal of which makes a superior product. In fact, milling 
actually facilitated the cleaning operation. 
It has been pointed out that seed of one accession of each species 
was used throughout these studies. This was done because it was 
found that there was some variation between lots, which might have 
affected the outcome of the tests. An example of the variation be- 
tween lots was obtained by processing four accessions of blue wild-rye. 
