DEDUCTIONS FROM SOME FIELD ARTILLERY PRACTICE. 515 
It seems clear that when laying is bad, minor corrections of elevation 
are not worth while, at any rate during ranging, for the dispersion of 
shell will cover such a lot of ground; such corrections may however 
serve to modify the fuze so as to give the proper proportion of “ highs ” 
to “ lows ” whenever battery fire is protracted. (The mean average 
laying error was 25 yards. Misobservation was only 2% and its effect 
was inappreciable). 
Fig. VI. shews that the best average effect was when there were at 
least 3 or 4 + 's and 3 or 4 — ’s within 51 yards of the final elevation 
ordered. As the average was rather under 2 of each sign we conclude 
that verification was underdone. It is to be noted however that the 
drill of the batteries may not have allowed the time necessary for suffi¬ 
cient verification. The figure seems also to show that there is latitude 
in verification. Beyond 2 of each sign ( + and — ) the curve rises very 
slowly, beyond 3, hardly at all. 
Fig. VII. shews how the error in the final elevaion decreased with 
verification, but how slowly it fell after the verificaion passed two of 
each sign (+ and —), thus confirming Fig. VI. Beyond 3 the curve 
nearly parallel to the base line and it is clearly seen there must be some 
upper limit to the verification that pays best. 
Fig. VIII. shows— 
(a) That the best average burst short for “ hits ” was 40 yards; 
for “ men hit ” 80 yards. With the head burster this latter distance 
was quite unexpected. We have seen that the laying mean error was 
25 yards, that fuze mean error was 35 yards ; the combined mean error 
was therefore _ 
^25 2 + 35 2 = 43 yards. 
(b) The comparatively rapid loss when the bursts are over; 50 
yards over seems as bad as 200 yards short. 
(c) That “ hits ” are a more pronounced measure of accuracy in 
determining the correct length of fuze than “ men hit.” At 80 yards 
burst short 40 men are hit, at 160 yards 32. At 40 yards burst short 
there are 240 hits, at 120 yards 140. In the first case the loss is 20%, 
in the second 42%. 
(d) That, with the same elevation, two fuze lengths give the same 
number of “ men hit,” viz., the one giving a burst short of S 5 yards and 
the other a burst short of 105 yards. Fig II. shows very good effect 
produced with an elevation of from 10 yards too high to 50 yards too 
short. Altogether, therefore, what with tolerations of burst short and 
of elevation there is a considerable margin for effect, and probably a 
number of combinations of fuze and elevation producing the same 
effect. 
In calculating the positions of the mean bursts short, the same num¬ 
ber of high shell have been cut out as there were grazes. Had this 
not been done, the best average short would appear even further back. 
Fig. IX. shows : — 
(a) That there should usually be grazes during Time Shrapnel 
fire. 
(b) That the best average proportion of grazes was 9% for " men 
hit” 
