222 
of the lamp. To fix it in this position, the wick-pipe ought 
to be flattened on that side, in the manner of a horizontal 
section, and thus instead of being circular, it will assume 
the shape of the letter D. 
In making this alteration, the small border at the extre¬ 
mity of the wick-pipe may be omitted on the flattened part, 
but it ought to be retained in the remaining extent. I have 
ascertained that the very remarkable beauty of the flame of 
this lamp depends essentially on its preservation, and its 
omission renders the flame less brilliant. 
The cause of this difference is easily explained, and will 
suggest an easy mode by which lamps may be highly im¬ 
proved, that of a double current of air. 
In comparing my portable lamp with its wick open at 
one side, with a very beautiful and large lamp, which had 
a current of air; I found that the flame of mine was the 
whitest and handsomest, and several persons present were 
of the same opinion. Measuring with a photometer, the 
quantities of light furnished by each at their greatest bril¬ 
liancy, that of the portable lamp was found equal to the 
light of four candles, of nine and an half lines in diameter, 
and that of the other equal to that of seven. 
During this experiment, the portable lamp consumed 
eighteen grammes* of purified oil (of Colza,) in the hour, 
and the one with a double current of air, double that quan¬ 
tity. 
When the portable lamp was furnished with a wick in 
the form of a closed cylinder, it gave no more light than 
the light of two candles, and notwithstanding consumed 
thirteen and an half grammes of oil during the hour. The 
following experiment however is still more extraordinary 
and appears to me highly interesting. 
Having given to the wick-pipe of the lamp with a double 
current of air, a small conical border about two lines broad, 
* A little more than half an ounce. 
