INSTITUTE AND COLLEGE, PUSA, FOR 1913-14. 27 



were five cows which produced more milk from 

 the one than from the other half udder. 



(it) In respect of fat percentage, with one exception, 

 the milk was of equal richness from either half 

 udder. The quality of the milk from the 

 several quarters of the udder was tested in 

 December and January for three cows, and here 

 characteristic differences were met with, for in 

 one case the milk was systematically richer 

 from the fore, than from the hind-quarter, in 

 another case it was systematically richer from 

 the hind, than from the fore-quarter, whilst the 

 third cow showed no such characteristic differ- 

 ences. Whilst the number of instances is so 

 small that one cannot say that such instances 

 are not exceptional, the tests were conducted 

 over so long a period, upwards of two months, 

 that the data are not to be ascribed to acciden- 

 tal error; they are real differences for the two 

 cows, however exceptional such cows may be. 



Errors involved in making milk tests. — When we 

 decided to make a series of milk tests of the Montgomery 

 cattle, it was apparent that an attempt should be made 

 either to avoid several errors which are liable to occur and 

 which very frequently minimise the value of such tests, or 

 to estimate their magnitude. These may be briefly 

 summarised. 



(i) The number of cows should not be very small. Only 

 three were included in the first period, but twelve and ten 

 cows respectively were employed in two subsequent periods. 



(id) If the periods elapsing between milkings are 

 unequal, it is known that this alone affects both the quantity 

 and quality of the milk. As there was no difficulty at Pusa 

 in making these periods equal, all the cows were milked at 

 12 hours' intervals, the variation in length of this period 

 being not more than 5 minutes, which is a negligible 

 quantity. 



