connected with the Boiling of Liquids. 443 



expense of a certain quantity of air." M. Gernez says: — "Each 

 bubble of vapour produced during the ebullition draws off with 

 it a certain quantity of air. This proposition is contrary to the 

 opinion of Mr. Tomlinson, who regards it as improbable. Never- 

 theless it may be seen a priori, from the very nature [disposition) 

 of the experiment, that the initial bubble of air occupying the 

 whole cavity of the bell [that is, a minute bell-glass full of air 

 attached to a glass rod and lowered into the water] expands 

 with the vapour formed at the surface of the liquid, so that it is 

 a bubble of mixed air and vapour that escapes. Whatever num- 

 ber of bubbles be disengaged, there will always remain a de- 

 terminate fraction, although infinitely small, of the quantity of 

 air that existed at the commencement/'' 



That is, every one of the half million of bubbles that escapes 

 draws upon the bubble of air of which the initial diameter is one 

 millimetre ; and, however long the process be continued, this 

 rapidly diminishing speck of air never vanishes altogether, but 

 continues to maintain the ebullition to infinity. 



Even if the small bell be filled with water instead of air, and 

 by a brisk application of the flame the water be made to boil, 

 the little bell becomes filled with steam, and this carries on the 

 operation for an indefinite time; but we are told that if the 

 water be left to cool, a minute speck of air will be found in the 

 bell. 



Of course with such ideas as these M. Gernez objects to my 

 definition of a liquid at or near the boiling-point, namely that 

 it is a supersaturated solution of its own vapour, on the ground 

 that it implies that in the same point of space there are at the 

 same temperature the same substance in the liquid state and in 

 the state of vapour. A liquid supersaturated with its own vapour 

 is as easy to conceive as a liquid supersaturated with a gas ; and 

 since the behaviour of the various nuclei is the same in both, I 

 infer that they are constituted alike. A fatty, or resinous, or 

 porous body to which the water will not adhere, but the gas or 

 the vapour will, is an efficient nucleus in separating gas or 

 vapour from solution, provided the one or the other already 

 exists in the liquid ; while there is no such separating action 

 if the solid be of vitreous structure and chemically clean. M. 

 Gernez admits that carbonic acid gas, ammoniacal gas, &c. do 

 exist in their gaseous solutions ; but he cannot admit that the 

 vapour of water, for example, exists in solution when the water is 

 raised to the boiling-point or superheated. What, then, is the 

 definition of such a liquid ? According to M. Gernez, it is 

 " a supersaturated solution of the gases of the air." Now, in the 

 case of recently distilled water we may dismiss carbonic acid. 

 If the water be cold, it is capable of dissolving only one fiftieth 



