THE PHYSICAL PROPERTIES OF INFECTIVE PARTICLES 307 



seems necessary as well. Since light-scattering techniques have improved in 

 recent years, the most reliable value is that obtained by Boedtker and 

 Simmons (1958), particularly since they provided other data helping to char- 

 acterize their preparations. 



The molecular weight of TMV has been obtamed by analytical electron 

 microscopy by "Williams et al. (1951). This technique involves direct particle 

 counting of the virus particles in a microdrop of known volume. From the 

 measured number of particles per unit volume and the dry weight concen- 

 tration the molecular weight is evaluated directly. Since some breakage of 

 the particles occurs during the preparation of the specimen for electron 

 microscopy, Williams el al. (1951) measured the total length of all the par- 

 ticles in the drop pattern and then calculated the mass per unit length. This 

 figure multiplied by the length of the virus particles, which was determined 

 independently, gave the molecular weight, 49 X 10^. No adequate explana- 

 tion for the discrepancy between this value and those given above has yet 

 been presented. Errors in the determination of the size of the polystyrene 

 latex particles (used in determining the droplet volume) may be responsible. 

 Evidence against this explanation is provided by the results of the length 

 measurements (discussed later). Alternatively, the virus sample may not 

 have been dried completely; this, too, seems an imlikely explanation since 

 the method used in this work is similar to that employed by others. Errors 

 in the determination of the dry weight would affect the accuracy of light- 

 scattering determinations as well as those of partial specific volume. A third 

 possible error, which would result in a high molecular weight by the counting 

 procedure, would be that impurities were present and not recorded as count- 

 able particles. However, examination of many fields containing thousands 

 of virus particles did not reveal such impurities, and it is unlikely that they 

 account for any appreciable increase in the apparent dry weights. 



It is also possible to determine the molecular weight of TMV by a com- 

 bination of X-ray and chemical analysis. The former has shown (Franklin, 

 1956a,b) that the virus rod is built of a number of crystaUographically 

 identical substructures. If these be identified with the chemical submiits of 

 the virus (Knight, 1954; Harris and Knight, 1955; Ramachandran, 1958), 

 and if a length of 3000 A is taken for the intact particle, a value of 40 X 10^ 

 can be calculated for the molecular weight. It should be pointed out that 

 this calculation is based on the imconfirmed assumption that the crystaUo- 

 graphic and the chemical submiits are identical. 



It is hoped that reinvestigation of the molecular weight of TMV by all 

 of the methods described above will produce individual results characterized 

 by higher precision and leading to a imique value that is accurate to a few 

 per cent. For the present, the tentative value of 4 X 10^ gm./mole is pro- 

 posed. 



