78 EXPERIMENT STATION EECOED. 



with tin lids will save the average creamery several hundred dollars per month. 

 Bottles with glass stoppers form the most satisfactoi-y seal, but are expensive. 

 Cork stoppers of good quality form a tight seal, but have a tendency to loosen 

 automatically. The increase in the test due to storing varied with the amount 

 of cream in the sample jar. With the size of the bottles unchanged, the smaller 

 the amount of cream, the greater the increase in the percentage of fat. Each 

 step in the process of testing cream is critically discussed, and suggestions are 

 offered for more accurate testing. 



" Ninety-six tests with cream and acid at temperatures ranging from 40 to 

 110° F. showed no visible changes in the percentage of fat and the clearness of 

 the test where the amount of acid used was govei'ned by the color of the mix- 

 ture. Where the amount of acid was not regulated by the color of the mixture, 

 but where the same amount of acid was used at all temperatures, the resulting 

 tests varied widely. In the case of cream and acid at 40°, it required more 

 shaking to effect complete action of the acid on the cream, but the tests were 

 clear and the percentage of fat correct. In the case of cream and acid at 100" 

 or above, the tests were very dark and charred, and the meniscus difficult to 

 read." 



" The average of the tests with two additions of water practically agreed 

 with the gravimetric fat estimation. The average of the tests with one addition 

 of water was 0.17 per cent higher than the gravimetric fat estimation." 



A series of 547 tests was made in an 18-inch tester. The speed used varied 

 from 100 to 1,000 revolutions. A reduction of speed from 1,000 to 400 revolu- 

 tions caused a gradual decrease in the test of 0.34 per cent of fat. 



A series of 26 test bottles, comprising 18 gm. 50 per cent, 18 gm. 40 per cent, 

 and 9 gm. 50 per cent bottles, were filled to the necks of the bottles with water 

 and weighed on analytical balances. Pure butter fat was added and after weigh- 

 ing the bottles were then whirled in the tester at full si>eed and the tests were 

 read at 135 to 140°, this being the temperature at which pure butter fat has a 

 specific gravity of 0.9. The meniscus amounted to exactly 1 per cent, except in 

 two of the 9 gm. 50 per cent bottles, where it was 1.25 and 1.5 per cent, respec- 

 tively. On repeating the experiment with cream the individual results lacked 

 uniformity; nevertheless, a comparison between these averages and those ob- 

 tained from the experiment with pure butter fat showed unmistakably that the 

 fat column of the ordinary cream test was not pure butter fat, but contained 

 impurities, mainly water and acid. A further study led to the following con- 

 clusions : 



" The percentage of Impurities contained in the fat column varies with the 

 speed of the tester and the amount of sulphuric acid used. Low speed and 

 excessive acid increase the impurities. 



"All of the fat in the cream does not appear in the fat column. The fat lost 

 in the test is the residual fat and a portion of the volatile fatty acids. 



" The loss of residual fat is greatest at a low speed and with insufficient or 

 excessive amounts of acid. The loss of volatile fatty acids increases with the 

 increase in the amount of acid used. 



" The losses due to residual fat and escape of volatile fatty acids are prac- 

 tically offset by the gains due to impurities in the fat column, provided that 

 the test is made under normal conditions." 



Tests made from a hot and moist tester and those from a cool and dry tester 

 showed no greater variation in the impurities of the fat column than tests from 

 the same tester. Samples of cream varying widely in richness were tested in 

 9 gm. and in 18 gm. 50 per cent cream test bottles, where the charges were 9 

 and 18 gm. respectively. The results showed that the richness of cream, style 

 of test bottle, and size of charge had no appreciable effect on the percentage of 



