﻿12 



BULLETIN 21, U. S. DEPARTMENT OF AGRICULTURE. 



The lots which averaged to weigh less than 1 f pounds per bird were 

 classed as broilers during the first part of the feeding season and the 

 broilers and roasters were separated and fed different feeding periods 

 after the 1st of October. 



Table 7. — Summary of Experiment C, 1912, Station 4, arranged according to length oj 



feeding period. 



EXPERIMENT D, 1911. 



The results of this experiment at Station 2 were quite even through- 

 out the season, except that during the month of November there was a 

 marked increase and great variation in the cost of gains. The lots 

 were handled like those in Experiment C, except that roasters were 

 fed 7 or 8 days, while broilers were on feed 14 days. This method is 

 open to criticism because cheaper gains are produced by gradually 

 decreasing the length of the feeding period on roasters, reaching 7 or 

 8 days about the middle of October, than by changing from 14 

 directly to 7 or 8 days as soon as the lots are separated into roasters 

 and broilers. However, much depends on. the weather conditions, 

 on the market, and on the economy of labor in the feeding station. 



