346 C. E. SCHWERDT 



spherical particles predominating in the preparation used by the advocates of 

 the "absolute assay," 



The hemagglutination method of counting influenza and Newcastle disease 

 viruses is not yet firmly established. Whether discrepancies between this and 

 the direct methods of counting reflect strain differences, technical difficulties, 

 or, in general, a lower efficiency of agglutmation than assumed requires 

 clarification by further study. Evidence, so far, suggests that these particles 

 are not fully effective or efficient in dimer formation. 



It would seem, then, that the direct procedures of enumerating morpholo- 

 gically identifiable virus particles by electron microscopy are the methods of 

 choice, since they demand that fewer assumptions be made and are applicable 

 to a wide variety of viruses without precluding viruses which do not possess 

 the special biological property of hemagglutmation. 



IV. Relationships between Total Particle Count and Biological 



Function 



A. Initiation of Infection by a Single Virus Particle 



Before the possibility existed of determining actual virus particle count by 

 direct observation, attempts were made by statistical means to test the 

 hypothesis that infection of a host may be initiated by a single virus particle. 

 It has been generally found in assays of the local lesion type that the con- 

 centration of the virus suspension (i.e., the number of physical particles per 

 inoculum) is directly proportional to the number of lesions produced. Thus a 

 linear curve is produced (ideally with a slope of unity), provided both variates 

 are plotted on the same scale, either arithmetic or logarithmic. This relation- 

 ship has been fomid over an appropriate dilution range for phage (Ellis and 

 Delbriick, 1939) as well as for various animal viruses assayed by the pock 

 count method on chick embryo chorioallantoic membranes, such as vaccinia 

 (Keogh, 1936; Burnet and Faris, 1942; Overman and Tamm, 1956a; Westwood 

 et al., 1957), infectious laryngotracheitis (Burnet, 1936), canary pox (Burnet 

 and Lush, 1936a), ectromelia (Burnet and Lush, 1936b), myxoma (Femier and 

 Mclntyre, 1956), and Rous sarcoma (Rubin 1955; Prince, 1958). Plaque 

 assays in tissue culture of western equine encephalomyelitis (Dulbecco, 1952), 

 vaccinia (Noyes, 1953), poliomyelitis (Dulbecco and Vogt, 1954a), influenza 

 (Ledmko, 1955), and foot-and-mouth disease (Sellers, 1955; Bachrach et al., 

 1957) have exhibited a similar proportionality. 



Luria (1940) and Dulbecco and Vogt (1954a) have concluded from statis- 

 tical reasonmg that the linear relationship between virus dilution and lesion 

 comit means that only one virus particle is necessary to mitiate a lesion. If 

 the production of a single lesion required the cooperative efforts of several 



