no country is likely for some time to possess sufficiently large air
forces to obtain a decisive victory, or at any rate an uncontested
superiority, at the outbreak of war. Though in air, as in every other
form of warfare, attack is more effective than defence, we cannot afford
to keep our air forces up to war strength in peace any more than our
Army or Navy.

The problem, from a military point of view, is therefore to ensure an
adequate reserve and to maintain our capacity for expansion to meet
emergencies. The number of units maintained at war establishment should
be the absolute minimum for safety and of the type immediately required
on mobilization, i.e. long-range bombing and naval reconnaissance
squadrons. The remainder should be in cadre form. We can, of course,
maintain a fixed number of machines and pilots in reserve for every one
on the active list, but, although some such system is necessary, on a
large scale it is open to many and serious objections. First of all,
even on a cadre basis, it means keeping inactive at considerable cost a
number of machines which may never be used and which, however carefully
stored, quickly deteriorate. Knowledge of aeronautics is still slender
and improvements are made so continuously that machines may become
obsolete within a few months. Moreover, the growth of service aviation
in peace must tend to become artificial and conventional rather than
natural, and this will react on design and construction, which will be
cramped, both technically and financially, within the limits imposed by
service requirements.

It is obvious therefore that the capacity of the construction industry
to expand cannot be fostered by service aviation alone; furthermore, in
the event of another war of attrition, expansion will be more essential
than any amount of machine reserve power immediately available, and in
the event of a war of short duration that power will win which has the
greatest preponderance of machines, service or civil, fit to take the
air. The asphyxiation of a large enemy city, if within range, can be
done by night-flying commercial machines, and it would require a
defending force of great numerical superiority for its successful
defence.

Whether, therefore, from this point of view, or others, which I will
mention later, another solution must be found, and this lies in the
development of civil aviation. An analogy in the Navy and the Mercantile
Marine has long been apparent. "Se