THE PHYSICAL PROPERTIES OF INFECTIVE PARTICLES 317 



favorably with the value, 9.0 X lO** (Oster, 1946), inferred from light- 

 scattering studies. It should be noted that this light-scattering deter- 

 mination was one of the earliest applications of this technique to the 

 study of macromolecules. The measurements were made before photo- 

 multiplier tubes of high sensitivity were used in light-scattering instruments. 

 Instead of detecting the laterally scattered light, as is customary now, Oster 

 measured the decrease in light transmission (or the turbidity). The mole- 

 cular weight of BSV was also determined by direct particle comiting in the 

 electron microscope yielding the value, 9.4 X 10'' (Williams and Backus, 

 1949). Although higher values have been calculated from early X-ray 

 diffraction data (Bernal et al., 1938; Carlisle and Dornberger, 1948; see 

 Markham et al., 1942, for a discussion of these data) it seems likely that the 

 molecular weight is close to 9 X 10^. 



Both electron microscopy (WiUiams, 1953b) and low-angleX-ray scattering 

 (Leonard et al., 1953) indicate that the virus particles are essentially 

 spherical. This was inferred also from the early crystallographic studies 

 which showed that the unit cell was body-centered cubic in structure. 

 Early electron micrographs yielded 2 60 A as the diameter of the particles 

 (Price et al. 1946) but more recent examination of air-dried specimens gave 

 a diameter of 300A for particles in crystal-like arrays (Williams, 1953b). 

 Individual particles, when the sample is air-dried, became somewhat flat- 

 tened and their volume is estimated to be about 20 % less than that of 

 frozen-dried particles. The latter particles are certainly not spherical but 

 bear, instead, the appearance of polyhedra (Fig. 11). Assuming them to be 

 spherical, the mean diameter of individual particles is 300A (Williams, 1953b). 

 It can be inferred from the careful examination and comparison of the isolated 

 particles in the frozen-dried specimen with both the individual particles and 

 those arrayed in microcrystals in the air-dried jDreparations that the gross 

 architecture of the particles in solution is preserved in the frozen-dried 

 sample. Comparison with the value from low-angle X-ray scattering is then 

 justified, since this technique measures the size of the particle in solution. The 

 value, 309A, obtained in this way (Leonard et al., 1953) is in excellent agree- 

 ment with the results of electron microscopy. From careful study of the 

 maxima and minima in the curve of the intensity of scattered X-rays as a 

 fmiction of angle, Leonard et al. (1953) concluded that the virus camiot be 

 described by a sphere of miiform electron density. 



The various types of measurements taken together indicate clearly that 

 the virus particles contain appreciable amoimts of water. This was suggested 

 by some of the early crystallographic studies which revealed shrinkage of 

 the crystals upon drying. Also the hydrodynamic behavior is incompatible 

 with a model of the virus as a compact, uniform sphere of density equal to 

 the reciprocal of the partial specific volume and molecular weight of 9 X 10^. 



