even where, as in the attack on an arsenal or magazine, the position is protected by counter aircraft artillery, it is fair to assume that the latter is mounted at a sufficient distance from the main object of attack not to be endangered unless bombs are wilfully diverted from their objective. Under these conditions a numerically great attacking force of aeroplanes will manifestly possess an advantage in that they will divide the limited fire capacity of the defending batteries and so suffer less individual punishment; we may take it that the actual injury inflicted on the attacking fleet will be constant and independent of its numerical strength. If, by the nature of the attack, the period during which the air fleet is under fire is lessened by a numerical increase, there is a gain to an extent proportional to the reduction of time the defending batteries can be brought to bear.
In any case the gain is clear, as for example if 10 machines can do in a given time 10 times the mischief of one machine, and if this is done at the same average total loss it is done 10 times as economically. In other words, if a given weight of bombs have to be dropped and this be done by 10 separately attacking aeroplanes, the protecting batteries will be able to "get off" a 10 times greater number of shells than if the attack were planned and executed by the 10 machines simultaneously.
It is of some interest to remark that in § 64 (originally published Nov. 6th, 1914) the suggestion is made of an attack by a "few squadrons" of aeroplanes as constituting a reasonably effective concentration; this is an almost exact forecast of the practice as it obtains to-day, since a "Squadron" may be taken as from 16 to 20 machines and in recent air raids it is reported that about 60 or 70 machines have been employed. We may confidently look to a substantial
184
