COLLECTION OP SEED. 23 



TESTING THE SEED. 

 SEED- TESTING LABORATORY. 



The Forest Service has in Washington a seed-testing laboratory, in 

 which germination and purity tests are made and the number of seed 

 per pound determined of all species collected on the National For- 

 ests. As each lot of seed is received it is passed through a seed- 

 sampling machine, which divides it into two approximately equal 

 quantities. This process is continued until one of the portions con- 

 sists of a few more than 1,000 seeds, with the usual amount of rub- 

 bish. This portion is weighed, then cleaned of all rubbish, and re- 

 weighed. In this way the percentage of purity of the seed by 

 weight is determined. From the same portion 1,000 seeds are counted 

 out, weighed, and the number of seed per pound determined by pro- 

 portion. Germination per cent is determined in two ways first, 

 by a cutting test on 200 seeds, and, second, by planting 200 seeds in a 

 sand-filled flat and subjecting them to a hothouse temperature of 

 about 60 F. The cutting test is, of course, only approximate, but it 

 serves as a good basis for comparison with the actual germination 

 secured by the planting of the seed. 



In addition to these tests, the seed laboratory is conducting experi- 

 ments to determine the relative merits of a number of methods of 

 storing coniferous-tree seeds. 



YIELD FROM CONES. 



The yield of seeds depends upon the quality of the cones, the thor- 

 oughness of drying and extracting, and the manner of cleaning. 

 There is a great variation in the yield of seed from a bushel of cones. 

 The cones of any species fill better during a " seed year " than during 

 " off years," so that in the former there is greater bulk, and especially 

 greater weight of seed. Table 3 shows by species the average quan- 

 tity of seed per bushel of cones or fruits. 



TABLE 3. Quantity of seed per bushel of cones or fruits of different species. 



5274 Bull. 9811- 



