of oxygen which it takes is definite, and the caloric is derived from it. 

 The denominator is affected, but the capacities of gases increasing 

 and that of charcoal also though in a lesser degree, the whole 

 iftust be increased, and therefore, the value of 6 diminished, so 

 that my result nmst be above the truth. 



The other objection depends on the observations of Wedgewood, 

 who supposed, that the extreme heat of a fiirnace was about 30,000* 

 degrees, but we cannot place much dependance on the method which 

 he employed ; there is no evidence that his pyrometers contract 

 uniformly, and the contrivance which he used to connect his scale 

 with Fahrenheit's is defective ; as silver is easily fusible, and 

 bodies expand irregularly near the points at which they change 

 their state, it is certain that the degrees measured by it are 

 much too large. Thus sulphuric acid will scarcely boil unless 

 the sand round the retort be red, yet it distills at 600, not 

 far from the point where Newton determined the commencement 

 of ignition. In the large reverberatories of the iron founderies, 

 it is stated by Mushet, that one ton of coal fuses one ton of 

 soft cast iron. Now it may be shewn from Watt's experiments, 

 that, one part of coal would, if no heat were lost, vaporise 10 of 

 water, therefore, to melt one part of iron, there is used as much 

 caloric as would heat one of water, 10,000°, reckoning the la- 

 tent heat of steam 1,000°. But the gases which pass through 

 the furnace are as hot as melted iron, and are twelve times the 

 weight of the coal, while their capacity is about .28, the caloric 

 must therefore be divided between them and the iron. Com-" 



