PREPARATION OF SAMPLES. 



11 



sentative of the hulk sample. Tn order that the small sample may 

 be representative, the hulk sample should he thoroly mixt and small 

 portions taken from dilierent i)arts of it. The greatest aceuracy . 

 can be secured thru the use of some form of mechanical mixer and 

 sampler, and the one shown on ])a<;es 12 and 13 of Circular ;U. Re- 

 vised. Office of Experiment Stations, modified so as to be adapted 

 for larger grain, is recommended. 



SIZE OF SAMPLE FOR THE MOISTt'RE TEST. 



The size of the sample to be taken for the individual moisture test 

 may l)e varied, but experience has shown that 100 grams of whole 

 kernels give the most satisfactory results, which quantity has there- 

 fore been established as the standard for the method and apparatus 

 for testing corn herein described. This is a sufficiently large quan- 

 tity to insure the securing, without difficulty, of a representative 

 sample. Moreover, when samples containing 100 grams are used, 

 every cubic centimeter of water expelled from the grain represents 1 

 per cent of moisture, and the readings in the graduated cylinders are 

 in percentages as well as in volume, thereby reducing the chances of 

 error to a minimum. 



WEIGHING THE SAMPLE FOR THE MOISTURE TEST. 



In makinir moisture determinations according to the method out- 

 lined in the foregoing pages, the use of delicate analytical balances 

 is obviated. An ordinary torsion 

 balance similar to the one shown as 

 fio-ure 1, which is sensitive to one- 

 thirtieth of a gram and can be pur- 

 chased for about $15, will serve 

 every purpose. A more delicate 



balance is entirelv unnecessary when fig. 2.— Aluminum weighing pan lor trans- 

 it is remembered that one kernel of f^"'"^ &^^*° ^^^p^^-^ ^« ^^^' elimination 



flasks. 



corn weighs approximately one- 

 third of a gram. The scale pans are inches in diameter, and the side 

 beam shows 5 grams, graduated in one hundred divisions, each one- 

 twentieth of a gram, and the total capacity of the balances is 2 pounds, 

 or 907 grams. Balances of this kind are easily operated, and the time 

 consumed in the weighing need not be longer than that reciuired for 

 the pharmacist to weigh out the prepared drugs for his medicines. 



A specially constructed scale pan, such as is shown as figure 2, 

 should be secured to facilitate the transferring of the weighed 

 samples to the distillation flasks, the opening in the end of the scoop 

 being of the same size as the neck of the flasks — 1 inch. If made of 



'to 

 99 



