Proceedings of the British Association. 97 



occasioned either by the conducting powers of the surrounding medium, 

 or by a chimney current. We thus see that one pound of anthracite will 

 communicate, to at least 12,000 times its weight of water, an elevation 

 of temperature of 1°, by Fahrenheit's scale. For the sake of brevity, 

 we may call this quantity, or energy, 12,000 unities of heat. One 

 pound of Llangennech, in the same circumstances, will afford 9,000 

 unities : one pound of good charcoal, after ordinary exposure to the air, 

 affords 10,500 : perfectly anhydrous charcoal would yield much more : 

 one pound of Lambton's Wall's-end coals affords 7,500 unities. It de- 

 serves to be remarked, that a coal, which produces in its ignition much 

 carburetted hydrogen and water, does not afford so much heat as a 

 coal equally rich in carbon, but of a less hydrogenated nature, because, 

 towards the production of the carburetted hydrogen and water a great 

 deal of latent or specific heat is required : indeed, the evaporation of 

 unburnt volatile matter from ordinary flaming coals abstracts unprofit- 

 ably a very large portion of their heat, which they would otherwise 

 afford. Hence, those chemists who, with M. Berthier and Mr. Richard- 

 son, estimate the calorific powers of coals by the quantity of carbon 

 which they contain, or the quantity of oxygen which they consume, 

 have arrived at very erroneous conclusions. The amount of error may 

 be detected by experiments on the cokes of flaming coals. M . Berthier 

 examines coals for their proportion of carbon, by igniting a mixture of 

 each, finely pulverized, with litharge, in a crucible, and estimates 1 

 part of carbon for every 34 parts of lead which is reduced. I have made 

 many researches in this way with both charcoal and anthracite, and 

 have obtained very discordant results. In one experiment, 10 grains of 

 pulverized anthracite from Merthyr Tydfil, mixed with 500 grains of 

 pure litharge, afforded 380 grains of metallic lead ; in a second similar 

 experiment, 10 grains of the very same anthracite afforded 450 grains of 

 lead ; in a third, 350 grains. In one experiment with good ordinary char- 

 coal, fresh calcined, 10 grains, mixed with 1,000 of litharge, afforded no 

 less than 603 grains of metal. The crucible was, in each case, covered 

 and luted. My future researches, which are intended to embrace every 

 important variety of fuel, natural and artificial, will be made with an 

 apparatus somewhat modified from that here described. Three furnaces 

 will be inclosed within each other, with a stratum of air or ground 

 charcoal between each, so as to prevent all loss of heat into the at- 

 mosphere, and thereby to transfer the whole heat disengaged by com- 

 bustion into a large body of water, of a temperature so much below 

 that of the atmosphere at the beginning of the experiment, as it shall be 

 above it at the conclusion. 



' On a method of filling a Barometer without the aid of an Air-pump, 

 and of obtaining an invariable level of the surface of the Mercury in the 

 cistern,' by Prof. Stevelly. — Prof. Stevelly said that it was very difficult 

 to fill a barometer tube so as to be quite free from air and moisture. Mr. 

 DanieU, in his Meteorological Essays, proposed to fill the barometer 



O 



