90 R. MARKHAM 



distantly related to tobacco. The physical properties investigated did not 

 differ greatly with the host, while the chemical properties studied were 

 somewhat limited. For convenience in analysis the proportions of only 

 3 amino acids, namely, tyrosine, phenylalanine and tryptophan, were 

 estimated. As it happens, the proportions of these amino acids differ between 

 the two \4ruses, but they are not modified by varying the host. (In fact, the 

 only differences which look particularly significant in this work are in the 

 phosphorus, and hence nucleic acid, content of the viruses, which was greater 

 when they were grown in Phlox by some 20 %.) It is, perhaps, unfortunate 

 that only these 3 amino acids of the 18 or so amino acids present in the 

 virus were studied. The choice was, of course, dictated by ease of analysis, 

 but as any difference which might be expected might be a small one, and 

 possibly involving only a smaU number of amino acids or even of one amino 

 acid, it is more likely to be detected by a complete analysis which includes, 

 in particular, those amino acids present in very small amounts. These are, 

 for example, histidine, lysine, cysteine, and, of course, tryptophan, the 

 values for which obtained in this work were rather higher than one might 

 expect from more recent determinations. 



In spite of the fact that the sample used was so small, and that only two 

 strains of the virus were studied, the work done on Plilox and tobacco has 

 caused nearly general acceptance of the dogma that the protein of virus has 

 a definite composition which is independent of the nature of the host on 

 which it was grown. This may indeed be so, but it is quite evident that this 

 matter should be remvestigated using modern techniques. 



There is some evidence that the host may in fact modify the virus which 

 is parasitizing it. A short time ago Bawden (1958) described a virus, related 

 to the tobacco mosaic virus, which was found growing on a leguminous host, 

 the cowpea, in Nigeria. This virus, unhke ordinary strains of tobacco mosaic 

 virus, was able to parasitize bean {Phaseolus vulgaris) plants systemically, 

 as well as tobacco, and when grown on the former host, it gave rise to very 

 small and characteristic lesions on Nieotiana glutinosa. On transfer to tobacco, 

 the virus changed its character, and gave larger lesions on the test plant, 

 more like those of the type virus. These changes are reversible, and are 

 accompanied by a definite change in the amino acid content of the protein, 

 the most spectacular manifestation of which is the presence of histidine in 

 the virus grown on bean plants and its absence in the virus grown on tobacco. 

 This variation in the histidine content is compensated by a complementary 

 variation in the lysine content of the protein. 



The most probable explanation of this phenomenon is that the cowpea 

 virus has a high mutation rate both from the tobacco type to the bean type 

 and back again, and that the host plant dictates the type which is predomin- 

 ant. There is no question of the cowpea virus being a mixture of two distinct 



