BY THE BOILING POINT OF WATER. 411 



the Elasticities of Steam adapted to the measurement of heights by the thermo- 

 meter, and no original observations. 



During a late journey in Switzerland (in 1842), I made several observations 

 on the boiling point of water at great heights. Having long since abandoned 

 WOLLASTON'S thermometrical barometer as practically useless, I was led to re- 

 sume the method in consequence of a very ingenious and compact apparatus for 

 chemical or culinary purposes having been shewn to me the preceding winter, by 

 Mr STEVENSON, instrument maker, under the name of a Russian furnace, and 

 which was, I believe, introduced into the country from Russia by Dr SAMUEL 

 BROWN. It consists of a very thin cylindrical copper-pan for holding water, Fig. 

 1, Plate XL, with three moveable wire-legs. The bottom is flat, so that the flame 

 of a spirit-lamp plays fully upon it. This lamp or furnace consists of two parts ; 

 a flat dish or saucer, Fig. 4, containing a little alcohol, which is set on fire, 

 and then covered by the double dome-shaped vessel, Fig. 5, also of thin copper, 

 with an air-tight plug a, by which a certain quantity of spirit of wine is intro- 

 duced, and the lower part communicating with a bent tube or nozzle b, by which 

 alcohol in ebullition is violently projected by the pressure of its own vapour, 

 when heated by the flame in the saucer. The jet of burning spirit thus thrown 

 up like a volcanic explosion through the aperture of the dome, has such force as 

 to resist the blast of a hurricane, and plays right upon the bottom of the cylin- 

 dric boiler or pan. Two fluid ounces of spirit of wine, will thus boil above a pint 

 of water in still air in four minutes ; and I have frequently first melted snow, and 

 then brought it to boil to the amount of a pint, with little more alcohol, but, of 

 course, in a longer time. 



The furnace and boiling apparatus, together with a reservoir of alcohol, packs 

 into the copper-pan, and that into a cylindrical leather case 4 inches high, and 

 6 in diameter. The thermometer, Fig. 2, is carried separately. It is 15 inches 

 long and the degrees measure fg inch, which is quite sufficient in practice. Paral- 

 lax is avoided, by having the scale repeated on each side of the tube on two 

 pieces of copper not in the same plane. 



Fig. 6 represents the spirit measure, Fig. 7 a reservoir for spirits, Fig. 8 a 

 water measure or cup, Fig. 9 a handle which opens all the plugs, and serves also 

 for lifting the lamp and pan when heated. 



I immediately saw the value of the apparatus for determining the boiling 

 point, and directed Mr ADIE to adapt a thermometer to it, graduated from 185 to 

 214 of FAHRENHEIT'S scale, divided to lOths of a degree, the divisions admit- 

 ting an estimation to 1 OOths. I am well assured, however, that in no circum- 

 stances, even the most favourable, is the observation true to less than go of a 

 degree. But this quantity corresponds to only 25 feet of elevation, and is there- 

 fore accurate enough for most purposes. The minute subdivisions of DELUC'S, 

 DE SAUSSUKE'S, and WOLLASTON'S instruments, are quite unavailing, as I have 

 found by using the instrument of the latter with every precaution. 



