Dr. Andrews's Report on the Heat of Combination. 527 



circumstances, it is not unlikely triat the numbers expressing the 

 quantities of heat disengaged in these reactions will not be found in 

 all other cases to he so nearly in the same ratio as in the foregoing 

 examples ; but it may be pi-esumed that the general results will be 

 the same, and that those metals which produce a greater amount 

 of heat by their combustion in oxygen will also produce a greater 

 amount of heat when dissolving in nitric acid. 



The heat produced by the solution of copper in nitromuriatic 

 acid is, according to the result of a single trial, about yth less than 

 that produced by its solution in nitric acid. 



Metallic substitutions. — I have lately treated this part of the sub- 

 ject at so great length in a paper published in the Philosophical 

 Transactions, that I shall here only transcribe the general result of 

 the investigation. It is thus expressed : — " When an equivalent of 

 one and the same metal replaces another in a solution of any of its 

 salts of the same order, the heat developed is always the same ; but 

 a change in either of the metals produces a different development of 

 heat." This is evidently an analogous law to that already stated 

 for the thermal changes which accompany basic substitutions. The 

 numerical results are however entirely different in their details. 



Combustions in Oxygen Gas. — Since the time when Lavoisier 

 published his celebrated experiments on the heat produced by com- 

 bustion, the subject has frequently engaged tlie attention of chemists. 

 But few results were obtained of any scientific value, till the post- 

 humous publication of Dulong's valuable researches, which have 

 formed the basis of all subsequent inquiries. More recently, Grassi 

 and Fabre and Silbermann have examined the same subject, and I 

 have myself lately published a set of experiments upon it, which 

 were made some years ago. With the exception of some of Grassi's 

 results, the numbers obtained by the different experimenters agree 

 very nearly with each other, and we may therefore consider the 

 quantities of heat developed by the combination of oxygen with the 

 more important simple bodies and with some of their compounds to 

 be determined with considerable precision. Fabre and Silbermann 

 have also examined the combustion of carbon in the protoxide of 

 nitrogen. A tabular view of nearly all the numerical results hitherto 

 obtained, will be found in the edition of Gmelin's Hand-book of Che- 

 mistry recently published by the Cavendish Society. I shall here 

 therefore confine myself to a iew general observations. 



The following bodies in their ordinary physical states, viz. hy- 

 drogen, carbonic oxide, cyanogen, iron, tin and antimonj^ disen- 

 gage nearly the same amount of heat in combining with an equal 

 volume of oxygen. The numbers which express the heat of com- 

 bination in these cases do not in fact differ from one another more 

 than J^th part of the whole quantity, — a difference which is nearly 

 within the limit of the errors of experiment. This observation ap- 

 plies only to the quantities of heat actually obtained by experiment. 

 But if we apply corrections for the heat due to the changes of phy- 

 sical state which occur in some of these reactions, the same agree- 

 ment will no longer be observed. Thus in the combustion of car- 



