109 



be only approximatoly coin))nte(l, liowever desirable tliey may be. 

 Could the sluiiikajie bo i)io[)erly accoiiiited for much of the apparent 

 difference in feeding ability would disappear. The proportions of 

 dressed weight to shrunic live weight do not vary greatly except in one 

 case. 



Had the experiment terminated a month before, the consumption of 

 food for each pound of gain would have been greatly different and alto- 

 gether more uniform, as between the breeds, viz., Durocs 4.57, Berk- 

 shires 4.84, and Polauds 5.15. The Polands were past their best, and 

 the Berkshires improving but slowly, while the Durocs showed every 

 sign of being able to feed much longer at the end of the experiment. 



The meat of the different breeds was cooked and was tested by a 

 number of persons, and it was their "unanimous opinion" that the 

 meat of the animals which were making the greatest gains was sweeter 

 and tenderer than that from the others. '^Oue of the most important 

 questions suggested is believed to be the one of the possible relation 

 between the quality of meat and the condition of growth of the animal 

 at the time of slaughter. If an animal should be slaughtered while still 

 making good gains let us know it." 



[On the whole these experiments do not answer the specific question 

 as to relative merits of the breeds for which an answer was desired by 

 the " leading breeders." Indeed, as urged by the author, and for other 

 reasons, such trials in the nature of the case can not answer questions 

 Avhich involve conditions so complicated. 



The differences in individuality of the animals require tests with 

 large numbers of each breed. The effects of different foods and methods 

 of feeding, the rapidity of growth, the character of the product, and 

 other factors are important. And to get accurate data regarding the 

 effects of the food, the variations in live weight of the animals, due to 

 differences in the contents of the alimentary canal, must be eliminated. 

 The latter from the meconium at birth, to the water and more or less 

 completely digested food and food residue at time of slaughter, are sub- 

 ject to notable variations. Two plans are available for the solution of 

 the questions. One consists in feeding experiments with a large num- 

 ber of animals, in which weighings and analysis shall be nmde of the 

 different parts and organs at the beginning and the end of each trial, 

 and of food and undigested residue. These may be carried out with 

 sufficient detail and thoroughness to bring decidedly useful results. 

 The other is by more refined methods, which are employed with the 

 respiration apparatus. Tliese are requisite for the highest accuracy of 

 results. I 



Tests of varieties of vegetables, E. Davenport, M. S. (pp. 

 8-10). — Tabulated notes on 37 varieties of potatoes, C> of carrots, 7 of 

 mangels, 2 of beets, 5 of rutabagas, and 14 of silage corn. The notes 

 on the potatoes include estimates on a scale of 10 of tlie color, grain, 

 mealiness, and flavor of the different varieties as tested by eight dif- 



