41 
cease to condense or produce water at that pressure. In 
exact agreement thereto, Mr. Roebuck says, This appear- 
" ance of water only showed itself occasionally, for it never 
" was observed, but either when the engine, after working 
" slowly, was made to work quicker, or after being stopped 
" for a few minutes, was set to work again." 
If, then, from such trifling deviations in the pressure 
and consequent temperature, a gallon of water per day 
could be collected, we may fairly infer that the whole 
quantity taken up by the additional pressure of 2, 2^, or 31b. 
per inch, and thrown into the furnace along with the air, must 
be very considerable indeed. 
Admitting a mean of the results, obtained from the 
tables of Saussure and Lambert, the additional quantity 
of vapour will be about one-fifth of that the air contains, 
with the atmospheric pressure at all temperatures; or if 
we reckon the ratio of vapour to be simply as the pressures, 
or as 30 to 35, l-6th will be the additional quantity taken 
up by compression, and consequently, following our preced- 
ing calculations, we may state the additional quantity of 
cokes lost, by making use of a water-house, at about 24 lb. 
weight per hour, on an average throughout the year, or 
about 100 tons per annum, in every such furnace. 
This result, which we believe exceeds not the truth, gives 
a very evident superiority to a dry blast, and which we have 
no doubt, when properly regulated, is due thereto. 
But our wish to be understood having led us out far 
beyond our intention, we must, at present suppress several 
circumstances which might be adduced in confirmation of 
this conclusion. We must also reserve some suggestions 
for extracting part of the vapour from air, by the use of 
quick lime, potash, sulphuric acid, &c., which we think 
might be found practicable in a dry blast or air vault ; 
and if so, well worth the attention of the iron master. 
We do not affirm this train of reasoning and computa- 
