628 SCIENCE PROGRESS 



the cyanogen compounds as analogous with the reversible 

 actions which Berthollet had studied in aqueous salt-solutions, 

 for instance — 



2KNO3 + CaCl, ^> 2KCI + Ca(NO s ) 2 



(Essai de Statique c/rimique, 1803, vol. i. p. 99), a case in which 

 by varying the proportions slightly it was possible to crystallise 

 out from the solutions either the chloride or the nitrate of 

 potassium, quite independently of whether the initial materials 

 were calcium chloride and potassium nitrate or calcium nitrate 

 and potassium chloride, and without the addition of any sub- 

 stance comparable with the sulphuric acid used to establish a 

 condition of equilibrium between the isodibutylenes. 



In order to elucidate his views, Butlerow also compared the 

 reversible isomeric changes of cyanic and hydrocyanic acids 

 with the conception of " molecular concurrence," which 

 Pfanndler had introduced in a paper on " The Struggle for 

 Existence amongst Molecules : a further contribution to 

 Chemical Statics" {Pogg. Ann. Jubelband, 1874, 182), in order 

 to describe the " simultaneous reciprocal reactions" which he 

 recognised as occurring in the double decompositions of 

 Berthollet and the dissociation phenomena studied by St. Claire 

 Deville. The comparison of dynamic isomerism with the 

 " simultaneous reciprocal reactions " of dissociation was par- 

 ticularly apt, as may be shown by reference to one of the actual 

 cases investigated by Deville. Nitrogen peroxide, he found, 

 could not be represented by any one single formula, since its 

 vapour density proved it to consist — at all temperatures between 

 20 and 150 — of a mixture of the dioxide and tetroxide in 

 equilibrium with one another ; the gas must therefore be 

 formulated as 



N.,0. t> 2 NO, 



In the case of the liquid this form of evidence is no longer 

 available ; but if the colour of the dioxide be used as a test for 

 its presence, the liquid also must be represented by a balanced 

 equation, but showing the tetroxide as the dominant constituent 



N 2 4 ^ 2N0 3 



Liquid nitrogen peroxide 



If then, as Butlerow suggests, the spontaneous occurrence of 

 a reversible isomeric change is a cognate phenomenon, hydro- 



