548 INFECTION AND RESISTANCE 



appear points of bright illumination, due to the presence of larger 

 particles. This is known as the Tyndall effect. The light which has 

 passed through such fields is found to have become polarized. 



It is evident that, in a solution whose particles are sufficiently 

 large to become individually visible as points of light under the ultra- 

 microscope, it immediately becomes possible to determine the size of 

 the particles on the assumption that these are all of the same size. 

 The procedure consists in determining the following quantities: (1) 

 the total number of particles in a given volume by the usual blood- 

 count method; (2) the weight of the dispersed substance in a given 

 volume by a chemical or other analysis; (3) the density of the dis- 

 persed substance, which is usually taken as equal to that in the undis- 

 persed state. This undoubtedly introduces an error in the computa- 

 tion, since, in all probability, the density increases in the dispersed 

 state, owing to increased compression by surface tension. This error 

 is probably small unless the degree of dispersion is very great. By 

 this method, particles in colloidal gold solutions have been observed 

 and counted whose diameters were as small as 10~ 6 mm. This rep- 

 resents about the limits of individual visibility under the ultramicro- 

 scope, that is, with particles much smaller than this the field appears 

 diffusely illuminated. This value is about one-hundredth that of the 

 wave-length of violet light, and about ten times that of the calculated 

 diameter of the ethyl alcohol molecule. 



The rate of settlement under the influence of gravity has also 

 been used to determine the size of colloid particles. By means of 

 Stokes' law for the fall-rate of bodies through a viscous medium, a 

 comparatively simple equation permits of the calculation of the diam- 

 eter of the falling body when the fall-rate, the viscosity of the me- 

 dium, and the densities, respectively, of the dispersed substance and 

 of the medium are known. Perriii r> used the same principle in the 

 preparation of suspensions in which the particles were all of more 

 or less the same size, using, however, regulated centrifugation in- 

 stead of simple settling under gravity. 



There has also been developed, largely by Bechhod another 

 method which throws some light 011 the relative sizes of particles, 

 and also offers a very interesting and valuable experimental weapon 

 for colloid investigation. This is the method of ultrafiltration. It 

 has been found possible to produce graded filters which allow of the 

 passage of particles below a certain size, and which restrain any 

 larger ones. These filters are made by impregnating ordinary filters 

 with gelatin and other colloidal solutions and drying with special 

 precautions. The permeability decreases with the concentration of 

 the gelatin or other substance used. 



5 C. #., 146, 967, 1908. 



6 Ztschr. Chem. Ind. Koll., 2, 3, 1907; Die Kolloide in Biologie u. Mcdizin, 

 Dresden, 1912. 



