As the name suggests, this explosive was originally in the form of powder or dust. The primitive formula burned slowly and gave low pressures--fortunate characteristics in view of the barrel-stave construction of the early cannon. About 1450, however, powder makers began to "corn" the powder. That is, they formed it into larger grains, with a resulting increase in the velocity of the shot. It was "corned" in fine grains for small arms and coarse for cannon. Making corned powder was fairly simple. The three ingredients were pulverized and mixed, then compressed into cakes which were cut into "corns" or grains. Rolling the grains in a barrel polished off the corners; removing the dust essentially completed the manufacture. It has always been difficult, however, to make powder twice alike and keep it in condition, two factors which helped greatly to make gunnery an "art" in the old days. Powder residue in the gun was especially troublesome, and a disk-like tool (fig. 44) was designed to scrape the bore. Artillerymen at Castillo de San Marcos complained that the "heavy" powder from Mexico was especially bad, for after a gun was fired a few times, the bore was so fouled that cannonballs would no longer fit. The gunners called loudly for better grade powder from Spain itself. How much powder to use in a gun has been a moot question through the centuries. According to the Spaniard Collado in 1592, the proper yardstick was the amount of metal in the gun. A legitimate culverin, for instance, was "rich" enough in metal to take as much powder as the ball weighed. Thus, a 30-pounder culverin would get 30 pounds of powder. Since a 60-pounder battering cannon, however, had in proportion a third less metal than the culverin, the charge must also be reduced by a third--to 40 pounds! [Illustration: Figure 16--GUNPOWDER. Black powder (above) is a mechanical mixture; modern propellants are chemical compounds.] Other factors had to be taken into account, such as whether the powder was coarse-or fine-grained; and a short gun got less powder than a long one. The bore length of a legitimate culverin, said Collado, was 30 calibers (30 times the bore diameter), so its powder charge was the same as the weight of the ball. If the gunner came across a culverin only 24 calibers long, he must load this piece with only 24/30 of the ball's weight. Collado's _pasavolante_ had a tremendous length of some 40 calibers and fired a 6- or 7-pound lead ball. Because it had plenty of metal "to resist, and the length to burn" the powder, it was charged with the full weight of the ball in fine powder, or three-fourths as much with cannon powder. The lightest charge seems to have been for the pedrero, which fired a stone ball. Its charge was a third of the stone's weight.