64 II. FUNDAMENTAL STRUCTURE OF PROTOPLASM 



a little longer than 0.3// or 280 mjn, a value regarded as the length of a 

 tobacco mosaic virus particle, although some insect viruses have been 

 claimed to have a length near 0.4/« as described above. 



Since the virus particles observed in electron micrographs are in 

 the dry state, some shrinkage should be expected in them. It has been 

 shown that protein crystals such as those of oxyhaemoglobin or insulin 

 undergo shrinkage on the dessication. In methaemoglobin and i5-lacto- 

 globulin even so remarkable a shrinkage as up to 50 per cent is proved 

 (35) (36). 



"Molecular weight" of tobacco mosaic virus has been claimed, as 

 mentioned above, to be 40,000,000 ; such a high molecular weight cannot 

 be attained unless approximately 200 globulin molecules with molecular 

 weight of 150,000 are put together in a bundle even when a consider- 

 able amount of nucleic acids is intermingled. The bundles are considered 

 to be originated from "elementary bundles", already described, from 

 which lipids have been eliminated, not directly from elementary bodies 

 themselves which tend to be decomposed into the elementary bundles. 



As stated above, insect virus particles are reported to be in length 

 approximately 0.3 ju. like some plant viruses, whilst in width they ap- 

 pear generally to be about two times as thick as plant viruses. The 

 elementary bundles of insects may be said, therefore, to be thicker 

 than those of plants, that is, in insects far greater number of protein 

 molecules appear to be united to make an elementary bundle than in 

 plants. 



According to the kind of cells, protoplasm protein molecules may 

 tend to form very long rods or filaments through end-to-end association. 

 In such a case, virus particles may have lengths several times as long 

 as 0.3 /u. The virus particles of latent mosaic of tomato have been 

 claimed to be 525 m/z in length (33) ; this may be an example of end- 

 to-end combination of two molecules. According to Takahashi (37) 

 Brassica nigra virus is about 0.7 jx long ; this may also be two molecu- 

 lar long. 



On the contrary, if the molecules split into shorter fragments, the 

 particle should naturally become shorter. Whereas the length of to- 

 bacco-mosaic virus protein is generally believed to be 280m/^ some 

 workers are disinclined to accept the existence of such definite sized 

 particle. For example. Crook and Sheffield (38) concluded from electron 

 micrographs of tobacco mosaic virus that the particle size varies 

 according to the method of preparation, showing no presence of basis 

 for assigning a length. Bawden and Pirie (39) stated likewise that the 

 average length of the particles in a virus preparation is determined by 

 both the past history and present environment of the preparation, and 

 that it is a compromise between the forces leading to end-to-end adhe- 



