CHAPTER VI. 



THE VARIATIONS OF THE COLLOIDAL NUCLEATION OF DUST-FREE AIR 



IN THE LAPSE OF TIME. 



97. Introductory. Above (Chapter I, section 26, et seq.) and in my 

 address* before the Physical Society, I gave an account of observa- 

 tions made several times daily since May 9, 1905, in a search for the 

 possible occurrence of an ultra-mundane radiation. The work was 

 there summarized as follows: 



Using the most sensitive condensation method, i.e., that depending on the depres- 

 sion of the limiting asymptote of non-energized, dust-free air, no change of the quality 

 of scrupulously filtered atmospheric air has thus far been detected. . . . Naturally [ions] 

 would vanish during the slow passage of air through the filter, but fresh ions should be 

 reproduced within the fog chamber by the same agency which generates them without. . . . 

 Probably, therefore, the coronal method is as yet inadequately sensitive to cope with 

 variations of the small nucleations specified. 



The ions, which are relatively large nuclei, withdraw much of the 

 available moisture which would otherwise be precipitated on the col- 

 loidal nuclei of dust-free air. Hence the size of the terminal corona is 

 diminished. The advantage of the method is its independence of the 

 drop in pressure if this exceeds a certain value. 



Since the discovery announced by A. Wood and A. R. Campbellf on 

 the probability of cosmical radiation as evidenced by the existence of a 

 daily period of the same, showing maximum ionization between 8 and 

 10 a. m. and 10 p. m. and 1 a. m., minimum ionization at about 2 p. m. 

 and 4 a. m., I have taken the subject up again. It seems possible that I 

 overestimated the sensitiveness of the earlier method. I have therefore 

 changed it in the present experiment, replacing the large terminal 

 coronas by the small coronas very near the fog limit. 



98. Method and data. The observations, in other words, are now made 

 with a drop in pressure, but just sufficient to produce coronal condensa- 

 tion on the larger colloidal nuclei of dust-free air (dp = 21 cm.). The 

 sizes of coronas vary rapidly with the pressure difference and hence with 

 the barometer, p, etc., and great care must be taken with these details. 

 This, however, has been done and the results obtained are given in table 

 54 and in the chart (fig. 64). 



* Physical Review, xxn, p. 105, 1905; also p. 109, on Radiant fields, 

 f Nature, vol. 73, p. 583, 1906. 



15s 



