12 Prof. R. Bunsen on a Method of determining the 



behind the cork, which would render the results of the experi- 

 ment completely illusory. The vapours from the liquid which 

 was boiled in the thermostat pass through the tube b into the 

 cooling-apparatus represented in fig. 3 ; and they condense at f, 

 in consequence of the radiation of heat from the platinum-wire, 

 and form a thin layer of liquid which prevents the entrance of 

 air into the vapour- vessel. When the evaporation is nearly 

 done, the bubbles of vapour pass through this layer of liquid 

 very slowly ; and when it is complete these bubbles cease entirely. 

 The layer of liquid is then drawn up between the glass and pla- 

 tinum-wire into the capillary of the vapour -vessel, and forms 

 there a thread of liquid which acts as a vapour-tight stopper. 

 There is no danger of this thread of liquid being pushed forward 

 and destroyed by the expansion of the vapour heated from its 

 boiling-point to the temperature of the thermostat, since all 

 vapours entering from the vapour-vessel into the capillary are 

 condensed there, and thus serve only to increase the length of 

 the thread of liquid. On the other hand, the thread cannot be 

 sucked back beyond the cork from cooling of the vapour, since 

 it must approach the heated part of the tube d, and cannot pass 

 that point without itself producing vapour which would condense 

 in its original position between c and/. If the thread is pushed 

 forward by degrees as the temperature rises, we may conclude 

 with certainty that the vapour is rendered impure by a per- 

 manent gas. This may result either from some atmospheric air 

 not having been expelled when the tube was filled, or from the 

 decomposition of the vapour into permanent gases. The tem- 

 perature to which a vapour may be heated without decomposition 

 is shown at once by this indication. 



The vapour-stopper which has been described renders the esti- 

 mation of vapour-densities a very convenient and certain opera- 

 tion. The vessel containing the liquid to be evaporated has 

 only to be placed in the thermostat next the one containing dry 

 air ; the apparatus may then be left without supervision until 

 the thread of fluid is formed in the capillary tube. The vessel 

 filled with dry air is then connected with a chloride-of-calcium 

 tube, and the apparatus again left to itself for an hour and a 

 half; the thread of liquid is now expelled as quickly as possible 

 by heating between d and c, and, lastly, the caoutchouc connec- 

 tors are fastened on. The vessels need now only be wiped, drawn 

 through the flame, allowed to cool for two hours, and the differ- 

 ence of weight from the vacuous vessel determined on a good 

 balance. 



The method which has been described has the advantage that, 

 by the use of the same vessels, the specific gravity of both gases 

 and vapours is obtained by simple estimation of two differences 



