SELECTED PAPERS 



microbes, and one might be inclined to conclude 'que tout etait pour 

 le mieux dans le monde des inframicrobes'. 



Nevertheless it is understandable that numerous investigators cher- 

 ished the desire to become more closely acquainted with the represen- 

 tatives of this invisible living world. Highly ingenious methods were 

 developed for this purpose, and slowly but surely the knowledge of 

 viruses of man, animal, plant, and bacterium increased. I shall delib- 

 erately neglect the historical development of this information, and 

 right away mention one of the most recent results obtained in this 

 field. The circumstances under which these most important observa- 

 tions were made are decidedly of great significance. It is but necessary 

 to imagine the situation of a chemist connected with the Rockefeller 

 Institute for Medical Research who decides to acquire a better under- 

 standing of the virus of a plant disease by means of the sober and clear- 

 cut methods of modern biochemistry. As experimental material he 

 chooses the causal agent of the tobacco mosaic disease, the classical 

 affliction that Iwanowski in 1892 first ascribed to a filtrable, hence 

 submicroscopic living agent, and which six years later led Beijerinck 

 to his bold hypothesis of a 'contagium vivum fluidum' . 



It was already known to Iwanowski that a tobacco plant suffering 

 from the mosaic disease can yield a press juice which, even after 

 filtration through filters of a normal pore size, is highly infectious for 

 healthy plants. Here, therefore, we have a clear liquid that contains, 

 besides a large number of soluble plant constituents, the causative 

 agent of the disease. To a biochemist with up-to-date training, W. M. 

 Stanley, the notion was obvious that one might attempt to separate 

 the agent from the other substances by fractional precipitation and 

 selective adsorption, checking the distribution of the causal agent in 

 the various fractions at each step by infection experiments on a quan- 

 titative basis. In order to accomplish the latter he used a method first 

 devised by Holmes in 1929. It is based on the fact that in leaves of 

 Nicotiana glutinosa, in contrast to what happens in JV. tabacum, the virus 

 causes strictly localized damage. Thus it suffices to smear half of a 

 leaf with a virus-containing liquid, using the other half as a control ; 

 this permits a straight-forward count of the number of virus particles 

 present in the solution. By extended trials Stanley succeeded in elim- 

 inating more and more inactive material, finally winding up with a 



334 



