owing to a few shots having a higher velocity than the bulk 

 of the charge. This is a point we shall refer to later on. 



The magnitude of Mr. Griffiths' work is shown by 

 the fadl that he fired with 1 sizes of shot at 1 2 different 

 ranges with 6 different charges, three shots being fired 

 under each of these 720 different conditions, or a total 

 of over 2,000 shots. The results of these experiments 

 were published in the "Field," December 4th, 1886. 



In order to calculate from the results of Mr. Griffiths* 

 experiments the striking velocity at any range, such as, 

 say, 40 yards, it is necessary to take the times of flight 

 over ranges respedllvely greater and less, say, 45 and 35 

 yards ; the difference between the two times representing 

 the time of flight between 45 and 35 yards. From this 

 the mean velocity between these two ranges can be cal- 

 culated, and this can be taken as approximately equal to 

 the inking velocity at 40 yards. The objedtlon to this 

 is that the conditions may vary between the firing at 35 

 yards and at 45 yards, so that the difference of the two 

 times may not truly represent the time of flight between 

 the two dl^ances. Any eiTors due to this cause may be 

 largely diminished by judicious averaging or by graphical 

 methods, such as are commonly employed in all physical 

 experiments. 



Whilst Mr. Griffiths' method of measuring the interval 

 of time between the shot leaving the muzzle and arriving 

 at the target is, owing to the spread of the shot, not as 

 accurate as measurements obtained by the breaking of 

 electric circuits in the ordinary way, yet the uncertainty 

 m the latter case as to whether the second circuit is broken 

 by the front, middle, or even the rear of the cluster of 



