The requirement that leaders and men should know one another I cannot accept as an indispensable condition of War-time efficiency. However desirable it may appear that such a relation should exist, it is one which has never been, and never can be, guaranteed in War. The practicable ideal rather consists in this, that the principles in accordance with which the commands are handled should be so thoroughly flesh and blood of both leaders and led, that under all circumstances a sufficient result is secure. To reach this ideal is the true purpose of our training. In the permanent existence of Divisions it seems to me there is great danger that such a guarantee for their successful employment would be sacrificed. We have seen that the demands likely to be made on the Cavalry require widely different arrangement of the disposable forces; that this requirement increases in importance as the Arm falls numerically beneath the needs of the situation, and that only a most adaptable organization can deal adequately with the emergencies this numerical insufficiency may entail. Hence it is to be feared that a permanent constitution in Divisions might lose this requisite adaptability, and, however highly we may appreciate the advantages of a firmly welded War organization, one should never allow the form to interfere with the practical application of the means--_i.e._, never allow the troops to become so rigid as to hamper their employment in the field. But this is just what would happen if the Divisions were maintained on a permanent War footing. Every application of Cavalry Masses requires a certain measure of drill control, because it depends always on the movement of closed bodies of troops, and if the Cavalry Divisions are constantly drilled together under the same Leader in Peace, there is at least a very great risk that this certain degree of drill control, which we recognise as indispensable, will degenerate into hard-and-fast prescription, since the Leader has always the same number of units at his disposal, and will thus by degrees habituate himself to consider these as invariable quantities in the solution of every tactical problem.