504 ANNUAL REPORT SMITHSONIAN INSTITUTION, 19 3 8 



hosts is a property that we have not hitherto ascribed to ordinary 

 molecules. We know, of course, certain autocatalytic reactions, such 

 as the conversion of trypsinogen into trypsin in vitro on the addition 

 of a little trypsin, in which a protein molecule appears to be reproduc- 

 ing itself, yet we also know that in such instances the precursor is 

 chemically quite similar to the end product. The situation prevailing 

 in the case of tobacco mosaic virus appears to be different, for here we 

 have found no evidence for the existence of a precursor similar to the 

 virus protein and the reaction has so far been carried out only within 

 living cells. Despite the fact that the net result of virus reproduction 

 may be regarded as similar to that of an autocatalytic reaction, there 

 is at this time no precedent for regarding the reaction itself as a true 

 autocatalytic reaction. I feel that it is better to admit our ignorance, 

 as long as we are forced to the inside of a living cell to achieve a set of 

 conditions in which to conduct our autocatalytic reaction. It is a 

 fact that we do not know all that goes on within a living cell and, 

 despite many attempts to duplicate it or the set of conditions pre- 

 vailing within, we have been unsuccessful. We are forced to conclude, 

 therefore, that, although tobacco mosaic virus protein has the ordinary 

 properties of molecules, it also has the ability to reproduce and to 

 mutate, properties not ordinarily ascribed to molecules, and hence 

 that tobacco mosaic virus protein represents an entity unfamiliar to 

 us. Now, we may speculate concerning its nature and the manner in 

 which it reproduces and mutates. I wish to impress upon you that 

 this portion of the discussion is not based upon fact, but upon fancy. 

 Let us first consider whether tobacco mosaic virus differs from ordi- 

 nary living organisms. The fact that it has proved impossible to 

 cultivate tobacco mosaic virus in the absence of living cells is fre- 

 quently cited as evidence that it differs in nature from ordinary living 

 organisms. However, this property of obligate intracellular para- 

 sitism is not peculiar to viruses, for there are certain organisms that 

 are known to be obligate parasites and to require the presence of 

 living cells for reproduction. I do not feel, therefore, that our failure 

 to cultivate viruses on synthetic media necessarily means that they 

 differ from bacteria. The fact that tobacco mosaic virus protein may 

 by crystallized has also been cited as evidence that it is not a living 

 organism. Yet we know that crystallization is merely orientation 

 and that orientation occurs in many living systems. Stretched muscle 

 is orientated and gives an X-ray diffraction pattern that is character- 

 istic of the crystalline state. At certain stages of cell division orien- 

 tation occurs within the nucleus, for it becomes highly doubly refrac- 

 tive. I do not feel, therefore, that the fact that we can cause tobacco 

 mosaic virus protein to line up in orderly fashion need necessarily 

 mean that it is nonliving. No evidence has been obtained that 

 viruses have cell walls, and this might be used as a basis for estab- 



