VARIATION IN VIRULENCE 239 



Francis, 1951). In some cases, there is clearly an increase in the final maximum 

 titer of virus in the lungs (Andrewes and Smith, 1937; Anderson and Burnet, 

 1947; Hirst, 1947b; Friedewald and Hook, 1948; Wang, 1948; Sugg, 1949, 

 1950; Davenport and Francis, 1951; Briody and Cassel, 1955; Ledinko and 

 Perry, 1955; Ledinko, 1956), and in some cases, not (Hirst, 1947b; Davenport, 

 1952); and in some cases the final yield alters only in that it has a higher 

 infectivity for the allantois in relation to its hemagglutinin content (Ledinko 

 and Perry, 1955) or for mice in relation to its allantoic infectivity (Sugg, 

 1949; Ginsberg, 1953b). In many cases, too, there was evidence that the 

 final yields of infective virus were high in the first one or two passages, 

 passed through a minimum, and then rose again as the virus became fully 

 adapted (Anderson and Burnet, 1947; Hirst, 1947b; Friedewald and Hook, 

 1948; Ledinko and Perry, 1955; Ledinko, 1956); this effect was seen even 

 when each passage employed the same dose of virus in terms of hemagglutinin 

 (Ledinko, 1956) and, at least in one instance, was associated with marked 

 formation of incomplete virus in the first few passages but not in later 

 passages (Ledinko and Perry, 1955). All these conflicting reports show that 

 unadapted strains are not always unadapted for the same reason. Some show 

 low multiplication rate, some produce low final yields, some mainly produce 

 incomplete virus, and some (probably most) show a combination of these 

 defects. 



Part of the complexity of the situation may be a manifestation of the 

 complexity of the mouse lung as a host for virus multiplication. First, 

 although the system can for many purposes be regarded as a simple sheet of 

 susceptible cells capable of adsorbing huge quantities of inoculated virus 

 (Fazekas de St. Groth, 1950), there is definite evidence, even in the case of 

 fully adapted strains, that virus multiplication fully exploits the system only 

 after very large inocula. Thus, intranasal instillation of saline, a day or two 

 after inoculation of virus, greatly increases the size of the lesions, so that mice 

 die which would otherwise survive (Straub, 1940; Taylor, 1941); also, the fact 

 that circumscribed lesions are produced by sublethal inocula of adapted 

 strains implies that the system is not normally fully exploited. The second 

 major complication to the interpretation of events is, of course, that the effect 

 of antibody production can be demonstrated from about the third day 

 (Donnelly, 1951). The effect of these two factors is that part of the change to 

 the adapted state could, on occasion, represent an increase in capacity to 

 spread through the system or a decrease in speed of evoking antibody 

 formation; either of these changes could be manifest simply as an increase in 

 final yield, without necessarily any alteration in infectivity for mouse lung 

 or multiplication rate. Indeed, there is some evidence that unadaptedness 

 to mouse lung may represent a very complex defect, for the process of 

 adaptation is markedly hastened either by keeping the inoculated mice at 



