BY H. S. HALCRO WARDLAW. 825 



The following data were also obtained at, the time of the collec- 

 tion of the samples: — 



(4) Tlie diet of the cows. The daily rations of the various cows 

 were as follows: — 



Cows, Nos. 1-22 — Pollard, 2 51bs. ; bran, 701bs. ; green barley- 

 chaft', ad lib. 



Cows, Nos. 23-34 — Bran, 81bs. ; green chaffed oats, (iOlbs. 



Cows, Nos. 35-46^Bran, Slbs. ; lucerne-liay, 301bs. 



Cows, Nos. 47-109 — Bran, Slbs. ; cow-meal, 21bs. ; lucerne-chaff, 

 161bs. 



In no instance had the diet of the cows been changed immedi- 

 ately before the collection of the samples of milk. The shortest 

 period, which elapsed between a change of diet and the collection 

 of samples, was three weeks. It occurred in the case of cows, Nos. 

 35-46. 



The period during which samples of milk were collected 

 extended from October 31st, 1916, to April 30th, 1917. This 

 period was one of six months, and included the latter half of 

 Spring, the whole of Summer, and the first half of Autumn. 



The cows examined were all milked twice each day. The 

 morning-milking occurred between six and seven o'clock, the after- 

 noon-milking between three and four o'clock. The period between 

 the milking on one morning, and that on the following afternoon, 

 was thus about half that between the afternoon-milking and that of 

 the following morning. It is to be expected, from what has been 

 stated already with regard to the proportionality between the time 

 of rest and the yield of milk, that the amount obtained at the 

 afternoon-milkings, from which the samples were obtained, would 

 be about half that obtained at the corresponding morning-milkings. 

 To test this supposition, the weights of milk obtained in tlie morn- 

 ing were measured, as well as the weights obtained in tlie after- 

 noon, in the cases of a number of cows. The following Table gives 

 the results obtained. 



