78 STORRS AGRICULTURAL EXPERIMENT STATION. 
The total cost of “each of the rations was large, although the 
second ration was slightly less expensive than the first. Quite 
a number of cows in the herd were well along in the period of 
lactation, and were no doubt being fed too heavily for the 
amount of product they were giving. It is very interesting to 
note that the experiment seems to corroborate results obtained 
at the Massachusetts and Pennsylvania Experiment Stations* 
in showing that heavy protein rations, if any, are the ones that 
tend to increase the percentage of fat in the milk. 
COMPARISON OF TESTS WITH WIDER AND NARROWER RATIONS. 
The experiments of the last three seasons—1 893-94, 1894-95 
and 1895-96—include nine cases in which comparative tests 
were made by feeding two different rations in succession to the 
same herd, in the manner described above, pages 53 and 54. 
In each case the ration actually being fed in the ordinary 
method practiced on the farm was determined by weighing the 
feeding stuffs on the spot as they were fed and taking samples 
for analyses, and at the same time weighing the milk of each 
cow and determining the percentage of butter-fat. As soon as 
the analyses of the feeding stuffs could be made, so as to calcu- 
late the amounts of nutrients, another ration was made up 
which was assumed to be a nearer approach to the accepted 
standards and a second test was made with this ration, the 
fodder and milk being weighed and analyzed as before. In 
eight of the cases the second ration was narrower than the first: 
in one instance the first ration was comparatively narrow, and 
the change was mainly from more to less expensive food mate- 
tials. The length of each test was twelve days. ‘The interval 
between the two tests of each comparative trial was from two to 
four weeks in the seven comparative experiments of 1893-94 
and 1894-95, and nine days in the two of 1895-96. 
The results of the eighteen tests with nine herds are sum- 
marized in the following table. ‘The rations fed each herd in 
the different tests, the cost of the rations, the daily milk and 
butter product, and the cost of food to produce 100 pounds of 
milk and one pound of butter, are given in such a way that the 
results from the two rations can be easily compared. 

* Massachusetts State Station Report, 1894, p. 43; Pennsylvania Experiment Station 
Report, 1895, p. 71. 

