The Coagulation of Smokes and the Theory of Smoluchowski
Proceedings of the Royal Society A
In a previous communication* it has been shown that smokes are unstable disperse systems which spontaneously coagulate from the moment of formation. In this respect they differ from hydrosols, which are stable normally and coagulate only on the addition of an electrolyte. For these latter systems there is evidence to show that the so-called " rapid coagulation " begins when the complete discharge of the double layer has been effected, and that the particles are brought together mainly by
... er mainly by Brownian motion. With smokes, too, chance collision brought about by molecular bombardment is probably the chief factor in aggregation, so that a close analogy between the rapid coagulation of hydrosols and the spontaneous coagulation of smokes would appear to exist. Further, if it is admitted that every chance collision between the particles in the two types of system results in a union, then the equations developed by Smoluchowskif for the rate of coagulation of colloids might be expected to be equally valid for smokes after making allowance for the difference in pro perties of the two dispersion media. Reliable experimental data on sol coagula tion are not easily obtained, for the process is completed in a relatively short period, but the measurements of Zsigmondy, J Westgren, § and others|| show a satisfactory agreement with theory. In a smoke, conditions are more favour able, for coagulation can be followed at much greater dilutions, so that the whole process is slowed down, and observations may be extended over several hours. On account of the interest presented by the above analogy, and the possi bility of testing it quantitatively, we have measured the velocity of coagulation of various smokes by a method which we believe to be free from any serious experimental errors. Experimental. In the previous work the particles were counted with the ultramiscroscope, using a modification of the slit instrument of Zsigmondy. Most of the number * W hytlaw -G ray and Speakm an, ' Roy.