70 DESCRIPTION orf a 
An eafy experiment will fatisfy any one of this con- 
ferving power of flame envelloping red coal. Take a {mall 
flick of deal or other wood the fize of a goofe quill, and 
hold-it horizontally and fteadily in the flame of the can- 
‘dle above the wick, without touching it, but in the body 
of the fame. The wood will firft be inflamed, and burn 
beyond the edge of the flame of the candle, perhaps a 
quarter ofaninch. When the flame of the wood goes out, 
it will leave a red coal at the end of the ftick, part of which 
will be in the flame of the candle and part out in the air. 
In a minute or two you will perceive the coal in the air 
diminifh gradually, fo as to form a neck; while the part 
in the flame continues of its firft fize, and at length the 
neck being quite confumed it drops off; and by rolling it 
“between your fingers when extinguifhed you will find it 
ftill a folid coal. 
However, as one cannot be always putting on frefh fuel 
in this ftove to furnifh a continual flame as is done ina 
candle, the air in the intervals of time gets at the red coals 
and confumes them. Yet the confervation while it lafted, 
fo much delayed the confumption of the coals, that two 
fires, one made in the morning, and the other in the af- 
ternoon, each made by only a hattull of coals, were fuffici- 
ent to keep my writing room, about fixteen feet fquare and 
ten high, warm awhole day. The fire kindled at feven in 
the morning would burn till noon; and all the iron of 
the machine with the walls of the niche being thereby 
heated, the room kept warm till evening, when another 
fmaller fire kindled kept it warm till midnight. 
Inftead of the fliding plate E, which fhuts the front of the 
box C, I fometimes ufed another which had a pane of glafs, 
or, which is better, of Mufcovy talc, that the flame might be 
feen defcending from the bottom of the vafe and pafling in a 
column through the box CG, into the cavities of the bottom 
plate, like water falling from a funnel, admirable to fuch 
as are not acquainted with the nature of the machine, and 
in itfelf a pleafing fpeQacle. Every 
