the commandant of the post deems it prudent to keep his ammunition outside, in chests.

I beg leave to entail upon you the trouble of reading the following remakes in reference to the alteration of guns from smooth to rifled, the banding of guns, weight of projectiles, the wear and terr of carriages, which is deemed important to the economy of ordnance, and of your consideration. The conclusions arrived at are based on the evidence of facts and experience in the fled, and will be placed before you as concisely as a clear expression will admit.

First. Some manufacturers of ordnance deny the fact that a gun is weakened by rifling, and attribute their frequent bursting to the heavier projectiles used. While there is some truth as regards weight of projectiles, it is a fact that the fractures in rifled guns follow the edge of the groove exactly as ice and granite fracture in lines cut upon the surface. It is known that acute re-entering angles upon the surface of guns are the usual lines of rupture, hence the present external form of guns without moldings. From these facts, no rifled guns should have acute or sharp-edged grooves, but flattened curve thus ----, as the parrott, which, though it odes not remedy the injury from rifling, has been proved to be the least injurious form.

Second. Banded guns, facts and experience prove, to be weaker at the breach than at the re-enforce, as of four which I have examined on the front of our defenses, all have fractured square at the vent, throwing the breach to the rear. If the breech is strengthened, explosions would to be so frequent. It would be economical, therefore, to use the wrought-iron shackles of Captain Brooke, although the expense of banding would be double. Some suppose this arrangement wakens the trunnion, but it is a mistake. A trunnion that is not previously fractured is never torn off by the recoil, which shock it is only subjected to, and to which the additional weight of the shackle would add but little momentum. An improvement on this plan would be to cast or turn off a hemispherical breech, over which fit a wrought-iron band accurately adjusted, and over this the trunnin shackle, or any other method that would secure the breech in a wrought-iron shell in close or firm contact with the cast iron.

Third. We are certainly in error as regards weight of elongated projectiles, which required immediate correction for effective service, as well as on the score of economy. We must have some safe, fixed limit determined for the weight of shot, beyond which weight it should be made penal to serve, for we cannot afford experiments in the field, excepting at the cost of dismantling our works, and this it would be more judicious, as well as economical, to leave to the prowess of the enemy.

In Battery Haskell we have 60-pound shells and 80-pound shots for 24-pounder rifled guns. The initial velocity of 1,600 fee per second has been fixed upon the experience of the past as a maximum for economy and efficiency for a 24-pounder and some other calibers. To double this velocity, if possible, would be straining the gun beyond a safe limit, yet it is a common practice here to use projectiles of twice the weight, which is equivalent to velocity * 2. To meet this additional strain, guns are banded, and the economy of the service demands that the banding should increase the strength of the piece to twice the resistance of the casting. This is not the case, however. Banded guns, fired with elongated shot, are not safe for over 300 rounds, when with round shot, guns not banded are