154 WISCONSIN STATE AGRICULTURAL SOCIETY. 
sheep, of one hundred each, were fed under like conditions, except 
that one was furnished a covered shed under which thev could run, 
while the other flock had no shed. The experiment was begun 
October 10th and continued until March 10th — five months. Dur¬ 
ing this time, the sheep having no shed ate one-fourth more than 
those having shelter, while at the same time the sheltered flock in¬ 
creased on an average three pounds more per head than the others. 
If shelter during the mild winters of England makes a difference 
equal to one-fourth the quantity of food consumed, it will prove a 
still greater economy to furnish warm stabling in our much colder 
climate. 
It is true also that food and water given while cold, besides re¬ 
quiring the expenditure of food to warm them, cool the stomach 
and check digestion, and if they are given in large quantities for 
a considerable time, tend to produce disease of the digestive organs, 
and hence seriously affect animals for any useful purpose. 
Several series of experiments have been made in France and 
Germany to ascertain the influence of the quality of food upon the 
richness of the milk produced. The result of these experiments 
in every case has been to show that, while the quantity of milk is 
greatly affected by the kind and amount of food given, the quality 
of milk is not appreciably affected by the food. For richness in 
either butter or cheese forming ingredients, the individual cow must 
be relied upon. 
Allow me to recapitulate the points touched upon which affect 
“The Cow as a Food Producer.” 
1st. Quality of food appears not to affect the quality of milk. A 
“butter cow” or a “cheese cow” must be so from her own inhe¬ 
rent qualities, and cannot be created by giving rich food. 
2d. Protection from cold and storms saves food, and enables cows 
to produce more milk than when unprotected. 
3d. An average milch covv requires for each 1,000 pounds of live 
weight, from 24 to 28 pounds of dry food substance per day. 
4th. The proportion between digestive albuminoids and carbo¬ 
hydrates, in the food of milch cows, should be 5|- of the latter to 1 
of the former. 
Mr. Anderson said he did not indorse the scientific experiments 
quoted in Prof. Daniells’ paper; believed that the experience of 
