256 THE CHLOR-HYDROGEN PHOTOMETER. [MEMOIR XVIII. 



hydrogen gas is passed in minute bubbles, it removes 

 with it a certain proportion of the chlorine. From this, 

 therefore, it is plain that hydrochloric acid thus decom- 

 posed will not yield equal measures of chlorine and hy- 

 drogen unless it has been previously impregnated with 

 a certain volume of the former gas. Nor is it possible 

 to obtain that degree of saturation by voltaic action, no 

 matter how long the electrolysis is continued, if the hy- 

 drogen be allowed to pass through the liquid. 



Practically, therefore, to obtain the photometric liquid 

 we are obliged to decompose commercial hydrochloric 

 acid in a glass vessel, the positive electrode being at the 

 bottom of the vessel and the negative at the surface of 

 the liquid. Under these circumstances, the chlorine as 

 it is disengaged is rapidly taken up, and the hydrogen 

 being set free without its bubbles passing through the 

 mass, -the impregnation is carried to the point required. 



Although this chlorinated hydrochloric acid cannot of 

 course be kept in contact with the platinum wires with- 

 out acting on them, the action is much slower than 

 might have been anticipated. I have examined the 

 wires of photometers that had been in active use for 

 four months, and could not perceive the platinum sen- 

 sibly destroyed. It is well, however, to put a piece of 

 platinum foil in the bottle in which the supply of chlo- 

 rinated hydrochloric acid is kept ; it communicates to it 

 slowly the proper golden tint. 



The liquid being impregnated with chlorine in this 

 manner until it exhales the odor of that gas is to be 

 transferred to the siphon a b c of the photometer, and its 

 constitution finally adjusted as hereafter shown. 



Thirdly, of the Voltaic Battery. The battery which 

 will be found most applicable for these purposes con- 

 sists of two Grove's cells, the zinc surrounding the plati- 

 num. 



