JUNB 6, 1902.] 



SCIENCE. 



913 



H,SO„ NaCl, CaCl„ FeCl,, FeSNO., A13N0,, 

 Ca2]S[03, (H,N)]Sr03, alum, Na,SO„ etc., and 

 neutral organic bodies like sucrose, glucose, 

 glycerine, urea, etc., I reached a number of 

 new results, to one of wliicli, in particular, I 

 will venture to call attention here. 



1. The number of nuclei produced under 

 identical conditions of agitation varies with 

 the violence of the agitation and the bulk of 

 solution used, and from a theoretical point of 

 view, particularly with the concentration of the 

 solution and its chemical nature. Thus under 

 given identical conditions of shaking one may 

 get from water about 30 nuclei per cubic cen- 

 timeter; from 1 per cent. CaCl^ solution, 240 

 nuclei; from 1 per cent. Na^SO^ solution, 450 

 nuclei, etc.; from 2 per cent, sugar solution, 

 157 nuclei; from 2.6 per cent, glycerine solu- 

 tion, 95 nuclei, etc. So far as concentration 

 alone goes, one may write for dilute solutions, 

 n=n„+^/(log (B/C)), where n is the num- 

 ber of nuclei produced per cu. cm. under other- 

 wise like conditions, n,„ the number in case of 

 infinite dilution (water), the concentration 

 and A and B constants. If absolutely pure 

 water were available, it is probable that Ji„ 

 would vanish. One may note that the degrees 

 of extreme dilution effective recall the sensi- 

 tiveness of electrolytic experiments. For or- 

 ganic neutral solutes, the number of nuclei 

 is not only smaller as a rule, but they are 

 characteristically fleeting. 



2. The point with which I am concerned, 

 however, is the rate at which the nuclei van- 

 ish in the lapse of time. From the marked 

 diffusion of these nuclei, their dimensions 

 must be comparable with molecular dimen- 

 sions. Subsidence is out of the question. If, 

 as I interpret it, the loss of nuclei in the lapse 

 of time is due to absorption at the solid walls 

 of the spherical receiver, one may write for 

 the absorption velocity, Ic (meaning that Icn 

 nuclei are absorbed per square cm. per min- 

 ute), A;== — (R/3n)(dn/dt). Computing k in 

 this way (essentially dQog n)/dt), one finds 

 from all the solutions, saline or neutral, an 

 important general result : For the case of solu- 

 tions of a few per cent. (1 to 3), fc is of the 

 order of .02 to .04 cm./min., though varying 

 from solute to solute; for the .01 per cent. 



solutions k is of the mean order of 08 cm./ 

 min. ; for the .0001 per cent, solutions, k is of 

 mean order of about 8 cm./min. ; for ordinary 

 distilled water, in glass vessels, k may reach 

 5 cm./min., etc. I have the specific data in 

 hand, but do not wish to weary the reader. 



It follows therefore in general, that not only 

 does the number of nuclei produced by shaking 

 (ccBt. par.) increase with the concentration of 

 the dilute solution, but the apparent rate of 

 decay of nuclei diminishes, i. e., their absorp- 

 tion velocity decreases with the strength of the 

 solution. For ordinary distilled water, these 

 velocities, if referred to three dimensions, are 

 already beginning to approach the ionic veloc- 

 ities. Again as the number of nuclei, n, is 

 greater, they vanish more slowly, so that an 

 apparent decay increasing with the density of 

 the nucleation is out of the question. The 

 whole, therefore, constitutes an entirely new 

 and striking corroboration of the isolated point 

 of view taken throughout my work.* 



3. The inference is therefore tenable that 

 the nuclei shaken out of stronger solutions are 

 larger. Since the nuclei are produced by 

 evaporation from a large diameter, it follows 

 that the dimensions at which evaporation 

 ceases at the surface of the particle are larger 

 for the stronger than for the weaker dilute 

 solutions. Naturally a given degree of con- 

 centration is reached in a larger globule in the 

 former case than in the latter. The theory for 

 the production of the nuclei here in question 

 is thus at hand. A particle of absolutely pure 

 water produced by shaking will either vanish 

 by complete evaporation, or it will grow and 

 eventually vanish by subsidence. If, however, 

 the evaporating globule is a solution, the incre- 

 ment of vapor pressure at the surface of in- 

 creasing convexity will gradually be compen- 

 sated by the decrement of vapor pressure due 

 to the increasing concentration of the solu- 

 tion. Hence there must be a critical diameter 

 at which the increased vapor pressure due to 

 surface tension just counterbalances the de- 

 creased vapor pressure due to concentration. 

 This is the stable diameter of the nucleus. A 

 smaller particle will grow because the concen- 



* Compare ' Experiments with Ionized Air,' p. 

 92, Smithsonian Contributions, Washington, 1891. 



