( 507 ) 
DEDUCTIONS 
FROM SOME 
EIELD ARTILLERY PRACTICE. 
BY 
MAJOR H. C. DUNLOP, R.F.A. 
Professor of Artillery at the Ordnance College , Woolwich. 
T HE whole of this paper is based on the analysis of old practice 
reports. There is no theory in it. The general conclusions, 
depending on some 5000 rounjds of 12-pr. B.L. Shrapnel with head 
burster , are placed first. Afterwards are given some details and diag- 
grams illustrating the manner of analysis. It is well at once to remark 
that there is no certainty about the deductions. They are merely prob¬ 
ably true as far as the number of rounds analysed is capable of disclos¬ 
ing the truth, and as long as the general conditions under which prac¬ 
tice takes place remain practically unaltered. 
Perhaps the most important deduction is this : that, 
deduction: just as in tactics, the best is the enemy of the good, 
possible maximum. so m shooting is the possible maximum the ene¬ 
my of the average maximum. Take for instance 
the length of fuze. Because an apparent possible maximum has 
been produced in a given series with a burst short of 20 yards, we have 
no right to conclude that that is the best average burst short. On that 
particular occasion the elevation may happen to have been exactly de¬ 
termined, and the whole shooting has, perhaps, been abnormally good. 
To arrive at the true state of the case, we must group all series with 
mean bursts short of 20 yards under one heading and compare the 
average effect then obtained with that resulting from other series with 
other distances of bursts short. We then arrive at the best average 
burst short which differs from the best possible. The variation is due 
no doubt to certain errors, errors in the final elevation ordered, errors 
in the laying observation and fuze-setting, but those errors are part 
and parcel of the shooting and will probably always remain so. In¬ 
stead of the fuze, let us consider the elevation. We should of course 
expect a maximum possible effect when the range was exactly deter¬ 
mined, but the best average result is found from the analysis to be 
when the elevation is low by an amount depending, as before, on the 
various errors inseparable from practice. To recommend that the best 
course for a Battery Commander to pursue is to try to order the exact 
//. VOL. XXVI. 
