The Explosion of Hydrogen and Chlorine. J 



three hours with the average current. The whole of the 

 generating and drying apparatus was protected from Hght 

 by being covered with black cloth, and the corridor was 

 only illuminated by one or two small gas flames, which did 

 not have the least perceptible action on the mixture. 



The method used to measure the rate of the explosion- 

 wave was similar to that used for other gaseous mixtures, 

 viz.: to make the explosion break two silver or platinum 

 strips or " bridges" stretched across the tube, one near each 

 end. These bridges carried currents, which actuated two 

 electro-magnetic styles making traces on a moving plate. 

 The bridges were fitted to two separate tubes, which could 

 be rapidly joined up to the long glass explosion tube by 

 means of steel flanges. These separate "end tubes" were 

 charged with a mixture of oxygen and hydrogen. When 

 the long tube was completely filled with the electrolytic 

 hydrogen and chlorine, the two "end tubes" carrying the 

 bridges were fastened to it, the electric connections were 

 made to the chronograph, and the gases fired by sending 

 a spark through the oxygen and hydrogen in the firing tube. 

 The flame travelled down this tube, which was about 

 4 feet long, breaking the silver bridge at the end of it, and 

 communicating the explosion to the hydrogen and chlorine. 

 On reaching the end of the glass tube, the explosion broke 

 the silver bridge fitted to the second end tube. In the few 

 seconds which elapsed between joining on the end tubes and 

 firing the mixture, but little diffusion of the gases was pos 

 sible ; and the silver bridges, being coated with paraffin, were 

 hardly acted on by the chlorine. Between each experiment 

 all the connections of the chronograph were reversed ; so 

 that whatever error was due to the defects of the chronograph 

 and connections was reversed in the second experiment. The 

 results are therefore grouped together in pairs ; the means 

 of the several pairs being closely concordant. The following 

 are the results obtained in the first series of experiments : — 



