98 CONDENSATION OF VAPOR AS INDUCED BY NUCLEI AND IONS. 



centimeter, since but little water is precipitated. In this series a second 

 large exhaustion was made to catch the nuclei left by the first exhaus- 

 tion in each of the four cases. But few nuclei were found, however, 

 perhaps because considerable time (5 minutes) was needed between the 

 exhaustions; but the reason for this is not clear. One may notice in 

 conclusion that the numbers found for the nucleation depend essentially 

 upon computation, as the coronas are large. There is one correction, 

 m/m 27 , to allow for the small quantity of water precipitated; another 

 for the volume increase on exhaustion; a third for temperature, etc. 

 The coronas themselves naturally increase as the expansion is larger, 

 but they do not keep pace with the corrections. 



62. Continued. In the experiments of table 42 the filter cock was 

 again left slightly open; but the vacuum chamber was kept at the 

 same initial pressure p dp'. The low drops of pressure were secured by 

 successively reducing the pressure of the fog chamber, as shown under 

 P P a- This is a much more convenient method of experiment, though 

 the computation is more troublesome. The final common pressure after 

 exhaustion is p dp 3 . All other data have the same meaning as before 

 and corrections are added for the precipitation of water, m' /m; for the 

 volume expansion v l /v and for temperature. The table contains six 

 series of results for different exhaustions and differently opened filter 

 cock. Data are reproduced in fig. 31. 



Naturally the same evaporation difficulties are again obtained, but 

 the curves as a whole are more definite. In series I and II the number 

 of ions which survive in the water nuclei is again about a third in each 

 case ; but if the filter cock is opened wider, about half as many water 

 nuclei occur relatively to the original number of ions. If radium is left 

 in place (series III, VI) the ions are still efficient in presence of the 

 increased number of nuclei. 



The curves corresponding to the distribution of water nuclei in series I 

 again suggest the distribution curve of ions and of vapor nuclei in dust- 

 free air. In other words, all sizes of nuclei within a certain range of 

 dimension seem to be present. Series II has not been carried far enough, 

 for the experiment places a lower limit at which the method necessarily 

 breaks down. Series VI, however, is of a similar character to series I. 



The distinctive feature of these experiments is the occurrence of 

 reduced nucleation at very much higher drops of pressure than above. 

 One would naturally associate this with the fact that the water nuclei 

 are stored before the precipitation of fog upon them, in a partially 

 exhausted vessel. Yet the evidence is not clear on this point. The 

 smallest nucleation occurs at the lowest pressure attainable, viz, 59.8 to 

 61.9; but in series II higher values of n appear at 62 . o to 62 . 4 cm. A 

 larger drop of pressure is here applied adapted to catch the smaller nuclei. 



