Management of Light in Illumination. 165 



in short, the combustion of the oil has been rendered 

 incomplete. 



Those who have attended to the striking effect pro- 

 duced by blowing wood fire with a bellows, in whiten- 

 ing the flame and increasing the light, will easily 

 conceive how much the beauty of the flame of a lamp 

 must depend on the manner in which the air is intro- 

 duced, which supplies the combustion. 



The glass chimney of Argand's lamp is useful, no 

 doubt, in defending the flame and preventing its being 

 agitated by the wind; but it is its usefulness as a 

 blower which renders this contrivance so highly inter- 

 esting. 



I have lately made several experiments with braided 

 wicks in the form of round whip-cords, which have pro- 

 duced a great deal of very pure white light ; and I am 

 almost inclined to think that these wicks will be prefer- 

 able to all others for portable lamps, and perhaps for 

 table lamps also, where not more light is wanted than 

 is emitted by three or four candles. 



These cord-wicks should be about two tenths of an 

 inch in diameter ; and, to stiffen them, they should be 

 braided round a very small cylinder of wood, of about 

 one twentieth of an inch in diameter, or round a small 

 slip of cane. This wood, which will be concealed in 

 the middle of the wick, will not only be useful to sup- 

 port that part of the wick which is on fire, but it will 

 also be very useful to prevent the ascent of the oil in 

 the centre of the wick, which will render it possible to 

 use cord-wicks of larger diameter than could otherwise 

 be used without danger of causing the lamp to smoke. 



When cord-wicks are employed, three of them must 

 always be used together ; and they must be fastened 



