RATIONS FOR POULTRY IN THE PACKING HOUSE. 3 
rations were calculated. In this way the variations resulting from 
differences in the weather, the physical condition of the birds, etc., 
were reduced. 
(2) Battery experiments. — The second series of experiments was 
conducted with a larger number of birds fed in batteries under 
packing-house conditions. Instead of securing the individual 
weights or feed consumption data, the total weight of the birds and 
their total feed consumption were recorded and the average gain 
and feed consumed by each bird calculated. Except that the con- 
trol ration of corn meal and buttermilk was not always fed with the 
experimental rations, the methods were practically the same as 
those employed in the 12-bird experiments. 
In the small-scale experiments, dressing and chilling records on 
different classes of birds were kept, to show the losses occurring be- 
fore and after feeding. The weights of the birds were recorded just 
before slaughter, after killing, and after cooling in a mechanically 
refrigerated chill room for 24 hours. These birds were then dis- 
sected into meat, skin, edible organs, crude gizzard fat, offal, and 
bones. Thus records were obtained of the edible and inedible por- 
tions of the different classes before and after feeding. The edible 
parts were analyzed for their moisture, fat, and protein content. 
RESULTS OF INVESTIGATION. 
Table. 1 shows the composition of the various poultry feeds em- 
ployed in these experiments. The results of the experimental work 
are given in Tables 2 to 16, inclusive. In the battery experiments 
all weights were obtained and recorded in avoirdupois units. In 
the 12-bird experiments the weights were obtained in metric units of 
weight (grams), but, for the convenience of the reader, they are 
recorded in avoirdupois units (pounds or ounces). Percentage fig- 
ures in the tables giving data on the 12-bird experiments were calcu- 
lated from the original weights expressed in grams. 
