485 
1910-11.] On some Nuclei of Cloudy Condensation. 
thing taking place in the sunning flask. The filter F was therefore added. 
As this filter stopped all dust, the air when it arrived at the test flask gave 
no condensation, so that any condensation which might now occur would 
be entirely due to what was taking place in the sunning flask. 
Salt and iodine, separately and together, were the first substances sub- 
mitted to the test. The apparatus was first fitted together, the flask E being 
omitted and the filter connected direct to S. The pump was now worked for 
a time to fill the apparatus with dustless air. The stopcock K was then 
closed and a stroke of the pump made to test the air in T. If any condensa- 
tion took place, the pump was worked till it ceased. If all condensation 
ceased, then the joints would be all tight and the filter right. The flask S was 
then sunned to see that there was nothing due to anything taking place in 
the flask. If the air kept free from all condensation, then a little of the sub- 
stance to be experimented with was put in S. The pump was again worked 
till all condensation ceased. The sunning flask with the substance in it was 
next sunned for one minute or more. A stroke of the pump was now made 
to bring some of the air from the sunning flask into the test flask, when 
the stopcock K was closed and the air in T expanded by means of the 
pump. While this is being done the air can be examined through a lens if 
necessary and the density of the condensation, if any, noted. The experi- 
ments with salt and iodine seemed at first quite satisfactory. It looked as if 
the explanation of these sunshine nuclei had been obtained when salt or 
iodine, or both, were in the sunning flask and it was exposed to sunshine ; the 
air when passed into the test flask gave a dense, foggy condensation, while, 
if the sunning flask were kept in the dark, no condensation took place in 
the test flask. This result seemed highly satisfactory. It was soon found, 
however, that the formation of this fogging was far from a certainty. 
Sometimes it would appear a number of times and then cease, and sometimes 
it did not appear at all. One might imagine that the salt had given off all 
of the something which was the fog-producer, and soon ceased, and certainly 
on adding new salt the action started again ; but this explanation will not 
hold for the iodine, which would continue to evaporate till it was all gone. 
This occasional activity of the sodium chloride may be explained by the 
fact that this salt is decomposed into soda and hydrochloric acid by heat 
in the presence of water vapour. This supposition is supported by tests, 
as it was found that on heating the salt slightly it recovered its activity. 
The great difficulty in investigations of this kind is the extremely 
minute quantities of matter which produce surprising results and make 
the work full of pitfalls for the hasty. As an example of this may be 
mentioned a part of this investigation. I wanted to introduce some 
