198 
meter, when it burnt sixty minutes, giving constantly 94^ 
of light. 
During this experiment, it consumed 2.7 grammes of 
wax, that is, 77 parts of wax per hour, for 64° of light. 
Whereas, from 100 parts of wax, we should only have 
had 83° of light instead of 100°, which an ordinary candle 
gives from the same consumption of wax. 
Ex}i. 15.—- A small watch lamp, seven lines in diameter, 
and two inches long, and which weighed 11.15 grammes, was 
placed standing and floating in a small vessel, in the manner 
that this kind of watch lamp is always arranged. After 
having burnt tranquilly for two hours and forty minutes, it 
was extinguished, wiped, and then weighed anew. 
It was found that it had consumed 4.75 grammes of wax, 
which gave 1.7812 grammes,=25 parts of wax, for the con¬ 
sumption of this small taper per hour. 
If it had yielded as much light in proportion to this con- 
sumption as an ordinary candle, we should have found its 
light —25°. 
Having measured its light by means of the photometer, I 
found it=1.52°, or a little more than a degree and a half^ 
instead of being 25°. 
Here then was a flame, and even the flame of a candle, 
which was sixteen times more faint than it ought to have 
been from the quantity of wax which disappeared. 
Notwithstanding the results of the preceding experiments 
and the reflections to which they gave rise, had prepared 
me to find the light of this small watch lamp very feeble, I 
was greatly surprised at finding it so extremely faint. I 
immediately repeated the experiment several times, and 
with the greatest care, and notwithstanding the little can¬ 
dle gave sometimes for an instant a little more light, it as 
often gave only half the quantity above stated, and I am 
persuaded that in estimating its light at a degree and a 
half, we give it its maximum. 
This result can be satisfactorily explained, admitting the 
