In peace, where there are no bullets, and prisoners may not be made, these methods lead to the best results, and to their being employed again and again, particularly if the superior commanders are inclined to praise such too complete information instead of condemning it. In truth, it is the worst system that could be conceived. Properly speaking, such protracted observation is only possible under certain circumstances; for instance, if the outer flank of the enemy's advance has been turned, then perhaps there might be a chance of sending reports round the flank of the hostile zone of security. It should, however, be quite inadmissible for a patrol to remain in this manner _between_ the hostile columns of the enemy's army. Unless it were unusually lucky, it would quickly be detected and captured; more particularly if the inhabitants were hostile. It is quite another matter where a hostile screen has to be broken through. This can generally only be accomplished by fighting; and it is the first duty of the reconnoitring squadron to break through the enemy's screen. Patrols also, which in such a case have succeeded in getting behind the enemy's outpost-line, cannot count upon sending back messages as they please. It will only be a question of a rapid offensive through the screen towards the main body of the enemy. The patrols will then be surrounded, and must at once proceed to effect their return, and only report what they have seen when they have successfully broken back again through the enemy's screen. Despatch-riders have, in such a case, small prospect of getting through. The less it is possible to observe and report, the more carefully must the points for breaking through be chosen, and attention must therefore be paid to reaching good points of vantage that command a view of places of probable importance. The time chosen for observation is also of great importance. The enemy can best be observed, and his strength and intentions appreciated, when he is on the move. Marches are generally undertaken in the morning, and towards the evening one may expect to find the enemy in quarters. Under such circumstances observation is difficult. The patrol leader must therefore arrange to reach in the morning those points from which he expects to be able to observe the enemy on the march. He will then be obliged in most cases to fall back before the advancing enemy, and will endeavour to ascertain his halting-place and the approximate line taken up by his outposts. If a distant patrol should be successful in obtaining such information it will generally have done as much as is expected of it. It is for the tactical close reconnaissance to send in information as to details. From the distant reconnaissance it is only required to form the foundations on which the Head Quarters can base its decisions. This fact should be borne in mind while reporting on the enemy.