418 MR. C. T. R. M ILSON ON THE CONDENSATION NUCLEI PRODUCED IN 
throughout both branches was obtained at once. On displacing the cadmium points 
so that the rays now passed along the other tube, and repeating the experiments, the 
tube which before was filled with fog now remained dark, and the other was filled 
with a white or coloured fog. 
These experiments prove that the nuclei produced by the action of ultra¬ 
violet light do not have their origin at the surface of the quartz. It might still be 
supjDosed that they are produced at the surface of the glass, where this is exposed to 
the ultra-violet rays. Lenard and Wolff, however, were able to detect no effect of 
this kind with glass. With the object of testing this 23oint, a T-shaped expansion 
a})paratus was now made (fig. 4). The length of tlie horizontal tube amounted to 
Fig. 4. 
ti SI 
27 centims., and the internal diameter was 1’3 centim. One end of this tube was 
closed by a quartz plate cemented on with shellac. 
The rays from a spark between cadmium terminals were sent axially along the tube, 
a quartz lens being inserted to make the rays converge to a point slightly beyond the 
far end of the tube. By observing the image of the cadmium points which was formed 
by the quartz lens, it was easy to test whether the light was passing axially, and also 
whether the points were sufficiently near together to give an image considerably smaller 
than the diameter of the tube. The length of the spark-gap was generally rather less 
than I millim. 
In the earlier experiments made with this apparatus, the fogs, which were produced 
on expansion under the influence of the ultra-violet light, although very dense near 
the quartz plate, diminished rapidly in denseness with increasing distance from the 
cpiartz, and did not reach the far end of the tube at all. This was at first interpreted 
as indicating either that the nuclei arose at the surface of the quartz, or that the 
active rays were al)soi“lied by a comparatively small tliickness of moist air. The latter 
view was easily disproved by interposing a layer of moist air (in an open tube 17 centims. 
long) between the source and the expansion apparatus. This exercised no appreciable 
absorption. The whole effect was finally traced to a deposit of fine dew on the Inside 
of the quartz j)late. On removing this by gentle warming, uniform fogs from end to 
end of the tube were obtained on expansion. There was never any indication of any 
increase in the density of the fog close to the far end of the tube, where the rays 
