THE PHYSICAL PROPERTIES OF INFECTIVE PARTICLES 261 



iv. Demonstration of Homogeneity. Among the modern tools now employed 

 in the study of large molecules, the ultracentrifuge possesses singular power 

 for investigations of the purity and homogeneity of the sedimenting sub- 

 stance. Despite the directness and sensitivity of the sedimentation velocity 

 method, there have been many ill-founded claims purporting to prove 

 homogeneity. Most of these erroneous conclusions have been based on a 

 cursory examination of ultracentifruge patterns which reveal a single, sharp, 

 sjonmetrical boundary. Such superficial observations more often than not 

 are likely to be misleading. Even those claims based on a much more detailed 

 investigation of the shape of the sedimenting boundaries are likely to 

 require revision. This is necessitated by the refinements in the theoretical 

 treatments which have occurred in the past few years. Thus the conclusions 

 from the thorough investigations of bushy stunt virus (Lauffer, 1942) and 

 T2 bacteriophage (Putnam, 1951) no longer can be accepted without question. 

 If, upon reinvestigation with the greatly improved techniques, the diffusion 

 coefficients are found to be slightly less than the values employed in those 

 studies, the conclusions regarding the homogeneity of those viruses are valid. 

 Alternatively, confirmation of the older data would lead to the inference 

 that the preparations contained particles of varying size and shape. Until 

 new data becomes available this matter of the homogeneity of bushy stunt 

 virus and T2 bacteriophage remains misettled; but all claims regarding ultra- 

 centrifugal demonstrations of homogeneity of virus preparations now must 

 be considered incomplete. 



The shape of a boundary, i.e., the distribution of concentration of solute 

 as a function of distance, in a sedimentation velocity experiment is con- 

 trolled by four factors. First of these is the broadening of the boundary due 

 to diffusion, a consequence of the concentration gradient formed in the 

 boundary region by the migration of the solute molecides. Second, broaden- 

 ing of the boundary occurs during the sedimentation of polydisperse material, 

 since the faster moving molecules become separated from the slower com- 

 ponents. In principle, the observed boundary can be considered as a com- 

 posite boundary resulting from the sum of the boundaries of the individual 

 components in the solution. Opposing these two effects which cause the 

 boundary to spread with time is, third, the so-called sharpening effect 

 resulting from the dependence of sedimentation velocity on concentration. 

 At the trailing, or solvent, side of the boundary the concentration is much 

 lower than on the solution side. Those molecules, falling behind as a result 

 of diffusion or a lower sedimentation coefiicient, find themselves in an 

 environment of lower concentration. Their sedimentation rate therefore 

 increases until they overtake the boundary. This self-sharpening of the 

 boundary occurs contmuously if the sedimentation velocity of the solute 

 molecules is concentration-dependent; and the greater the concentration 



