74 II. FUNDAMENTAL STRUCTURE OF PROTOPLASM 



hexagonal, plates are produced. These begin to form after a few 

 days, but growth continues for months and will give rise to single 

 crystals several milimeters long. In this case the preparations lost 

 most of their infectivity by the time crystallisation was completed, 

 indicating the occurrence of some change in the protein structure, 

 during the lipid expulsion, unfavourable for the virus activity. As 

 previously discussed, the expulsion of lipids may be caused by changes 

 of the protein structure generally unfavorable for the virus acti- 

 vity, though the virus may become more stable after the expulsion. 

 In the above example of tobacco-mosaic virus, it has also been 

 stated that the formation of rods was never followed by the in- 

 crease in infectivity, but on the contrary by the reduction by about 

 one-half. 



The presence of an intimate correlation between the viruses and 

 such rods or crystals could be demonstrated by the fact that both 

 rods and crystals had the serological property specific to the respec- 

 tive virus. In general, the serological property of a virus is more 

 stable than the infectivity, so that in the case above cited only the 

 infectitivity seems to be lost, the serological property being left 

 infact. 



In the writer's opinion (22), virus particles having the property of 

 the coagulated elementary bodies, if the coagulation is complete, fail* 

 to show the agglutinability by their antibody, because of the disap- 

 pearance of the active polar groups on the folding of the protein 

 molecules through the coagulation. However, if the protein molecules 

 are unfolded and expanded, the reactive groups will reappear with the 

 recovery of the agglutinability. Such an unfolding results from the 

 expulsion of lipids, and as a consequence the rod formation is accom- 

 panied by the appearance of the agglutinability, w^hereas since serolo- 

 gical property of plant viruses is usually recognized by the presence 

 of the agglutinability, the appearance of the property is occasionally 

 thought even to be the production of the virus protein itself, thus 

 making the matter fall into the most confusion. It is stated that the 

 formation of rods, in the above cited case of tomato-mosaic virus pre- 

 parations, was actually accompanied by the revelation of the serological 

 property. 



3. The Size of Virus Particles as Coagulated 

 Elementary Bodies 



Among the virus particles isolated by our method those of phage 

 are found to be the smallest and the dry particles are estimated to 



