THE BIOCHEMISTRY OF PLANT VIRUSES 63 



mind that all, or nearly all, workers in the field make the tacit assumption 

 that all, or nearly all, of their virus particles are viable. This assumption may 

 be unjustified. 



From a study of the properties of crude virus in sap it was realized at an 

 early stage that the tobacco mosaic virus was probably protein in nature; 

 in 1933, Takahashi and Rawlins, by an inspired experiment, deduced that 

 the virus was rod-shaped. The very fact that they were able to do this, using 

 a simple technique involving the observation of the sap flowing through 

 capillary tubes in polarized light, should, of course, have indicated that the 

 quantity of virus material in plant sap was fairly large. About the same time, 

 Bechhold and Schlesinger (1933) made an estimate of the size of the virus 

 particles. They did this by a centrifugal method, which depended upon the 

 estimation of virus activity, so that their results could be attributed directly 

 to the virus in the plant sap, rather than to material, nonviral in nature, 

 but associated with virus infection. Unfortunately, they assumed that 

 the virus particles were spherical, but their experiments showed that the 

 virus could be sedimented in ordinary laboratory high-speed centrifuges. 



Shortly after this, Stanley (1935, 1936), who had been making an intensive 

 study of the properties of the virus in impure sap and of the effects of various 

 treatments on its infectivity, came to the conclusion that the infective agent 

 was probably protein in nature, thus resembling the enzymes, and so he 

 applied the purification methods which had been developed for the isolation 

 of the latter. Using ammonium sulfate and acid as precipitants, he was able 

 to isolate a protein having the characteristics of a globulin and having 

 high infectivity. The material also had a crystalline appearance. Repeated 

 recrystallization had no effect on the general properties of this material, 

 so Stanley concluded that his preparations were essentially pure and con- 

 sisted of the virus itself, even though the amounts which he isolated were so 

 much greater than had been anticipated by other workers. 



As is frequently the case in scientific investigations of this type, several 

 others were working in the field at the same time. Notable among these 

 was Best (1936) in Australia, who noted that a globulin-like protein, which 

 had virus activity, was precipitated when the sap from infected plants was 

 acidified. Also about this time Bawden and Pirie (1936), who were interested 

 in the antigens present in virus-diseased plants, made partial purifications of 

 potato virus X (potato latent mosaic virus) and found that their preparations 

 consisted largely of protein. 



The next advance made was that it was found that the tobacco mosaic 

 virus was a nucleoprotein. Largely owing to the adoption of misuitable 

 methods for the drying of the virus preparations for analysis, Stanley had 

 overlooked the presence of some 0.5 % of phosphorus on a dry weight basis 

 in his preparations. This phosphorus, which is present as ribonucleic acid, 



