954 Summary of Agricultural Experiments, [march, 



The reliability of milk records based on samples taken at various 

 intervals was also examined. The estimate based on the occasional 

 samples was under the actual amount when the whole of the cows 

 were considered, when the interval was seven days, but above it when 

 the interval was longer. The possibility of error in the estimate 

 increased with a longer interval, the error on the average for the 

 seven-day interval being apparently doubled at fourteen days, trebled 

 at twenty-one days and quadrupled at twenty-eight days. 



Milk Records (Highland and Agric. Soc, Transactions, Fifth Series, 

 Vol. XX., 1908). — A report is given on the results obtained in 1907 

 in the experiments in milk testing-, which are now in their fifth season. 

 These experiments were described in an article by Mr. John Speir in this 

 Journal, September, 1907, p. 320. 



Milk Records (Lanes. Educ. Comm., Agric. Dept., Bull. 5). — This 

 experiment was intended to ascertain the degree of accuracy attained 

 by sampling at intervals of one, two, three, or four weeks. The con- 

 clusion reached is that the total yield of milk and the percentage of 

 fat contained in it can be estimated with approximate accuracy by 

 weighing and testing the morning- and evening milk on one day each 

 fortnight, and multiplying the average of the two results by fourteen. 

 When the test was made on one day in three weeks, although the average 

 result from twelve cows was approximately correct, the error was 

 considerable in the case of individual cows. A note on these experiments 

 was given in this Journal, March, 1908, p. 740. 



Milk Records (Essex Educ. Comm., Notes on Agric. Analyses, 

 1903-6; Harper-Adams Agric. Coll., Field Expts., Report, 1907). — 

 Records are given in the Essex Report of the composition of the milk 

 of a herd for two years, and of another for three years. 



The Harper-Adams Report contains the results obtained from the 

 samples submitted by farmers, and also records of production and 

 percentage of fat in milk from nine cows in the College herd. A note 

 on the variations observed at this centre in milk taken from the top 

 and from the bottom of a churn during delivery was given in this 

 Journal, June, 1908, p. 213. 



Effect of Milking Intervals on the Composition of Milk (Camb. 

 Univ. Dept. of Agric, Guide to Expts., 1907). — Three cows were milked 

 for a fortnight at equal intervals, and for a fortnight at unequal 

 intervals (16 and 8 hours), and diagrams are given showing the 

 composition and yield of the milk of each cow during the period. 

 The composition of the milk of the individual cows varied greatly from 

 day to day, but with milking at equal intervals the percentage of fat 

 was generally above 3, while with unequal intervals the percentage 

 in the morning's milk was usually (in the case of one cow always) 

 below 3. With equal intervals the average percentage was, morning 

 3*59, evening 3*43 ; with unequal intervals, morning 2*40, evening 4*41. 



Effect of Milking Intervals and of Rain on the Composition of Milk 

 (Lanes. Educ. Comm., Agric Dept., Bull 5).— In the course of the 

 experiments comparing different concentrated foods referred to above, 

 some causes of the variations in the percentage of fat in milk were 

 investigated. The intervals between the times of milking were io| 

 and 13 hours, and the morning's milk contained from '6 to "8 

 per cent, less fat than the evening's. On Sundays, however, the cows 

 were milked f hour earlier in the afternoon, the intervals then 



