521 
DEFENCE OF ESTUARIES, HARBOURS, ETC., AGAINST 
TORPEDO-BOAT ATTACK . 1 
With reference to the above paper, which appeared in the March 
number of these “ Proceedings/’ I venture to submit the following 
remarks :— 
Electric Search-lights. 
Firstly, with regard to the form of light to be employed—the dis¬ 
persed beam, as shown in the cut, is seldom if ever used ; the usual 
practice being to place the arc in the focus of the parabolic reflector, 
and so produce a parallel beam which gives the maximum intensity 
of light at the maximum distance. 
If we consider the coned beam, two cases arise :— 
1st, the cone of dispersion ; and 
2nd, the cone of concentration. 
With the first, the illuminated area is increased, but the intensity, 
and, consequently, the distance at which an object could be distin¬ 
guished, is correspoudingly diminished. 
In the second case, which would never be employed until the object 
had been discovered, the illuminated area is diminished but the in¬ 
tensity is increased. 
The intensity evidently varies inversely as the cross section of the 
beam ; so that if— 
I = intensity of parallel beam at distance x from mirror. 
i ,, ,, coned ,, ,, ,, ,, ,, 
a = area of cross section of coned beam at distance x from 
mirror. 
A = area of cross section of parallel beam, 
and 0 — angle of cone, 
I A d? 
then r- = — = where d and D are the diameters respectively of 
Q 
the areas a and A and d = D ± 2 x tan The positive sign for 
the cone of dispersion and the negative for the cone of concentration. 
1 See R.A.I. “ Proceedings,” p. 109, No. 3, Vol. XXI. 
11. VOL. XXI. 
69 
