THE PHYSICAL PROPERTIES OF INFECTIVE PARTICLES 299 



4. The point of adding this hypothetical mixture to the list of virus pre- 

 parations is to show that it would be homogeneous with respect to morpho- 

 logical and hydrodynamic properties, but would probably be inhomogeneous 

 with respect to other properties. In respect to infectivity, however, it might 

 appear homogeneous, because the presence of one type of virus might com- 

 pletely suppress the expression of infectivity of the other type. 



It is abmidantly clear from the foregoing that the characterization of a 

 virus suspension in terms of homogeneity is significant only if the terms of 

 reference of the characterization are specified. It is safe to say that no virus 

 exists whose particles are known to be miiform, one to the other, in terms of 

 all the tests that can be applied to detect uniformity. The T-even bacterio- 

 phages are possibly so, within fairly wide limits of uncertainty in their 

 hydrodynamic properties. Bushy stunt virus, poliovirus, and tobacco 

 ringspot virus are examples of those whose particles may be completely 

 uniform; it is only because of their low specific infectivity that we cannot 

 be fairly certain. 



In the light of the above discussion another glance should be given the 

 word "pure." The difficulty with this word is that it is a catchall one, imply- 

 ing too much in too general a sense. It might be argued that a virus susj)ension 

 can be called pure only if it contains particles that are identical in every 

 measurable and conceivable respect. But short of tliis extreme point of view 

 the words "purified" and "purification" are useful ones, and even necessary 

 ones. The process of increasing the over-all homogeneity of a suspension of 

 virus particles is conveniently and properly called "purification." We are 

 also faced with the problem of how to describe simply a virus preparation 

 from which all known contaminants are removed, and which demon- 

 strates homogeneity in several respects. To call it "homogeneous" is to 

 require that its tyjjes of homogeneity be described. Like most preparations 

 of TMV, it may not even be physically homogeneous, or it may be that 

 some aspects of homogeneity have not been ascertained. What are we to 

 call such a preparation, to distinguish it from one that is known to contain 

 many kinds of particles besides virus? The latter kind of preparation is un- 

 hesitatingly called "impure." It would seem that the most satisfactory way 

 to describe a generally homogeneous preparation is to call it "purified," in 

 accord with common practice. 



B. Identification of Physical Particles as Infective Agents 



Physical measurements made on objects as small as viruses must involve 

 great numbers of particles; consequently one can hope to identify physically 

 only a particular class of particles with an infective entity. The process of 

 identification consists in the separation of a virus-containing suspension 



