NATURE OF VIRUSES 7 



is. According to the i)resent all-pervading genetic doctrines, the 

 attenuation is the selection of a particular virus strain from a 

 mixture. Thus, any fresh virus preparation is thought of as 

 perhaps containing thirty or so different strains. By serial 

 passage through a well-chosen host, the projjortion of these can 

 be so altered that the virus is predominantly of a virulent strain. 

 This way of explanation is not necessarily what one would choose 

 ab initio from the facts, inasmuch as modification can still be 

 observed after 300 serial passages. 



Chemical Composition of Viruses 



In order to determine this, quite pure preparations are neces- 

 sary. This is relatively easy in the case of larger viruses, but 

 very hard for smaller. So no great claim to precision should be 

 made. A table, taken from Stanley and Lauffer (1948), of the 

 elemental c()m})osition, and of the proportion of protein, lipid, 

 and nucleic acid, is given below. It is nearly true that only 

 animal viruses contain any lipid. 



In the case of tobacco mosaic virus (TMV) and cucumber 

 virus, an analysis of the amino acid content of several strains 

 has been made. This is given in the table on page 8. (See 

 Knight 1947.) 



It can be seen that the differences between the TMV strains 

 are not marked. 



It is renuirkable that 90% of the surface amino groups on 

 tobacco mosaic can be acetylated without affecting the infec- 



