508 A. TSUCITA AND H. !• HAKXKEL-rOXHAT 



proceeded to tryptic splittinp;, and aiialyzcfl the i-csultaiit peptides 

 f^oparatc'ly. Ho has never reported in detail how he jmrified each peptide 

 and it is our experience that some, and particuhirly the I-iiei)tide (41 

 residues), can he frcH'd oidy witli (Hfliculty from traces of adsoi'hecl p(^p- 

 tides. The po.ssil)ility thus exists that a seeming (hfference in the composi- 

 tion of the T-peptide is due to this technical difficulty. The results 

 reported hy W'ittniami ai'e in part in line with our obser\'ations, hut 

 show some unexpected differences. His observed i)roportion of mutants 

 of unchanged comi)osition (70%) is higher than ours, as would he 

 expected from the fact that he relied on random sampling only, while 

 we fi'equently used a method which selects for protein changes. 



A brief discussion of this group of mutants of unchanged pi'otein 

 composition might here be interjected. These nmtants are of considerable 

 interest since their presence clearly signifies that part of the RNA 

 carries information which is not related to the stnicture of the virus 

 coat protein. This is not a surprising fact, but had nevertheless not 

 been previously demonstrated, and it is important to evaluate the 

 validity of the evidence on which it is based. It would seem that the 

 failure to detect a difference between two proteins might well be due 

 to methodological causes. Thus, interchanges of two amino acids within 

 one peptide, and translocation of amide groups have not been ruled out 

 by the methods used, although the latter occurrence would probably 

 be revealed upon chromatographic separation of the peptides. In view 

 of these possibilities, the failure to detect an actual difference is a 

 definite possibility for any one strain protein, even though it is not 

 very probable. However, the prol)ability of this having occurred 7o 

 times in one laboratoiy, and about 30 times in another is as good as nil, 

 and thus the conclusion that different strains may share the same pro- 

 tein appears on statistical grounds firmly established. 



Besides this group, Wittmann also observed mutants showing 1 or 2 

 amino acid replacements, but none showing the many alterations 

 observed six times in our laboratoiy. His exchanges were also often the 

 same, but oddly enough his most frequent exchanges did not coincide 

 with those of our laboratory. Yet in certain instances (e.g., serine #138 

 — > phenylalanine) the same exchanges were detected in both laboratories, 

 an occurrence which would be statistically most improbable if it were 

 due only to chance. 



C. VIRULENCE AND ^■IAHILITY OF STRAINS 



The isolation of a number of new strains of TM\' calls forth Cjues- 

 tions about their comparative virulence and viability. As jireviously 

 mentioned, manv of tlie new strains are more harmful to the host than 



