108 ANNUAL REPORTS OF DEPARTMENT OF AGRICULTURE. 



medium, soft, or soft and oily. Chemical tests of samples of fat 

 were made and compared with the committee's grading. The agree- 

 ment between the committee's grading and the refractive index and 

 iodin number of back fat was found to be quite close, while the 

 melting point of back fat was found to be the i^oorest chemical 

 indication of the firmness of the fat. 



Summarizing the results so far obtained in tliree years' investiga- 

 tions, it has been shown conclusively that when hogs starting at a 

 weight of approximately 100 pounds are fed on peanuts in the 

 dry lot or grazed in the field for a period of 60 days or more a 

 soft carcass is produced, and that it is impossible to produce a hard 

 carcass by feeding corn and tankage or com and cottonsee<:l meal 

 to these soft hogs for a subsequent period of 60 days oi* less. This 

 statement is not meant to discourage in any way the feeding of 

 peanuts to hogs in those sections of the South that are well adapted 

 to the production of this crop. The experimental data compiled 

 by the southern stations indicate very clearly that peanuts are one 

 of the most economical feeds known for hogs and can be fed with 

 profit in many sections of the South, even though soft hogs are 

 discriminated against on the market. Much work remains to be 

 done on certain phases of the soft-pork problem. 



SWINE BREEDING. 



To determine the effects of purebred sires in the grading up of 

 common hogs on costs of production and feeding qualities, a series 

 of tests have been begun at Beltsville, Md. The sows include 

 tyi^ical Piney Woods rooters from the South and sows of mixed 

 breeding representing at least three or four known crosses. It is 

 the purpose to determine the effect of feed and management apart 

 from breeding, and as thoroughly as possible the effect of breeding 

 apart from feeding and management. 



SHEEP AND GOAT INVESTIGATIONS. 



FAUM SHEEP INVESTIGATIONS. 



Four main lines of sheep investigations were carried on at the 

 bureau's experiment farm at Beltsville, Md., namely, (1) effects ot 

 '•flushincr" and various other methods of feedmg and management 

 upon laSb yields; (2) a practical system of forage grazing; (3) 

 studies in growth bv means of weighing lambs weekly; (4) hxmg 

 tvpe of purebred Southdown, Shropshire, and Hampshire sheep by 

 means of selective breeding. The breeding flock at Beltsville at 

 the close of the fiscal vear consisted of 256 sheep of the Southdown, 

 Shropshire. Hampshire, and Corriedale breeds. 



At the United States Morgan Horee Farm, Middlebury, \ t., the 

 breedin^T flock of sheep consisted of 251 animals, western ewes and 

 Shropshire and Southdown rams and their progeny. Here the 

 experimental work included (1) effect of "flushing'' upon amb 

 vields- (2) early versus late lambing for market lamb production ; 

 Y3) 'Studies in ^owth; (4) a practical system of building up farm 

 flocks of sheep for lamb and wool production by mating western 

 ewes of the Lincoln-Rambouillet type with purebred Southdown 

 and Shropshire rams and further top crossing with purebred rams 



