48 
MINUTES OF PKOCEEDINGS OE 
There are three ways in which the target may be placed, and which will 
be at once explained by the following figure. 
In the first place, the target may be a horizontal one, i.e. the shot may be 
allowed to fall on some level surface, as is the usual way when practice is 
carried out on the sands. 
Or secondly, the target may be placed vertically , which is the usual way, 
I believe, in small arm practice. 
Or thirdly, it may be placed at right angles to the line of trajectory, so 
that the paths of the various shots (which may for this purpose be assumed 
parallel at that point) will be normals to the plane of the target. It is clear 
that this (although 1 am not aware that it has been before proposed) is the 
most correct position, as being the only one in which the errors of the shot 
in the direction of range are represented on the target without distortion. 
I call it the normal target , and propose throughout this paper, unless other¬ 
wise stated, always to consider the target as placed in this position. 
It is right however to state that, for low elevations, th ^-vertical target may 
practically be considered as coinciding with the normal one, the distortion 
being very small; and also that in many cases it is perfectly indifferent 
which of the three targets is used, inasmuch as having the positions of the 
shots on one, we can, if we know the angle of descent of the shot, easily 
adapt them to the others. Assuming as befote the paths of the various 
shot, for this small distance, to be parallel, and the angle they make with 
the horizontal to be represented by 0 , then we shall have for errors in the 
direction of range, 
Error on vertical target — error on horizontal target X tan 0 . (1) 
Error on normal target — error on horizontal target x sin0.(2) 
The errors in the othetf direction may, without sensible error, be assumed 
the same on all. 
8. Suppose theii, a normal target to be set up, sufficiently large to 
catch all the shots, and a gun to be fired at it a great number of times* with 
its axis always in precisely a similar direction. Assuming for the present, as 
the simplest case, that there is no tendency to err more in one direction that! 
another, We shall find the marks distributed on the target somewhat in the 
following manner; that is clustered round a central point in numbers 
gradually diminishing as their distance from this point becomes greater. 
