fifteen cow!S is 55 pounds., or over 3^ pounds butter per cow for the 

 month. Some of this may be due to an advance in lactation, or some to 

 errors in calculation, but the evidence seems to point very strongly to- 

 wards quite a marked decrease in milk and butter during- the first month 

 after the installation of the machines. However, we should expect a 

 decrease with most cows after making- so radical a change, as from hand 

 milking to a method new and strange to the cows. 



Special Tests Comparing Hand and Machine Milking. 



From time to time during the past year special trials comparing 

 hand and machine milking have been made. These tests, however, were 

 not altogether satisfactory, as it is almost impossible to get an exact 

 comparison. There is no way of ascertaining what a cow might have 

 given under other circumstances and conditions. The most we could do 

 was to keep feed, water, temperature of stable, etc., as nearly alike as 

 possible during the periods of tests, and assume that any differences in 

 milk yield were due to methods of milking. This may or may not have 

 been correct, but it was about the only thing which we could do under 

 the circumstances. While most of these tests were for but short periods 

 of time, the difficulties of comparison are even greater when we compare 

 one lactation period with another. How can we say that differences in 

 milk yield for one lactation period as compared with another are due to 

 methods of milking or to any other one factor? We may tliink and say 

 it is so, but it would be very difficult to prove our assertion. 



For the sake of comparison, we give the results of some of these tests 

 which have been made during the year. The machine was installed dur- 

 ing the last weeks of December, 1905, and we began using it January 

 1st, 1906. It was thought advisable to allow the cows at least one month 

 to become accustomed to the machine before making any special tests. 



The first comparison was made on four cows, Nos. 15, 56, 106, and 

 107, beginning February 1st, 1906. The milk of each cow was kept 

 separate, weighed and sampled for testing with the Babcock test every 

 seven days for the two weeks. From February 15th to 28th, inclusive, 

 these four cows were milked by hand. The milk from each was weighed 

 and sampled daily for the two weeks. From March 1st to 14th the same 

 cows were milked with the machine. Milk was weighed and sampled 

 same as for the previous machine milked period. If we average the 

 machine period before hand milking, with the machine period after hand 

 milking, we shall balance as nearly as can be the effects of the period 

 of lactation. 



The following table shows the average yields of milk and fat for the 

 two machine periods and also for the period of hand milking: — 



