TESTING SEEDS. 



35 



statement certifying to the fairness of the sample, its source, 

 etc. Blanks for this purpose will be furnished by the Station 

 upon application. In case of guaranteed seed, the sample 

 must be taken in accordance with directions given above. 



Purity test. — All purity tests shall be made by weight from 

 fair, average samples of seed. The minimum quantities to be 

 used for this determination are named below and must be so 

 drawn as to secure a thoroughly representative sample. 



One gram: Agrostis spp., the Poas, yellow oat grass, tobacco. 



Two grams: Bermuda grass, velvet grass, timothy, meadow 

 foxtail, crested dog's tail, orchard grass, sweet vernal grass, 

 alsike clover, white clover, Umbelliferae, and all the fescues 

 except meadow fescue. 



Three grams: All grass seed not enumerated above. 



Five grams: Melilotus, Medicago spp., millet, lettuce, and 

 all species of clover seed except white and alsike. 



Ten grams: Cruciferae, flax and lespedeza. 



Thirty grams: Buckwheat, Vicia spp., Lathyrus spp., beet 

 "balls," sunflower, serradella, cucurbits, and all cereals except 

 corn. 



Fifty grams: Peas, beans, corn, lupines, cotton, and cow- 

 peas. 



Amounts to be taken of seeds not enumerated shall be the 

 same as those required for seeds named which are of similar 

 size. 



Keeping Samples. — A sufficient amount of each sample should 

 be kept in well-corked vials in a dark, dry and cool place for six 

 months, to be used in case a retest is found necessary. 



Record. — The record of seed tests shall include name of seed, 

 source of sample, weight of sample, date of tests, percentage by 

 weight and, as far as practicable, character of impurities. 



Report Blank. — The following form used by the Station is rec- 

 ommended for reports of tests. 



