McCLELiiAND AND M'Henry — Uncharged Nuclei. 291 



surface. Thus in one experiment the uranium current was increased by the 

 spark five-fold in ordinary air, and seven-fold in air passed over a water 

 surface. The air was filtered in both cases as usual. 



Nahire and Origin of the Nuclei. 



It has been pointed out that the slower the moist air passed the spark the 

 greater were the nuclei produced. If moist air at rest be exposed to strong 

 ultra-violet light for about thirty seconds, the nuclei grow so large that they 

 become visible, and the path of a strong beam of light through the vessel 

 reveals a dense blue cloud, the individual particles of which can be seen in a 

 microscope with a dark-field arrangement. Those particles are evidently 

 drops of H2O, and thus the nuclei which we dealt with in the preceding 

 experiments were in all probability minute drops of water formed by the 

 light, and growing in size with the intensity of the light and the time of 

 exposure. 



C. T. E. Wilson first attributed the formation of the nuclei to the 

 production of some substance such as HjOa, which when dissolved in water 

 lowered the equilibrium vapour-pressure, thus causing drops to be stable and 

 grow which would otherwise evaporate. Other experimenters have proposed 

 oxides of nitrogen as the parent substance of the nuclei. 



In our experiments the difference between the formation of nuclei in dry 

 and moist air has been shown to be very remarkable. The number and size 

 of the nuclei were found to depend upon the quantity of water-vapour present 

 in the air. Experimenters who used Wilson's expansion apparatus to detect 

 the nuclei have stated that carefully dried and moist air behave similarly as 

 to the production of nuclei (St. Sachs. Ann. der Physik, 34, 1911, and 

 Saltmarch, Proc. Phys. Soc, London, 27, 1915). In view of our experi- 

 ments quoted above, this must mean either that the expansion apparatus is 

 30 very sensitive a method of detecting nuclei that those formed on the slight 

 traces of moisture left in the air after careful drying cause the formation of 

 an appreciable cloud in the cloud-chamber, but yet would not be numerous 

 enough to affect appreciably the value of the currents in our experiments, or 

 else that the hygroscopic parent substance (e.g., oxides of nitrogen) is formed 

 in dry air, and forms the large nucleus when admitted into the saturated 

 atmosphere of the cloud-chamber. On the latter supposition, if dried air is 

 exposed to the ultra-violet light in our experiments and then passed over a 

 water surface before entering the ionization chamber, large nuclei should be 

 formed and detected by the increase in the current C2 in the usual way. This 

 experiment was performed, but no nuclei were detected. 



A further experiment was performed in which dried air was passed over a 



