A MOISTURE TESTER FOR GRAIN. 



13 



graduated glass gauge and a stopcock fitted to a cylindrical brass cup 

 of a size to hold sufficient oil for at least a dozen tests, as shown in 



figure 12. 



DRAWING AND HANDLING SAMPLES. 



.. — Oil measure. 



It is admitted that to secure a representative sample is often very 

 dilhcult, but the necessity of sparing no trouble in this respect is 

 most strongly urged. 



In securing samples of grain from cars, etc., it 

 must be remembered that, as a rule, grain from the 

 surface will show less moisture than that in the body 

 of the shipment. In corn in cars it 

 is not uncommon to find differences 

 of several per cent between the 

 moisture content of the layer of 

 corn immediately at the surface and 

 that of the body of the grain di- 

 rectly beneath. In an experimental ^'°- "--^^ 

 shipment of three lots of corn contained in a single 

 car in which special care was taken to see that each 

 lot was uniform througliout, moisture tests at des- 

 tination, fifteen days after shipment, showed 14.1 

 per cent, 14.3 per cent, and 14.1 per cent for corn 

 from the surface of each o( the three lots, as com- 

 pared with 17.3 per cent, 16.9 per cent, and 19.2 

 per cent, respectively, for the samples taken with 

 a grain trier from 2 to 3*^ feet below the surface, 

 the latter percentages being practically the same 

 as at the time of shipment. 



Likewise, the same care must be observed in 

 handling the samples after they are drawn from 

 the bulk or weighed for the test. If exposed to the 

 air they will lose moisture in accordance with the 

 Fig. i2.-improved oil condition of the atmosphere and the percentage of 

 ZZlJZ^Z water in the sample. A preliminary test with 100 

 represents 150 cubic grams of wlicat containing 14.5 per cent of moisture 

 centimeters of oil. ^^^^^^.^^ ^ loss of 0.20, 0.35, 1.10, and 1.90 per cent 

 when exposed to the air of the laboratory for periods of one-half, one, 

 four, and twenty-six hours, respectively. A second lot of wheat con- 

 taining 16.1 per cent of moisture lost 0.50, 0.75, 1.75, and 3.20 per 

 cent, respectively, during the same periods under identical treatment. 



BALANCE FOR WEIGHING SAMPLES. 



While not a part of the apparatus, a balance is necessary for weigh- 

 ing the samples. One of simple construction, having a sensitiveness 

 of one-twentieth or even one-tenth of a gram has been found quite 



tCir. 72] 



