82 



EXPERIMENTAL FARMS. 



period of 8 weeks (December 6 to January 31), and some of them for a period of 12 

 weeks (December 6 to February 28). 



Table XT. 



No. of Swine. 



Breeding. 



Date of 

 Birth. 



5 Crossbreds . . 



5 Purebreds. 

 5 Crossbreds 



4 Grades 



5 Crossbreds 

 5 do 



2 do 



Improved Large Yorkshire sire, and 

 Berkshire dam 



Improved Large Yorkshire 



Essex sire and Improved Large York- 

 shire dam 



Tamworth sire and Berkshire grade 

 dam 



Berkshire sire, and Improved Large 

 Yorkshire dam 



Essex sire, and Improved Large York- 

 shire dam 



Berkshire sire, and Tamworth dam. . . 



1893. 



June 9. 

 June 15 . 



May 31. 



July 3. 



June 6. 



May 31. 

 May 7. 



Average of 



live weight per 



head. 



Dec. 6. 



Lbs. 



86 



82 



98 



117 



108 



83 

 173 



Lbs. 



Jan. 31, 150 

 Feb. 28, 191 



Jan. 31, 169 



do 202 



Feb. 28, 223 



do 

 do 



192 



225 



Feed consumed 



per lb. 



of increase. 



Meal + 

 Lbs. 



3-53 

 2 



2 



3 



3 



3 

 3 



Milk. 

 Lbs. 



3 -56 



31 



32 



95 



17 



53 

 45 



Table VIII. 



Conclusions. From these four series of tests it appears that : — 

 (1.) The breeding of the swine which gave the largest increase per pound of 

 feed consumed was different in each of the four tests, viz. : — 



Crossbreds, Berkshire sire and Poland China dam ; 



G rades, Improved Large Yorkshire and Berkshire Grade dam ; 

 " IX., Crossbreds, Improved Large Yorkshire sire and Essex dam ; 

 " X., Grades, Tamworth sire and Berkshire grade dam; 

 " XL, Crossbreds, Improved Large Yorkshire sire and Berkshire dam ; 



(2.) The breeding of the swine which gave the least increase per pound of feed 

 consumed was : — 



Table VIII., Pure breds, Improved Large Yorkshire; 



" IX., do do do 



" X., Crossbreds, Essex sire and Improved Large Yorkshire dam ; 



" XL, do Berkshire sire and Tamworth dam ; 



(3.) There was no constant or appreciable superiority in the breeds and 

 breeding tested, in respect to the quantity of feed consumed per pound of increase in 

 live weight; 



(4.) The difference in the thriftiness, or power to increase in live weight per pound 

 of feed consumed, was greater between different animals in the same litter than 

 between breeds or breeding as such, in different litters ; 



(5.) On the whole for fattening purposes, crossbred swine and grades gave 

 better results than pure breds. 



SECOND SERIES. 



Test in feeding skim-milk. 



An experiment was begun to gain further information on the quantity of skim-niilk 

 which may be equivalent to one pound of mixed grain in the fattening of swine. 



Six Swine of a litter of crossbreds of Improved Large Yorkshire sire and Poland 

 China dam (farrowed Sept. 25th) were put into three lots as nearly even as possible. 

 There were ten pigs in the litter from which they were selected. They were fed on a 

 mixture composed of equal parts by weight of pease, wheat and rye, all ground and 

 soaked in cold water for an average of 18 hours, plus skim-milk to two of the lots. 



