THE TESTING OF SEED 63 



to the Editor and Chief of the Division of Publications, De- 

 partment of Agriculture, Washington, D.C. Its many 

 illustrations show both the pure seed and its adulterants, 

 which are chiefly weed seeds that cause much trouble. 



When we have thus separated the sample, we can estimate 

 its value. This can be done roughly, by guessing at the 

 amount in the three piles. Or it can be done accurately, by 

 weighing. The scales should be delicate ; yet if the student 

 does not own or care to buy a set, he can, if he is clever with 

 his hands, make one himself by studying Exercise 12 in 

 Farmers' Bulletin 408. 



The combination of the germination test with the purity 

 test of seed is very valuable. After one has separated out 

 the pure clover seed from the sample, he may find that 

 many of the seeds are small, dull in color, or shriveled. 

 Let him count out a hundred of the seed, taking them just 

 as they happen to come, and test them for sprouting. 

 . From the figures that these various tests give him, he can 

 make very valuable records. For each test the student 

 should put in his note-book first the weight of the sample, 

 theh the weights of pure seed, weed seed, and waste matter. 

 Next he can calculate the percentage of pure seed, by dividing 

 its weight by the weight of the sample, and multiplying by 

 one hundred. He has already been shown how to find the 

 percentage of germination of the seeds tested. But most im- 

 portant of all is the percentage of good seed in the whole 

 sample. This is found by multiplying the two percentages 

 already found, and dividing by one hundred. Thus if in a 

 sample there is but seventy per cent of pure seed, and if eighty 

 per cent of that germinates, then in the whole sample there 

 is but fifty-six per cent of pure seed. 



A few tests of this kind will show which is really cheap 

 seed and which is not. Seed which costs less may often be 



