348 Missouri Agricultural Report. 



properly balanced the end of the needle will stand at rest directly under 

 the pin. It probably will not balance until a staple of wire (n) is placed 

 over the pencils in proper position on one side of the knife blade or on. 

 the other as a counterweight. This completes the construction, and whert 

 properly mounted the balance should oscillate freely by the slightest 

 touch. As the trays are likely to be interchanged in use it is advisable to 

 mark each, placing corresponding marks on the ends of the pencil at. 

 which the trays preferably belong. 



In making seed tests we may use common BB shot (whole and frac- 

 tional) for weights. This is because we wish to know only the compara- 

 tive weights of the pure seed and of the foreign seed and other impurities- 

 in the sample. Thus, if we test an amount of seed, balancing ten shots^ 

 and find that the weed seeds it contains just balance one shot, it is evi- 

 dent that one-tenth of the original seed, or 10 per cent, consists of im- 

 purities. In other words, 90 per cent (or 90 pounds of each 100 pounds) 

 of the original seed is pure seed. In using a balance so sensitive as the- 

 one described a single BB shot is too heavy for use as the lightest weight. 

 We need a weight to be balanced by only a few clover seeds at most. 

 Very small shot is troublesome to handle and count, so we use the larger 

 BB shot, flattening the whole ones to prevent them from rolling, and. 

 cutting some into halves, quarters, eighths and sixteenths. By careful 

 selection according to weight a fairly uniform series (o) of whole and 

 fractional shots can be provided. Now, the weight of one-sixteenth shot 

 is 1 per cent of the weight of 6i/4 shots, because 61^4 equals 100-^16. 

 So if we test a sample of seed balancing 614 shots any impurity balancing- 

 the Vig-shot weight represents 1 per cent of the sample tested. If the sam- 

 ple is twice as heavy, balancing 12i/> shot, the Vig-shot weight repre- 

 sents one-half of 1 per cent of the whole. 



It is evident, therefore, that the means described enables one to de- 

 termine the quantity of pure seed or of impurities in a sample to within 

 1 per cent, or even one-half of 1 per cent of the true quantity. This is 

 sufficiently close for the practical seed testing under discussion. 



TJie Forceps. — A pair of forceps is very useful in picking up the 

 small weights used with the balance, also individual seeds. Suitable 

 forceps may be made of two thin pieces of hickory wood separated by a 

 piece of wood to which one end of each piece is fastened. The free ends 

 are flattened and pointed. A piece of spring wire bent in U shape and 

 having flattened and pointed ends serves very well as forceps. 



The Magnifiers. — After the seed to be tested has been properly 

 weighed it is to be separated into pure seed and foreign seed or other 

 impurities. This requires a magnifier. Very coarse seed — such as that 



