IV. THE CRYSTALLINITY OF VIRUSES 65 



sion and those, such as thermal agitation, that tend to break this ad- 

 hesion, breakable rod, when broken, leaving polar ends which can join 

 together again. 



As already mentioned, plant protoplasm proteins show a high 

 tendency to unite with one another ,while at the same time it is known 

 that the process of purification causes aggregation of the particles 

 initially present in the fresh infective sap. In connection with this 

 fact, Bawden (25) stated that since the available evidence on tobacco- 

 mosaic virus and virus X suggests that any process that concentrates 

 them or that removes contaminations from them is likely to cause 

 agglutination, agglutination may be indispensable from a high degree of 

 purity, and claims to have isolated homogeneous, unagglutinatod par- 

 ticles may be equivalent to claiming incomplete purification. At pre- 

 sent it seems generally accepted that tobacco-mosaic virus particles can 

 occasionally exist in unusually long threads under the electron micros- 

 cope. Stahmann et al. (40) found that Wisconsin pea-streak virus par- 

 ticles, on standing in distilled water for several weeks, associated end- 

 to-end and side-to-side to produce bundles in parallel alignment which 

 were often coiled about each other. 



Alteration in the particle lengths appears to occur not only in vitro 

 but also in the plant cells in situ. On an electron microscope study of 

 tobacco-mosaic virus extracted from pulp and juice after various periods 

 of infection, Takahashi and Rawling (41) found that when the virus is 

 extracted from finely macerated pulp and juice at the pH of plant 

 juice, the proportion of short particles increases between the 4th and 

 16th days after infection. They claimed that the short particles did 

 not result from fracture of longer particles during maceration. More- 

 over, according to Black et al. (42), electron micrographs of sections 

 through tobacco leaves infected with tobacco-mosaic virus sometimes 

 showed in the cytoplasm of the diseased cells thread-like filaments much 

 longer than 280 m/j, indicating the natural occurrence of end-to-end 

 association of the rods. 



Joly (43) showed by the method of streaming double refraction, 

 that the prolonged cooling of tobacco-mosaic virus produces a partially 

 reversible shortning of the virus particles. Furthermore, according 

 to Schramm (44), in alkaline buffers, tobacco-mosaic virus particles 

 dissociated into smaller components having molecular weight of about 

 360,000 as compared with about 40,000,000 for the original intact protein, 

 but readjustment of the solution to pH 5 yielded a material practically 

 indistinguishable from the original native protein in molecular weight 

 and shape, in crystal form and electrophoretic behaviour. 



Thus, there seems no doubt that at least tobacco-mosaic virus par- 

 ticles can on the one hand split into smaller particles and on the other 



