The nature of this oily vapour is most clearly exhibited when a candle is blown out by a sudden current of air ; this removes the flame sideways from the wick, and carries off so much heat from it, that the gas which it still continues to afford, in the form of a thick white smoke, will not com bine with the oxygen of the atmosphere, in sufficient de gree to produce either soot or flame. This white smoke, when it cools, forms a rancid tallow, as is evident from its unpleasant odour. This is less obvious in a lamp, because oil seems to require a greater heat to volatilize it than tal low.
These particulars being understood, we may deduce front them the requisite properties for a perfect lamp.
1st, It must be supplied with carbonaceous matter, and with oxygen ; 2d, It must convert the former into a gase ous state ; and, 3d, it must bring the gas, so produced, in contact with oxygen, at such a temperature, that the car bon will combine with the oxygen, in the fullest degree, to produce the greatest quantity of flame, without any smoke.
With respect to the supply of oxygen, we have no means sufficientiy simple for common use, in any greater degree of concentration, than as it exists in atmospheric air ; hence all that can be attained in lamps is, to cause a constant and liberal supply of atmospheric air to be given to the flame.
The proper supplying of oil to the flame, depends prin cipally upon the wick. The best material for a lamp-wick, is coarse cotton thread, very loosely twisted ; but a coarse article has, of iatc years, been manufactured from the re fuse of the tow of hemp and flax. The most common form of a wick is several threads, put together, :and twisted slightly round each other; this is called the round wick. It burns well, when of a small size, but does not answer so well for large flames, because the vapour produced from the central part of the wick is not sufficiently supplied with air.
The best form of a wick, is a flat baud, composed of a number of threads, disposed parallel, and united by a weft of very slight cotton. Ha very great body of light is requir ed, the flat wick is bent into a cylindrical form, so as to form a circle of flame, and the air is admitted to the inside of the ring, as well as the outside.
To make a lamp burn with a clear and steady light, the length of wick standing up above the tube or nozzle, which supports the wick, must be regulated according to its ca pacity for conducting the oil. otherwise the whole of the
oil which it supplies will be volatilized by that portion of flame which surrounds the lower part of the wick. In this case. the upper part of the wick, which is in the centre of the flame, will burn to a charcoal or snuff, which will he an extraneous body in the centre of the flame. This is an unavoidable evil in candles, and 'oust be corrected ,by the constant application of the son lines. In a lamp, a snuff must he avoided, by keeping the wick to a proper height.
A snuff in the centre of the flame, produces a smoke, and a great diminution of light, which it occasions, by in creasing the body of the wick ; and as the air cannot have free access to the wick in the centre of the flame, the gas there produced cannot burn imnt( diatcly, but must rise up wards, and will inflame when it gets high enough to meet the air, provided it then retains a sufficient heat. This ascent of the vapour before it inflames, is the cause of the pyramidal form of the flame, the vapour arising from the central parts of tl,c wick being obliged to rise to a greater height before it can meet with sufficient oxygen to burn. A large snuff to a lamp or candle, occupies the space that should he filled with flame, and the, eby diminishes the heat ; it also divides the flame into several small points, instead of being in one pyramid. The real poilo of the py ramid is wanting ; and great part of the gas which should form this point rises in smoke, for want of sufficient heat to make it burn when it has risen into the open air. On the other hand, a wick, which is too little raked above the nozzle, or tube which holds the wick, win no veiatilize the oil so fast as it draws it up, because the 1.:atc.1 body of wick through which the oil is diffused, is too small in pro portion to the. quantity of oil curtained in is : nence the ac tion of the flame is exerted upon too large a quantity at once na volatilize it perked). Still what is volatilized is burned, and smoke is not peodt.ced ; but the flame is din.inutit e, and liable to be exunguish:d by any shock or current of air.