'I'ino STiMKrruTiE or tiik nttcleus. loy 



nuclei sliaken out of the more conceutivited dilute solutions, but not so small in 

 general as the nuclei shaken out of pure water or any other pure solvent, in which 

 the amount of solute is an actually vanishing cpiantity. To carry out this compari- 

 son one should eliminate the i)eculiar features of water and use the same neutral 

 solvent in both cases. Hence, the results for benzol have been brought together in 

 table 13. Water and alcohol have been added because of the greater purity of 

 the solvent. The same infeiences follow. 



20. Conclusions. — A survey of the results obtained as a whole, attests the 

 occurrence of the diffusion beyond a doubt. The consistency with which 

 the fog bank in case of benzol, for instance, is found at a definite height aftei' a 

 definite time under given conditions, no matter how often the exhaustion may be 

 repeated, the obvious tendency of the plane of demarcation to rise I'egularly in the 

 lapse of time, in every case, etc., — all these observations can not be the result of 

 chance convections but point to some more definite underlying cause. 



But while the occui-rence of diffusion and its average rate is thus established 

 by dii'ect observation, the amount of motion in case of any given vapor has not 

 been clearly made out. Water vapoi-, from its peculiar behavior except in the case 

 of shaken nuclei, would naturally be left aside here for further research; but the 

 other vapors frequently give evidence of very diffei'ent rates of motion, u, of the 

 nuclei contained, depending on accidental conditions like the presence of moisture, 

 the existing pressui'e and tempei'ature, etc., not yet made out. Here again, how- 

 ever, the i-ate for a given charge of vapor is a definite quantity and the ol)servations 

 are consistent among themselves. Witness, for instance, the remarkable results for 

 ethyl and for methyl alcohols, figures 7 and 8, in each of which two definite veloc- 

 ities are given in two independent series of measurements with the same body. 

 Something definite is moving in each case, but that something diffei's for the same 

 vapor and the same emanation under different accessory conditions. 



If the diagram, Chapter V, figure 20, be resumed, in which vapor pressure is 

 represented in terms of the I'adius of the nucleus, it may again be argued that the 

 curves hms and hm's' correspond, respectively, to stronger and to weaker solutions, 

 while the curve he shows the vapor pressure of the solvent. With the same mass 

 of solute entrapped in the nucleus, however, the curves will depend on the soln- 

 tional affinity, as evidenced by the vapor depression of strong solutions. Thus if 

 in the weaker solution bm's\ the solute were I'eplaced by the same mass of a body 

 having gi'eater cohesional affinity for the solvent, the conditions expressed bj' cui've 

 bms might be I'egained. As the curves are di'awn this corresponds to a stable 

 nucleus of larger size. In other words the nucleus is largest when the affinity is 

 greatest. If we compare the I'esults found in Cha])ter V, for nuclei shaken out of 

 inorganic saline and of neutral organic solutions, respectively, this I'esult seems to 

 hold good, the latter being much the more fleeting. As it does not follow that the 

 conditions of the figure which is purely qualitative and inferential, hold good, all 

 that need be stated is that in addition to the mass of solute (emanation) entrapped, 

 in each nucleus, the cohesional affinity of the solute and solvent further determines 

 the size of the nucleus. 



