in closed Fire-places. 495 



course downwards into the canal destined to receive it, 

 and to circulate in the flues of the ovens ; but, to my 

 astonishment, I found that the ovens, instead of being 

 heated, were barely warmed. An accident, however, 

 very fortunately for me, discovered to me the real cause 

 of the ill success of the experiment. Throwing a piece 

 of paper on the top of the coals that were burning in 

 the grate, in order to see if the whole of the large 

 flame which I knew the paper must produce would be 

 drawn downwards into the horizontal opening of the 

 canal, situated behind the back of the grate, I was 

 surprised to find that this flame was not only drawn 

 into this opening, but that it appeared to be violently 

 driven downwards to the very bottom of the canal. 



In short, every appearance indicated that there was 

 a very strong vertical wind that was continually blow- 

 ing directly downwards into the opening of the canal ; 

 and it immediately occurred to me that, as this wind 

 consisted of a stream of cold air, this air must neces- 

 sarily cool the ovens almost as fast as the flame heated 

 them ; and I was no longer surprised at the ill success 

 of my experiment. 



On considering the subject with attention, I saw how 

 impossible it must be for the current of hot vapour, 

 flame, and smoke that rises from burning fuel, to be 

 made to pass off horizontally, or to deflect considerably 

 from its direct ascension in contact with the cold air of 

 the atmosphere, without drawing after it a great deal of 

 that cold air; and I now saw plainly why so much time 

 and fuel were required to heat the boiler in the kitchen 

 of the Foundling Hospital, in the experiments that 

 were made with its first fire-place. 



The cold air which entered the fire-place at its door, 



