120 Prof. Dixon and Mr. Harker on the 



and was then sealed up with phosphoric acid, and kept 

 several weeks, it could not be exploded by the flash at 

 i^ feet. We may state, as a general result, that a mag- 

 nesium flash (which will invariably explode a wet mixture 

 of hydrogen and chlorine at 5 feet distance) has to be 

 brought within one foot in order to explode a well dried 

 mixture of the same gases ; in other words the intensity of 

 the light must be increased 25 times to produce the same 

 effect on the dry as on the moist gases. This difference in 

 the action of light is so well marked that we have con- 

 tinually made use of it as a test of the dryness of the 

 electrolytic gas used in our experiments on the volume- 

 relation of hydrogen and chlorine to the hydrogen chloride 

 gas formed from them. 



To determine the law which regulates the division of 

 hydrogen between chlorine and oxygen, in an explosion of 

 the mixed gases, it is essential that the gases should be 

 accurately measured. All our attempts to measure chlorine 

 in a eudiometer over a liquid have failed. Oil of vitriol 

 (used by Botsch) absorbs chlorine ; liquid chlorinated hydro- 

 carbons absorb chlorine rapidly. We have had to measure 

 chlorine by filling a glass explosion vessel of known capacity 

 with it under known conditions ; and have then introduced 

 into the same vessel known volumes of hydrogen and oxy- 

 gen. Our latest apparatus consists of two strong bulbs (of 

 iSoc.c.and 140 c.c. capacity), which can be put into communi- 

 cation through a wide tap. Each bulb can be filled separately ; 

 for instance, one may be filled with electrolytic hydrogen 

 and chlorine and the second with oxygen, or one may be 

 filled with electrolytic hydrogen and oxygen and the second 

 with chlorine. The two bulbs are then put into communi- 

 cation, and the gases allowed to mix by diffusion. A pressure 

 tube is attached to the bulbs, and is shut off by a stopcock. 

 This tube is filled with an inert gas, e.g. nitrogen, and dips 

 into a mercury trough. When all is ready for an explosion 



