1 8 An Inquiry concerning 



been so various and contradictory that no reliance 

 whatever can be placed on them. 



When a hot body is suspended in the air to the arm 

 of a balance in order to its being weighed, as it con- 

 tinually gives off heat to the fluid in contact with it, 

 this communication of heat occasions a strong ascend- 

 ing current of air to be formed over and by the sides 

 of the hot body, which current cannot fail to affect the 

 result of the experiment, and render the conclusions 

 drawn from it fallacious. To prevent, if possible, these 

 causes of error, the following experiments were con- 

 trived. 



The hot body to be weighed, which was a small 

 metallic ball, heated red-hot, was placed in the scale of 

 the balance in a small hemispherical porcelain cup, 

 which had a slender foot, or stand, about one inch 

 high ; and this cup, with the hot ball in it, was covered 

 over by a porcelain coffee-cup, turned upside down, 

 which, without touching the hot ball, confined the 

 heated air which surrounded it. This coffee-cup and 

 the porcelain stand were very exactly balanced, by 

 weights in the opposite side, before the ball was intro- 

 duced. 



The following experiment was made at Munich on 

 the 2oth of April, 1785. The weather being cloudy, 

 with intervals of sunshine, the thermometer in my 

 room stood at 52 F., and the barometer 26 inches 4 

 lines, French measure. 



At 30 minute* after noon, a small bullet, or grape- 

 shot, of cast-iron, very well formed, and apparently 

 solid, having been well washed and cleaned by scouring 

 with sand, and thoroughly dried, was exposed in a 

 clean vessel of porcelain in the midst of a mixture of 



