No. 6 * DEPARTMENT OF AGRICULTURE. 217 



iain that the samples sent it are correctly taken. Hence we are care- 

 ful in our reports to those sending us samples to disclaim all re- 

 sponsibility as to the accuracy of sample taking. I think, however, 

 that the bulletin to which I have referred', which was sent to be 

 posted at every creamery and cheese factory in Vermont, which was 

 mailed by thousauds (hroughout our State to the station uuiiling li»t, 

 aud which concludes this article, does help to make the samples that 

 come to us more uniform and trustworthy. 



HI. VARIATION IN TESTS WITHIN THE SAME HERD. 



Why is it that Smiths mill) or cream talcen to BrowrCs creamery 

 varies one month with another? AYhy does it not test evenly? 

 ScA'eral of the reasons cited under the first head apply here. 



LACTATION CHANGES. 



The change in lactation of the cows is one important reason why 

 there should be variation. The general tendency of the herd will be 

 as the cows go along in lactation to give somewhat richer milk. 

 While there are many exceptions, the general rule is that cows com- 

 ing in in the spring will give a fairly even grade of milk for the first 

 live months in their lactation, and then increase in quality until they 

 go dry. If they are farrow cows, quality changes but little as time 

 goes on. If an all-the-year-round dairy is kept there should be less 

 change on this account. 



These same changes pertain to the cream. The richer milk is apt 

 to make richer cream for reasons hitherto pointed out, if it is handled 

 I.."^ a deep-setting device. Centrifugal separators, however, are no 

 respecters of rich or of poor milks. A rich cream may be made from 

 one and a poor cream from the other, according to the setting of the 

 cream-screw or regulating device. If, however, this remains un- 

 altered and the same proportion of milk is taken as cream from the 

 rich and from the thin milk, creams will vary accordingly. 



For example, if one dairyman has 1,000 pounds of new milk testing 

 3 per cent., and the other, 1,000 pounds of stripper milk, testing 5 

 per cent, and each takes 100 pounds cream and 900 pounds skim milk, 

 the former would have a cream testing nearly 30 per cent, and the 

 latter, one containing approximately 50 per cent. fat. 



WEATHER. 



stress of weather is another cause of variation. We have given 

 much time at the Vermont station to the study of the effect of tem- 

 perature upon the milk-flow. Our results indicate that the quality 

 of a cow's milk niters inversely to temperature changes. When the 

 temperature rises the tendency is for the quality of the milk to drop; 

 when the temperature falls the tendency is for the quality of the milk 

 13 



