66 Report oF DEPARTMENT OF ANIMAL HUSBANDRY OF THE 
When the nutritive ratio was narrowed but kept wider than 
1:6, no change occurred in the average amount*of digestible dry 
matter consumed for each pound of milk solids produced. When 
the ratio was made wider there was an increase (over 4 per ct.) 
in amount of digestible dry matter for each pound of milk solids. 
When rations with a narrower ratio than 1:6 were made still 
narrower there was the same increase in the amount of digesti- 
ble dry matter required for each pound of milk solids as when 
corresponding rations were made wider but kept narrower 
than 1:6. 
It must be remembered that these summarized results apply 
only to the immediate effect on milk production of the specified 
changes in the ration. It is not unreasonable to assume, how- 
ever, that those modifiations of the ration which at once lead-to 
increase in milk flow point toward the composition of a ration 
adapted to more permanent advantage, and that those modifi- 
cations which are immediately followed by diminished product 
point in the direction of a ration more likely to prove inefficient 
for extended periods. 
INTRODUCTION. 
In the feeding of milch cows as well as that of other animals 
many difficult problems are involved. The solution of most of 
these cannot be accomplished without numerous investigations 
and studies under specially arranged conditions where factors 
commonly uncontrolled can be directly accounted for and their 
value considered. Many years must necessarily elapse before 
much of the positive knowledge sought can be secured. This 
knowledge will come in time. But every day the animals must 
have food, and any information relating to the commonly prac. 
ticed methods of feeding is worth considering. Any careful-y 
collected data should, therefore, repay attention; though they 
may not light up certain of the complex problems of nutrition, 
and may show only in a circumstantial way relations between 
the milk and the food. j 
While the fixing of an absolute standard stated in the terms 
we now use is not possible, it is still probable that with wider 
