EVOLUTION AND VIRUSES 



considered them fundamentally different from even the simplest 

 organism. This conflict was resolved in favour of the minority 

 in the 1930's when tobacco mosaic and several other viruses 

 were isolated and characterized as liquid crystalline or crystal- 

 line nucleoproteins. At the time it was their ability to crystallize 

 that excited most attention, but this was mainly important in 

 showing a uniformity of particle size that would be unusual 

 with organisms and seems incompatible with multiplication 

 by binary fission. It also had the great practical value of making 

 the virus preparations amenable to study by the technique of 

 X-ray crystallography, which showed that the individual virus 

 particles were composed of many similarly sized units grouped 

 in a fixed and regular manner that contrasts strikingly with 

 the fluidity and variability of structure in cells. In their manner 

 of structure and in their chemical simplicity, consisting of 

 only two major components, protein and nucleic acid, they 

 clearly resembled macromolecules, single components of or- 

 ganisms, rather than whole organisms. 



Viruses then seemed to be self-replicating nucleoproteins, 

 and the temptation to introduce these apparently simple 

 reproductive systems into discussions on the origin of life was 

 irresistible to some people, although as always it meant 

 neglecting the fact they multiply only in living cells. The 

 temptation was increased by the fact that aminoacids, the 

 component parts of proteins, can be made by the action of 

 electrical discharges on a reducing atmosphere containing 

 methane, ammonia, water and hydrogen, the conditions often 

 assumed to have existed when life originated. If aminoacids, 

 why should not these condense into proteins and become self- 

 perpetuating systems? Perhaps, but recent work shows that 

 this has little relevance to viruses. The nucleoprotein particles, 

 which for 20 years seemed to be the minimal infective unit and 

 the replicating system, have now lost this position. Tobacco 

 mosaic virus will suffer the indignity of being taken apart and 

 put together again, and this fact has allowed the roles of its 

 two components to be tested. It is now clear that the protein 

 as such does not reproduce itself; virus fragments that consist 

 predominantly, and possibly exclusively, of nucleic acid can 

 be infective. The infections they initiate produce not only 

 more nucleic acid but complete virus particles containing their 



14 



