AND CHLORINE UNDER THE INFLUENCE OF LIGHT. 
99 
direction.” In fact, any operation with the gas mixture is sufficient to alter its 
properties. The only method of procuring reliable results is to evolve the impurity 
required from a second electrolytic vessel and to pass it continuously through the 
apparatus together with the normal mixture for some days, using the same current 
for producing the hydrogen and chlorine and the impurity. However, even under 
Fig. 14. Abscissae, time from beginning of action in minutes. 
Ordinates, HCl formed. 
these circumstances, the results obtained are not in any very close agreement. The 
only conclusion arrived at is that the effect of impurity is large at first, l)ut the 
effect increases in a much less ratio than the amount of impurity. 
There is one form of impurity, however, which is not subject to such experimental 
difficulties as occur with others, and that is the uninsolated gas mixture itself. The 
o 2 
