20 BULLETIN 21, U. S. DEPARTMENT OF AGRICULTURE. 
meat and bones from local butcher shops was fed at Stations 2, 3, 
and 4 at irregular intervals during the season, but no consistent 
effect was noticed from this special feeding. Several lots at Station 
3 were fed specially prepared mixed feeds which were claimed to pre- 
vent feather picking, but the results were inconsistent. The feather 
picking broke out during a period of cool weather, while the birds 
were eating ravenously, but stopped quite suddenly when the 
weather became warm and the birds were not so eager for their food. 
There appeared to be less loss due to this trouble where the largest 
per cent of buttermilk was fed in the ration, but feather picking can 
not be entirely controlled by regulating the proportion of buttermilk 
in the feed. Less heating rations, or those containing a large per 
cent of shorts and mixed feed and a small per cent of corn meal, make 
the best feeds for use in hot weather where feather picking is preva- 
lent. The mixed feeds, however, produced chickens covered with 
small pin feathers, which resulted in a poorer grade of dressed prod- 
uct, and therefore made the feeding of the mixed feeds unprofitable 
as well as undesirable. 
FATTENING HENS. 
The results of fattening over 20,000 hens are shown in Tables 10 
and 11, the feeding having been done at Stations 1, 2, and 4. All 
the lots were fed during November, 1911, and November and De- 
cember, 1912. 
The hens at Station 1 were fed coarse corn chop, or cracked corn 
with the meal left in, and 15 per cent of shorts, mixed with butter- 
milk. The shorts were added to facilitate mixing the feed, otherwise 
the corn chop would sink to the bottom of the mixer. The feed was 
mixed with considerable buttermilk and fed in a wet state. 
The regular chicken mixture was fed to the hens at Stations 2 and 
4, which, while producing shghtly smaller gains, was apparently more 
efficient, as the average gain was produced with a pound less grain 
than with the corn chop, shorts, and buttermilk in 1911. It should 
be stated, however, that the increased cost of gain at Station 1 was 
due partly to the increased cost of buttermilk at this station, as the 
cost of grain in the rations was about the same, the regular chicken 
rations being slightly cheaper than the corn-chop ration. The com- 
parative difference in cost of labor is due to the condition explained 
under Experiment B. A comparison of the results in 1912 does not 
show any marked advantage of one ration over the other 
