THE PHYSICAL PROPERTIES OF INFECTIVE PARTICLES 241 



Tlie theory for the polarization fluorescence method permits the calculation 

 of the rotational diffusion coefficient from measurements of the polarization 

 (Perrin, 1926, 1936). Many macromolecules of biological interest are not 

 fluorescent for exciting light with wavelengths which are readily accessible, 

 and fluorescent derivatives are usually employed in which dyes are coupled 

 chemically to the macromolecules. This method has not as yet been applied 

 to viruses although it has been used extensively with proteins. 



Each of the methods described above has certain virtues and drawbacks. 

 No single one is likely to be useful for all tyj)es of materials. A choice among 

 them depends on the tjrpe of problem under investigation and restrictions 

 imposed by the nature of the macromolecule, such as its size, shape, and 

 solubiHty. Some of the methods are severely limited to solutions of low 

 conductivity, thus precluding experiments at high ionic strength and over a 

 range of pH values. 



4. Diffusion 



a. General Considerations. The spontaneous transfer of material from a 

 region in a solution in which there is a certain concentration of a given 

 species of solute molecules to a neighbouring region of lower concentration 

 is known as diffusion. This mass transfer of solute molecules continues until 

 the concentration everywhere in the solution is uniform. Individual mole- 

 cules continue to move under the influence of thermal energy after a uniform 

 state is achieved, of course, but this random movement is not directed and 

 no net transport of material results from this Brownian motion. As already 

 indicated in the discussion of rotational diffusion, measurement over a long 

 time period of the individual movements of a particle can provide a satis- 

 factory value for the diffusion coefficient if the material in solution is amen- 

 able to direct observation by microscopy. This is rarely the case, of course. 

 Even when such observation is feasible, experimentation is difficult because 

 the movements are very rapid and in all directions. Fortunately, the detailed 

 description of the path of a single particle is not required, since only the net 

 motion over a relatively long period, such as a second, is needed. The net 

 displacement is, of course, much less than the distance represented by the 

 complete path. Thus, a determination of the rate of diffusion by this method 

 consists of the evaluation of the position of a particle every second and the 

 many movements of the particle between observations within the one-second 

 intervals are immaterial. Owing to the lack of techniques for the direct 

 visualization of individual macromolecules in solution, practically all 

 diffusion measurements of interest to biologists are based on the fact that 

 the motions of the particles are directed preferentially when concentration 

 differences are estabHshed. Such a gradient of concentration can be formed 

 initially by suitable experimental techniques or it may be the indirect 



