VARIATION IN VIRULENCE 245 



111 Australia and in France, myxomatosis is predominantly a summer 

 mosquito-borne disease; in Britain, there is no marked seasonal incidence and 

 the rabbit flea appears to be the important vector. The virus strains originally 

 introduced in Australia and Europe differed in their passage histories and in 

 the symptomatology produced in Oryctolagus rabbits, but both almost 

 invariably produced lethal infections. 



A single introduction of virus was made in Europe, whereas in Australia 

 there are annual inoculation campaigns with the highly virulent virus, but in 

 both continents the virus has established itself enzootically. Samples of virus 

 from different parts of Australia and from Europe have been collected each 

 year, inoculated into test rabbits, and grouped into one of five categories 

 according to the mortality rates or the mean survival times seen in laboratory 

 rabbits (Fenner, 1958b). 



In spite of the annual reintroduction into Australia of highly virulent 

 virus (causing a mortality rate of over 99 %) the majority of strains recovered 

 from natural cases have been attenuated to a slight or high degree. Within a 

 year of the first introduction of the virus, strains with a 90 % mortality rate 

 had appeared. These have remained dominant ever since, but still more 

 attenuated strains (causing mortalities sometimes as low as 20 %) have been 

 recovered since 1955. In Europe the fully virulent strain has persisted longer 

 and on a wider scale, but here, also, attenuated strains are becoming common. 

 When viruses are passed repeatedly in a particular host in the laboratory 

 they often become more virulent for that host, although this may be accom- 

 panied by attenuation for some other (perhaps the natural) host. Attenuation 

 for the passage host is virtually unknown. Yet in both Australia and Europe 

 (and in many widely separated areas in each continent) the tendency has 

 been toward a moderate degree of attenuation. 



The explanation lies in the method of transfer of virus from one host to 

 another. In the laboratory, the usual procedure is to select the animal show- 

 ing the first signs of infection and to use material from this animal for the 

 next passage; this procedure naturally tends to select variants with maximum 

 virulence. In nature, however, the rabbit most likely to act as a source of 

 infection for others is the one which offers large virus-rich skin lesions as 

 feeding grounds for mosquitoes for the greatest length of time — that is, a 

 rabbit infected by a strain of virus which does not terminate the infectivity 

 of its host for others by killing it rapidly. This survival advantage of less 

 virulent strains is presumably greatest when spread of the disease is least 

 efficient — that is, during winter. In short, selection in the laboratory is for 

 rapid multiplication and high final titer; in nature, it is for these properties 

 and also for the property of persistence of the high final titer. 



The results of titration of the superficial cells of the skin of rabbits infected 

 with several strains of virus of different virulence showed that all naturally 



