produced in Aii^ by Ultra-Violet Light. 485 



than that transmitted by quartz. Thin quartz was found to 

 transmit down to a wave length A, 1450 while some specimens 

 of clear colourless fluorite transmitted down to A, 1230. 



It therefore seemed more promising to use fluorite instead of 

 quartz. Several crystals of clear colourless fluorite were obtained 

 and plates about 2 cms. thick were cut from them. Most of them 

 in the process of cutting and grinding broke up into bits too 

 small to be of any use. One plate, however, perfectly free from 

 flaws was obtained with a clear space of 2 cms. in diameter. This 

 was the piece used in the following experiments. 



The apparatus used is shown in fig. 1. The light is obtained 

 from the discharge tube L. The two electrodes were separated 

 by a glass tube with a nan'ow opening facing the fluorite plate. 

 In this way, great concentration of luminosity was obtained at 

 the end of the inner tube. The electrodes were connected to the 

 secondary of a small induction coil used as a transformer. The 

 primary was connected to the electric light mains (alternating) 

 using as a resistance a few lamps in parallel. With hydrogen 

 in the tube at a pressure of 2 or 3 mm., a steady source of light 

 was easily obtained. 



The air is drawn in through a wide plug of cotton-wool and 

 passes up a narrow tube until it reaches the fluorite window. 

 The end of the tube is within 1 mm. of the fluorite window, which 

 insures that all the air which passes through the apparatus comes 

 under the action of the easily absorbed short wave length ultra- 

 violet light. The air then passes into the ionisation cylinder 

 through a short glass tube suitably bent to prevent reflection of 

 the ultra-violet light on the electrodes. For the mobility experi- 

 ments two electrodes iV and F along the axis were used. 



It was found that when the light was produced in L, the 

 stream of air flowing through the ionisation cylinder was highly 

 conducting and contained both positive and negative ions, the 

 latter predominating. Control experiments were made (1) with 

 the air in motion but without the light, and (2) with the light 

 but no motion of the air. Under these conditions there was no 

 leak in the ionisation cylinder. 



To make sure that the positive carriers were not positively 

 charged dust particles, the air after leaving the ionisation cylinder 

 was drawn through a Wilson expansion apparatus where samples 

 could be tested. With expansions of less than 125 no nuclei 

 were obtained, while above that expansion there was no increase 

 over the ordinary effect obtained with dust-free air. This showed 

 that the cotton-wool plug was efficient in removing dust particles. 



There seems to be no way of explaining the presence of these 

 positive ions other than by an actual ionisation of the air by the 

 ultra-violet light. 



32—2 



