THE PROBLEMS OF VIROLOGY 3 



every phase of fundamental biology. It is axiomatic that aU available 

 techniques are being, and must be, used to define those qualities of viruses 

 which are capable of being usefully expressed in terms of physical and chem- 

 ical concepts. This approach is directly applicable to the specification of the 

 infective units. There is now such a wide range of chemical and physical 

 techniques for the purification or isolation of molecules, macromolecules, and 

 larger aggregates that the obtaining of smaller or larger quantities of "pure" 

 virus is no longer the formidable task that it was only a few years ago. As in 

 any other type of biochemical work, perhaps the most important single factor 

 is a reasonably convenient means for the bioassay of fractions. A large enough 

 supply of crude starting material and facilities for carrying out the necessary 

 manipulations are, of course, necessary. For most viruses the ultracentrifuge 

 is probably the most important tool for purification but will always need to 

 be assisted by other types of manipulation. Where small viruses must be 

 separated from particular fractions of host cells of similar size and density, 

 the problem of purification may be very difiicult and may require the use of 

 all the empirical artifices that the ingenious chemist can devise. The progress 

 of purification will normally be followed by assessmg the biological activity 

 in relation to the amount of material in the fraction (measured as dry weight, 

 N content, or in some other fashion). Where viruses have biological activity 

 other than their power to infect and cause demonstrable lesions, this is often 

 more convenient to use in combined chemical-biological work. Influenza 

 viruses wiU act as hemagglutinins, and many types of virus can be detected 

 and titrated by complement fij^ation tests with appropriate antisera. In aU 

 cases it must eventually be established that the property chosen is one 

 specific for the infective particle and not, for example, associated with some 

 nonviral product of infection, such as a soluble hemagglutinin or complement- 

 fixing antigen. 



With purified material it is possible to use a variety of physical methods 

 to define size, shape, and density of the infective units and, if necessary, to 

 elaborate in regard to electrostatic charge, electron density of different 

 regions, and adsorption to various types of surface. Some of the physical 

 methods are applicable to unpurified virus preparations; the first reliable 

 estimate of the range of sizes presented by bacteriophages and animal 

 viruses was obtained by Elford's use of graded collodion membranes in 

 filtration experiments with relatively crude materials (Elford, 1932). 



Many of the smaller plant viruses and two of the animal viruses have now 

 been crystallized or at least been shown to take up regular two-dimensional 

 arrays. It is doubtful whether this means more than that the virus particles 

 are nearly or completely uniform in size and composition and are geometric- 

 ally capable of regular packing. The possibihty of crystallization, however, 

 does open up an important new approach to deeper understanding of virus 



