452 C. Barus — Colloidal Silver. 



sarily remained as a result of differences in the size of particles 

 in the two cases, the silver being much more finely com- 

 minuted. For this reason we can not admit that Mr. Lea's 

 recent experiment is decisive. A similar test had already been 

 made by Prange* and it was repeated by us with the remark 

 that the Tyndall experiment is no valid criterion unless precise 

 statements are made as to the size of particles which just 

 appreciably interfere with optical clearness. In other words 

 the dimensions below which Tyndall's experiment practically 

 fails are here vitally in question, and data for these have not 

 been forthcoming. 



Suppose a solid is dropped into an excess of its solvent. In 

 order that the system may become a solution, the disaggrega- 

 tion must at least reach the molecule. In electrolytes it may 

 even go further as is evidenced by Arrhenius's celebrated fac- 

 tor 2. But, under other circumstances, may not the separation 

 stop short before the molecule is reached ; or conversely, when 

 a precipitate is being formed out of individual molecules, may 

 not the process of growth be arrested in virtue of an equi- 

 librium of forces when the particles formed consist of 2, 10, 

 100 or even 1000 molecules ? To answer affirmatively is to 

 find a home for the family of colloidals, and they will more 

 nearly resemble solutions in proportion as the particles are 

 smaller. Certainly the beam of light is no longer an available 

 criterion, for the whole phenomenon is mapped out on a scale 

 which is small even in comparison with the wave length of 

 light. 



2. About a year and a half ago I incidentally made an 

 experiment in connection with certain meteorological ques- 

 tions, which has a precise bearing on the point here at issue. 

 In the endeavor to pass compressed air through a wet porous 

 porcelain septum into water, I was struck by the magnitude of 

 the pressures necessary. Supposing I waited long enough to 

 insure the transpiration of liquid, no flow of gas through the 

 septum occurred for pressure excesses of even above 100 lbs., 

 excepting at isolated points which were obviously the seat of 

 fissures. Now let T be the surface tension of water in dynes 

 per linear centimeter, a the angle of capillarity, r the mean 

 radius of the pores of the septum, and x the pressure of the 

 gas in atmospheres. Then (very nearly) 10 8 aOTT 2 = 2nrTcosa, or 



r = 2 7l-oso'/10 6 a: 



If therefore in the above experiment, x = 8 atm., T= 71 

 (Everett's tables, p. 50), cosa = 1 (say, for the superior limit is 

 in question), 



r— 18X10-" cm, 



* Prange : Rec. des Trav. chim. des Pays. Bas, ix, p. 125. 



