V. FINER STRUCTURE OF VIRUS PARTICLES AND ITS SIGNIFICANCE 73 



are remarkable differences between the virus isolated from polyhedral 

 bodies and that from the blood. For instance, polyhedrosis virus 

 having- a rod-shape isolated by Bergold (33) from polyhedral bodies 

 was sedimented completely by ultracentrifugation at 10,000 r.p.m. and 

 its "molecular weight" was calculated to be 916x10^, while the virus 

 particles isolated by Glaser and Stanley (61) from the infected insect 

 blood was not sedimented at 10,000 r.p.m. but only at 27,000 r.p.m. and 

 its "molecular weight" was estimated to be about 3,000,000. Moreover, 

 the virus from the blood was readily inactivated in solutions of more 

 acid than pH 5, whilst rod-shaped virus from polyhedral bodies was 

 so stable as to stand unchanged for 24 hours at pH 2.0 ; and again no 

 rod-shaped particles similar to the virus was found in normal in- 

 sects, whilst globular particles similar to those of the virus isolated 

 from the blood were obtained from normal ones (62). These facts 

 strongly suggest that the particles from the blood are nothing but the 

 coagulated elementary bodies containing lipids, just as in usual animal 

 viruses. 



The expulsion of lipids appears to occur not only immediately 

 after or during the change of the protoplasm protein by virus infec- 

 tion, but sometimes may occur gradually even after the change. The 

 particles present in plant saps infected with tobacco-mosaic virus were 

 separated into various fractions by Bawden and Pirie (63). The most 

 rapidly sedimenting fractions consisted mostly of rod-shaped nucleo- 

 protein, whilst only about a half of the total material in the most 

 slowly sedimenting fraction was virus nucleoproteins ; these fractions 

 showed no anisotropy of flow and had low infectivity. Electron mic- 

 rographs of the most slowly sedimenting fractions showed particles 

 most of which are little longer than they were wide, while with 

 inci^easing sedimentation rate the numbers of obvious rods increased. 

 All the fractions were unstable and rapidly passed into forms that 

 sedimented rapidly, showed intense anisotropy of flow, the change 

 being most striking in the preparations that previously sedimented 

 most slowly. Electron micrographs showed that these changes were 

 accompanied by an increase in the number of rods and in their average 

 length. 



This fact can be interpreted as the gradual occurrence of lipid 

 expulsion in vitro. Since the formation of rod-like particles never 

 means the generation of viruses, it is only natural that no increase in 

 the infectivity was observed despite the increase in the number 

 of rods. 



Bawden (26) cited another interesting fact : When concentrated 

 salt-free solutions of a tobacco-necrosis virus are kept at O'C, crystal- 

 lization occurs and thick triclinic prisms and the hexagonal, or pseudo- 



