1907.] 



Milk TESTS and RECORDS. 



91 



are required in order to give an accurate indication of the 

 average richness of the milk. Cows vary so much in the 

 amount and quality of their milk from one milking to another 

 that exact results cannot be obtained by testing the milk 

 from one milking at distant intervals, while at the same time 

 it is obvious the testing of the milk of every cow in a herd is 

 too laborious or expensive a process to be undertaken at very 

 close intervals by the ordinary dairy farmer. 



In Denmark the testing of milk is done co-operatively by 

 a number of farmers combining to employ a man who goes 

 from farm to farm and weighs, tests, and records the milk pro- 

 duced at each in twenty-four hours, and in these cases the 

 milk i^ usually tested fortnightly.* The Wisconsin Experi- 

 ment Station, which has devoted much attention to this 

 subject, recommends sampling once a week. The Illinois 

 Experiment Station recommends sampling eacli cow's milk 

 for fourteen consecutive milkings every seventh week, and it 

 has been proved that so far as average results are concerned, 

 figures deduced from tests on this basis are substantially 

 correct. 



The number of tests required during a lactation period is 

 ill um rated further by an experiment made at the Illinois 

 m.i lion, in which the milk of each of six cows was weighed 

 and analysed daily during the whole period of lactation. 

 These daily records enabled the actual results to be compared 

 with the results which would be obtained from estimates cal- 

 culated from the tests at intervals of 7, 10, 15 or 30 days. 

 Testing the milk every seventh day gave 98 per cent, of the 

 actual total butter-fat; testing every two weeks gave o< s '3 

 per cent., and testing every month gave 07 per cent, or an 

 error of 3 per cent. 



The Vermont Experiment Station also has given some atten- 

 tion to thequestion when a cow should be tested in order to give 

 a correct idea of the year s production if only one or two t sts 

 are made during the lactation period. This station recom- 

 mends, when only two tots <>t each cow's milk are to be made 

 during the same lactation period, that samples should be 



taken as h >llows : — 



• Milk Testing and Control in Denmark.— Journal, April, 1905, vol. \ii. p. 21, 



