2 BULLETIN 1062, U. S. DEPARTMENT OF AGRICULTURE. 
packer buys the surplus of this seasonal product when it is available, 
puts it in condition for the market, and holds it in cold storage until 
needed. 
The ration and methods of feeding must be designed to accom- 
plish the desired results. For instance, the kind of ration and length 
of feeding period should vary with the age and class of the birds. In 
order to secure such information, poultry-fleshing experiments were 
begun by the Food Research Laboratory in 1916. In connection 
with these experiments, data on the losses due to dressing, such as 
blood, feathers, and offal, and on the loss caused by chilling were se- 
cured. Representative lots of birds were selected before and after 
feeding for dissection into their edible and inedible components. 
Chemical analyses of the various edible portions were made to de- 
termine the composition of the range or unf attened birds as compared 
with that of similar birds after fleshing. 
METHODS OF PROCEDURE. 
Two types of experiments were conducted: 
(1) Twelve-bird experiments. — The metal batteries, commonly used 
in poultry-feeding houses, were partitioned into individual com- 
partments, 12 by 18 inches. Each bird was supplied with an in- 
dividual cup so constructed as to eliminate all possible waste of feed. 
Thus an accurate record of each bird's feed consumption was ob- 
tained. As far as possible all variations in size and vigor of the 
birds were eliminated, so that the results indicate the effects of the 
rations on normal birds, rather than the gains which can be made 
with the rations fed under packing-house conditions. The birds 
were fed twice daily, at 8 a. m. and at 4 p. m. Each was weighed 
at the beginning of the experiment, and again on the fourth, eighth, 
eleventh, and fourteenth day. After selection, the birds were held 
for a preliminary period of 24 hours, during which time they re- 
ceived only a light feed of corn meal and buttermilk, in order that 
the contents of their digestive tracts might be uniform at the be- 
ginning of the experiment. In conformity with the usual practice, 
they were fed sparingly for the first three days, the amount of feed 
being gradually increased to full feed according to their desire until 
about the sixth day. As the object at all times was to maintain a 
keen appetite, any feed remaining at the end of 20 or 30 minutes 
was removed. At the time of feeding, an experimental ration, con- 
sisting of corn meal (40 parts) and buttermilk (60 parts), was fed 
to similar birds selected as controls. The results secured with each 
experimental ration were compared directly with those secured with 
the control ration. The efficiency of the ration fed to the control 
was given the value of 100, and the values of the experimental 
