115 
L A 
ways keep the oil at the fame height, or very near; the 
weight of the water being to that of the oil as 20 to 19, 
which in two or three inches makes no great difference. 
If the water runs falter than the oil waftes, it will only 
run over at the top of the pipe, and what does not run 
over will come under the oil, and keep at the fame height. 
Phil. Tran/. No. 245. 
From experiments made in order to afcertain the ex- 
penfe of burning chamber-oil in lamps, it appears, that 
a taper-lamp, with eight threads of cotton in the wick, 
confumes in one hour '3250/. of fpermaceti oil, at 2s. 6d. 
per gallon ; fo that the expenfe of burning 12 hours is 
4-57 farthings. This lamp gives as good a light as the 
candles of eight and ten in the pound ; it feldom wants 
fluffing, and calls a Itiong and fteady light. A taper, 
chamber, or watch, lamp, with four ordinary threads of 
cotton in the wick, confumes 0M6640Z. of fpermaceti oil 
in one hour; the oil at 2s. 6d. per gallon, makes the ex¬ 
penfe of burning 12 hours only 2^34 farthings. 
Argand’s Lamp, 1784. This is a very ingenious con¬ 
trivance, and the greateft improvements that has yet been 
made. It is the invention of a citizen of Geneva ; and 
the principle on which the luperiority of the lamp depends, 
is the admiffion of a larger quantity of air to the flame 
than can be done in the common w’ay. This is accom- 
plilhed by making the wick of a circular form; by which 
means a current of air ruflies through the cylinder on 
which it is placed with great force; and, along with that 
which has accefs to the outlide, excites the flame to fuch 
a degree that the lmoke is entirely confumed. Thus both 
the light and heat are prodigioully increafed, at the fame 
time that there is very confiderable faving in the expenfe 
of oil, the combuftion being exceedingly augmented by 
the quantity of air admitted to the flame ; and that what 
in common lamps is diffipated in fmoke is here converted 
into a brilliant flame. See the article Chemistry, vol. iv. 
P- 333 - 
Thislamp is now very much in ufe; and is applied not 
only to the ordinary purpofes of illumination, but alfo to 
that of a lamp-furnace for chemical operations. It con- 
fifts of two parts, viz. arefervoir for the oil, and the lamp 
itfelf. The refervoir is ufually in the form of a vafe, and 
has the lamp proceeding from its fide. The latter confifts 
of an upright metallic tube about one inch and fix-tenths 
in diameter, three inches in length, and open at both ends. 
Within this is another tube about an inch in diameter, 
and nearly of an equal length ; the fpace betwixt the two 
being left clear for the paflage of the air. The internal 
tube is clofed at the bottom, and contains another fimilar 
tube about half an inch in diameter, which is foldered to 
the bottom of the fecond. It is perforated throughout, 
fo as to admit a current of air to pafs through it; and the 
oil is contained in the fpace betwixt the tube and that 
which furrounds it. A particular kind of cotton cloth is 
ufed for the wick, the longitudinal threads of which are 
much thicker than the others, and which nearly fills the 
fpace into which the oil flows; and the mechaniftn of the 
lamp is fuch, that the wick may be raifed or depreffed at 
pleafure. When the lamp is lighted, the flame is in the 
form of a hollow cylinder; and, by reafon of the ftrong 
influx of air through the heated metallic tube, becomes 
extremely bright, the fmoke being entirely confumed for 
the reafons already-mentioned. The heat and light are 
Hill farther increafed, by putting over the whole a glafs 
cylinder nearly of the fize of the exterior tube. By dimi- 
nilhing the central aperture, the heat and light are pro- 
portionably diminifhed, and the lamp begins to fmoke. 
The accefs of air both to the external and internal furfaces 
of the flame is indeed fo very neceflary, that a fenfible 
difference is perceived when the hand is held even at the 
diftance of an inch below the lower aperture of the cylin¬ 
der ; and there is alfo a certain length of wick at which 
the effect of the lamp is lirongelt. If the wick be very 
fuort, the flame, though white and brilliant, emits a dif- 
M P. 
agreeable pale kind of light; and, if very long, the upper 
part becomes brown, and fmoke is emitted. 
