668 SCIENCE PROGRESS 



bubbled the gases through water into the insolation vessel 

 after they had been exposed to light. It was this discovery, 

 coupled with the fact that the least expansion capable of pro- 

 ducing a cloud in moist chlorine is less in the illuminated than 

 in the darkened gas, that caused Bevan to infer that the first 

 action of light on a mixture of chlorine and hydrogen is to 

 produce a soluble " addition" compound of chlorine and water 

 and that the molecules of this compound can act as condensation 

 nuclei for water vapour. The further action of light, according 

 to Bevan, causes these molecules to combine with molecules of 

 hydrogen and form a compound (similar to that postulated 

 by Mellor) which finally breaks up into hydrogen chloride 

 and water. 



J. J. Thomson (Proc. Camb. Phil. Soc. 1900-1, 11, 90) has 

 proved with the aid of an electroscope of the C. T. R. Wilson 

 type that either no free ions are formed in a mixture of hydrogen 

 and chlorine when it is exposed to light or that, at any rate, the 

 number over and above those contained in the darkened gas is 

 negligible. A very slow interaction of chlorine and hydrogen 

 takes place in the presence of those radium rays which will 

 penetrate thin layers of glass and aluminium (W. P. Jorissen 

 and W. E. Ringer, Bench te der deutsch. Chem. Gesell. 1905, 

 38, 899). There is therefore very little experimental evidence 

 favouring the view that the interaction is necessarily preceded 

 by dissociation. 



The theories of photochemical induction existing at the time 

 that Burgess and the writer began to work on the subject can 

 be arranged under three heads, namely : 



(1) Draper's theory : That the light increases in some unknown 

 way the electro-negative properties of the chlorine and thereby 

 exalts the affinity between chlorine and hydrogen. 



(2) Bunsen and Roscoe's theory : That the light gradually 

 removes the resistance which opposes the force of affinity. 



(3) The intermediate compound theories : That chlorine and 

 hydrogen cannot interact directly but that the change must 

 be considered as consisting of several distinct chemical 

 transformations. 



It will be seen that both Draper and Bunsen and Roscoe 

 approached the problem in a way essentially different from that 

 pursued by the later investigators : they regarded the induction 

 period as something essentially new and inexplicable by known 



