15 



These results suggest some diflEicult questions. Why, for instance, do Berk- 

 shires, Yorkshires, Du roc- Jerseys, and Poland-Chinas range all the way from the 

 top to the bottom of the list in the different tests; and why should an average of 

 the Ontario tests give a rating of the breeds which is entirely different from an 

 average of the Iowa tests? The averages have been purposely omitted, because 

 they are entirely misleading in a case of this kind. For example, one breed may 

 suffer from some unfavorable circumstance in one or more of the tests which is in 

 no way related to or influenced by the breeding of the animals, yet this circum- 

 stance may seriously affect the standing of the breed in question. 



Eliminating averages and looking over the individual tests with an un- 

 prejudiced mind, we can scarcely escape the conclusion that the factor which placed 

 a certain group at the top in any of the tests was in no way related to the breed 

 represented by that group. 



The point is further emphasized by a test reported by Professor Burns in 

 Texas Bulletin 131. In this test "razor backs" or scrubs were fed against good 

 average Poland-China grades. The Poland-China grades made more rapid gains 

 and sold for a higher price per pound, but th6 cost of producing 100 pounds of 

 gain was practically the same for both, being approximately $6.02 for the scrubs 

 and $5.94 for the grades, a difference of two twenty-fifths of a cent per pound. 



If any person wishes to test the question further, let him take 8 or 10 pigs 

 of the same litter, divide them into two groups as nearly even as possible, and feed 

 the two groups exactly the same. The question of breed cannot enter into such a 

 comparison, but in almost any such test it will be found that there is a difference 

 in the amount of feed required for 100 pounds gain in the two groups. 



Bacon and Fat Types Compared.— Another interesting point brought 

 out in these breed tests is the fact that the bacon type is able to hold its own against 

 the lard type in economy of production. Take the two bacon breeds, Yorkshire 

 and Tamworth, in the Ontario tests, and compare their standing with such breeds 

 as thp Poland-China, Chester White, and Duroc- Jersey ; or take the Yorkshire in 

 the Iowa tests and compare it with the other breeds. So far as breed tests go, 

 therefore, they fail to demonstrate that it costs any more to put a pound of gain 

 on a hog of bacon type than it does to put a pound of gain on a hog of lard type. 



Conclusion.-Breed tests, therefore, have served a useful purpose in dem- 

 onstrating that no one breed is superior to all other breeds in ability to make cheap 

 gains. A healthy, thrifty hog will make economical gains no matter what breed it 

 represents. 



PART III. 



INVESTIGATIONS WITH SWINE 



It is out of the question to review in detail all the work of experiment stations 

 in swine feeding, but there are certain phases of the work which may be dealt with 

 profitably in a somewhat general way. The problems which face the swine feeder 

 are numerous, and the experiment stations have been working for years to find 

 solutions for some of them. To solve any problem in stock feeding is a tedious 

 matter, because animals differ so much individually in their ability to utilize feed, 

 that it requires many repetitions and the employment of large numbers of animals 



