DEDUCTIONS FROM SOME FIELD ARTILLERY PRACTICE. 511 
verification. It is with much less confidence that verification is 
dealt with. The analyses seem to show that, when shooting against 
time, there was sometimes too much verification, occasionally so much, 
that there was no time left for effect. But, on the other hand, they 
also show that, when the score was by effect per shell, verification was 
insufficient. 
The degree of verification that will give the best average effect, 
seems to be three times as much when the effect is judged per shell, as 
when it is judged per minute. It seems to vary inversely as the root 
of the range, directly as the time or as the ammunition still remaining 
for effect. 
A formula is given below, also a table, but I must confess they are 
next door to pure speculation, the analyses being far too few for any 
reliable deductions. 
The formula for time effect is— 
Sh 
for effect per shell , it becomes— 
30y 
75 
Where 
®-;S 
Sh = The number of Time Shrapnel that should be fired for effect. 
R = The range in thousands of yards. 
V = The degree of verification that will give the best average effect. 
This quantity is measured by the least number of rounds of 
the same sign (+ or —) within 51 yards of the most likely 
elevation. 
Thus supposing we had 
2000 yards 2100 yards 
+ 
the most likely elevation would be 2050 yards and V would equal 1 
30 x 1 
In such a case, for time effect, Sh = —= 21 ; that is to say that, 
v/2 
21 
if the firing is at 6 rounds a minute and we had -g- = 3.5 minutes left, 
the formula tells us we may accept the 100 yards bracket as sufficient 
verification. But, if we were shooting for effect per shell, 
07 10 x 1 * , 
Sh = —= 7 onl 7 
and we see that, by the formula, we have only sufficient verification for 
seven Time Shrapnel. In both cases, as already explained, the eleva¬ 
tion selected for effect would be 2000 yards. 
If, when firing for effect per shells we know that after ranging we are 
to fire 21 Time Shrapnel at 2000 yards, we have 
10y 