The faving of expenfe in the ufe of this inftrument for 
common purpofes is very confiderable. By foine experi¬ 
ments it appears that the lamp will continue to burn three 
hours for the value of one penny ; and the following was 
the refult of the comparifon between the light emitted by 
it and that of a candle. The latter, having been fuffered 
to burn fo long without fluffing, that large lumps of 
coally matter were formed upon the wick, gave a light at 
24 inches diftance equal to the lamp at 129 inches; whence 
it appeared that the light of the lamp was equal to 28 
candles in this ftate. On fnuffing the candle, however, 
its light was fo much augmented, that it became neceflary 
to remove it to the diftance of 67 inches before its light 
became equal to that of the lamp at 129 inches ; whence 
it was concluded that the light of the lamp was fomewhat 
lefs than that of four candles frefh fnuffed. At another 
trial, in which the lamp was placed at the diftance of 1313 
inches, and a candle at the diftance of 55 inches, the 
lights were equal. In thefe experiments the candles made 
ufe of were 10-J inches long, and 2^ inches in diameter. 
When the candle was newly fnuffed, it appeared to have 
the advantage; but the lamp foon got the fuperiority ; 
and on the whole it was concluded, that the lamp is at 
lealt equivalent to half a dozen of tallow candles of fix ii* 
the pound ; the expenfe of the one being only twopence 
halfpenny, and the other eightpence, in feven hours. 
The belt method of comparing the two lights together 
feems to be the following : Place the greater light at a 
confiderable diftance from a white paper, the fmaller one 
being brought nearer or removed farther off as occafion 
requires. If an angular body be held before the paper, it 
will project two ffiadows: thefe two fliadows can coincide 
only in part; and their angular extremities will, in all po~ 
fitions but one, be at fome diftance from each other; and, 
being made to coincide in a certain part of their bulk, 
they will be bordered by a lighter fliadow, occafioned by 
the exclufion of the light from each of the two luminous 
bodies refpeftively. Thefe lighter fliadows, in faff, are 
fpaces of the white paper illuminated by the different lu¬ 
minous bodies, and may eafiiy be compared together, be- 
caule at a certain point they actually touch one another. 
If the fpace illuminated by the fmaller light appear brighteft, 
the light mult be removed farther off, but tile contrary if 
it appear more obfcure. 
On cutting open one of Argand’s wicks longitudinally, 
and thus reducing the circular flame to a ftraight-lined 
one, the lights appeared quite equal in power ; but the 
circular one had by far the greateft eft'eft in dazzling the 
eyes; though, when the long flame was made to fliine 011 
the paper, not by the broadlide, but in the direction of 
its length, it appeared more dazzling than the other. On 
placing this long flame at right angles to the ray of Ar¬ 
gand’s lamp, it projected no fliadow; but, when its length 
was placed in the direction of the ray, it gave a fliadow 
bordered with two broad, well-defined, and bright, lines. 
The broad-wicked lamp feems to have the advantage of 
the other, as requiring lefs apparatus; and indeed by this 
contrivance we may at the mod trifling expenfe have a 
lamp capable of giving any degree of light we pleafe. The 
only diiadvantage attending either the one or the other is, 
that they cannot eafiiy be carried from one place to ano¬ 
ther; and in this refpect it does not feem poffible to bring 
lamps (in general) to an equality with candles. We ffiafi 
however in the courle of our article fpeak of a lamp that 
may be moved. 
The moft economical method of lighting up large apart¬ 
ments by means of different lamps and candles, as it is of 
great importance, has occupied the attention of many 
ingenious men, particularly of count Rumford and M. Haf- 
fenfratz. The following is the fimple and accurate me¬ 
thod propofed by the count, for meafuring the relative- 
quantities of light emitted by lamps differently ccmitnuft- 
3 ed« 
