Against brick walls the breaching effect of percussion shell is certainly as great as that of solid shot of the same caliber. They do not penetrate as far by 20 to 25 per cent., but by bursting they make a much broader crater. Such shell would doubtless break against granite walls without inflicting much injury.

Sir W. Dennison, from a comparison of the several sieges in Sprain during the Peninsular war, estimated that a practicable breach at 500 yards could be made in a rubble wall backed by earth by an average expenditure of 254,400 pounds of metal fired from smooth-bore 24-pounders for every 100 feet in width of breach - equal to 2,544 pounds of metal for ever linear foot in width of breach.

Before we can draw any comparison, however imperfect, between this estimate and the results obtained at Fort Pulaski, it is necessary to make certain deductions from the amount of metal thrown form the breaching batteries used against that work, as follows:

First. For the shots expended upon the barbette guns of the fort in silencing their fire.

Second. For 10 per cent. of Parrott projectiles, which upset from some defect which I know from personal observation has been entirely removed by the recent improvements of the manufacturer.

Third. For nearly 50 per cent. of the 64-pound James shot, due to the fact that one of the two pieces from which they were thrown had by some unaccountable oversight been bored nearly one-fourth of an inch too large in diameter, and gave no good firing whatever. Making these deductions, it results that 110, 643 pounds of metal were fired at the breach.

The really practicable portion of the breach was of course only the two casemates that were fully opened, say 30 feet in aggregate width; but the scarp wall was battered down in front of three casemate piers besides, and had these piers not been there, or had the scarp been backed by earth alone, as was generally the case in Spain, the practicable portion of the opening would have been from 45 to 50 feet wide. Calling it 45 feet, the weight of metal thrown per linear foot of breach was 2,458 pounds, against 2,544 per linear foot in the Peninsular sieges. Had the fort held out a few hours longer this difference would have been much greater, for the wall was so badly shattered to the distance of 25 or 30 feet each side of the breach that the opening could have been extended either way with a comparatively trifling expenditure of metal. On repairing the work 100 linear feet of the scarp wall had to be rebuilt.

It must be borne in mind that at Fort Pulaski only 58 per cent. of the breaching metal was fired from rifled guns, the balance being from smooth-bored 8-inch and 10-inch columbiads (68 and 128 pounders) of Battery Scott.

It may therefore be briefly and safely announced that the breaching of Fort Pulaski at 1,700 yards did not require as great an expenditure of metal, although but 58 per cent. of it was thrown from rifled guns, as the breaches made in Spain with smooth-bores exclusively at 500 yards. In the former case the wall was good brick masonry, laid in lime mortar, and backed by heavy piers and arches; in the latter, rubble masonry, backed by earth.

A knowledge of the relative value of heavy round shot, 10-inch for example, and elongated percussion shells from lighter guns, say James 64-pounders (old 32-pounders), in bringing down the masses of brick masonry cracked and loosened by the elongated solid shot, is a matter of some importance, considering the vast difference in the amount of labor required to transport and handle the two kinds of ordnance. The