THE STliUCTUKE OE THE NUCLEUS. 141 



frequently associated with electrical charges that it seemed expedient to examine 

 the occurrence of the I'eniurkable diffei'euces of diffusion fi-oni the electrical point 

 of view first. The advantages of the simple mechanical hypothesis already stated 

 in the last chapter/ may now be reaffirmed for reference. 



6. lliermodynamic agencies. — After this general statement of the phenomena 

 in question from the electrical [)oiiit of view, I may revert to certain advantages of 

 the mechanical theory for the i)i'oduction of nuclei in a medium of saturated vapor. 

 It is here supposed that the increment of vapor pressure at the surface of increasing 

 convexity is finally compensated for by the decreased vapor pressure due to the con- 

 centration of an evaporating droplet. It is inferred that a case of equilibi'ium will 

 eventually result, corresponding to a particidar radius and degree of concenti'ation 

 of the nucleus, which is thus merely a concentrated solution. Incidentally, the 

 nucleus may or may not be electrically charged. Hence this theory accounts at the 

 outset for nuclei, whether ionized or not. It accounts for nuclei produced by any 

 kind of radiation in the manner discussed in the preceding cha[)ter (§ 47). Cor- 

 relative electrical phenomena (§21 below) are not wanting. 



Nuclei produced spontaneously by certain liquids like concentrated sulphuric 

 acid, in presence of water or other vapor, by carbon disulphide in presence of its own 

 vapor, by dissolved naphthalene in presence of benzol vapor, etc., admit at once of 

 the same explanation, assuming that these bodies emit a slightly volatile constitu- 

 ent, soluble in the liquid from which the vapor arises. A similar explanation 

 holds for the nucleation from phosphorus, flames, glowing bodies, ignited metals, 

 etc. Finally, the nuclei produced by shaking have suggested the present hypothe- 

 sis. The terapei-ature and Ijygrometric conditions at which nucleation in case of a 

 phosphorus 'source is a maximum diffei's decidedly by Chapter I from the tem- 

 pei-ature, etc., at which ionization is a maximum. 



EXPERIMENTAL EQUIPMENT. 



7. Method. — For the reasons stated, the method of determining the rate of 

 diffusion of nuclei is simple and direct. Nuclei are introduced into the bottom of 

 a tall cylindrical receiver containing saturated vapor issuing from a flat pan 

 of liquid, and the height to which they have diffused after a measured length of 

 time is tested by adiabatically cooling the vapor. The height of the fog bank 

 seen immediately after exhaustion shows the limiting altitude which diffusion has 

 reached, and there is usually a sharp plane of demarcation I'ising regularly in the 

 lapse of time between the nucleated air below and the air free from nuclei above. 

 It is assumed that when diffusion of a sluggish relatively heavy medium takes 

 place against gravity in this way in a room of reasonably constant temperature, the 

 influence of conviction is not a serious factor. 



The position of the advanced front of diffusion at the end of a given period of 

 time is thus, as a rule, well determined. The position of the plane of demarcation 

 immediately after the nuclei are introduced {i. e., the fiducial zero) is not detei-min- 

 able with the same degree of cei'tainty. In the case of the slowly diffusing vapors, 



' Chapter V., §§ 44, 47-49- 



