RACING IN INDIA. 141 



lead to disappointment to allow for an error that, for 

 all the owner knows, may be either for or against his 

 horse's powers. 



I believe that, even with the greatest care and under 

 the most favourable circumstances, the time test alone 

 cannot be relied upon, with a smaller margin for error 

 than 10 Ibs. in 1 mile, which, allowing for difference in 

 horses, and for the different way races are run, we may 

 assume would be equivalent to from 20 to 30 yards in 

 that distance, or from 1 J to 2 seconds, which does not 

 say very much for timing, beyond affording an ap- 

 proximate idea of a horse's powers. 



On a heavy course, a horse will take 4 or 5 seconds 

 longer to do a mile, than on a light one. 



Another thing to be considered is, that we must 

 allow at least half a second for the timer's own 

 individual error ; while there is still more to be allowed 

 for the way in which horses get off, whether from a 

 flying start, or from a walk. 



It does riot at all follow that because a horse cannot 

 be got to do good time in private, he is, on that 

 account, a moderate animal ; for many particularly 

 stayers require the stimulus of company to make 

 them extend themselves. 



When timing in private, the trainer should be most 

 careful to observe how his horses finish ; for an animal 

 who finishes strong, will always provided he runs 

 honest beat another that can do even slightly better 

 time, but is " all out" on Hearing the winning-post. 



The following tables furnish instances of first-class 

 timing. For uniformity sake, and because distances 

 and time are carefully measured there, I have given 



