SECT. IT. BY GASES AND VAPOURS. 43 



Thus when atoms of hydrogen and nitrogen are 

 mixed in the proportion of three to one, the absorption 

 of the mixture is represented by unity ; but when they 

 are chemically united in ammonia, the absorption is 1190 

 times as great. Atoms of hydrogen and oxygen mixed 

 in the proportion of two to one absorb very feebly ; when 

 chemically united into a molecule of aqueous vapour 

 the absorptive power is enormous. The absorptive 

 power of nitrous oxide ?> a chemical compound of oxygen 

 and nitrogen, exceeds that of dry air 250 times ; a con- | 

 vincing proof that the atmosphere is a mixture and not^ 

 a compound gas. Olefiant gas at five inches of tension 

 absorbs 1000 times that of its constituent hydrogen. In 

 fact all the compound gases and vapours far surpass 

 the simple elementary gases and dry atmospheric air in 

 their capacity for absorption. 



Chlorine and bromine, which have so many singular 

 properties in common, have this peculiarity also, that 

 though simple substances respectively formed of homo- 

 geneous atoms, their absorptive powers are similar to 

 those of compound substances, for the absorptive power 

 of chlorine is 60 times that of the elementary gases, and 

 that of bromine 160 times. This high absorptive power 

 is ascribed by Professor Tyndall to their atoms being 

 united into groups which act powerfully as oscillating 

 systems, instead of the feeble action of single atoms. 



Ozone is an analogous instance of the presumed union 

 of homologous atoms into oscillating groups. By com- 

 paring the absorptive effect of ozonized oxygen obtained 

 from the electric decomposition of water with that of 

 the same oxygen deprived of its ozone by passing it over 

 a very strong solution of iodide of potassium, Professor 

 Tyndall found that ozonized oxygen possesses an absorp-> 

 tion force 136 times greater than that of pure oxygen.| 

 The quantity of ozone producing this astonishing effect 

 was too small even to admit of estimation, far less 



