48 E. MAKKHAM 



possibly oversimplified assumption that all subunits are identical or closely 

 similar, one might find out interesting and, possibly, important facts. The 

 reader should, however, be referred at this stage to a review by Pirie (1956b) 

 relevant to this matter. 



Following upon the work on tobacco mosaic virus, Crick and Watson (1956) 

 pointed out that the most economical and rational way of constructing a 

 "spherical" virus was also from submiits, all of one kind fitted together in an 

 array which was based upon the geometry of the regular solids, and the 

 model which they favored was the solid figure, the icosahedron (Kaesberg, 

 1956), which has 20 equilateral triangular faces. Placing 3 subunits, which 

 may, of course, have their own substructure, in every equilateral triangle, 

 one then has a soUd figure having 60 identical subunits. This gives a poly- 

 peptide subunit having a molecular weight of about 100,000, a value which 

 is reasonable, and also gives a structure which is more or less compatible with 

 the known features of some of the smaller "spherical" viruses. It should be 

 remarked here that this type of regular figure is likely to be associated with 

 those viruses which crystallize in a cubic or hexagonal lattice, or an approxi- 

 mation to the latter. There are, of course, viruses such as the southern bean 

 mosaic virus, which have shapes incompatible with this concept of high 

 symmetry, while the very recent observations of the alfalfa mosaic virus 

 (Bancroft and Kaesberg, 1958) are even less compatible with this type of 

 model. However, for those viruses, such as the turnip yellow mosaic and 

 bushy stmit viruses, which show a high degree of symmetry, it is not unlikely 

 that they possess a regular subunit which is one-sixtieth of the whole, and 

 which itself may consist of a number of subunits. The latter may, of course, 

 not all be identical, but the whole structure, if made up in this way, would be 

 susceptible to investigation by orthodox chemical methods with some 

 possibility of success. Whatever transpires, it is certain that the next few 

 years will yield exciting results. 



C. Methods for the Determination of the Structure of Polypeptide Chains 



The methods which are at present available for studies of this kind are 

 essentially of three types. These are: 



1. Determination of Amino Acid Composition 



Given a pure peptide, the determination of its amino acid composition is 

 easily performed by chromatographic methods. If it is a small polypeptide 

 containing, say, 4 residues, simple chromatography on paper wiU suffice. If it 

 is very large, the Moore and Stein column chromatography will have to 

 be applied, bearing in mind the loss or destruction of such amino acids as 

 cystine and tr5^tophan during hydrolysis. 



