90 
Steam JSngine. 
this was followed by a proportionate destruction of steani 
on refilling the cylinder. 
Mr. Watt now perceived, that to make an engine in 
which the destruction of steani should be the least possi- 
ble, and the vacuum the most perfect, it was necessary 
that the cylinder should condense no steam on filling it^ - 
and that, when condensed, the water, forming the steam, 
should be cooled down to 100 degrees, or lower. In re- 
flecting on this desideratum, he was not long in finding 
that the cylinder must be preserved always as hot as the 
steam that enters it ; and that, by opening a communica- 
tion between this hot cylinder when filled with steam, and 
another ^^essel exhausted of air, the steam, being an elas- 
tic fluid, would rush into it, until an equilibrium was e^ 
tablished between the two vessels ; and that if cold water, 
in sufficient quantity, were injected into the second ves- 
sel, the steam it contained would be reduced to water^ 
and no more steam would enter until the whole was con- 
densed. 
But a difficulty arose~-How was this condensed steam 
and water to be got out of the second vessel without let- 
ting in air? Two methods presented themselves. One 
was, to join to this second vessel (which, after him, we 
shall call the condenser) a pipe, which should extend 
do'vnwards more than 34 feet perpendicular, so that the 
column of water contained in it, exceeding the weight of 
tlie atmosphere, would run out by its own gravity, and 
leave the condenser in a state of exhaustion, except in so 
fiir as the air, which might enter with the steam and injec- 
tion water, should tend to render the exhaustion less per- 
fect : this air lie proposed to extract by means of a pump. 
The second method which occurred, was to extract both 
air and water by means of a pump, or pumps ; which 
would possess the advantage over the other, of being ap- 
plicable in all situations. This latter contrivance was. 
