A CALCULATION OF PROBABILITY OF FIRE. 59 
As above, the whole error 
AX =N da? + dy? 
The calculation of the error dy is rather lengthy, so I have omitted 
it and merely give the result, 
=dXN2 2 ; 
where 7’ is the time of firmg and ¢ the number of seconds between each 
observation. 
Le. 7’=time of flight + Fort time. 
t=time taken by objective to increase or diminish range by 
50 yds. when ranges are called every 50 yards, 
T? 
dy? =2dX? @? 
and 
0, AM Wan ee NORE belo ltste nt Mid (5). 
T12 
The value pcan easily be solved by reference to a F. C.’s correction 
table. 
2 
e.g. if the correction is 50, then ¢=7' and Z =1. 
He 2 
If the correction is 25, =27 and = —— =]: 
If it is 100, t=47 and = = 
1 3L 1 NG 
it SS, BS my and re Ty 
The value dX can be ascertained for any particular instrument and 
observer at fixed known ranges, and for all except Mark JD instrument 
should not exceed 1 per cent. at medium ranges with well-adjusted 
D.R.F. and well-trained observers. The values of AX can then be 
tabulated for different ranges and speeds. 
AP, the error in reading the hydro-clinometer, or other means of 
reading Q#, should not exceed 24’ and can be tabulated in yard values 
for any particular gun. 
AT is found in the range tables. 
From the above I can be found. 
To find WV or breadth of zone of dispersion. 
By similar reasoning, 
IB) Wy Dy HPAES, TESTING saaobovensnabooasasesooosee+e6 (6). 
A,T is the breadth of the 50 per cent. zone. 
A,P is the lateral error in laying and is usually so small that it may 
be taken as ‘1 per cent. of the range. 
