282 Dr. A. Richardson on the Measurement of the 



on again exposing the tube to light, after two days, an ex- 

 pansion of 5 centim. occurred. This is a point of some im- 

 portance where the expansion of chlorine as against air is 

 made a measure of the actinic intensity of light, as the 

 diffusion of chlorine from the sulphuric acid into the air-bulb 

 would introduce a gradually increasing source of error*. 



As it seemed possible that the sensitiveness of chlorine to 

 light might be influenced by the temperature of the gas 

 during exposure, an experiment was made in which the two 

 tubes A and B, fig. 2, Plate III., were filled with chlorine, 

 these were connected together by a tube which could be 

 closed by means of a stop-cock F ; changes in volume being 

 measured as before by means of the gauge C containing sul- 

 phuric acid and bubbles of air. The tube A was kept at a 

 temperature of 14° 0. by a current of water circulating 

 through the outer tube D, whilst B was heated in the jacketing 

 tube G Gr, by the vapour from chlorobenzene boiling in E at a 

 temperature of 132° C. The pressure in the two tubes was at 

 first equalized by opening the stop-cock F ; and when the 

 temperature was constant in the two tubes (as shown by the 

 index remaining stationary when F was closed), the tubes 

 were alternately exposed to and shaded from the light. The 

 following is the mean of a series of experiments made : — 



Both tubes shaded, zero = 0. 



Heated tubes exposed = 6*0 centim. expansion. 



Both tubes shaded, zero = 0. 



Cooled tubes exposed = 6*7 centim. expansion. 



Both tubes shaded, zero = 0. 



In the heated tube the illumination of the gas was interfered 

 with by the partial refraction of the light, due to the con- 

 densation of the chlorobenzene on the sides of the jacketing 

 tube, and further it is possible that the vapour of chlorobenzene 

 may absorb a portion of the actinic rays. This probably 

 accounted for the slight difference observed in the expansion 

 of the gas in the two tubes. From this experiment it is 

 concluded that the expansion of chlorine by light is practically 

 unaffected within a range of temperature between 14 c and 

 138° C. 



Having made these preliminary experiments, an apparatus 

 was next devised whereby the automatic registration of the 

 actinic intensity of light was effected by the expansion of 

 chlorine. This was done by suspending a differential appa- 



* In the chlorine actino meter it is found desirable to cut off com- 

 munication between the two bulbs, when not in use, by means of a 

 stop-cock. 



