74 Dr. S. J. Plimpton on the Recombination of 



If when the discharge occurs the key F is closed or opened 

 to a greater distance, no spark will be seen at F. A spark 

 is seen at F only when the key F is so situated in the path 

 of the pendulum with regard to the key S that the contacts 

 of the key F are separated by a distance less than 0*5 milli- 

 metre and greater than millimetre during the flash of 

 X-rays. If it opens near the beginning of the flash, only a 

 small spark will appear since the discharge potential has not 

 then attained its maximum, so that when the key has opened 

 only a short distance, the resistance becomes too great and 

 the spark ceases. On the other hand, if the key F begins 

 to open so near the end of the duration of the discharge that 

 when its contacts have separated only a very small distance, 

 the discharge has ceased, then again only a small spark will 

 be observed. By placing the key F at different positions 

 relative to S, it was found that the discharge in the X-ray 

 tube began practically simultaneously with the opening of 

 the key S, increased to maximum and then fell to zero, 

 the total duration of the flash being 0*0035 second. The 

 limiting position of the key F at which the discharge ceased 

 could be determined within a distance corresponding to 

 0*0001 second. 



The calibration of the pendulum, as determined by the 

 condenser method described above, was corrected for the 

 time occupied by the flash by subtracting 0*0015 second 

 from each value. It was necessary to know the total 

 duration of the flash accurately also for a special experiment 

 to be described. 



Preliminary Experiment. 



In order to ascertain the effect of secondary and corpus- 

 cular radiation from the electrodes A and B of fig. 1, the 

 lead screens MM' and NN' were removed, allowing the 

 Rontgen rays to fall on A and B as well as in the space 

 between them. The plate A was charged continuously to 

 160 volts. Air was introduced into the chamber K and a 

 series of deflexions noted for different pressures using single 

 flashes of X-rays of constant intensity. The air was then 

 pumped out and similar readings taken for hydrogen. The 

 results are plotted in fig. 4, where the abscissa3 represent 

 pressure in millimetres of mercury, and the ordinates electro- 

 meter readings in scale-divisions. It will be seen that each 

 curve beginning at the left has a marked curvature but 

 eventually becomes a straight line, and that these two 



