428 EXPERIMENTAL INVESTIGATIONS. 



and that under such circumstances, the boiling took place by 

 soubresauts. I have, I trust, shown that when the vapour 

 liberated by boiling is allowed to condense, it does not alto- 

 gether collapse into a liquid, but leaves a residual bubble of 

 permanent gas, and that at a certain point this evolution be- 

 comes uniform. 



Boiling, then, is not the result of merely raising a liquid to 

 a given temperature, it is something much more complex. 



One might suppose that with a compound liquid the 

 initial bubble by which evaporation is enabled to take place 

 might, if all foreign gas were or could be extracted, be formed 

 by decomposition of the liquid, but this could not be the case 

 with an elementary liquid whence then comes the oxygen 

 from bromine or the hydrogen from phosphorus and sulphur ? 

 as with the nitrogen in water, it may be that a minute 

 portion of oxygen, hydrogen, or of water is inseparable from 

 these substances, and that if boiled away to absolute dryness, 

 a minute portion of gas would be left for each ebullition. 



With water there seems a point at which the temperature 

 of ebullition and the quantity of nitrogen yielded become 

 uniform, though the latter is excessively minute. 



The circumstances of the experiments with bromine, phos- 

 phorus, and sulphur, did not permit me to push the experi- 

 ment so far as was done with water, but, as far as it went, the 

 result was similar. 



When an intense heat, such as that from the electric spark 

 or voltaic arc, is applied to permanent gas, there are, in the 

 greater number of cases, signs either of chemical decomposition 

 or of molecular change ; thus compound gases, such as hydro- 

 carbons, ammonia, the oxides of nitrogen, and many others 

 are decomposed. Phosphorus in vapour is changed to allo- 

 tropic phosphorus, oxygen to ozone, which, according to 

 present experience, may be viewed as allotropic oxygen ; 

 there may be many cases where, as with aqueous vapour, a 

 small portion only is decomposed, and this may be so masked 

 by the volume of undecomposed gas as to escape detection ; 

 if, for instance, the vapour of water were incondensable, the 

 fact that a portion of it is decomposed by the electric spark or 

 ignited platinum would not have been observed. 



