Combination of Bodies with Oxygen. 331 



These numbers cannot be considered perfectly accurate, 

 but are probably a little below the truth, in consequence oF 

 the formation of the small quantity of carbonic oxide to which 

 reference has already been made. The results of M. Dulong 

 differ from one another still more than the preceding; no 

 doubt from the operation of the same cause, namely the for- 

 mation of variable quantities of carbonic oxide. His nuuibers 

 for one litre of oxygen consumed vary from 3770 to 4004 

 units, the mean result indicating 7'288 units for the heat pro- 

 duced during the combustion of one gramme of carbon. The 

 number given by Despretz for the same is 7912 units. The 

 ancient experiments of Lavoisier present a remarkable coin- 

 cidence with the results now obtained ; and considering the 

 form of apparatus employed, and the infant state of the science 

 at the time they were performed, are deserving of special re- 

 ference, as furnishing a singular example of the accuracy and 

 ability by which so many of the works of that eminent philo- 

 sopher are distinguished. He found ihat 1 lb. of carbon in 

 burning melted 96-5 lbs. of ice, which corresponds to 7624 

 units. In deducing the latter number, the latent heat of 

 water has been taken to be 79°, in accordance with the expe- 

 riments of Provostaye and Desains. The results of Crawford 

 and Dalton on the heat evolved during the combustion of 

 charcoal are altogether erroneous*. 



* Since the above was written, the results of an extended inquiry into 

 the same subject have been communicated to the French Academy by MM. 

 Fabre and Silbermann. (Comptes Rcndits, xx. 1565, and xxi. 944.) They 

 find that the development of heat is considerably affected by the physical 

 state in which tiie carbon exists before combustion. Thus the mean quan- 

 tity of heat disengaj;ed, according to their experiments, by the diamond, 

 amounts to 7824 units, by natural graphite to 7796, by artificial graphite 

 to 7760, and by wood-charcoal to 8080 units. In these experiments, the 

 quantity of carbonic oxiile formed during each combustion was determined, 

 and tlie result obtained by direct experiment was corrected accordingly. 

 Considering the great importance of the subject, I have long intended to 

 resume the inquiry, and have indeed already obtained some new results, 

 but they are still in a very imperfect state. From a rude estimate of the 

 quantity of carbonic oxide which was formed in the experiments described 

 in the text, I inferred at the time that the true quantity of heat disengaged 

 during the conversion of carbon into carbonic acid amounts to about 7900 

 units, or nearly the number already obtained by Despretz. On repeating 

 these experiments, I found that if the charcoal be suspended in a cage 

 formed of fine platiua wire instead of being placed in h cup, the combustion 

 |)rocceds with such vivacity that not more llian .^.V.dih part of the carbon is 

 converted into carbonic oxide. In a single experiment performed in this 

 way with wood-charcoal (not however perfectly purified), I obtained 7860 

 units of heat, which, corrected for the c.n-bonic oxide formed, would cor- 

 respond to 7881 units, as expressing the entire heat [iroduced by the con- 

 vurAiun of carbon into carbonic aeiil. 'i'his nearly agrees with the experi- 

 ments of Fabre and Silbermann on the combustion of the diamond and 



