630 SCIENCE PROGRESS 



which has been abundantly justified by the observations of 

 later workers, followed as a logical result from the con- 

 sideration of the analogous cases discussed above. Thus 

 the double decomposition of pairs of salts is usually confined 

 to cases in which at least one of the interacting compounds 

 is in the dissolved or liquid state, and dissociation is equally 

 a characteristic of the fluid and not of the solid condition. 

 Of this, nitrogen peroxide affords an excellent illustration, 

 for whilst the gas is brown, and the liquid straw-yellow, the 

 solid which separates on further cooling forms an absolutely 

 colourless ice, and therefore consists entirely of the tetroxide ; 

 similar considerations doubtless apply to the conversion of 

 water into ice, although the evidence for the homogeneity 

 of the solid is less conclusive than in the previous case : in the 

 cases of ammonium chloride and phosphorus pentachloride, 

 however, a direct analytical proof is possible, since the solid 

 that separates has a constant composition even when the 

 vapour contains a variable excess of one of the dissociation- 

 products. The application of those observations to the 

 analogous phenomena of dynamic isomerism is obvious, namely, 

 that whilst oscillatory isomeric changes may occur freely, 

 as Butlerow suggested, in gases and liquids, the separation 

 of crystals will usually result in the production of a homo- 

 geneous solid in which no further change can take place. 



Arrest of Isomeric Change. — One other important result 

 follows from Butlerow's close association of reversible isomeric 

 change with chemical actions of other kinds, namely, that 

 any development of the theory of chemical change must apply 

 equally to this as to other cases. When, therefore, it 

 became evident from the works of Dixon (Chem. News, 1882, 

 46, 151) that the presence of moisture was an essential factor 

 in the explosive combustion of carbonic oxide, and from 

 the observations of Baker {Phil. Trans. 1888, 179, A 571) that 

 similar considerations applied to the combustion of carbon 

 and of phosphorus in oxygen, it became likely that the presence 

 of a third substance or catalyst might be necessary in all cases 

 of chemical change, including those now under consideration. 

 This probability ripened almost into certainty when Baker 

 showed that moisture was necessary to promote both the 

 union of ammonia and hydrogen chloride and the decomposition 

 of ammonium chloride by heat (Trans. Chem. Soc. 1894, 65, 



