478 Mr Wood, On the Effect of Screening 



across the insulation. This brass tube was fitted into a vulcanite 

 stopper which served to insulate it from the vessel, and the 

 electrode and insulation could thus be moved from one vessel to 

 another, and the ionisation in each directly compared, the vessels 

 being of the same size and the capacity of the electrode system 

 remaining the same. To the end of the brass wire which carried 

 the gold leaf of the electroscope was attached a small mercury 

 cup into which dipped the electrode wire. This cup and the 

 connecting wires were shielded from electrostatic influences by 

 a screen of wire gauze. In this screen a small hole was made 

 exactly above the mercury cup, and through it could be in- 

 serted into the cup one end of an insulated wire, the other end 

 of which was connected to a potentiometer arrangement. By this 

 means unnecessary insulation usually associated with a key was 

 avoided, the insulation being confined to one sulphur plug where 

 the electrode passed into the metal vessel and another where the 

 wire entered the electroscope. The method of measuring the 

 ionisation was as follows. The potentiometer wire was inserted 

 into the mercury cup and the electrode system brought to zero 

 and then insulated. After twenty minutes had elapsed the 

 position of the gold leaf was read by means of a microscope. The 

 potentiometer wire was reinserted in the mercury cup and the 

 resistance adjusted until the gold leaf was brought back to the 

 same position on the scale. The potentiometer reading then gave 

 the final potential of the electrode and this was a measure of the 

 leak in the vessel and consequently of the ionisation of the air. 

 This method of reading the electroscope gave results independent 

 of small changes of zero and of variations of sensitiveness in 

 different parts of the scale. With the volume of air used — about 

 400 c.c.s — it was found that the electrode was charged to about 

 0*4 volt in twenty minutes, an amount which could be measured 

 with a considerable degree of accuracy. The potential of the 

 vessel was 360 volts, which was found to be largely in excess of 

 that required to saturate the enclosed air. In the actual experi- 

 ments five sets of four readings each were taken — the first, third, 

 and fifth sets being taken in the ordinary way, and the second 

 and fourth being taken while the vessel was surrounded with 

 a lead or other screen. From these determinations the mean leak 

 with and without the screen was calculated and the percentage 

 reduction found. The first experiments were of a purely pre- 

 liminary character and the results obtained were in agreement 

 with those previously published by Rutherford and M c Lennan, as 

 also with those published by Cooke* while the present investi- 

 gation was in progress. Screens of lead, iron, wood, and paper 

 were tried. The lead screen was about "6 cm. thick and gave a 



* Phil. Mag. Oct. 1903. 



