of Hydrogen and Chlorine under the influence of Light. 399 



temperature by a thermostat. The inner bath was stirred and 

 kept at a temperature varying not more than 2 or 3 degrees. 

 The inner bath was of glass, and was covered almost entirely 

 with a frame of wood which served to keep all extraneous light 

 from the bulb and also to stop loss of heat. This wooden frame 

 had a hole cut in it through which light could be admitted to 

 the bulb. For a constant source of light an ordinary gas burner 

 was used ; in front of this was an iron screen with five holes cut 

 in it at the corners and the centre of a square. The centre hole 

 alone served for transmitting light to the bulb, a lens sending 

 light transmitted through this hole in parallel rays on to the 

 bulb. The other four holes with the lens in position gave images 

 on the side of the frame over the water bath, and by adjusting 

 the size of the flame these four images could be made of equal 

 brightness, thus insuring that the light through the centre hole 

 was from the brightest part of the flame, and that the flame was 

 of practically constant size. The constancy of the light thus 

 obtained was quite satisfactory. The constancy was tested at 

 various times, and rates of action consistent to within four or 

 five per cent, were obtained, individual experiments ranging over 



yjtKmwMM.'M.v 



Fig. 1. 



a bulb containing gas mixture. 



b stirrer. 



c inner water bath. 



d index tube. 



e wooden case. 



a period of three months. Fig. 1 shews the arrangement of the 

 bulb and baths. 



When the temperature of the bath is changed, changes in the 

 composition of the gas mixture take place, and several days are 



