The distance from our nearest batteries to the Spanish lines when the siege began was 1700 to 2000 yards,[544] and at this range our fire was ineffective. Many of the mortar shell burst at the muzzle from the heavy charges required for these long ranges, a gunner losing his life on one occasion from this cause.[545] The shell that withstood the shock flew wildly; the fuzes were “in general faulty”;[546] many good shell were smothered in the sand of which the Spanish works were constructed; those that burst produced but little effect;[547] and round shot were of no avail against sandbanks twenty-two feet high. As fire against the Spanish works was useless, it only remained to direct it on the working parties. Against them our mortar fire was as ineffective as against the works, and what was to be looked for from guns provided only with round shot and case? Case would not carry one-sixth of the range, and round shot against handfuls of men, scattered here and there, were as worthless as shell. The difficulty was still unsolved when Captain Mercier, 39th Regiment, suggested firing the 5.5-inch shell of the royal mortars, with _short_ fuzes, from the 24-pounder guns which had the same calibre as the mortars, 5.8-inch. A trial was made on the 25th September 1779 with (I believe) the “Rock gun,” which was a 24-pounder; the “calculated fuzes,”[548] it was found, “often burst (the shell) over the heads of the working parties,”[549] and Merciers brilliant proposal was officially adopted. When the siege was over, and men had time to think, it became clear enough that excellent as was Captain Mercier’s plan as a makeshift during the stress and strain of a siege, it had its weak points. The strong charge necessary to burst the common shell tended to scatter the fragments here and there in all directions, and the fragments were few in number. Experiments were carried on in Prussia in 1761 to determine the bursting charges which broke (mortar and howitzer) shell into the greatest number of pieces. It was found that royal mortar shell (_maximum_ bursting charge, 1 lb. 2 oz.) broke into eight pieces, with a bursting charge of 1 lb., and into nineteen pieces with a bursting charge of 14 oz., these figures being the means of six trials.[550]