A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 125 



The case may be stated by computation as follows : The velocity at which 

 force is annulled by the resistance is 



v = (io 4 d/i8) 2 = .3 id 2 X io 6 

 from which the following table has been computed. 



TABLE 12. 



If the results of this table are laid off graphically in a chart, the column 

 marked "constant io'X^" in tables 9 and io above may be obtained from it. 

 It will then be seen that in most of the instances of small coronas in table io, the 

 observed velocity is less than the limiting velocity. Hence these data are to be 

 rejected, and it follows that the subsidence method is scarcely applicable until 

 the middle g-b-p corona (s = 6.2-6.5) ^ as been reached. The mean value of a 

 computed from the admissible data of table io is in successive decades, .00284, 

 .00290, .00293, -00290, giving a mean value of a = . 00289. This agrees closely 

 with the datum (.0029) accepted in the above coronal tables, but is much below 

 the value in table 9. Hence the following experiments were made with large 

 coronas in the cubical apparatus. 



25. Further results. The data of the following table 13 were obtained by 

 observing the time of fall for three successive distances, of 2 cm. each (usually), 

 with three stop-watches. The results taken were those cases only where the 

 three intervals observed are nearly the same. The fog was usually precipitated 

 in the long vessel for observing atmospheric nucleation on air nuclei. If the fog 

 line became billowy before the last observation was reached, this was discarded. 

 The mean results are all given at a temperature of about 20 centigrade. 



The new results like the above fail to suggest a smooth curve; though it 

 would be difficult to obtain them under more generally trustworthy conditions 

 than in this large vessel, showing a fog line half a meter long. The slightest 

 variation of temperature, etc., gives rise to an undulatory fog line or produces a 

 washed upper surface to the fog-bank. After a lapse of time of about i min. 

 the demarcation is rarely available. In the very large coronas above g-b-p, the 

 color and the character of the coronas is fleeting. In fact, a total change is liable 

 to occur within a fall of one centimeter. Such results have no meaning. 



If these data be compared graphically with the results of tables i, 8, and 13, 

 Chapter VI, they will be found to agree with them as well as these results 



