AND CHLORINE UNDER. THE INFLUENCE OF LICIIT. 
81 
temperature of the whole apparatus alter the solubility of the chlorine, and produce 
slight changes in the relative amounts of chlorine and hydrogen in the insolation 
bulb. It was found also necessary not to alter greatly the rate of evolution of the 
gases. Such alteration produces effects which are probably to be attributed to the 
alteration of concentration of the solution round the electrodes in the generating 
vessel, which causes the gases for a time not to be given off in the right proportions. 
If the rate of evolution is not fast, however, this difficulty gives rise to no trouble. 
Following Bunsen and Roscoe, a small current was kept decomposing the acid 
continually during a series of experiments, so that there was always a slight excess 
of pressure inside the apparatus above that outside; in this way all danger of small 
quantities of air being drawn into the apparatus was avoided. 
In the first experiments performed the light of the sky was used, but this proved 
to be much too variable, so the Welsbach burner was employed. In some of the 
later experiments an incandescent electric lamp was used, and, for instantaneous 
illumination, a spark from a Ruhmkorff coil with several Leyden jars in circuit. 
If a not too powerful source of light be used, the phenomena observed when the 
light falls on the mixture are as follows :—For some time the water index remains 
stationary ; but after a certain time, depending on the intensity of the light, the 
water index begins slowly to move towards the insolation Imlb, indicating a decrease 
m the volume of the gas under observation caused by the absorption in the water at 
the bottom of the bulb of the hydrochloric acid formed. The rate of this motion 
increases at first slowly, that is, the acceleration is at first small. The acceleration 
increases to a maximum and then decreases to zero, when a maximum rate of action 
is reached which I'emains constant. The curves (fig. 4) represent the results of an 
experiment of this kind, showing the general phenomena of the “ Period of Induction.” 
The curve I has for abscissae the time in minutes from the beginning of illumination, 
and for ordinates the observed contraction from the beginning of the experiment, 
measuring the amount of hydrochloric acid formed. The curve II is the velocity 
curve for the same experiment. 
If the light be more intense the phenomenon known as the Draper effect appears. 
Instead of the period of inertness when the water index does not move, the 
index, immediately on illuminating the gases, moves away from the insolation bulb, 
showing an increase in the volume of the gas mixture. With very intense light this 
expansion can be made to increase in magnitude practically without limit, until the 
point is reached when explosion takes place. And this suggests at once that the 
phenomenon is a heat effect. The expansion may continue for as long as half-a- 
minute (there seems no reason why it should not continue for a longer period, the 
only difficulty being that with more intense light the whole apparatus is likely to 
explode). But after a time the expansion stops and at once tlie gas begins to 
contract; and then the general phenomena are the same as when no expansion 
occurs. If the illumination is stopped before the expansion is comjjlete, contraction 
VOL. ecu.—A. 
M 
