570 



STATE BOARD OF AGRICULTURE. 



The average time of souring, pH score, reduction time, and bacterial count 

 of samples appearing in Table VII is retabulated in Table VIII. The high 

 bacterial counts of several samples with scores of 75 to 100 make it appear 

 that the majority of these samples have much higher counts then is actually 

 the case. One count of a sample of milk souring in 28 hours has been disre- 

 garded because it raised the average of the samples in the group by over 

 1,000,000. 



Table VIII 



The average time of souring, pH score, reduction time and bacterial count of 



samples appearing in Table VII. 



*One sample with a count of 16,000,000 was ignored in making average. 



An examination of Table VII shows that all of the methods studied are 

 satisfactory indicators of samples of milk with very good keeping quality 

 or samples with very poor keeping quality. The real test of the various 

 methods, however, is; which one wall indicate the proper position of the 

 samples just on the border line between the good and the ,bad? 



The writer with assistance from members of the department has selected 

 24 hours as the time that milk should be expected to remain sweet after de- 

 livery in the home and storage at room temperature, 70 °F. The samples 

 of milk souring in 22 hours, 24 hours, and 26 hours then, may be considered 

 on the border line between good and bad milk. 



The correlation ratio* between the actual keeping quality of the samples 

 of milk selected as above, and their pH score, reduction time and bacter- 

 iological count is as follows: 



pH score 75 



Reduction time 52 



Bacteriological count — .25 



When used to place the samples of milk as to their keeping quality the 

 reduction time is thus seen to be subjected to about 100 percent greater error 

 than is the pH score. The bacteriological count is subject to about 300 per- 

 cent greater error. 



Hastings and Davenport (7) in a recent comparison of the relative value 

 of the methylen blue reduction test, the brom-thymol blue test and the brom- 

 cresol purple test arrive at the following conclusions. 



"In a comparison of the three tests, methylene blue reduction, brom- 

 thymol blue, and brom-cresol purple for the determination of the keeping 

 quahty of milk, it is shown that the methylene blue reduction test is prefera- 

 ble because of its greater sensitiveness to biological differences in milks; 



*See Elements of Statistical Methods — King. Introduction to Mathematical Statistics— West. 



