86 
MR. P. V. SEVAN ON THE COMBINATION OF HYDROGEN 
with the one preceding it. The effect of this is to make the acceleration of the 
expansion less than it should be. There remains no doubt that the acceleration 
exists. The reason that these experiments were made in such detail was that 
PniNGSHEiM"^ concluded that the amount of the expansion for a given quantity of 
light is independent of the state of the induction. It was partly on this conclusion 
that he based his theory that the first action of the light is to cause a dissociation of 
hydrogen and chlorine molecules into atoms. 
§ 2. The Relation to the Hydrochloric Acid Formed. 
The next point to be considered is the relation between the initial expansion and 
the hydrochloric acid formed. Experiments on this subject are considerably more 
difficult than experiments on the exj^ansion alone. For the complete contraction, 
defining the amount of hydrochloric acid formed, takes a considerable time, and so 
very small and slow changes of temperature of the whole apparatus may cause 
relatively large errors. The expansion is complete in seconds, while the contraction 
requires minutes to come to an end. For this reason numbers expressing contraction 
are not very regular. It is also difficult to obtain a long series of comparable 
experiments, since small changes of the conditions of temperature and pressure alter 
the behaviour of the gas mixture by changing the relative proportions of hydrogen 
and chlorine in the mixture. In this way experiments from day to day may differ, 
even when the most careful precautions are taken for complete saturation of all the 
liquids involved. 
The curves in fig. 6 represent the results of one experiment on the relation between 
the initial expansion and subsequent contraction representing the hydrochloric acid 
formed. In curve A the abscissae represent the contraction (hydrochloric acid 
formed), the ordinates the initial expansion. In B are plotted the contraction and 
the ratio of initial exjjansion to the contraction. C is the early part of A enlarged. 
The result obtained from such exjDeriments is that the initial expansion is always 
followed by contraction, showing the formation of hydrochloric acid. The ratio of the 
expansion to the final contraction below the original volume increases as the final 
contraction diminishes. This is to be expected if the expansion be due to heat 
arising from combination, for the shorter the time of illumination the less heat can 
escape. 
With tlie object of obtaining as short exposures as possible, the mixture was 
illuminated by the light from electric sparks. The sparks were obtained from an 
induction coil with several Leyden jars in circuit. No special care was taken to 
obtain sparks of constant length, the object being to see if, with instantaneous 
illumination, any hydrochloric acid at all was formed. The size of the spark could 
however be regulated so as to give expansions from 1 millim. of the capillary tube 
^ ‘ Wied. Ann.,’ 1887, vol. 32, p. 412, 
