14 PHYSIOLOGICAL TRIGGERS 



TMV replication, its prime function being the transmission of the disease to a 

 new host. 



In considering a model for the continuous virus spread in N. glutinosa we are 

 faced with the problem of the cell wall. When intact, it presents an impenetrable 

 barrier both to extracellular as well as intracellular virus. In fact, the only 

 method for extracting TMV involves destroying this cell wall. Nothing compa- 

 rable to lysis or the gradual exodus of virus particles through a cell membrane 

 has been observed in this plant system. In order for the virus to spread, then, 

 from cell to cell, it most probably is obliged to use the only pathway available, 

 and that is via the plasmodesmata. These fine protoplasmic strands that join 

 each cell with its neighbor can serve as adequate bridges for the virus to cross. 

 Thus a simple model to describe the continuous spread of necrosis is one based 

 on the chance migration of infectious units to adjacent cells. The rate of spread 

 would depend upon the quantity of infectious particles per cell, the velocity of 

 cyclosis, and the number of available cell exits in the form of plasmodesmata. 

 Admittedly, the model may be too simple, for it does not consider such possi- 

 bilities as active transport of particles or other host responses. 



Once the virus is extracted from its cellular environment, it is characterized 

 by a very low infectivity. This low infectivity has been ascribed at least in part 

 to the inefficiency of inoculation methods. Data from certain of our serological 

 studies suggest that of equal importance is the fact that only a small fraction of 

 the rods actually possess the necessary information for starting a new infection. 

 We found that by mixing a series of low virus concentrations with antiserum, 

 different sized aggregates formed. The infectivities associated with these 

 clumped rods nevertheless remained alike. Part of the data bearing on this point 

 appear in table i. 



Although the concentrations of serum and virus employed were varied, their 

 ratios were maintained constant. After a suitable incubation period the mixtures 

 were centrifuged moderately, at a speed which did not sediment single virus 

 particles from normal serum controls. The upper portions from the centrifuged 

 samples were assayed for virus activity along with the control uncentrifuged 

 mixtures. The level of survivors before centrifugation showed little change, 

 despite the fact that the average aggregate size increased. This is seen most 

 clearly in the last column of table i, where the percent of virus infectivity re- 

 maining unsedimented decreased markedly with increasing concentrations. Our 

 interpretation of the data is simply that the number of infectious particles pres- 

 ent initially is minute compared to the total number of rods in the population. 

 Then, despite the increment in mean aggregate size, the probability of trapping 

 more than one viable particle per clump, under these conditions, is very small. 



The low infectivity of the virus inoculum is puzzling. We do not know 

 whether most of the infectivity is destroyed while extracting the virus, or 

 whether out of the hundreds of thousands of rods the cell makes, it impresses on 



