STATISTICAL STUDY OF BODY WEIGHTS 325 



the initial measurements with gains in live weight. This sug- 

 gests the possibilities of using these measurements in the selec- 

 tion of feeding steers, at least for experimental purposes, as a 

 means of reducing the experimental error caused by individu- 

 ality of animals. 



The measurement of initial weight shows no relationship 

 with gains, although this is usually considered one of the most 

 important factors in selection of steers for experimental 

 purposes. 



Measurements that show intermediate relationship as indi- 

 cated by coefficients of correlation are width of loin, depth of 

 chest, width of fore flank, length of head, height of withers, 

 circumference of paunch, and width of shoulders. 



Measurements in which slight or no relationship exist as 

 shown by correlation coefficients are initial live weight, width 

 of rear flank, height at shoulder point, and the distance :>f rear 

 flank and chest from the ground. 



A close relationship of circumference of chest with height of 

 withers is indicated by r ± 0-621 ± 0-034. Likewise, the width 

 of chest and depth of chest by r = 0-365 ± 0-072 show a 

 definite relationship. 



The hind quarters of a steer are more important than the 

 fore quarters in determining the gaining capacity of a steer, 

 with the exception of the circumference of chest as shown by 

 a correlation coefficient of 0-224 ± 0-054 for width of thurls, 

 0-271 ±0-053 for distance of hip to buttock, and 0-221 ±0-053 

 for circuniference of rear flank. 



The circumference of chest and rear flank are more impor- 

 tant in ascertaining gains than feed capacity as indicated by 

 the circumference of paunch. 



