299. =Davy's Safety-Lamp.=--One of the most beautiful applications of the conduction of heat we have in the Safety-Lamp of Sir Humphrey Davy, an invention which has been the means of saving the lives of multitudes of miners. It is represented in Fig. 213. With this lamp one can go into the midst of the most explosive gases with impunity. Now all that prevents the flame within from setting on fire the gases without is a covering of wire-gauze. This, being a good conductor, conducts off the heat of the flame within so rapidly that it can not go through the openings _as flame_, and so does not set fire to the gas without. The fact upon which the construction of this lamp was based was discovered by trying many experiments. Among them were the following: A piece of wire-gauze was held over a candle so that its flame struck against it. The smoke issued above, but no flame. Then a stream of gas was allowed to pass through the gauze, as seen in Fig. 214, and was set on fire above. It burned without inflaming the gas below.[4] [Illustration: Fig. 214.] 300. =Relation of Density to Conduction.=--Generally the more dense a substance is the better is its conduction of heat. Thus the metals are better conductors than wood, marble than brick, the solids than liquids, and liquids than aeriform substances. We have frequently a good illustration of the difference between stone and brick as conductors in the melting of snow on sidewalks. If a light snow fall in the spring, after the earth has become somewhat warm, you will see it melted from the stone walks much before it is from the brick ones. This will be especially the case if the snow be melted mostly by the warmth of the earth without the agency of the sun. The explanation is obvious. The stone is a better conductor than the brick, and therefore the heat of the earth comes up through the former more rapidly than through the latter. [Illustration: Fig. 215.] [Illustration: Fig. 216.] 301. =Conduction in Liquids.=--That liquids are poor conductors of heat may be shown by an experiment or two. If a thin glass tube closed at one end be filled with water, and the heat of a spirit lamp be applied to its upper portion, as seen in Fig. 215, though the water at this portion may be made to boil, there will not be the least movement in the lower part. This will be very obvious if you have some amber-dust in the water. Again, let a little water be frozen in the lower part of the tube by placing it in a freezing mixture, and introduce a little oil, and then over that some alcohol. Hold now the tube over the chimney of a lamp, as represented in Fig. 216, until the alcohol boils. The ice in the bottom of the tube will not be in the least affected, and the oil will be but slightly heated. If the heat were to be applied in either of the above cases at the lower part of the tube the result would be different, because convection would then operate in the diffusion of the heat.