MR. C. T. R. WILSON ON CONDENSATION OF WATER VAPOUR 
2;'»2 
After use, the chlorine was absorbed by driving KOH into the expansion chamber. 
(This was done by expelling the water from B, and drawing in potash in its place, 
and then opening Tj, while was raised sufficiently high to overcome the pressure 
in A.) The bubble which remained was then drawn up into the capillary tube for 
measurement. 
Out of the three samples of chlorine used, two were analysed in this way. In both 
cases the volume of gas unabsorbed amounted to between 1 and 2 parts in 1000 of 
the whole. 
Results .—Chlorine shows both kinds of condensation, each requiring a definite 
minimum exjaansion. With expansions greater than was necessary to produce cloud¬ 
like condensation, colour phenomena w^ere observed like those exhibited by air and 
oxygen. 
The solubility of the chlorine introduced difficulties of the kind experienced with 
COg, interfering considerably with the exactness of the measurements. The final 
pressure had to be measured after each expansion, no water being run in till after 
this measurement was completed. 
The drops when the condensation was of the rain-like form appeared to the eye to 
be much more numerous in chlorine than in the other gases. 
Expansion required to produce Rain-like Condensation in Chlorine. 
1 
i Date. 
B. 
f C. 
77. 
Ih- 
August 20 
761 
21-5 
19 
443 
„ 20 . . 
761 
21-5 
19 
420 
August 24 . 
760 
20 
17 
560 
„ 24 . . 
760 
20 
17 
549 
R- 
Vi- 
Po. 
1 
Result. 
1184 
166 
907 
1-306 
Rain. 
1161 
153 
894 
1-299 
0 
1302 
257 
9P9 
1-304 
Rain. 
1291 
252 
994 
1-299 
i 
0 
Rain-like condensation begins when P 1 /P 3 is about 1'30. 
Expansion required to produce Cloud-like Condensation in Chlorine. 
Date. 
B. 
t°C. 
j 
TT. , 
Pi- 
Pi- 
P-:- 
P1/P2- 
Augu.st 28. 
768 
18 
15 ' 741 
1490 
280 
1029 
1-448 
„ 28. 
768 
18 
15 660 
1 
1409 
223 
972 
1-449 
Cloud-like condensation begins when Pj/Pg is about 1’45. The second observation 
was made with tlie condensing lens removed, the readings given being those corre- 
