40 
a vault, excavated out of a rock, and containing upwards 
of 13,000 cubic feet of air. He, with his companion, 
entering the vault, while the engine was silent " found 
" a dampness and mistiness in it, which disappeared after 
*' the door was shut, and the engine recommenced working:" 
" and when he gave the signal to stop the engine, as soon 
as it ceased to work, and the condensation abated, and 
" before the door of the vault was unscrewed, the whole 
" vault in a few seconds was filled with a thick vapour, so 
" that they could hardly see the candles at four or five yards 
distance." 
This singular experiment contains much information, and 
corresponds exactly with the preceding statement. Mr. 
Roebuck does not indeed seem to be aware of the cause 
of this appearance and disappearance of vapour, as neither 
of his supposed causes are sufficient. The first is evidently 
inadequate; and if the second be admitted, if *' the air in 
" a state of condensation be capable of holding a greater 
" quantity of water," (in a state of vapour) how happened it, 
that a gallon of water per day, was collected in the wind 
chest, where it is certain the air was equally condensed ? 
The fact appears clearly to be, and Mr. Roebuck thus, 
though indirectly, acknowledges it : The very small 
quantities of water we at times discovered, proceeded from 
" nothing else but this vapour, in its passage to the furnace 
along with the blast, being condensed into water by the 
coolness of the eduction pipe and iron wind chest." But 
if its condensation depended upon cold, certainly its for- 
mation depended upon heat, not any heat proceeding from 
the furnace, but the heat given off" by the compression of 
the air. 
We may further observe, that for every portion of vapour 
thus condensed, the pipe and chest would receive the heat 
of that vapour, and consequently would soon become of a 
temperature equal to the compressed air; and of course, 
