TRUE AVERAGE TEST OF CREAM 323 



composite sample would show the same as the direct average of 

 these four, or 29.25 per cent. That this test is too high is readily 

 seen when we divide the actual quantity of fat 127.4 by the 

 quantity of cream, 490. Performing this we see that the true 

 average test, instead of being 29.25 per cent is only 26 per cent. 

 The direct or false average, therefore, is 3.25 per cent too high. 

 An error of this magnitude operating upon 490 pounds of cream 

 throws an error of 15.925 pounds of fat. If fat is being bought 

 at the rate of 30 cents per pound an over-payment of $4.78 has 

 occurred (490 times 0.0325 equals 15.925; 15.925 times 30 

 equals $4.78). It is evident, therefore, that any butter maker 

 who makes composite samples of cream delivered by his patrons is 

 almost certain to make errors of this sort. While in the long 

 run and on the various patrons there may be a balance or com- 

 pensation it certainly is an inaccurate and unreliable method of 

 computation. Likewise, a farmer, in endeavoring to check up 

 the accuracy and honesty of the creamery operatives, is more 

 likely than not to make an error of this kind, either one way or 

 the other, if he tries to make composite tests. Each and every 

 can or shipment of cream should be weighed, sampled, and tested 

 by itself. It should be remarked in passing that this is the 

 practice with the larger and better-conducted creamery plants. 



Testing Skim Milk. — Skim milk is tested in the same man- 

 ner as whole milk, except that the skim milk bottle is used. 

 The essential difference between the skim milk bottle and 

 the whole milk bottle is in the size of the tube or bore of the 

 neck. At present most skim milk bottles are so made that the 

 entire length of the graduated scale equals one-fourth of one 

 per cent (0.25 per cent), and each of the five grand divisions 

 of the scale has a value of five one^hundredths per cent (0.05 

 per cent). In testing skim milk care must be taken to make 

 sure that no drop of water or other plug remains in the 

 graduated neck at the time of pouring the milk, or the acid, into 

 the bottle. A drop of water lodged in the neck, very frequently, 

 indeed will cause the air to bubble back against the acid or milk 

 while being poured down the large neck, causing same to spill. 

 If any portion of the sample is wasted, the whole is lost. 



