During all exercises, especially when a long advance is being made and distant patrols are not actually sent out, the commanding officers should be furnished by the directing staff with such information as these patrols would in all probability have obtained. They should also be told which of the distant patrols may be assumed to be in touch with the enemy, and which have either returned or been captured or wiped out. The distant patrols, which are to be in touch with the enemy, might with advantage be despatched by the directing staff, before the manoeuvre commences, in time to procure quarters and receive instructions as to the situation. The troops should, of course, be informed of the despatch of these patrols, and all further reconnaissance will be furnished by the cavalry commander. Every cavalry detachment must also know exactly what area it is to reconnoitre, and what reconnaissances have been, or are assumed to have been, carried out by neighbouring detachments. It is a mistake to indicate the direction in which a near reconnaissance should be made, without limiting the distance of it. It is the duty of the distant reconnaissance to locate an opponent who is advancing from a distance. As long as the enemy is under the observation of the distant patrols, the close reconnaissance should not be pushed forward to meet him, but should be advanced from one position to the next, within definite limits. When these limits have been reached, the close patrols should be drawn in by the squadrons furnishing them, and fresh patrols be despatched to the next position. It would also be well if the directing staff were to lessen the work by stopping and sending back to their units such patrols as, by reason of their direction, cannot possibly come in contact with the enemy; or it may even suffice, for the purposes of the manoeuvre, to assume the despatch of patrols in such directions. They might also be given sealed orders, only to be opened at a certain place, containing the data necessary for negative reports and instructions regarding their return to their unit. I believe that if such methods were adopted, and if the patrol leaders confined themselves to sending such reports as would be sent in real war, which would include a clear and concise statement regarding the configuration of the country, it would be possible to avoid the unwarrantable weakening of the squadrons now in vogue, which does not even produce a correspondingly efficient service of communication. It is, however, true that the art of sending a few, but good reports, and of sending them at the right time requires, in the leader of the patrol, sound tactical judgment, and a training that is nowadays but seldom obtained.