8 EXPERIMENT STATION RECORD. 



A complete seed analysis includes determinations of (1) genuineness, 

 ('_') purity, 3) weight, and (4) germination. If partial examinations are 

 desired covering one or more points, the fact that such examinations 

 alone have been made, and that the certificate does not give any guar- 

 anty for the general quality of the seed, is to be expressly stated. 



(1) Genuineness.— The genuineness of the seed is determined by the 

 aid of the standard seed collection of the control station or the seed 

 manuals. Both popular and Latiu names of the seed examined are 

 given in the certificate, and, so far as practicable, also the variety and 

 cultural form. 



(2) Purity.— The purity of the seed is determined in average weighed 

 samples by hand picking or screening. As the lower limit for the size 

 of the sample used for this purpose the average dry weight of 1,000 

 seed multiplied by 10 is generally taken. An exception is made in case 

 of seed a hectoliter of which weighs more than 25 kg. For such seed 

 the lower limit is found by multiplying by 5. The sample must never 

 weigh less than 2 gm. and for large seeds like beans, etc., at least 300 

 gm. is required. 



In purity determinations the seed is separated into (1) pure .seed and 

 (2) impurities. The latter include (a) offal (fragments of plants and 

 seeds, dirt, sand, stones, etc.), (b) foreign culture seed, and (c) weed 

 seed. The quantities of these different components are determined by 

 weighing, aud are given on the certificate issued, calculated to per cent 

 of the seed. The weed seeds are separated into harmless and noxious 

 seed. Under the latter are included the following II kinds in addition 

 to rust spores and sclerotia: Agrostemma githago, Bromus secalinus, 

 Bunias orientalis, Centaurea cyanus, Chrysanthemum segetum, Cirsium 

 arvense, Cuscuta trifolii, Matricaria inodora, Ranunculus repens, Rapha- 

 nus raphanistrum, Rhinanthus crista galli, Sinapis arvensis, Brassica 

 campestris, and Sonchus arvensis. 



(3) Weight. — For the determination of the dry weight of 1,000 seeds 

 thiee lots of 200 seeds of the thoroughly mixed sample of pure seed are 

 counted and weighed. The fresh weight of 1,000 seeds is found by mul- 

 tiplying by 5. If the parallel determinations differ more than 5 per 

 cent for small seed (i. e., average dry weight per 1,000 seeds below 5 gm.) 

 and 10 per cent for huge seed (i. e., average dry weight per 1,000 seeds 

 over 5 gm.) the determinations are repeated. A dry-matter determina- 

 tion is then made of the sample of pure seed by grinding 10 to 20 gm. 

 of large seed in a small mill and weighing out 2.5 gm. of the ground 

 seed. In ease of small seed 1 gm. of whole seed is weighed out. These 

 quantities are dried for 1 hours in a steam oven. The percentage of dry 

 matter thus found, multiplied by the fresh weight of 1,000 seeds, gives 

 the dry weight of 1,000 seeds, which is reported with 3 decimals in 

 the analysis certificate. 



The determination of the hectoliter weight of cereals and legumes is 

 obligatory in case of complete analysis, the result being expressed with 

 3 decimals. 



