Feb., 1918] Cattle Temperament 135 



explained, that Three H is more variable than Seven H under 

 the same condition. The mean of Three H is 7.21 and that of 

 Seven H is 5.34, which shows that Three H had a large number 

 of respirations about 7 mm. in length, while Seven H had a 

 large number of respirations about 5 mm. in length. Their 

 standard deviations, 3.04 and 2.20 respectively, are more 

 significant, since they show that Three H deviated more from 

 its mean, than Seven H did from its mean. Now combining 

 the respective means and standard deviations of each individual, 

 it is seen that the mean and standard deviation of Three H 

 increased in about the same ratio, and hence the coefficients 

 of variability are practically alike, since by formula: 



^ . ^ • 1 .,• standard deviation ^^^ 



— coefficient of variability = .100. 



mean 



But the fact remains that Three H is more variable than Seven 

 H, under the given condition, even though their coefficients of 

 variability do not show as marked difference as the polygons 

 themselves indicate. 



Further, the fact that the mean and standard deviation of 

 Three H increased in the same ratio, is explained by the condi- 

 tion under which the tracings were made. This was not con- 

 stant. Above it was stated that the animals were subjected to 

 like conditions, but these conditions were not constant so far as 

 the time element is concerned. For sufficient reasons, as will be 

 explained later, the after feeding condition was not always 

 taken at a specific interval after the regular feeding time. 

 This, as will be seen, affects the means of the polygons, and in 

 turn the coefficient of variability. As the pneumographic trac- 

 ings of the after feeding condition were taken farther from or 

 nearer to the regular feeding time, the resulting polygons here 

 plotted moved farther from or nearer to the "o" ordinate, due 

 to the increasing or decreasing depths of breathing as the case 

 may be. The reason that the mean should vary directly with 

 the time of taking the after feeding tracings, is probably due to 

 the metabolism of the animal which also varies as the after 

 feeding condition draws near to or farther from the actual 

 feeding time. Consequently, the time element not being con- 

 stant, the means varied to such an extent as to make the 

 coefficients of variability misleading, and so the standard 

 deviation is of more consequence than the mean in interpreting 



