main security against raids on our communications consists in keeping the enemy so well occupied in taking care of himself that he will have no time or opportunity to trouble us. Hence the importance of thorough preparation and perfect concert of action among the different corps. Suppose it is determined to concentrate the forces near Shreveport, preliminary to a movement into Texas. This point is the principal depot of the enemy west of the Mississippi. There are some machine-shops and dock-yards there and the place is fortified by a line of works with a radius of 2 or 3 miles. The position is a strong one, being on a bluff and commanding the eastern bank. The point suggests itself at once as a proper one for such a concentration.
The most direct and only reliable line of supply to this point would be the road from Vicksburg to Shreveport, railroad as far as Monroe, 52 miles, and a graded road the rest of the way, 96 miles. It would be necessary to put the road in running order and procure materials for completing the road. The security of this road requires that the enemy be driven out of Northern Louisiana and Arkansas. This line could be held much more easily than the Red River, which is very narrow and crooked and which has, in many places, high bluff banks where field artillery could be placed to enfilade the channel and have no fear of gun-boats. Such a point is Grand Ecore, where the bluff is 120 feet high. This point, I have been informed by spies, is fortified. Concerning the mode of uniting the forces near Shreveport I will mention no details, as it will depend much on the enemy's movements and the character of the routes in Southern Arkansas, which I have not had time to examine fully; our forces there have, doubtless, the information necessary to arrange this matter.
These movements, however, should be so arranged as to drive the enemy out of Arkansas and Northern Louisiana. I anticipate no danger from any large force moving on New Orleans from Texas. In case of this movement our forces would then immediately come in on the rear of this force and cut it off. The enemy will, I think, be unable to interfere seriously with our concentration of troops, and will then mass his whole force, except that at Galveston, near Shreveport, when he will fight or retire on the line he may select.
Suppose our forces to be united at Shreveport, which could probably be effected during the season of high water, and that arrangements have been perfected to supply the army by the road from Vicksburg via Monroe, Arkansas and Louisiana clear of rebels, and the enemy in retreat. I assume that he will do this, as our force should be much larger, than this and that he will continue to retreat, knowing that we will be weakened thereby, while he can select a defensive position far from our base. Whatever way he takes we must follow and expect to have our path disputed at every point, as he will be driven to desperate efforts. The numerous streams with high banks will afford him a favorable opportunity to retard our progress and effect a secure retreat to any point he may select. Our subsequent movements cannot well be foreseen. It does not seem probable that the enemy will return to Houston unless his force is large and he should propose to draw us into a trap.
It is more probable that he will retire farther west and use his cavalry to harass our flanks and rear, a species of war peculiarly adapted to Central and Western Texas. We should then be prepared for a most active campaign, and our force of cavalry should