334 C. E. SCHWERDT 



The reproducibility of plant virus assay by local lesion counts will vary with 

 the experimental situation (Section II, C). Differences in virus activity of 

 from 50 % (Beale, 1934) to as low as 10 % (Loring, 1937; Spencer and Price, 

 1943) have been detected for tobacco mosaic virus preparations under 

 carefully controlled conditions. The accuracy and general statistical analyses 

 of the local lesion method for measuring comparative infectivities have been 

 re\'iewed in detail by Lauffer and Price (1945), Price (1946), and Bald (1950). 



The relationship of total virus mass or numbers to lesion count will be 

 considered later (Section IV, B). It might be well to point out at this time, 

 however, that, in general, it is difficult if not impossible to translate differences 

 in counts to differences in plant virus mass. 



2. Bacterial Viruses 



The classic example of a local lesion type of virus assay is the highly 

 accurate plaque count of bacterial viruses first described by Gratia (1936). 

 The procedure as it is used currently is presented in detail by Adams (1950) 

 and is carried out as follows: Dilutions of virus are mixed with a bacterial 

 suspension in nutrient agar and poured over the surface of ordinary agar 

 plates. Clear areas or plaques are produced by the virus in the confluent 

 growth of bacteria in the agar overlay after a suitable incubation period at 

 37°C. The plaque comit per plate divided by the volume and dilution of the 

 virus inoculum gives the titer in plaque-forming units per milliliter of original 

 suspension. 



A direct proportionality is observed between plaque count and relative 

 virus concentration when both variates are plotted on the same scale 

 (usually logarithmic); thus the curve is linear with a slope of 1. Because of 

 this linear relationship, each plaque is considered an infective center initiated 

 by a single virus particle infecting a susceptible bacterium (Luria, 1940). The 

 plaque count does not necessarily represent the total virus particle count but 

 only that fraction capable of adsorbing to and infecting viable bacterial cells. 

 This fraction is called the "efficiency of plating" and can be estimated in a 

 relative sense by comparing the plaque count with the infectivity of the 

 virus for bacterial cells in broth suspension where conditions are usually more 

 favorable for virus adsorption. For example, a bacterial virus preparation in 

 high dilution is added to a suspension of susceptible bacteria. Small aliquots of 

 the mixture, designed to yield on the average less than one infected bacterium 

 per sample, are incubated until the infected bacteria have burst, liberating 

 virus. Each aliquot is plated and the number of plates yielding no plaques as 

 well as those yielding large numbers of plaques is noted. From the proportion 

 of plates with no plaques one can estimate the average number of virus 

 particles, n^, per aliquot from the Poisson formula: 



