Effects of Heat modified ly Compression. 2,07 



experiments with gun-barrels, by which I occasionally ob- 

 tained very fine results : but I was at last convinced that 

 their thickness was not sufficient to ensure regular and 

 steadv success. For this purpose it appeared proper to em- 

 ploy vessels of such strength, as to bear a greater expansive 

 force than was just necessary ; since, occasionally (owing to 

 our ignorance of the relation between the various forces of 

 expansion, affinity, tenacity, &c.), much more strain has 

 been given to the vessels than was requisite. In such cases 

 barrels have been destroyed, which, as the results have 

 proved, had acted with sufficient strength during the first 

 stages of the experinients, though they had been unable to 

 resist the subsequent overstrain. Thus my success with gun- 

 barrels depended on the good fortune of having used a force 

 no more than sufficient to constrain the carbonic acid, and 

 enable it to act as a flux on the lime. I therefore deter- 

 mined to have recourse to iron barrels of much greater 

 strength, and tried various modes of construction. 



I had some barrels executed by wrapping a thick plate of 

 iron round a mandrel, as is practised ia the forin-ition of 

 gun-barrels ; and likewise by bringing the two flat aides to- 

 gether, so as to unite them by welding. These attempts, 

 however, failed. I next thought of procuring bar> of ivoji, 

 and of having a cavity bored out of the solid, so as to form 

 a barrel. In this manner I succeeded well. The first barrel 

 I tried in this way was of small bore, only half an incli : its 

 performance was highly satisfactory, and such as to con- 

 vince me, that the mode now adopted was the best cjf any 

 that I had ti'ied. Owing to the smallness of the bore, a py- 

 rometer could not be used internally, but was placed unon 

 the breech of the barrel, as it stood in the vertical muffle. 

 In this position it was evidently exposed to a much less iieat 

 than the fiducial part of the apparatus, which was ahvays 

 placed, as nearly as could be guessed, at the point of greatest 

 heat. 



On the 4lh of April an experiment was made in 'his way 

 with some spar, the pyrometer on the breech givmg 33°. 

 The spar came out clean, and free from any contamination, 



adhering 



