492 A. TSL'GITA AND H. I- UAKX KKL-CONHAT 



one or the othiT manner. The local lesion response observed in Xicotiana 

 glutinosa and the artificial hybrid, -V. talxu-uin var. Xanfhl, nc, serves 

 for the quantitative assay of the virus, whih' the spi'cadin^ of the \ini> 

 throughout the growing plant in A', tabacuni is most favorable for i.-ohi- 

 tion of the virus in jireparative scale. Apart from the pui'poses of (juanti- 

 tation and isolation, the host response plays a crucial role in the detection 

 and cliaracterization of strains or mutants of tlic viius. Thus, many 

 strains ^ive characteristically small and non-spreading lesions and can 

 in this manner be differentiated on the local lesion host. Many strains 

 give local lesions in N. sylvestris and Java tobacco, varieties on which 

 connnon TMV spreads systemically. Finally, the nature of the systemic 

 response of A'', tabacum and other varieties can be markedly different. 

 Virus sti-ains are known which do not show any detectable symptoms, 

 even though they build uji to high concentrations throughout the plant. 

 Others give mild symptoms (mottling), or the typical blistered effect 

 of the mosaic disease, which may be accompanied by more or less yellow 

 discoloration of leaves. Finally, necrotic areas may develop upon infec- 

 tion with certain strains, and leaves or even the whole plant may die as a 

 consequence of the infection. Thus, the observation of the response of 

 several varieties of tobacco usually permits the identification of a 

 known strain or the characterization of a new strain. The same tech- 

 niques serve to test the biological homogeneity of a strain or to free it 

 from a contaminant. Repeated passage through a local lesion host at 

 limit dilution gives the best assurance that one is dealing with a geneti- 

 cally homogeneous population of virus particles. Further, the use of a 

 differential host i.e., one which responds systemically to one strain and 

 locally to another can rule out the simultaneous presence of both types 

 of strains. 



An additional criterion has been proposed for the differentiation of 

 strains which also supplies data concerning the viability of such 

 strains. This method consists in homogenizing individual lesions (8-day- 

 old, on A', tabacum var. Xanthi, nc) and testing these homogenates after 

 appropriate dilution on new Xanthi plants. The number of lesions which 

 can be evoked by one parental lesion has been termed productivity and 

 was found to vaiy over tenfold for different natural strains (Veldee and 

 Fraenkel-Conrat, 1962). 



Since mechanical wounding is required to initiate infection, the virus 

 solution is usually rubbed onto the surface of the leaves in presence of 

 an abrasive (i.e., carborundum). Any trace of virus accidentally de- 

 posited on the leaf surface at any time prior to this rubl)ing through 

 transmittal by agents such as the nozzle of a water hose can upon 



