266 T. FRANCIS, JR. 



strain, which was originally closely related. Little alteration of the second 

 strain was noted with adaptation to the mouse lung. Sugg (1949) also noted 

 distinct serological variation in the Cam strain adapted to mice; this difference 

 was more that of an increased antigenic and antibody combining potency of 

 the mouse line than any evidence of a major antigenic deviation. A second 

 passage series carried out with the same strain became pathogenic for the 

 mouse lung without any evidence of serological alteration. No antigenic 

 variation was observed with certain other strains systematically studied in 

 the process of adaptation to mice (Wang, 1948; Davenport and Francis, 1951). 

 On the other hand, in 1947 a serological variant of the PR8 mouse line was 

 detected in virus which had been transferred to and maintained in tissue 

 culture. The variant was still clearly PR8 in character but differed serologic- 

 ally and was sufficiently distinct so that mutual cross-protection did not 

 occur in vaccinated mice. Extended studies have shown, however, that 

 serological differences are clearly detectable between strains newly isolated in 

 eggs or other hosts from the same outbreak, so that the antigenic variations 

 observed in the earlier antigenic studies cannot be attributed primarily to 

 animal adaptation but seem properly to represent inherent differences in the 

 strains themselves. In 1952 Hirst considered the modification he noted 

 earlier during adaptation to mice to be of little moment. In fact, the mouse 

 lines often appear to display the proper antigenic structure of a strain more 

 completely. Mulder et at. (1956) have pointed to the fact that numerous 

 antigens may be hidden in egg passages which show up in the mouse lines. 

 The behavior of the same initial strain of virus in a variety of hosts was 

 studied by Jensen et at. (1957). Lines of virus from a single human gargling 

 were initiated and maintained in eggs, ferrets, mice, hamsters, and tissue 

 cultures of allantoic membrane. Alterations were looked for at intervals in 

 the course of adaptation by using the different lines as antigens for production 

 of antiserum in the homologous and other species of animals. A variety of 

 variations was observed but the most prominent anomalies were in the 

 tissue culture and mouse or hamster lines. The tissue culture line was the 

 most line-specific serologically, combined in hemagglutination-inhibition 

 tests least readily with antibody even to homologous serum, was the poorest 

 antigen, and exhibited the least ability to absorb antibody from pooled serum. 

 The mouse and hamster lines were the most efficient in these respects. The 

 possible antigenic variations which were observed were unstable and passage 

 of the tissue culture line in mice or of the mouse line in eggs resulted in restora- 

 tion to the original basic pattern. Review of the results strongly suggests that 

 the differences are related to the nutritional adequacy of the medium in 

 which the line was propagated. The behavior of the tissue culture and egg 

 strains may well be expression of a starved or poorly developed virus, while 

 the mouse and hamster lines appear well fed with much more complete 



