3. PROTEIN REQUIREMENTS. 
The experimental feeding of the dairy herd for the winter 
1901-2 was planned to secure more data for the further 
study of protein requirements in milk production. The sub- 
ject is one of great pecuniary importance to the feeder because 
food stuffs containing a high per cent of protein are expen- 
sive and if cows require less protein than the standards have 
prescribed, the fact should be definitely determined so that 
farmers will not be led into incurring unnecessary expense in 
providing feed for their cows. 
The basement in which the dairy stock is kept is parti- 
tioned into four divisions, each of which will accommodate 
20 cows, and hastwo alleys, onerunning lengthwise through 
the middle and another crosswise, dividing the cows into 
four groups of five each. In the 1st division groups 1 and 2 
were fed ration I; groups 3 and 4, ration IJ; in the 2nd divi- 
sion groups 5 and 6 received ration III, and groups 7 and 8, 
ration IV. The cows were not arranged with reference to 
this experiment, but were left for the winter just as they 
happened to stand when the experiment began. Since some 
were farrow, some far advanced in their period of lactation, 
some to come in during the experiment and others fresh in 
milk, the records of all are not included in the principal ta- 
bles from which deductions are made. For tables cows were 
selected that were doing normal work. In the first division 
no disturbance of a serious character took place during the 
winter. But one cold night the door leading trom the sec- 
ond division to the third, which is a covered runway, was 
left open accidentally, and the cows were subjected to a 
draft, which gave some of them severe colds, proving fatal 
in one case—Rose—and causing serious inflammation in the 
udder with another—Tricksey. But fortunately, with Rose, 
the cold did not take aserious turn until toward spring, 
when it was necessary to close her record a week before the 
end of the experiment. Tricksey’s condition was more crit- 
