308 H. K, SCHACHMAN AND R. C. WILLIAMS 



c. Homogeneity and Length. There has been considerable disagreement over 

 the question of the "length" of the particles of TMV, particularly as measured 

 by electron microscopy. Part of the source of this disagreement lies in the 

 fact that the word "length" has more than one connotation when applied 

 to TMV. It may mean: (1) the length distribution that is fomid by electron 

 microscopy of virus suspensions that have been prepared by any one of 

 several purification methods. It may mean: (2) the length distribution that 

 is found when the jireparative procedure, from infected plant to the electron 

 microscope, has been performed in ways such as to minimize (or account 

 for) the effects of breakage and aggregation of the TMV particles. But the 

 word "length" may also mean: (3) the minimal length of the particle which 

 is infectious; (4) the length distribution of the particles that exist in the 

 crude juice from infected plants immediately after extraction; and (5) the 

 length distribution of the virus particles as they exist within infected cells. 

 Failure to distinguish which type of length is being investigated or dis- 

 cussed is an obvious source of confusion. 



Most TMV preparations that have been examined by electron microscopy 

 have exhibited a rather broad length distribution (Bawden, 1950), with the 

 mode usually fallmg in the vicinity of 3000 A. The breadth of the distribu- 

 tion curve, the position of its mode, and the type and severity of its skewness 

 can be influenced by the extraction procedures, the method of purification, 

 the pH conditions of storage, and the preparation of the suspension for 

 electron microscopy (Pirie, 1957). In two published cases, however, a high 

 degree of uniformity of length has been reported. In the earlier one (Williams 

 and Steere, 1951), the virus was prepared for observation by extracting the 

 plant juice by homogenization, followed only by heating to 50°C. for 5 

 minutes and clarification by low-speed centrifugation. Samples of the entire 

 virus suspension w^ere sprayed upon the electron microscope grids at dilu- 

 tions such that the lengths of all the particles in each droplet pattern could 

 be measured, and the data handled m a way such that the effects of breakage 

 and aggregation due to drying could be corrected for. The results showed 

 that: (1) before correction of measured lengths, about 70 % of the particles 

 had lengths between 2900 and 3100 A; (2) after correction, about 95 % of 

 the particles had lengths within this range, or had lengths exactly twice this. 

 Recently, a measurement of length distribution has been made of TMV 

 particles tliat had been purified by gentle centrifugal methods (Hall, 1958). 

 The distribution fomid was practically identical with that fomid by Williams 

 and Steere from their measurements, before corrections for breakage were 

 applied. 



The experimental results just described could have conceivably arisen 

 from one or more of the following situations: (1) the virus in the crude juice 

 was polydisperse, consisting of numerous short particles as well as long ones, 



