Jink. 1912. 



KNOWLEDGE. 



intimately upon the problcni of solution in general. 

 Many investigators have been inclined to the idea 

 that they represent perfect solutions, whilst others 

 identify them as suspensions. The intervention of 

 the ultramicroscope has contributed largely to the 

 development of the question, if not to its final 

 solution. 



With one or two exceptions, all colloidal solutions 

 on ultrainicroscopic examination are found to con- 

 tain distinct particles, which reveal themselves, as 

 alread\- stated, in diffraction-patches on a gre\- 

 background. These particles can be counted and 

 their size calculated. There is thus strong evidence 

 for regarding a colloidal solution as merely a 

 limiting case of a suspension. A crystalloidal solution, 

 i.e., the solution of a substance like sodium chloride, 

 gives an absolutely clear field, and one, therefore, 

 infers the absence of even the tiniest particles in a 

 solution of this t\pe. Such a conclusion, however, 

 is questioned by the observations of van Calcar and 

 de Hruyn. who found, for example, that rapid 

 centrifugalisation of a sodium sulphate solution 

 induced partial separation of the salt. From the 

 abo\-e reasoning, and further observation but confirms 

 this, it seems impossible to draw any sharp line of 

 demarcation between suspensions, colloidal and 

 crystalloidal solutions, the one class merging im- 

 perceptibly into the other. Ultramicroscopic study 

 shows over how wide a range the size of the particles 

 in a colloidal solution varies and renders any attempt 

 at classification of solutions still more unsatisfactory. 

 Indeed, it seems probable that in a solution the 

 size of the '" dissolved " particle can var\- gradually 

 from that of the ordinary chemical molecule to the 

 dimensions of a visible suspensoid particle, the 

 properties of the resulting solutions varying with 

 the molecular forces called into pla\-. 



But more interesting developments even than 

 these can be attributed to the ultramicroscope, for 

 results recently obtained by its aid go far towards 

 strengthening the probabilit\- of the Atomic and 

 .Molecular Theories. Over a century ago, the 

 naturalist Brown observed that the particles of a 

 fine suspension, when observed under the microscope, 

 are in a continual slate of agitation, but the observa- 

 tion was shelved and almost forgotten. It is only in 

 recent years, and mainl\- by the application of the 

 ultramicroscope, that the further investigation of this 

 interesting phenomenon has been rendered possible. 



The ultramicroscope has demonstrated that the 

 smaller the particle the greater the activity, until in 

 colloidal solutions we get a movement so violent as 

 to resemble, in the words of Zsigmondy, " a swarm 

 of dancing gnats." ■ There seems to be some 

 connection between this rapid movement of small 

 bodies and the slower movement of the heavier 

 particles observed by Brown. Indeed, it almost 

 suggests that, if further diminution of the particles 

 be imagined, the attainment of molecular dimensions 

 might give a value for the molecular velocity of an 

 order comparable with_ that calculated on the 

 theoretical assumptions of the Kinetic Theor\- ; in 



other words, that tlie kirutic energies of a molecule 

 and of a colloid or suspensoid particle are equal. 

 This view has developed almost into a certainty, 

 particularly by Perrin's remarkable investigations, 

 and an experimental verification of the Kinetic 

 Theory is thus forthcoming. 



A brief outline of Perrin's work will serve not onl\- 

 as a development of the above brilliant idea, but also 

 as a fitting demonstration of the methods employed 

 in ultramicroscopic research. 



In the first place, his investigations required a 

 colloidal solution, the particles of w hich were of the 

 same size, and this was rendered possible b\- the 

 method of " fractional " centrifugalisation. The 

 uniform solution, after dilution and standing for 

 some time, was then examined ultramicrosco[)ically, 

 the microscope being focussed on an extremelj- 

 shallow beam of light capable of vertical movement. 

 At positions corresponding to different heights of the 

 solution, an estimate of the number of particles was 

 accurately made. This, of itself, is an extremely 

 laborious undertaking. The particles illumined by 

 the beam are visible in the field of the microscope, 

 but are executing their Brownian dance, and it is 

 impossible to make even an approximate computation 

 of their number. The difficulty is overcome by the 

 use of one of two methods. Either the field of view- 

 is instantaneously photographed several times and 

 the mean number taken ; or a stop is inserted in the 

 microscope so as to limit the field to contain only a 

 few particles, and then, by means of a shutter, 

 instantaneous "peeps" are obtained at small 

 intervals, the number of observed particles being noted 

 at a glance, and subsecjuently the mean of several 

 thousand of these readings taken. Both methods 

 give very concordant results. Having ascertained 

 the number of particles corresponding to different 

 heights in the solution, examination showed that 

 these were, within the limits of experimental 

 accuracy, exactly in geometrical progression. 

 Thus, at heights 100 75 .tO 25 microns, 



the numbers 200 170 1 4() 116 were obtained, 

 whilst 201 169 142 119 arc in e.xact 

 geometrical progression. Such results demonstrated 

 beyond doubt that the particles reach a state of 

 equilibrium where their distribution corresponds to 

 that of the molecules of a fluid. Thus, in the 

 atmosphere, the concentration of the constituents 

 gradually diminishes as we ascend, the diminution 

 obeying the above exponential law. The only differ- 

 ence between the two cases is that whilst the 

 atmosphere requires a rise of six kilometres for the 

 halving of the concentration, a colloidal solution 

 requires about one-tenth of a millimetre. 



Having proved that such an experimental law was 

 applicable to colloidal solutions, a mathematical 

 expression was easily deduced by which, given the 

 density of a particle, its mass and the number per 

 unit volume, the Kinetic Theory constant — which 

 denotes the number of molecules per gram-molecule 

 of any compound — is calculable. For the present 

 purpose it is not necessary to go into the mathematics 



