The Variation m Milk Tests. 



25 



a single sudden variation would not be recorded against him, the con- 

 clusion being that some accident had occured to the sample. Needless 

 to say, in such a case the manager would like two or more tests to ■ 

 ascertain which was correct. 



There are many causes of the variation in the fat contents of cow's 

 milk which sometimes show astonishing changes. In Bulletin No. 24, 

 Illinois, U.S.A., is the following record of the variation of one cow's 

 milk during a period of nine months : — 



Average for Month. On Any One Day. 



The largest variation in two successive months is 0'5 per cent., 

 July and August. The greatest difference in any two months is I'l 

 per cent., February and August. The highest test on any one day is 

 in August, 7'9 per cent., and the lowest in January, 2"7 per cent., 

 difference 5'2. The largest variation on any one day was 7*9 to 2*9, 

 a difference of 5"0 per cent. 



The Journal of the Highland and Agricultural Society of Scotland 

 gives the results of upwards of 700 analyses of the milk of eighteen 

 cows over a period of 21 days, each cow's milk having been collected 

 and analysed separately morning and evening. Amongst the con- 

 clusions arrived at it is stated that the proportion of fat in genuine 

 cow's milk is liable to greater variation than is generally admitted. 

 Samples containing less than 3"0 or even 2*0 per cent., may often occur 

 while the same animals may, within a few days, yield milk with 

 6'0, 6'0, or even 7'0 per cent, of fat. Of course, the average tests of 

 whole herds will not vary to so great an extent, but as a general rule, 

 it may be taken that what will affect a single cow will also, more or 

 less, affect the herd. The tests of the Jersey herd at the World's 

 Fair, Chicago, 1893, will illustrate this. 



July 



The greatest difference was fi-om 19th to 20th, 0*4 per cent. 

 According to Fleischman, one of the greatest authorities on dairy- 

 ing, the test of a herd may go more than 30 per cent, above or 

 below the average for the year. For example, if the herd's average 

 test is 5 per cent, on some days it may be as high as 6"5 and others 

 as low as 3*5. The same thing has been proved in England by 

 Lloyd when conducting investigations for the Imperial Government 

 in 1895 at Hazelbury in Somersetshire. The average test of the 

 herd was : — in April 3*7 per cent., May 3'39, June 3'5, July 3*6, 



