TESTING FOR MOISTURE. 211 



It is not wise to weigh less than five grains of butter 

 to be used for testing for moisture, and it is much 

 more safe to use ten or fifteen grams of butter for 

 this purpose. A slight error does not affect the re- 

 sult so much when fifteen grams are used as when 

 only five grams are used, supposing the same error 

 is made in each case. To illustrate : 



I. A lot of butter is weighed, and is supposed to 

 weigh 850 lbs., but by some error in handling it 

 weighs only 848 lbs. The loss in pounds here is 2 

 or in per cent is 2~850=--.002353Xl00=.2353— or 

 .23+ per cent. 



II. A lot of butter is weighed as 48 lbs. when 

 the actual weight is 50 lbs. Here an error of 2 lbs. 

 is made. Expressing the loss in per cent it will be 

 2-^-50X100=4. per cent. 



A loss of 2 lbs. on 850 lbs. is .23 -|- per cent or .23 

 of one pound on every hundred pounds. 



A loss of 2 lbs. on 50 lbs. is 4. per cent, or 4. lbs. 

 on every hundred pounds. By losing 2 pounds on 

 every 50 lbs. the loss is 17.+ times as great as when 

 the loss is 2 lbs. on every 850 lbs. This can be found 

 by dividing the loss in per cent: 4.-^2353 — =.17X 

 100=17. or 850-^50=17. 



This shows that Avhen an error is made in weigh- 

 ing out only 2 grams the result obtained is increased 

 enormously, while the same error made when 10 or 

 15 grams are weighed out would be very small, and 

 in many cases not noticeable in the results. 



The smaller the sample used for testing the 

 greater must be the care taken in its preparation 

 and handling, as well as in calculating the result. 



