The first course is correct in principle, and will always give the best results when a competent Cavalry officer is kept sufficiently acquainted with the views of his superior, and made personally responsible for results.

Finally, as concerns practice in conjunction with the other Arms, these, too, must lead into different paths from those at present pursued. Detachments of Infantry and Artillery need Cavalry for reconnaissance, security, and orderly duties; but for the Cavalry itself these exercises (of small bodies) have little importance, except in so far as they familiarize the men with the organization and development of the other Armies, and thus learn to form an opinion when viewing these from a distance. The combined action of Cavalry with companies, battalions, and regiments has no importance at all, and is often the merest waste of time. It is much more important to secure practice for the troops in judging the appearance of considerable bodies of Infantry, either on the march or in position, and thence deducing their probable numerical strength.

The former class of exercises will, therefore, be on as restricted a scale as possible, whilst more time is given to the former, and by arrangement with neighbouring garrisons opportunities can often be made. In this way much can be done in the garrison and the Brigade to further the training of Cavalry for the work of a great War, although such exercises can in no way replace the actual manœuvring of large bodies.

It must rather be insisted on that the whole Field-Service training of the Arm can only be brought to its legitimate conclusion in the strategical exercises of large and changing combinations of units. Even the manœuvres, important as they undoubtedly are, can never replace them. They must, therefore, be placed on a footing of equal importance with the great drill practices of the higher units, since the strategical value of the Arm to the supreme Commander of the Army depends on its intelligent handling in this particular field, and it is here that the difficulties to be overcome are especially great.

Again, system must be inculcated if clearness of apprehension of the several problems is to be produced.

It will be of the utmost importance that baggage and, if possible, trains corresponding to the conditions on mobilization should be employed, and that men and horses should be fed from their contents; otherwise we would only too easily drop into habits of under-estimating the difficulties with which in real War we shall have to contend. But to guard against this self-deception should be the end and object of all our efforts.

It will, therefore, be necessary to calculate distance and the width of areas to be swept over, with reference to modern conditions, and not to work on the scale that sufficed us in France in 1870. The telegraph must also be taken into account, and it must be strictly insisted on that it is only to be used for such reports and in such cases in which it would be available on service. It should, therefore, be laid down beforehand which side is to be considered as in an enemy's country. For the representation of following portions of the Army, flag columns and peace garrisons may be employed, as this is of the utmost importance for the systematic practice of the patrol service.

The formation of territorial Cavalry Inspections and Sub-Inspections, already advocated above, would give the necessary foundation for our purposes. After finding the Divisional Cavalry necessary on a War strength, the remaining regiments in their several districts could be formed in Divisions and Corps of always varying composition in order to carry out Independent Cavalry manœuvres.

To practise the participation of the other Arms, corresponding arrangements with the adjacent Corps Commanders could be made, so that their annual manœuvres could work in with the Cavalry scheme, but the strategic side should have the preference. Operations would not always culminate in a great Cavalry engagement, but a skilful conduct of the manœuvres in practical country would generally insure such encounters without too great a charge for agricultural compensation.

I do not believe that such a scheme, though it would certainly mean a complete break with established Routine, would materially increase the cost of the manœuvres. The damage, now disseminated by the many regiments without any corresponding gain to their efficiency, would now be concentrated on the same district, but the gain to the Arm itself, as well as for the whole Army, would be both decisive and epoch-making.