Mr Tredgold on the Principles and Practice 
same mariner as a steam-engine boiler, and of the same propor- 
tions. It should contain as much steam as will fill all the pipes, 
or other vessels, for affording heat, besides about an equal space 
for water. From the boiler the steam flows into pipes, which 
convey it to the places where heat is required, and where it 
flows into larger pipes, or into other appropriate vessels for af- 
fording surface to give out heat. From these pipes, or vessels, 
the condensed w'ater is returned to the l)oilcr, provided the pipes 
or vessels be situate above the level of the water in the boiler ; 
but if this should not "be the case, the condensed water is .al- 
lowed to run off’ by an inverted syphon, where a column of 
about nine feet of water is opposed to the force of the steam. 
Sometimes the same thing is effected by an apparatus termed 
a sUam-trap^ which acts by means of a hollow ball, similar to a 
ball-cock. And, in both these methods, it is necessary to have 
a small outlet, for clearing the pipes of air when the «t earn is let 
in. The valve by vvdiicii the air is let out, and admitted at, 
when the pipes are clear of steinn, is often made self-acting ; the 
motion being produced by the expansion and contraction of the 
pipes. ' When the pipes are cool, the valve is open ; but whesi 
; they become heated, by l lie steam being admitted^ they expand 
in length, and close the valve. 
In some cases, the condensed water may be let off’ by a com- 
mon cock ; which, when tiie apparatus is at work, should be on- 
ly opened just so as to let the condensed water escape. For hot- 
houses, this answers very well, and requires very little more at- 
lention than the other, modes. The conveying pipe should as- 
- cend as nearly in a vertical direction as possible fi’om the boiler, 
and then descend to the vessels for containing steam to afford 
heat by this arrangement, the steam is not interrupted by the 
return of condensed water down the pipe. It must be obvious, 
that the condensed water should be let out at the lowest part of 
the pipes or vessels ; but it is not equally obvious, that the air 
in the pipes should be let out at the same place; and, from a 
want of attention to this circumstance, there has sometimes been 
a difficulty in expelling the air from the pipes. Common air is, 
however, heavier than steam, and should be let out at the lowest 
part of the pipes. 
The heating surface may be obtained in various ways. For 
