ANIMAL PRODUCTION. 571 



were measured. They showed to what extent the true as compared with the 

 theoretical probable error was reduced by mixing the sample before drawing 

 the subsamples for testing. 



After measuring the stretch on 5,000 fibers it was also concluded that mixing 

 before drawuig subsamples in measuring stretch would have no more effect in 

 making the facts coincide with the theory than it did in the case of breaking 

 strain. The frequency constants showed that stretch was more variable than 

 breaking strain, that SO per cent of the deviations exceed the theoretical prob- 

 able error, and that the difference between the highest and the lowest means of 

 hundreds is over 40 per cent, all of which shows an exceedingly high degree of 

 variability in stretch of wool fiber. 



The author states that results of these studies are for the most part negative 

 and in many points self-contradictory, but they at least show the necessity of 

 painstaking experiments in developing and trying out methods for use in wool 

 investigations. It shows that under ordinary room conditions and mixtures the 

 probable error calculated from the standard deviations of the individual fibers 

 is not a good measure of the variability of the means of hundreds and thousands 

 for both breaking strain and stretch ; that the variation of the means of hun- 

 dreds is so great that the mean of this small number of tests is a very inaccu- 

 rate measure of the mean of a sample of wool containing only a few thousand 

 fibers ; and that the means of thousands can scarcely be used for anything more 

 than the most general work. 



"There are, moreover, certain classes of work, e. g., the comparison of the 

 strength of wool from 2 breeds of sheep, or even of 2 sheep, in which it 

 will be necessary to measure the diameter of the fiber, but before this is at- 

 tempted there should be a careful study made of the variability of the diameters 

 of single fibers by taking measurements at small intervals throughout the length 

 of fibers as well as on different fibers. Such a study as this might throw some 

 light on the question of changes in cross-section due to permanent elongation 

 before rupture, which is one of the difiiculties of finding the relation between 

 breaking strain and cross-section." 



Report on seven experiments on the feeding of pigs, W. Stevenson (West 

 of Scot. Agr. Col. Bvl. 57, 1911, pp. 255-334 ) .—This is a report of a series of ex- 

 periments intended to illustrate the best methods of utilizing skim milk and 

 whey for feeding pigs. 



The supremacy of the American hog, J. O. Armour {Breeder's Gaz., 60 

 (1911), A'o. 25, pp. 1290, 1291, 1337, 1338, 13J,0, 1342, figs. 10).— A statistical 

 account of the swine industry in the United States. 



Slaughtering operations and comparisons (Cincinnati Price Current, 69 

 (1912), No. 1, pp. 1, 2). — A statistical review of the growth and extent of hog 

 slaughtering operations' in the West from 1872 to 1911. 



Horses and horse breeding, H. K. Bush-Brown (Amer. Breeders Mag., 2 

 (1911), Nos. 2, i)p. 85-97; 3, pp. 175-188, figs. 25).— A discussion of the Celtic, 

 Prezewalskii, forest, Arab, and Thoroughbred types of horses, with remarks 

 concerning their use for foundation stock and breeding. The inheritance of cal- 

 losities is discussed. 



The book of the horse, Sa'adat Yar, trans, by D. C Phillott (London 

 1911, pp. XX+83; rev. in Nature [London], 88 (1911), No. 2197, pp. 172-173).— 

 A translation of a classic work on the horse by a native of British India. 



The horse: His breeding, care, and use, D. Buffum (Neiv York, 1911, pp. 

 160, figs. 8). — A popular work on buying, breeding, and managing horses. 



Breeding poultry for egg production, R. Pearl (Maine Sta. Bui. 192, pp. 

 113-176, figs. 9). — A review is given of the history of the work at the Maine 

 35693°— No. 6—12 6 



