E N E H G Y 209 which allowed the passage of the arms of the paddle but prevented the water from rotating as a whole. The paddle was driven by weights connected with it by strings which passed over friction rollers, and the temperature of the water was observed by thermometers which indicated T7^yth of a degree Fahrenheit. Special experiments were made to determine the work dotio against resistances outside the ves sel of water, which amounted to about -006 of the whole, and corrections were made for the loss of heat by radiation, the buoyancy of the air affecting the descending weights, and the energy dissipated when the weights struck the floor ith a finite velocity. From these experiments Joule obtained 772 - 692 foot-pounds in the latitude of Manchester as equivalent to the amount of heat required to raise 1 ft of water through 1 Fahr. from the freezing-point. Adept- ing the centigrade scale, this gives 1390 846 foot-pounds as the mechanical equivalent of heat. With an apparatus similar to the above, but smaller, made of iron and filled with mercury, Joule obtained results vary ing from 772 814 foot-pounds when driving weights of about 58 Ib. were employed to 775 - 352 foot-pounds when the driving weights were only about 19^ ft. By causing two conical surfaces of cast-iron immersed in mercury and contained in an iron vessel to rub against one another when pressed together by a lever, Joule obtained 776 045 foot pounds for the mechanical equivalent of heat when the heavy weights were used, and 774 93 foot-pounds with the small driving weights. In this experiment a great noise was produced, corresponding to a loss of energy, and Joule endeavoured to determine the amount of energy necessary to produce an equal amount of sound from the string of a violoncello and to apply a corresponding correction. The close agreement between the results of these experi ments, differing widely as they do in their details, at least indicates that " the amount of heat produced by friction is proportional to the work done and independent of the nature of the rubbing surfaces." Joule inferred from them that the mechanical equivalent of heat is probably about 772 foot-pounds, or, employing the centigrade scale, about 1390 foot-pounds. Previously to determining the mechanical equivalent of heat by the most accurate experimental method at his command, Joule established a series of cases in which the production of one kind of energy was accompanied by a disappearance of some other form. In 1840 he showed that when an electric current was produced by means of a dynamo-magneto-electric machine the heat generated in the conductor, when no external work was done by the current, was the same as if the energy employed in producing the current had been converted into heat by friction, thus show ing that electric currents conform to the principle of the conservation of energy, since energy can neither be created nor destroyed by them. He also determined a roughly ap proximate value for the mechanical equivalent of heat from the results of these experiments. Extending his investiga tions to the currents produced by batteries, he found that the total voltaic heat generated in any circuit was pro portional to the number of electrochemical equivalents electrolysed in each cell multiplied by the electromotive force of the battery. Now, we know that the number of electrochemical equivalents electrolysed is proportional to the whole amount of electricity which passed through the circuit, and the product of this by the electromotive force of the battery is the work done by the latter, so that in this case also Joule showed that the heat generated was proportional to the work done. During his experiments on the heat produced by electric currents, Joule showed that, when a platinum wire was heated by the current so as to emit light, the heat generated in the circuit for the same amount of work done by the battery was less than when the wire was kept cold, proving that when light is produced an equivalent amount of some other form of energy must disappear. In 1844 and 1845 Joule published a series of researches on the compression and expansion of air. A metal vessel was placed in a calorimeter and air forced into it, the amount of energy expended in compressing the air being measured. Assuming that the whole of the energy was converted into heat, when the air was subjected to a pressure of 21 5 atmospheres Joule obtained for the mechanical equivalent of heat about 824 8 foot-pounds, and when a pressure of only 10 5 atmospheres was employed the result was 796 9 foot-pounds. In the next experiment the air was compressed as before, and then allowed to escape through a long lead tube im mersed in the water of a calorimeter, and finally collected in a bell jar. The amount of heat absorbed by the air could thus be measured, while the work done by it in expanding could be readily calculated. In allowing the air to expand from a pressure of 21 atmospheres to that of 1 atmosphere the value of the mechanical equivalent of heat obtained was 821 89 foot-pounds. Between 10 atmospheres and 1 it was 815 - 875 foot-pounds, and between 23 and 14 atmospheres 761 74 foot-pounds. But, unlike Mayer and Seguin, Joule was not content with assuming that when air is compressed or allowed to expand the heat generated or absorbed is the equivalent of the work done and of that only, no change being made in the internal energy of the air itself when the temperature is kept constant. To test this two vessels similar to that used in the last experiment were placed in the same calorimeter and connected by a tube with a stop-cock. One contained air at a pressure of 22 atmospheres, while the other was ex hausted. On opening the stop-cock no work was done by the expanding air against external forces, since it expanded into a vacuum, and it was found that no heat was generated or absorbed. This showed that Mayer s assumption was true. On repeating the experiment when the two vessels were placed in different calorimeters, it was found that heat was absorbed by the vessel containing the compressed air, while an equal quantity of heat was produced in the calorimeter containing the exhausted vessel. The heat absorbed was consumed in giving motion to the issuing stream of air, and was reproduced by the impact of the particles on the sides of the exhausted vessel. 1 The more recent researches of Dr Joule and Sir William Thomson (Phil. Tram., 1853, p. 357, 1854, p. 321, and 1862, p. 579) have shown that the statement that no internal work is done when a gas expands or contracts is not quite true, but the amount is very small in the cases of those gases which, like oxygen, hydrogen, and nitrogen, can only be liquefied by intense cold and pressure. It is worthy of note that mixtures of nitrogen and oxygen behaved more like theoretically perfect gases than either of the gases alone. For the other contributions of Joule to our knowledge of energy, and for those of Sadi Carnot, Eankine, Clausing, Helmholt/., Sir William Thomson, James Thomson, Favre, and others, we must refer the reader to the articles on the several branches of physics, especially to HEAT. Though we can convert the whole of the energy possessed by any mechanical system into heat, it is not in 1 Joule s papers will be found scattered through the Philosophical Magazine from 1839 to 1864 ; also in the Memoirs of the Manchester Society (2) vii. viii. ix. and (3) i. ; the Proceedings of the Manchester Society, 1859-60. 175; Phil. Trans., [1850] i. 61, [1853] 357, [1854] 321, [1850]!)!, [1859] 133, [1863] 579 ; Proceedings of Roy. Soc., vi. 307, vi. 345, viii. 41, 178, viii. 355, viii. 556, viii. 564, ix. 3, ix. 254, ix. 496, x. 602 ; and the Reports of the British Asso ciation [1859] ii. 12, and [1861] ii. 83.
VIII. 27