740 



Milk Tests and Records. 



[march, 



In this connection it may be mentioned that it is stated in 

 the United States Consular Report (No. 3075) that there are 

 many indications that the State of Minas Geraes is determined 

 upon a course of expansion in agriculture. The administra- 

 tion has begun a systematic campaign for the introduction of 

 modern farming implements, and as an experiment has recently 

 ordered 150 disc ploughs from an American manufacturer, 

 and also 20 farm wagons. These are to be furnished to 

 colonists or established farmers on easy terms in the hope of 

 awakening a widespread interest in up-to-date methods of 

 agriculture. This State has also arranged for an exhibiton 

 of domestic animals, and offers free transportation from all 

 parts of the State, with free forage during the exhibition. 



The frequency with which the milk of cows should be tested 

 in order to obtain an accurate indication of the average richness 



of the milk is a matter of considerable 

 Milk Tests and importance to any dairy farmer who 

 Records. makes a practice of keeping dairy records. 



Particulars of some American experiments 

 on this point were given in an article in this Journal, May, 1907, 

 and these may now be supplemented by a comparison of actual 

 with estimated results obtained at the Lancashire County 

 Council Farm (Bulletin No. 5). At this farm the milk of 

 twelve cows was tested daily for a period of ten weeks, and this 

 gave an opportunity of judging of the error which would arise 

 by estimating the fat content of the milk at intervals of one, 

 two, three and four weeks. The tests for the percentage of fat in 

 the milk showed that the estimated results for the whole twelve 

 cows taken together agreed very closely with the average actual 

 results obtained by testing daily, as the errors in the case of in- 

 dividual cows compensated one another. The errors in the in- 

 dividual results, on the other hand, were in some cases con- 

 siderable. With samples taken at intervals of seven days the 

 error in six cases was 3 per cent., in two cases 5 per cent., 

 and in one case it was 11 J per cent. In three cases the 

 sample gave a correct result. The position was much the 

 same when the test was taken fortnightly, but when it 

 was taken at intervals of three weeks or a month, the errors 



