Ohfervntions on the ComltiJI'ion of Oil. 69 



we may judge from the difltrent odour in blowing out a candle of each fort, there is feme 

 reafon to fufpetl that the decompofition of the oil is not effeiled precifely in the fame 

 manner in each. We have alfo forae obfcure accounts of prepared wicks for lamps, which 

 are dated to polTefs the property of facilitating the combuftioa of very impure oils, fo that 

 they fliall burn for many hours without fmolie or fmell *. 



The accefs of air is of the laft importance in every procefs of combufi:ion. When a lamp 

 is fitted up with a very flender wick, the flame is fmall, and of a brilliant white colour : if 

 the wick be larger, the combuftion is lefs perfeiS, and the flame is brown : a ftill larger wick 

 not only exhibits a brown flame, but the lower internal part appears dark, and is occupied 

 by a portion of volatilifcd matter, which does not becor.^e ignited until it has afcended to- 

 wards the point. When the wick is either very large or very long, part of this matter 

 efcapcs combuftion, and fhevvs itfelf in the form of coal or fmoke. The diiVerent in- 

 tenfity of the ignition of flame, according to the greater or lefs fupply of air, is remarkably 

 feen by placing a lamp with a fmall wick beneath a fhadc of glafs not perfeflly clofed below, 

 and more or Id's covered above. While the current of ait through the glafs {hade is per- 

 fectly free, the flame is white; but in proportion as the aperture above is diminilhed, the 

 flame becomes brown, long, wavering, and fmoky ; it inllantly recovers its original 

 whitenefs when the opening is again enlarged. The inconvenience of a thick wick has 

 been long fince obferved, and attempts made to remove it ; in fome inftances by fubilituting 

 a number of fmall wicks inflead of a larger ; and in others, by making the wick flat in- 

 ftead of cylindrical. The mofb fcientific improvement of this kind, though perhaps lefs 

 fimple than the ordinary purpofes of life demand, is the well-known lamp of Argand. In 

 fhis the v.-ick forms a hollow cylinder or tube, which Aides over another tube of metal, fo 

 as to afl'ord an aJjuftiment with regard to its length. When this wick is lighted, the 

 flame itfelf has the figure of a thin tube, to the inner as well as the outer furface of which 

 ■ the air has accefs from below. And a cylindrical fliade of glafs ferves to keep the flame 

 fteady, and in a certain degree to accelerate the current of air. In this very ingenious ap- 

 paratus many experiments may be made with the greateft facility. The inconvenience of 

 a long wick, which fupplies more oil than the volume of flame is capable of burning, and 

 ■which conlequently emits fmoke, is feen at once by raifing the wick ; and on the other 

 hand, the efl^efl of a fhort wick, wliich affords a diminutive flame merely for want of a 

 fufEcicnt fupply of combuftible matter, is obfervable by the contrary procefs. 



The mod obvious inconvenience of lamps in general, arifes from the fluidity of the 

 combuftible material, which requires a veflel adapted to contain Lt, and even in the beil 



• The economical wicks of M. Legcr, concerning which a Report was prefentcd to the Academy at 

 Parit in i;82 by Condorcet, Lavoificr, and De Milly, were compofed of cotton of different fizes and forms, 

 namely, round nnd fiat, according to the ufc ihty were intended to fcrvc. They were covered with a. 

 fat fubftance, of a fmell not difaijrecable, but feebly aromatic. From the trials of thtfe commiffaries it was 

 afcertaincd ; i. That they afforded a clearer flame, with lefs undulation. 2. That, they confumed fomewluic 

 left oil -, and 3. That they pofreffcd the remarkable property of affording neither fmell nor fmoke, however 

 commco the oil made ufe of. 1 find a difficulty refpe£ling thel'e experiments, where it is faid that the flame with 

 the prepared wick was only ten lines long, and very fteadyi while that of the unprepared wick of compiiifou was 

 fdjr inches and a quarter long, of a conical figure, and vibrated much. It I'etms as if this laft had been dil'ad- 

 vantagcoufly trimmed, or too much raifcd, in order to have produced fuch an extravagant Hame. It foon 

 bUckcncd a filvtr plate at upwaids of a foot dill^iice. 



conlirutlcd 



