- 3 8 - 
than 3-32 of an inch. The pods were full, the seeds press¬ 
ing one upon the other. The analysis of these seeds is 
given in the table. When well dried the seeds absorb mois¬ 
ture readily. Fifteen portions of one gram each were 
weighed off, after thoroughly shaking the sample, and 
counted. The average was found to be, 456 seeds to the 
gram ; lowest number per gram, 450 ; highest number 463 ; 
number of seeds per pound, 206,837. 
Sample No. 2, purchased in the market, gave 458.6 seeds 
per gram ; 208,021 seeds to the pound. These seeds were 
not so even in size as the first sample. There were a few 
shrunken seeds. The sample was clean, containing less 
than one per cent, by number of foreign feed. Sample 
No. 3, also purchased in the market, was of a brownish yel¬ 
low color; sample contained 8 per cent., by number, of 
foreign seed, mostly of an amaranthus. The average 
number of seeds to the gram of this sample was 504.46. 
The seeds were very even in size ; minimum number to the 
gram, 503 ; maximum, 505. The number of seeds to the 
pound was 228,818. 
Sample No. 4, consisted of first quality screenings, 
furnished by J. E. Gauger, Rocky Ford, Colo., about 65 per 
cent, of which was immature when cut. The seed was 
shriveled and dark brown in color. The sample was 
quite free from grass seeds, weed seeds, and stems, and 
contained 259,340 seeds to the pound. 
Sample No. 5, first quality screenings from the same 
source as No. 4, was dark and contained many shriveled 
seeds, in which by weight there was 23 per cent, of impuri¬ 
ties—grass and weed seeds. This sample contained 344,123 
seeds to the pound. 
Sample No. 6, first quality screenings (J. E. Gauger) 
seed evidently well cured, many seeds green and immature, 
contained 266,233 to the pound. 
Sample No. 7, second quality screenings (J. E. Gauger j, 
containing more stems and weed seeds, especially of an 
amaranthus, than any other sample, contained 331,383 seeds 
per pound. 
Sample No. 8, third quality screenings (J. E. Gauger), 
was quite clean. The seeds were large, but shriveled, 
numbering 312^385 to the pound. 
VVe may assume that a pound of first-class seed con¬ 
tains 210,000 seeds; first quality screenings, 260,000 and oc¬ 
casionally many more on account of shriveled seeds ; and 
for second and third quality screenings, about 320,000 seeds 
to the pound. 
