NATURAL PHILOSOPHY. 195 



to it, the effect which it produces is very great; but if it is at all loose, it is 

 much less efficacious than gunpowder. If a few pounds of powder explode 

 in a room, the devastation must necessarily be great; but if the same weight 

 of gun-cotton is strewn loosely on the floor and then ignited, a sudden " puff " 

 takes place, but the articles in the room are uninjured. The Austrian Gov- 

 ernment has expended a great deal of time and much money in making 

 experiments; but it is beginning to discover that powder is preferable to 

 cotton. 



A SIMPLE MEANS OF DEMONSTRATING THE WORKING OF LIQUID 



FIRE-SHELLS. 



The bi-sulphide of carbon is first poured into the shell, and then small bits 

 of phosphorus are dropped in; the mouth of the shell is then closed with a 

 cork, partly projecting, like the cork in a wine-bottle. The shell may then 

 be laid on canvas, or other combustible matter; and in about ten minutes, 

 the fermentation of the mixture will force its way through the pores of the 

 cork, and, meeting the oxygen of the atmosphere, will become ignited, the 

 cork acting like the wick of a candle, and the liquor underneath feeding it. 

 A leaden shell thus charged, and adapted to a military rifle, will continue to 

 burn for ten minutes with an intense flame, which cannot be extinguished 

 by water. London Mech. Magazine. 



ON THE FREEZING-POINT OF WATER IN CAPILLARY TUBES. 



Many years*ago, M. Donne showed that water enclosed in narrow tubes 

 of a substance capable of being welled up by it, might be raised to a tem- 

 perature considerably above 212 without boiling. Mr. H. C. Sorby, in a 

 note communicated to the London, Edinburgh, and Dublin Philosophical 

 Magazine, completes these researches by showing that in capillary tubes the 

 temperature of water may be lowered far below 32 without freezing, even 

 when the tubes are shaken. In tubes of from ^^o or "5^ v mcncs in diame- 

 ter, the water may be reduced to 5 D Fah., without freezing, provided it be not 

 in contact with ice. These experiments go to show that these phenomena 

 are caused by the adhesion of the water to the walls of the tube interfering 

 with its change of state. 





