21 



The term "Hjirinless Impui-ity" includes: 



1. All pure seeds being broken or damaged so as to have probably lost their germinating 

 capacity'). , , 



2. Seeds of other cultivated plants. 



3.- Dea^seeds, empty glumes of grass-seeds, straws, glumes, earth, saiid, residues of seeds 

 and fruits, also of weed-seeds and of parasites which are esteemed incapable -of 

 development, and 



4. Seeds surely containing some special kinds of parasites, the quantity of which is how- 

 ever liable to speedy changes and which are not of a very noxious nature^). 

 The term "Noxious Impurity" includes: 



1. Seeds of plants being in general or very nearly in general considered as "weeds". 



2. Seeds of plants which — though cultivated — may become troublesome in some way 

 or other, or of which the growth is not recommendable, &c., such are f. i. . Aira flexuosa, 

 Alopecnrus agrestis, Anthoxanthum Puelli, Avena fatua, Brassica- and Sinapis-species, 

 Bromus, small-seeded Vicia-species, &c. 



On the contrary the seeds of Daucus, found among grass- and clover seeds, are 

 reckoned to the Noxious Impurity, while originating from proper weeds, notwithstanding 

 Daucus is frequently cultivated. 



3. Sclerotia, seeds of blight and Tylenchus. Here it should be observed that for the 

 determination of the "purity" of corn and leguminous plants, a special procedure is 

 practised in case the seeds are infected with some disease. This procedure will be 

 described below. 



4. Insects, larvae and insect-eggs, including what has been mentioned sub 3 and 4 (Noxious 

 Impurity). 



Details concerning the application of the method. 



1. Weighing the quantity to be tested. 



The required quantity is to be taken from the sample so as to guarantee that its 

 properties and blend represent the main sample most accurately. For this purpose the 

 sample is carefully blended by means of suitable implements as: spatulas, mixing-basins, 

 &c. The seed after being carefully blended and spread out smoothly must not be 

 shocked or shaken before being weighed ; should this happen, the blending and spreading 

 out must be done over again. The weighing consists in scooping little quantities off 

 at least five different places, until the proper quantity has been reached. Its extent 

 depends on the size of the grain and on a rough estimation of the purity, for very 

 pure samples as well as very impure ones allow the quantities to diminish. The grain- 

 size is important, so far as it determines the number of grains, found in the quantity 

 that is to be sifted and this number must not be too small with a view to unavoidable 

 errors. Formerly it was assumed as a principle that the quantity to be tested should 

 contain no less than 5000 or 1000 genuine seeds, but this principle cannot be carried 

 out into practice because the quantity of the samples might become much too large. 

 In Wageningen, in the course of time, the experience has led to the average quantities 

 stated in Appendix D. 



2. Parallel-test. 



Each determination of purity is niade in duplicate (at the least). The average 

 results are calculated in tenths of percents. The sifting of the seed must be done as 

 quickly as possible and possibly without intervals. Loss of material may not occur. 

 The difference between the sum of weights of the pure seed and the impurities, and 

 the weight of the commencing quantity, may not exceed the weight figured out for 

 the loss of water. As the waterpercentage of the seeds will diminish considerably 



') Slight damages f. i. those of Red-Clover seed — as generally known — don't necessarily cause 

 the loss of germinating capacity. Small cracks or abrasions or little cracks in the seed-lobes withont 

 loss of tissue, do not make the seed valueless, resp. impure. 



>) Notably the seeds of Alopeourus pratensis containing seeds of OKgotrophfls Alopecuri. 



