Ultra-Violet Light on Moist Air. 309 



to their velocities being widely different. This was so with or 

 without the field. 



A most curious phenomenon not apparently very directly 

 concerned with this subject was observed while carrying out these 

 experiments. Occasionally the ordinary ultra-violet light clouds 

 were invaded by wisps and bursts of dense clouds which at 

 times showed iridescent colours. These were produced without 

 any ultra-violet light and were due to small sparks passing over 

 the glass inside the chamber. These clouds were charged and 

 behaved very similarly to the ultra-violet light clouds. They 

 need further investigation. 



When a cloud produced by ultra-violet light is allowed to 

 remain in the chamber for some time without the light or the 

 field being on, it loses its charge, and little motion can be ob- 

 served if tested after a few minutes. An uncharged cloud can 

 be readily obtained by keeping the field on after the light has been 

 cut off. The charged droplets are simply removed and the un- 

 charged ones remain behind. 



Carbonic acid gas and oxygen give clouds like those in air. 

 Contrary to the experience of C. T. R. Wilson I found that the 

 cloud is formed more readily in carbonic acid gas than in air. 

 Hydrogen does not give these clouds ; the gas must however 

 be fairly pure or they are formed readily. 



A remarkable circumstance about the formation of these 

 clouds is that they are generated even in a strong field. Thus 

 I find in my notes, " The way in which cloud comes even with field 

 on is very remarkable. Thus this morning with 840 volts on 

 9 cm. got cloud in one minute. The cloud thus formed is always 

 a discrete cloud. It moved towards the negative plate but seemed 

 finally to stop moving. On reversing the field it moved quickly 

 towards the new negative plate." This does not mean that the 

 cloud was composed entirely of positively charged drops. It 

 means simply that the arc light was concentrated on the negative 

 plate and that thus the positively charged particles were under 

 observation. By a discrete cloud I mean one whose particles are 

 big and far apart. 



If ultra-violet light be allowed to act on a cloud for a long 

 time (say 10 minutes) in the absence of any electric field, the 

 cloud sometimes shows colours. This growth ceases after a 

 certain time. Thinking that this was owing to the heavy drops 

 falling out of the clouds and that this effect limited the life of 

 any particular drop I tried to devise some way of keeping the 

 drops in the ultra-violet light. This I did not succeed in doing. 

 I arranged a glass tube 5 cm. in diameter and 25 cm. long 

 as a cloud chamber. The tube was placed vertically and was 

 closed at the top by a quartz lens and at the bottom by a 



