get a few submarines through the Heligoland Bight into open waters. Several were unquestionably destroyed in the operation, yet the activity of submarines in the Atlantic showed that these mine-fields had by no means succeeded in proving more than a harassing measure. It was estimated that the North Sea barrage would require about 400,000 mines, far more than existed in the world at that time, and far more than all our manufacturing resources could then produce within a reasonable period. I have already made the point, and I cannot make it too frequently, that time is often the essential element in war--and in this case it was of vital importance. Whether a programme is a wise one or not depends not only upon the feasibility of the plan itself, but upon the time and the circumstances in which it is proposed. In the spring of 1917 the situation which we were facing was that the German submarines were destroying Allied shipping at the rate of nearly 800,000 tons a month. The one thing which was certain was that, if this destruction should continue for four or five months, the Allies would be obliged to surrender unconditionally. The pressing problem was to find methods that would check these depredations and that would check them in time. The convoy system was the one naval plan--the point cannot be made too emphatically--which in April and May of 1917 held forth the certainty of immediately accomplishing this result. Other methods of opposing the submarines were developed which magnificently supplemented the convoy; but the convoy, at least in the spring and summer of 1917, was the one sure method of salvation for the Allied cause. To have started the North Sea barrage in the spring and summer of 1917 would have meant abandoning the convoy system; and this would have been sheer madness.