206 Mr. M. Ishino on Velocity of Secondary Cathode Rays 



The actual measurements were made at the pressures between 

 0'003 and 0*0008 mm. of mercury. At this low pressure, 

 the current of the secondary corpuscles was diminished, but 

 was quite measurable owing to the form and position of the 

 electrometer-electrode and to the highly sensitive electro- 

 meter employed. The sensitiveness of the electrometer was 

 1500 scale-divisions per volt at 1*5 metre distance when the 

 needle potential was 200 volts. 



The different discharging potentials in the bulb were 

 obtained by slightly varying the pressure, or by changing 

 the primary current of the induction-coil. The potential 

 was measured with a spark-gap micrometer having two 

 brass spheres of equal diameters of 3' 5 cm. connected with 

 the discharge-bulb in parallel. 



Hydrogen gas was prepared by the electrolysis of a solu- 

 tion of barium hydroxide, and the ionization-chamber was 

 rinsed with a stream of the hydrogen for a considerable 

 time. The purity of the hydrogen gas filling the chamber 

 was examined. This was done by connecting a small spec- 

 trum tube to the ionization-chamber, passing a discharge 

 through the tube, and measuring the wave-lengths of the 

 spectral lines from the gas by a'Hilger wave-length spectro- 

 meter. The examination of the spectrum showed that the 

 ionization-chamber was filled with quite pure hydrogen 

 gas. The same examination also proved that the cooling by 

 means of U-tubes was absolutely necessary in order to free 

 the gas in the chamber from any wax vapours. Measurement 

 of the current due to the secondary rays was made to at 

 least half an hour after applying the cooling material to 

 the U-tubes. 



A great difficulty experienced, however, was in geting a 

 constant deflexion of the electrometer, which measured the 

 ionization current inside under a definite condition. Much 

 time was spent in trying to effect this. The induction-coil 

 worked uniformly, and the electrometer very satisfactorily. 

 The McLeod gauge showed that the pressure did not suffer 

 marked change. But the variation of the electrometer de- 

 flexion amounted sometimes to 15 per cent. The cause of 

 this variation might be either a small fluctuation of the 

 induction-coil, or a small trace of wax vapour inside the 

 ionization-chamber. 



A method of taking readings of the electrometer de- 

 flexion illustrated in the following table was adopted to 

 overcome the difficulty above mentioned. Here v was the 

 potential applied to one of the gauzes, and was varied to 

 any desired value, while the other gauzes were kept at 



