The second consists in placing one or two guns under a metallic cupola, the embrasures in which are as small as possible.  The cannon, in a vertical aim, revolves around the center of an aperture which may be of very small dimensions.  As regards direct aim, the carriages are absolutely fixed to the cupola, which itself revolves around a vertical axis.  These cupolas may be struck in three different ways:  (1) at right angles, by a direct shot, and consequently with a full charge—­very dangerous blows, that necessitate a great thickness of the armor plate; (2) obliquely, when the projectile, if the normal component of its real velocity is not sufficient to make it penetrate, will be deflected without doing the plate much harm; and (3) by a vertical shot that may strike the armor plate with great accuracy.

General Brialmont says that the metal of the cupola should be able to withstand both penetration and breakage; but these two conditions unfortunately require opposite qualities.  A metal of sufficient ductility to withstand breakage is easily penetrated, and, conversely, one that is hard and does not permit of penetration does not resist shocks well.  Up to the present, casehardened iron (Gruson) has appeared to best satisfy the contradictory conditions of the problem.  Upon the tempered exterior of this, projectiles of chilled iron and cast steel break upon striking, absorbing a part of their live force for their own breakage.

In 1875 Commandant Mougin performed some experiments with a chilled iron turret established after these plans.  The thickness of the metal normally to the blows was 231/2 inches, and the projectiles were of cast steel.  The trial consisted in firing two solid 12 in. navy projectiles, 46 cylindrical 6 in. ones, weighing 100 lb., and 129 solid, pointed ones, 12 in. in diameter.  The 6 inch projectiles were fired from a distance of 3,280 feet, with a remanent velocity of 1,300 feet.  The different phases of the experiment are shown in Figs. 4, 5, and 6.  The cupola was broken; but it is to be remarked that a movable and well-covered one would not have been placed under so disadvantageous circumstances as the one under consideration, upon which it was easy to superpose the blows.  An endeavor was next made to substitute a tougher metal for casehardened iron, and steel was naturally thought of.  But hammered steel broke likewise, and a mixed or compound metal was still less successful.  It became necessary, therefore, to reject hard metals, and to have recourse to malleable ones; and the one selected was rolled iron.  Armor plate composed of this latter has been submitted to several tests, which appear to show that a thickness of 18 inches will serve as a sufficient barrier to the shots of any gun that an enemy can conveniently bring into the field.

[Illustration:  Fig. 7.—­Casemate of chilled iron after receiving Ninety-six shots.]