50 Dr. R. T. Beatty on 



The liquid resistance was contained in a glass tube whose 

 uniform portion was 26 cm. long and 2'5 cm. internal 

 diameter. The electrodes were two parallel copper disks, 

 the lower one fixed, while the upper could be raised or 

 lowered by a vertical copper rod screwed into the disk. 



When the electrodes were at their greatest distance apart 

 the resistance between them was 4*3 x 10 1 ohms, and the 

 resistance for intermediate positions was determined, so that 

 any desired value could be obtained by altering the position 

 of the movable electrode. 



One electrode was also connected to the gold leaf of a 

 Wilson tilted electroscope and the other to the case and brass 

 plate inside. The position of the gold leaf was observed 

 i through a microscope, with a scale in the eyepiece which 

 was so arranged that when a difference of potential of 

 130 volts existed between the case of the electroscope and 

 the gold leaf, the latter was focussed on the zero of the 

 eyepiece scale. 



The current through the Geissler tube was accordingly to 

 be calculated by dividing this difference of potential by the 

 resistance of the liquid between the copper electrodes. To 

 obtain a given current these electrodes were set at the proper 

 distance apart, and the primary current adjusted till the image 

 of the gold leaf coincided with the zero of the microscope-scale. 

 The current was kept constant by an observer who controlled 

 the sliding copper bar while looking into the microscope. 



The tube became hot during the passage of the dis- 

 charge, and in order to prevent the pressure of the gas from 

 changing, a bulb of 1 litre capacity was sealed on as described 

 by Guild *. This arrangement also diluted any gaseous im- 

 purities which might be given off by the electrodes or glass 

 walls of the tube. Nevertheless, it was found that the 

 intensities of the spectrum-lines increased with time when 

 a steady current was maintained, nor was it possible to attain 

 to a final steady condition!. Accordingly a clockwork 

 arrangement was included in the primary circuit, so that the 

 current was on for six seconds and off for fourteen. In this 

 way the tube was kept in a constant condition, and readings 

 could be repeated time after time in a satisfactory manner. 

 The advantage of using an electroscope to measure the 

 current now became strikingly evident. The motion of the 

 gold-leaf is dead-beat, and the position of equilibrium is 



* Phys. Soc. Proc. xxviii. pp. 60-71, Dec. 1915. 



t A reference to fig. 9 will give the explanation of this phenomenon. 

 The heated gas in the Geissler tube will have a density equivalent to 

 cold gas at a lower pressure. Hence if the gas is at a pressure greater 

 than 1*3 mm. the intensities of the lines will increase corresponding to a 

 .movement along the curve in the direction of decreasing pressure. 



