THE PHYSICAL PROPERTIES OF INFECTIVE PARTICLES 319 



current concepts of small viruses as organized structures composed of a 

 nucleic acid core surrounded by a shell of specifically aggregated protein 

 subunits. 



Like other plant viruses, TYMV has been purified by a variety of methods. 

 First of these was the fractionation procedure employing ethanol and am- 

 monium sulfate (Markham and K, M. Smith, 1949). Later, centrifugal 

 methods were substituted for salting-out techniques, and finally treatment 

 with butanol and chloroform replaced the initial step that employed ethanol 

 (Cosentino et al., 1956). 



The first preparation of TYMV that was examined in the electron micro- 

 scope appeared to contain particles of only a single type with a diameter of 

 220A (Markham and K. M. Smith, 1946). Further, electrophoretic examina- 

 tion revealed only a single boundary, and analysis of the diffusion behavior 

 did not show evidence of inhomogeneity. These observations, plus the success 

 in crystallizing the purified material, served to heighten the surprise evoked 

 when two boundaries were observed in the ultracentrifuge (Markham, 1951). 

 The slower of the two boundaries represented about 20 % of the material 

 and it was tempting to view this as a dissociation product of the main com- 

 ponent. However, centrifugal fractionation of the mixture, effected by the 

 separation cell (Tiselius et al., 1937), quickly eliminated this hypothesis, 

 since the slowly sedimenting component was found to be free of nucleic acid. 

 Moreover, this material was not infectious. It did, however, have a sero- 

 logical specificity identical to the faster migrating component which had 

 been shown to be the viral agent. These results of Markham and K. M. Smith 

 (1949) have since been confirmed, and purification of the so-called "top" 

 and "bottom" components has by now been accomplished in different ways. 

 The unusual physical-chemical data seem consistent with only one inter- 

 pretation, that given by Markham (1951) and discussed in detail below. 

 Subsequent work in other directions has substantiated in all important 

 details his hypothesis. 



Diffusion studies of the individual components gave almost identical 

 values, about 1.5 X 10^'' cm.^/sec, showing clearly that the frictional co- 

 efficients for the two materials were essentially equal (Markham, 1951). 

 Since electron micrographs of isolated particles and microcrystals of the 

 mixture of both components showed that the particles were essentially 

 spherical (Cosslett and Markham, 1948), the diffusion data were used to 

 calculate the radius of the hydrated particles, giving the value 280A. To 

 explain the differing sedimentation velocities, 49 and 106 S for the "top" 

 and "bottom" components, Markham (1951) examined the partial specific 

 volumes. For the "top" component he found 0.74 cc./gm., a value expected 

 for proteins; the "bottom" component had a partial specific volume equal 

 to 0.666 cc./gm. On the basis of the additivity of the volumes of protein and 



VOL. 1—22 



