The question then arises whether it would not be as well to sacrifice a part of the tactical training of the Divisional Cavalry in the interest of the proposed strategic manoeuvres, and whether the advantages we anticipate from these latter might not, at any rate partially, be attained in another manner. It seems to me that to a certain extent this may well be possible, if we can only make up our minds to break with our existing arrangements regulating the present exercises, and order a certain number of garrisons, detailed in groups, to operate one against the other. If this grouping is carried out without reference to Corps boundaries, and the exercises are so managed that the troops need only spend one night out of quarters, during which they can bivouac, very great advantages at very small cost would be derived, because, since in these operations it is not at all necessary to carry them through to their culmination in an engagement, but only to concentrate them for the purpose, when necessary, in a practical manner, and to set all the machinery for reconnoitring, for transmission of orders, and reports, in operation, the damages to cultivation might be kept within very reasonable limits. An example will help to make the idea clearer. If from the regiments in Metz, Thionville, and St. Avoid on the one side, and of those in Saarburg, Saargemund, Saarbrucken on the other, two opposing forces are constituted, it would be easy to draw up a general idea by which each element of the group considered as an independent Cavalry screen covering the advance of an Army had reached on a given night the points at which they are actually quartered. The distances of the places named one from another are such that they fairly represent a possible situation in War, and a single day's march might well bring them into collision. Inexpensive bivouac places could easily be found in the wooded districts of Lorraine or elsewhere, and the Infantry in the respective garrisons might represent the heads of the following Armies' columns without undue interference with their programme of training. If the Cavalry march out with four squadrons only per regiment, the fifth can find horses for a part of the train, the point being not so much the number of such waggons provided as the service loading of those that are taken. In the afternoon and night preceding the march, outposts could be established, and the service of exploration commenced. Thus in two or three days a strategic exercise for the Cavalry on a large scale could be arranged at a very low cost and with very little difficulty. Such operations would naturally be conducted by the Inspectors-General of Cavalry, and, as the example shows, they could be generally, if not everywhere, easily organized.[30]