PROCESSING SEED OF GRASSES AND OTHER PLANTS 5 
half capacity reduced the deawning and materially increased the 
amount of injured seed. The exact amount of seed lost by grinding, 
under the conditions of the tests, could not be determined, but ex- 
amination of the cleanings indicated that a very small percentage was 
ground when the mill was operated at 600 r. p.m. The material re- 
moved by cleaning consisted chiefly of awns, portions of the hull, and 
other inert matter. The amount of oround seed increased at speeds 
above 600 r. p. m., especially when the mill was fed at half capacity. 
TABLE 2.—The effect of cylinder speed and rate of feeding during milling on processed 
seed of Canada wild-rye (Elymus canadensis) 
[The data are averages of three trials in each of which 30 pounds of seed was used.] 
Rate of milling Test 
De- Loss in Germi- Live Seeds weight 
awned weight nation pure Der per 
seed pound bushel 
Speed of 
cylinder 
(r. p.m.) |Relative|) Amount 
capacity per hour 
Injured Purity 
Thou- 
Pounds | Percent | Percent | Percent | Percent | Percent | Percent | sands | Pounds 
@heckeess ta | eae 6.5 0 |S 93.5 97. 75 91. 40 67.6 5.8 
600 {Hat es 250 79.0 3. 37 22. 0 98. 1 98. 00 OG 14s pew ane 
mee oR Mulla = 600 96. 0 152 28. 0 96. 8 98. 75 95. 59 79.9 28.55 
790) fale 300 83.5 11. 87 36. 0 97.3 89. 00 SO5633 | Sees sae al | eer es 
Ce tea Wille 700 96. 5 5. 25 32.0 98. 2 95. 50 O37 88 ee see Ses | eo aes 
830 ae 350 86.5 24. 87 45.0 97.4 73. 30 GELS G in | pee ee re | iae T eap 
SaaS Bulle 800 96. 5 9. 87 34. 0 98. 4 90. 55 SOFA Oe) Reece Ol ener 
Processed seed is easily planted with ordinary machinery, whereas 
untreated seed is very difficult to plant. This is reflected in table 
2, which shows that processing increased the number of seeds per 
pound from 67,600 to 79,900, or 18.2 percent. This increase is great, 
but it is to be expected because the large awns in the untreated sam- 
ple make up a considerable part of the weight. 
Table 2 also gives the results of tests to determine the changes in 
quality of Canada wild-rye seed associated with changes in the method 
of milling. The data show that processing improves the purity only 
about 4 percent, which is not a significant difference. This is to be 
expected because this seed is easy to clean and the percentage of 
purity of the original lot was high. Germination tests indicated that 
the treatment affected viability, reducing it measurably when cylin- 
der speeds above the optimum ‘for the species were used and still fur- 
ther with low rates of feeding. The germination of seeds processed 
at 600 r. p.m. was slightly improved, but the difference is not regarded 
as significant. There is an inverse relationship between percentage 
of 1 Injury and germination, oermination decreasing as the percentage 
of injured seed increases. ‘‘Live pure seed,” which integrates purity 
and CHE ON, also shows this same relationship. 
The data also indicate that the loss in weight due to milling at 
optimum speed and rate is largely offset by increased quality as 
measured by live pure seed and the number of seeds per pound. The 
totai loss in weight at 600 r. p. m. and full rate of feeding was 28.0 
percent, but live pure seed increased 4.19 percent and seeds per 
pound 18.2 percent. Another improvement is the reduction in the 
volume of the processed seed to nearly one-fifth that of the original 
sample, as indicated by the reciprocal of the increase factor for change 
