OF PROJECTILES. 22Y-

pfojected from the eprouvette-mortar , when it reaches thevertex of the curve, deflects either to the right or left, thoughit be very dense, and its surface very smooth. 7 he deflec-tion of shells, whose specific gi avity is less, and centre ofgravity distant from the centre of the figure, and which fromthe ears and the head of the fuse present an uneven surface tothe air is greater: the fuse, from the rotation of the shellround its axis, describes a kind of picciloide.

This deflection does not much diminish the force of theshock on a flat, but on a convex superficies its effects arevery considerable; for, however small the deflection may beto the right or left, the angle of incidence instead of beingright, becomes very acute. This circumstance merits amore particular examination :

239. Suppose in the first place, that the superficies B to beimpinged be flat and horizontal like the ramparts of for-tresses, (PI. 6, Fig. 22) under which are calcinated barracksor magazines, covered with thick layers of earth well ram-med : in order to break through these casemates, the mor-tar should be placed in very near to the point B, and laidat a very high elevation, as 75 0 or 85°, that the angle ofincidence may be nearly a right angle. If the distance fromB to be short, and the plane of the mortar much higherthan B, the compound velocity of the shell will be greater,since the angle of incidence approaches nearer to a rightangle, and the force of the shock will be greatly increased :the reverse will happen if the mortar be much lower thanthe object; insomuch, that if the shell on reaching thevertex of its curve, meet a horizontal plane, it will rollupon it without penetrating (236). This first supposition isthe most favourable to.the effect of shells, the vertical planecutting the plane of the object at right angles ; and the givensuperficies being generally of great extent, the mortar may beplaced so, that the shell will always fall on it, in spite ofarty trifling irregularity in the length of the range.

240. Suppose in the second place, that the plane to beimpinged, be y B Z, inclined to the horizon AB; the mor-tar must then be removed so much the farther from Btowards A, as the angle A By is the greater; the axis ofthe mortar produced should pass through a vertical plane,supposed to cut Y Z at right angles. The diuance betweenthe mortar and object, and the proper elevation for giving a

P 3 direct