THE COMMERCIAL FATTENING OF POULTRY. 11 
Table 6 shows the common custom of separating roasters and 
broilers, -feeding the former for short periods and the latter for a 
longer time. The broilers, produced cheaper gains with less feed 
than the roasters, the average total cost per pound of gain being 1.89 
cents less. Weather conditions were more favorable when most of 
the broilers were fed, which gives them an advantage over the 
roasters. Broilers fed during cool weather hi summer produced the 
cheapest gains, but the gains later in the season, though cheaper than 
those produced by roasters, were much higher than earlier in the 
season, because a large number of the broilers were stunted and the 
weather conditions unfavorable. 
The very marked increase in cost of gains in this experiment 
during November and December shows plainly the effect of weather 
conditions on the birds and the unprofitableness of feeding when 
this happens. It may be seen from Table III of the appendix that 
an unusually large proportion of dead birds are recorded in this 
experiment' toward the close of the season. Comparing the results 
at this station with those obtained in Experiment D at Station 2, 
we find that the average gain and the amount of feed per pound of 
gain was the same for the season, while the cost was slightly greater 
at the latter station, due to the higher cost of the buttermilk. Con- 
densed buttermilk diluted with H parts of water was used in Experi- 
ment D, while the regular buttermilk, which was used in Experiment 
C, cost only 1^ cents per gallon. The proportion of corn meal in the 
ration was increased hi cool weather without any injurious effects, 
but a study of the results indicates that a smaller per cent of corn 
meal in the ration produced cheaper gams. 
EXPERIMENT C, 1912. 
The ration at Station 4 in 1912 varied considerably during the 
season but on the whole was quite similar to that used in 1911, except 
that a smaller proportion of shorts was used throughout the season 
while a larger proportion of low-grade wheat flour was used during 
the latter part of the season. From 1 to 2 per cent of meat and 
bones was fed during the last half of June, throughout August and 
during the first hah of September. The ration by months was as 
follows: July, 1 part of shorts, 3 parts of low-grade wheat flour, 6.5 
parts of corn meal, mixed with 65 per cent of buttermilk; August. 1 
part of shorts, 2 parts of low-grade wheat flour, 4 parts of corn meal, 
mixed with 67.5 per cenj: of buttermilk; September, 1 part of shorts, 
4 parts of low-grade wheat flour, 7 parts of corn meal, mixed with 
62 per cent of buttermilk; October, 1 part of shorts, 5 parts of low- 
grade wheat flour, 6.5 parts of corn meal, mixed with 62 per cent of 
buttermilk; November, 1 part of shorts, 6.5 parts of low-grade wheat 
flour, 11 parts of corn meal, mixed with 62 per cent of buttermilk. 
