146 ALICK ISAACS 



standard virus. These findings imply that preparations of virus showing 

 varying degrees of incompleteness are heterogeneous with regard to the total 

 nucleic acid content of their individual virus particles. In view of the known 

 significance of nucleic acids in the infectivity of bacteriophages and tobacco 

 mosaic virus, it seems reasonable to conclude from Ada and Perry's findings 

 that the probability of a virus particle being able to initiate infection may be 

 a function of its nucleic acid content. This conclusion gains further signifi- 

 cance when we consider the mode of production of incomplete virus. 



B. Production of Incomplete Virus 



The all-important factor in producing incomplete virus seems to be that 

 large doses of seed virus are essential. As is discussed below, seeds obtained 

 after two, three, or more serial passages of undiluted virus are more effective 

 in producing incomplete virus than seed which has not been passaged in this 

 way. Nevertheless, the same fact holds with all these seeds, i.e., if they are 

 passaged in undiluted form, further incomplete virus is produced, whereas if 

 they are passaged diluted 1/100 or more, standard virus is once more produced 

 (von Magnus, 1951c). It is natural to suppose that the critical factor in 

 producing incomplete virus is the multiplicity of infection. This point was 

 first discussed by von Magnus (1951c) who found that about 10 2 agglutinating 

 doses of virus, or more than 10 9 virus particles (Donald and Isaacs, 1954a) 

 was the minimum amount which would produce incomplete virus. Recent 

 counts of the number of surface allantoic cells in 10- and 11-day eggs are 

 between 1.8 X 10 7 (Tyrrell et al., 1954) and 3 to 4 X 10 7 cells (Cairns and 

 Fazekas de St. Groth, 1957); this would correspond to a multiplicity of 

 between about 10 and 100 virus particles per cell. Since incomplete virus is 

 not produced when seeds are diluted 1/100 or more, these figures support the 

 idea that multiple infection of cells is an important factor in the production 

 of incomplete virus. On the other hand, Cairns and Edney (1952) concluded 

 from their results that incomplete virus was produced when only 1 % of the cells 

 was infected initially. However, their calculations were based on the assump- 

 tion that a multiplicity of 1 occurred when 10 2 * 2 agglutinating doses of virus 

 were taken up by the allantoic cells; more recent techniques of counting virus 

 particles and allantoic cells would suggest a much higher figure for the 

 multiplicity after this inoculum. From Cairns and Edney's experimental 

 findings, incomplete virus production is first detectable when between 10 and 

 100 agglutinating doses of virus are taken up (10 8 to 10 9 virus particles) and 

 the corresponding figure given by Horsfall (1954) is 3 X 10 7 "hemagglutina- 

 ting" particles, which is equivalent to 3 X 10 8 virus particles (Tyrrell and 

 Valentine, 1957). There is, therefore, nothing in these estimates to refute the 

 idea that multiple infection of cells is a critical factor in producing incomplete 



