INSTITUTE AND COLLEGE, PUSA, FOR 1913-14. 29 



similar records which have been obtained elsewhere. The 

 principal deductions depended on the arithmetical means of 

 all the tests of any one period, and the probable error then 

 fell to rather less than + -1 both in the case of yield and per- 

 centage of fat. We were thus enabled to decide with great 

 accuracy what differences were due to error of experiment 

 and what were real differences in yield or quality. 



Error due to milking in the ordinary way. — During a 

 part (28 days) of the third period, September to Nov- 

 ember, the ten cows were hand milked as for profit, that 

 is, the calf was allowed as little as possible, the object being 

 to ascertain the magnitude of the errors which such a 

 method would incur. The yields by hand were about 50 

 per cent, greater during this time, which was to be expected. 

 It is also obvious that under such conditions the real 

 quantity of milk secreted could not be ascertained. But 

 in addition to this, the probable error of an individual 

 milking rose from + -26 to + -74, that is, it became three 

 times as great. 



Detection of added water in milk. — During recent years 

 the reliability of the freezing point of milk as a criterion 

 of its freedom from added water has been recommended, 

 chiefly by Brownlie Henderson in Queensland. It has been 

 found by him and others that the freezing point of pure 

 milk is so constant that a comparatively small addition of 

 water, e.g., 2 per cent., can be detected. At present the 

 method is purely empirical and would only become depend- 

 able if a large number of tests showed what the variation of 

 freezing point of pure milk is. To this end the freezing 

 point of the milk of a number of individual cows and buffa- 

 loes at Pusa, at the Lyallpur Agricultural College and at 

 the Military Dairy Farm, Peshawar, was determined, and 

 it seems certain that even among individuals the varia- 

 tion in freezing point is only small. At Pusa it varied 

 from --543° to--577°C, at Lyallpur from --527° to 

 -•562°C, and at Peshawar from --529° to --564°C. 

 The variations between individuals are naturally greater 

 than between samples of the mixed milks of a herd; thus 



