360 DEDUCTIONS FROM ANALYSIS. 



with one of the numerous arrangements for keeping the cream 

 mixed automatically ; the action of these is, too, aided materially 

 by the motion the milk receives from the movement of the 

 cart or barrow when drawn along the streets. A steady current 

 of about 1 foot per hour is enough to keep milk mixed, and, 

 except in very hot weather, milk is not churned by gentle stirring. 

 To ensure that the composition of the milk will not vary the 

 minutest fraction perhaps demands more skill and attention 

 than the average milk distributor possesses, but a practical 

 compliance with the judgment mentioned above can and oughb 

 to be obtained. The author has calculated from a large number 

 of results that the probable variation of fat in milk due to cream 

 rising is only 0-11 per cent. 



It cannot be insisted on too strongly that the calculated 

 standards apply only to the mixed milk of a number of cows ; 

 the milk of a single cow may be below these figures to a serious 

 extent. As this case is one which occurs but rarely the sale of 

 milk being confined almost entirely to the product of herds 

 it is not necessary to make any allowance for the greater 

 variations of quality of the milk of individual cows. 



Appeal to the Cow. In cases of doubt it is advisable to 

 resort to what is known as. " appeal to the cow," or the " stall 

 or byre test." This consists in having the cow or cows 

 milked in the presence of a responsible witness who can certify 

 to the absolute genuineness of the milk, which is analysed and 

 compared with the suspected sample. It is desirable, if possible, 

 that the milk of the morning and evening meals both should be 

 examined. To make the test as fair as possible, the cows should 

 be milked by their usual milkers at the same time of day as the 

 previous sample, and under the same conditions ; the test should 

 be carried out at as early a date as convenient, and care should 

 be taken that the meteorological conditions are nearly alike, as 

 a poorer milk is yielded in warm, damp weather than if it is clear 

 and frosty. The test should not be carried out on a Sunday or 

 Monday, or on a public holiday or its morrow, unless the previous 

 sample was taken on a similar day, as it has been shown that the 

 irregularity in the time of milking which occurs on such days 

 affects the quantity and quality of the milk ; and serious 

 divergence from the average quantity of milk yielded may be 

 looked upon as throwing doubt on the reliability of the test. 

 The witnesses must be specially careful in seeing that the cows 

 are milked out, and that nothing occurs likely to disturb the 

 equanimity of the cows, such as undue commotion or noise. If 

 the milk is cooled, it is the duty of the witnesses to satisfy them- 

 selves that the refrigerator does not leak, as well as to see that 

 all vessels into which milk is received are clean and dry. 



