L oa. J 
pipe a, Fig. 3, which communicating with the refervoir 4, fup- 
-plies oil gradually to the lamp, through an aperture contrived 
for that purpofe. The lamp, which 1s contained in the body of 
the furnace, is made according to Argand’s conftruction, with an 
oil ciftern, which is an hollow cylinder. The diameter of the 
wick-holder, in the clear, is one inch and jths; the diameter of 
the circular air aperture within (A, Fig. 1.) 1s one inch and ths. 
The lamp is fupported by two crofs ftays, (a, Fig. 1.) which are 
fixed to the top of the tube 4, Fig. 1, This tube rifes and falls 
on the ftem @, and is fixed at different heights by means of the 
{pring catch ¢, Fig. 1, which paffing through the holes 4 4, pailes 
into correfponding holes of the ftem. The tube, in rifing and 
falling, carries with it the lamp, which, by this means, may be 
fupported at different diftances from the veffels in the labora- 
tory. The furnace itfelf anfwers the purpofe of a chimney to 
the lamp. In the body of the furnace is an opening (d, Fig. 2.) 
for trimming the lamp; this may be clofed by a flide. When 
this is clofed, the heat of the lamp is confiderably increafed, for 
reafons too obvious to be infifted upon. The bottom of the lamp, 
to make it more fteady, is, loaded with lead. © 
I was employed in making fome experiments with this lamp, 
aflifted by my ingenious friend Mr. Healy, when he fuggefted that 
the heat produced by it would probably be greater if the internal 
air aperture of the wick-holder were diminifhed. This conjecture 
I thought probable, as, in the ftate of the lamp which I have de- 
{eribed, the central air was fo far removed from the flame as not 
to be heated confiderably by it, and this, we imagined, might 
eounterbalance any advantage which might refult from the more 
liberal 
