260 T. FRANCIS, JR. 



appear, nevertheless. Each strain may exhibit individual characteristics, but 

 those from an epidemic season have great similarity. For a period of years 

 after the appearance of a major antigenic variant, there is a tendency to 

 divergence in the strains of ensuing epidemics. This appears to be a random 

 variation within the group rather than an orderly progression. The variations 

 of themselves are of general biological and immunological interest, but their 

 epidemiological significance must clearly be gained from study of the disease 

 in man. 



2. Studies with Human Sera 



a. Response to Infection. After the demonstration of the development of 

 specific neutralizing antibody to influenza virus in convalescent patients 

 (Francis and Magill, 1935), attention was directed to the specificity of the 

 response to infection with known strains and also to the antibody content 

 of serum from the general population. Andrewes and associates (1935) 

 observed that antibodies to swine influenza virus were present in the serum of 

 100 % of a group of human subjects over 15 years of age, but in none under 

 10 years, whereas antibody to the human WS strain was found in both 

 children and adults. Similar results were obtained with the PR8 and swine 

 strains in serum from the American population, and with WS, Melbourne, and 

 swine strains in Australian subjects (Francis and Magill, 1936; Shope, 1936; 

 Burnet and Lush, 1938). The suggestion was made that the swine antibodies 

 resulted from infection with virus of the 1918 pandemic, which had then 

 become established in swine (Laidlaw, 1935; Shope, 1936). On the other hand, 

 it had been well established that repeated exposures of experimental animals 

 to human strains gave rise to antibodies to swine virus. In St. Helena, where 

 the 1918 epidemic was said not to have occurred, sera of adults who had never 

 left the island contained, in 1935, little antibody to swine virus. When an 

 epidemic of a WS-like influenza infection occurred in 1936, antibodies to 

 WS strain promptly rose and, in addition, swine antibodies also appeared, 

 even in persons who showed none previously (Stuart-Harris et al., 1938). This 

 seemed to indicate a response induced by common antigens in adults with 

 broad serological reactivity. Rickard and associates (1945) noted, moreover, 

 the development of swine antibodies in infants recovering from presumably 

 their first infection with influenza A, although Hare and Riehm (1941) had 

 pointed out that it was more likely after the first decade of life. 



As knowledge of strain variation grew, there was increased attention to the 

 use of multiple strains in the measurement of antibody response to infection 

 and vaccination. It was recognized that the response of children was more 

 strain-specific than that of adults. In 1947, however, vaccination of adults 

 with the PR8 and Weiss strains enhanced titers to those strains but not to 

 the oncoming A-prime strains, but patients recovering from infection with 



