on Chlorine and Bromine. 293 



apparatus was emptied and refilled, the anomaly in its working 

 disappeared. The displacements were rather slow, and some- 

 times took several minutes to attain their maximum; but this 

 is satisfactorily accounted for by the viscosity of the fluid; for 

 (1) the same slowness of motion is observed when the instru- 

 ment is used as a common differential thermometer in the dark; 

 and (2) when chloride of carbon, C CI 4 , was substituted for sul- 

 phuric acid, the index attained its final position in a few seconds. 

 Hence it may be assumed that the state of equilibrium in inso- 

 lated chlorine very quickly follows any change in the (chemical) 

 intensity of the illumination. The light I used was sunlight of 

 only middling brightness, which was reflected by a bad mirror 

 and decomposed by a small glass prism; the " ultra-violet " 

 above-mentioned only encompasses the rays not absorbed by 

 glass. 



Incontestably proved by these experiments is the fact of the 

 existence of a substance which apparently behaves to " actinic " 

 as most other known bodies do to " thermic " rays. I have 

 little doubt that in my experiments it really was the chemical 

 individual chlorine which produced the effect. In order to get 

 further conviction, control experiments were made. 



(1) An ordinary but very sensitive differential thermometer 

 (with an indicator probably consisting of coloured spirit) was 

 treated quite as the former one ; in the more refrangible part 

 of the spectrum no increase of temperature could be observed. 



(2) A differential thermometer charged with carbonic acid 

 and ether as an indicator behaved in the blue light like the one 

 with air and spirit ; according to Tyndall, the contents of the 

 bulbs should absorb heat far more largely than chlorine does. 



(3) A differential thermometer charged with chlorine (and 

 vitriol) was kept in a water-bath and exposed to direct sun- 

 light. By alternately shading the one and the other of the 

 bulbs I produced displacements of the index amounting to seve- 

 ral centimetres, which I am inclined to ascribe essentially to 

 the action of the chemical rays, because 



(4) A CO 2 thermometer, under the same circumstances, ex- 

 hibited no action ; and 



(5) On shading the bulbs with a plate of blue cobalt-glass, 

 about one quarter of the effect of the insolation remained. 



The mere fact that there is a body which shows the phe- 

 nomena in question is of great interest, suggesting, as it does, 

 the possibility of constructing an " actinometer " which could 

 be read off as easily as an ordinary mercury thermometer. I 

 purpose undertaking experiments in this direction. It is not 

 very likely that chlorine (and bromine) should be the only 

 substance to show that remarkable behaviour in actinic rays. 



