430 EXPERIMENTAL INVESTIGATIONS, 



tube filled with water, the other with hydrogen, the platinised 

 platinum plates in each of these tubes were connected with a 

 galvanometer, and a deflection took place from the reaction of 

 the hydrogen on the air dissolved in the water. After a time 

 the deflection abated, and the needle returned to zero, all the 

 oxygen of the air having become combined with the hydrogen. 

 If now the stopper were taken out, a deflection of the galvano- 

 metric needle immediately took place, showing that the air 

 rapidly enters the water, as water would a sponge. Absolute 

 chemical purity in the ingredients is a matter- for refined 

 experiments, almost unattainable ; the more delicate the test, 

 the more some minute residual product is detected ; it would 

 seem (to put the proposition in a somewhat exaggerated form) 

 that in nature everything is to be found in anything if we 

 carefully look for it. 



I have indicated the above sources of error to show the 

 close pursuit that is necessary when looking for these minute 

 residual phenomena. Enough has, I trust, been shown in the 

 above experiments to lead to the conclusion that, hitherto, 

 simple boiling, in the sense of a liquid being expanded by 

 heat into its vapour without being decomposed or having per- 

 manent gas eliminated from it, is a thing unknown. Whether 

 such boiling can take place may be regarded as an open ques- 

 tion, though I incline to think it cannot ; that if water, for 

 instance, could be absolutely deprived of nitrogen, it would 

 not boil until some portion of it was decomposed ; that the 

 physical severance of the molecules by heat is also a chemical 

 severance. If there be anything in this theoretic view, there 

 is great promise of important results on elementary liquids, if 

 the difficulties to which I have alluded can be got over. 



The constant appearance of nitrogen in water, when boiled 

 off out of contact with the air almost to the last drop, is a 

 matter well worthy of investigation. I will not speculate on 

 what possible chemical connexion there may be between air 

 and water ; the preponderance of these two substances on the 

 surface of our planet, and the probability that nitrogen is not 

 the inert diluent that is generally supposed, might give rise 

 to not irrational conjectures on some unknown bond between 

 air and water. But it would be rash to announce any theory 



