REPORT OF THE CHEMIST. 149 



THE BABCOCK TEST. 



In June of this year, a bulletin (No. 13 of the Dairy series) was issued, in which 

 I gave full instructions for working this test in the dairy, for single and composite 

 samples— the latter by the use of potassium bichromate as a preservative. Informa- 

 tion was also added, necessary for the calculation of the payments to patrons of 

 creameries and cheese factories in which the percentage of fat in the milk has been 

 adopted as a basis of remuneration. 



Since that date, many questions have been received regarding the manipulation 

 of the composite test. Chief among these have been those that referred to the max- 

 imum and minimum quantities of potassium bichromate that can be used without 

 affecting the reading of the fat obtained. Several series of experiments were ac- 

 cordingly instituted to establish the limits between which this preservative can be 

 employed with certainty, and also to ascertain the length of time a composite sam- 

 ple so treated may be kept without showing a diminution in its fat contents. 



Composite samples were prepared in duplicate from the milk of a herd which, 

 during the week that the samples were taken, gave the average of 3"40 per cent of 

 buttei'-fat as the mean of the daily tests. 



While preparing the composites (from Monday to Saturday) and until the test 

 was concluded, the bottles were gently shaken every day to incorporate the risen 

 cream and to prevent the latter from sticking to the sides of the bottle. The tem- 

 perature of the samples throughout the test would average about 63°F. 



Experiment A, 



Date of composite samples. May 8th-15th, 1893. 

 Mathematical average of daily tests : 3*40 per cent fat. 

 Composite sample A. =600 c.c. milk-j- '3 grms. potassium bichromate. 

 «* " B. = 600 c.c. " + -6 " 



{Note. These quantities of potassium bichromate are equal to 4J grains and 9 

 grains to the pint respectively.) 



The percentage of fat was ascertained in these samples on the subjoined dates, 

 with the following results : — 



A. B. 



Fat, Fat, 



Date. per cent. per cent. 



May 15th 3-4 3-4 



" 20th 3-4 3-4 



" 22nd 3-4 3*4 



" 27th 3-4 3-4 



June 3rd 3-4 3-4 



" 10th 3-4 3-4 



" 17th 3-4 3-4 



These tests were made with the usual quantity of acid and gave clear readings 

 throughout. 



It may fairly be concluded from these results that (1) any quantity of potassium 

 bichromate between 4|- grains and 9 grains to the pint serves equally well in pre- 

 serving the milk in a fluid condition without interfering with the accuracy of the 

 Babcock test, and (2) that milk so treated, if kept carefully shaken and moderately 

 cool, shows the same percentage of fat for at least one month. 



