But the piece of work by which Rumford will be best known to future generations is that described in his paper entitled "An Inquiry concerning the Source of the Heat which is excited by Friction." It was while superintending the boring of cannon in the arsenal at Munich that Rumford was struck with the enormous amount of heat generated by the friction of the boring-bar against the metal. In order to determine whether the heat had come from the chips of metal themselves, he took a quantity of the abraded borings and an equal weight of chips cut from the metal with a fine saw, and, heating them to the temperature of boiling water, he immersed them in equal quantities of water at 59-1/2 deg. Fahr. The change of temperature of the water was the same in both cases, and Rumford found that there was no change which he could discover _in regard to its capacity for heat_ produced in the metal by the action of the borer. In order to prevent the honeycombing of the castings by the escaping gas, the cannon were cast in a vertical position with the breech at the bottom of the mould and a short cylinder projecting about two feet beyond the muzzle of the gun, so that any imperfections in the casting would appear in this projecting cylinder. It was on one of these pieces of waste metal, while still attached to the gun, that Rumford conducted his experiments. Having turned the cylinder, he cut away the metal in front of the muzzle until the projecting piece was connected with the gun by a narrow cylindrical neck, 2.2 inches in diameter and 3.8 inches long. The external diameter of the cylinder was 7.75 inches, and its length 9.8 inches, and it was bored to a depth of 7.2 inches, the diameter of the bore being 3.7 inches. The cannon was mounted in the boring-lathe, and a blunt borer pressed by a screw against the bottom of the bore with a force equal to the weight of 10,000 pounds. A small transverse hole was made in the cylinder near its base for the introduction of a thermometer. The cylinder weighed 113.13 pounds, and, with the gun, was turned at the rate of thirty-two revolutions per minute by horse-power. To prevent loss of heat, the cylinder was covered with flannel. After thirty minutes' work, the thermometer, when introduced into the cylinder, showed a temperature of 130 deg. Fahr. The loss of heat during the experiment was estimated from observations of the rate of cooling of the cylinder. The weight of metal abraded was 837 grains, while the amount of heat produced was sufficient to raise nearly five pounds of ice-cold water to the boiling point.