VIRUS PARTICLES AND THEIR FUNCTIONAL ACTIVITY 343 



particle. This value divided into the dry weight of an aliquot of purified virus 

 suspension yields a figure for the number of particles per miit volume. Such 

 an approach requires (1) objective criteria for judging virus purity, (2) the 

 assumption that the particle is spherical in shape, and (3) reasonably 

 accurate methods of estimating virus particle diameter and density values 

 which are used to calculate particle mass. 



Purity as used in this context is analogous to physical homogeneity, i.e., 

 all particles are similar in size, shape, density, and surface charge, as judged 

 by their behavior upon ultracentrifugal and electrophoretic analysis. That 

 these criteria may be inadequate has been pointed out by Smadel et al. (1940), 

 who demonstrated that a mixture of washed vaccinia elementary bodies and 

 graded collodion particles coated with vaccinia antigen will migrate as a 

 single component in a centrifugal or electrical field. In spite of these short- 

 comings and in lieu of an independent method of testing for biological purity 

 (i.e., each j)hysical particle equivalent to one infectious unit) physical 

 uniformity has generally served as the criterion of purity. 



The diameters of spherical or nearly spherical virus particles are usually 

 estimated by ultrafiltration, electron microscopy, or ultracentrifugal sedi- 

 mentation analyses. Kesults obtained by ultrafiltration yield a rough 

 approximation of particle diameter at best. The surface tension forces of 

 evaporating solvent flatten the virus particles in specimens prepared for 

 electron microscopy. This effect can be minimized if the particles are spheres 

 of uniform size and align themselves in two-dimensional crystalline arrays on 

 the collodion membrane. Thus, an average diameter can be computed from a 

 linear array of about six or more particles. Preparation of specimens by 

 freeze-drying (Williams, 1953) completely eliminates flattening of individual 

 particles, since the ice is sublimed away in this technique, thereby avoiding 

 the distortive effects of surface tension forces. The critical point method of 

 Anderson (1951) also avoids structural artifacts of drying although speci- 

 mens prepared in this way usually cannot be shadowed. 



Sedimentation constants can be determined with a high degree of precision 

 but an estimation of particle diameter from such data requires equally precise 

 measurement of particle density which is more difficult to achieve. The 

 partial specific volume (i.e., the reciprocal of the dry weight density) of 

 purified virus particles can be determined by the pycnometric method. 

 Because the amount of available purified virus is usually quite small, the 

 pycnometer is being displaced by the ultracentrifuge as the tool of choice for 

 virus density measurements (Sharp et al., 1945, 1950; Schwerdt, 1957). 

 Sedimentation constants of aliquots of the purified virus are estimated in 

 media of successively increasing density, corrected for viscosity, and then 

 plotted as a function of the density of the suspending medium. An extrapola- 

 tion of the best-fitting curve to the density value corresponding to zero 



