ctors, we should be able to judge the
extent to which the change was justified. It was a change for which,
under pressure, I bore a large share of responsibility, and it involved
replacing, in the middle of a great war, an organization built up by
experts well acquainted with naval needs by one in which a considerable
proportion of the personnel had no previous experience of the work. The
change was, of course, an experiment; the danger lay in the fact that,
until technical and Admiralty experience has been gained, even men of
the greatest ability in other walks of life may find it difficult to
produce satisfactory results even if there are no limits imposed on the
size of the Staff which assists them.

The question of production is best examined under various headings and
the results under the old Admiralty organization compared with those
under the new, although comparison is admittedly difficult owing to
changing conditions.

WARSHIP PRODUCTION

Under the Admiralty organization existing up to May, 1917, the Third Sea
Lord--as the Controller was termed when changes were introduced by Mr.
Churchill in 1912--was head of the Departments of the Director of Naval
Construction and Engineer in Chief, and of that part of the work of the
Director of Naval Ordnance which dealt with the design and production of
guns and gun mountings. Under the new organization a civilian Controller
became responsible for production, the Third Sea Lord being associated
with him on technical matters of design.

A special department for warship production and repairs was set up under
a Deputy Controller, the Third Sea Lord having no authority over this
department except by his association with the Controller.

Under the old organization it had been the custom during the war for the
Third Sea Lord to give to the Board and to the Commander-in-Chief of the
Grand Fleet a personal forecast of the anticipated dates of completion
of all warships under construction. My experience whilst in command of
the Grand Fleet had been that this personal forecast was generally
fairly accurate for six months ahead.

As an example it may be stated that in the first four months of 1917 the
delivery of destroyers _was within one of the forecast_ made in October,
1916, four vessels of the class being slightly behind and three ahead of
the forecast. Of thirteen "E" class submarines forecasted in October,
1916, for delivery by March, 1917, all except two were deliv