DECOMPOSITION BY LIGHT. 97 



the lamp, were coincident. In the first experiments 

 the two ends of the tube were closed by plates of rock- 

 salt, and subsequently by plates of glass. For the sake 

 of distinction, I call this tube the experimental tube. 

 It was connected with an air-pump, and also with a 

 series of drying and other tubes used for the purifica- 

 tion of the air. 



A number of test-tubes, like F, fig. 2 (I have used 

 at least fifty of them), were converted into Woulf's 

 flasks. Each of them was stopped 

 by a cork, through which passed 

 two glass tubes: one of these 

 tubes (a) ended immediately 

 below the cork, while the other 

 (b) descended to the bottom of 

 the flask, being drawn out at its 

 lower end to an orifice about 

 0.03 of an inch in diameter. It 

 was found necessary to coat the 

 cork carefully with cement. In 

 the later experiments corks of 

 vulcanised india-rubber were in- 

 variably employed. 



The little flask, thus formed, 

 being partially filled with the 

 liquid whose vapour was to be 

 examined, was introduced into 

 the path of the purified current 

 of air. The experimental tube 

 being exhausted, and the cock 

 which cut off the supply of 

 purified air being cautiously 

 turned on, the air entered the flask through the tube ft, 

 and escaped by the small orifice at the lower end of 

 6 into the liquid. Through this it bubbled, loading 



