132 I The Process of Evolution 



(homozygous for the common allele), meclionigra (heterozygous), 

 and bimacida (homozygous for the rare allele). The frequency of 

 the rare gene was .012' prior to 1928, .092 in 1939, and .111 in 1940. 

 After 1940 it dropped rapidly, leveled off around 1947, and since 

 then has fluctuated between .011 and .037. There seems to be little 

 doubt that these changes are due largely to fluctuating selection 

 pressures. Among other things, the gene is known to affect color 

 pattern, mating behavior, fertility of the males, and larval viability. 

 There has been some controversy over the possible role played by 

 drift in this situation; the degree to which random changes interact 

 with selection has not been determined. 



Dowdeswell, Fisher, Ford, and others have made a long series of 

 studies of the frequency distribution of spot number on the under- 

 side of the hind wings of the satyrine butterfly ManioJn jiirtina. 

 Over most of southern England the spot distributions are remark- 

 ably uniform, in spite of the great diversity of environments. How- 

 ever, the spot distribution in females was found to be sharply difiFer- 

 ent between populations in Devon and Cornwall. Furthermore, the 

 change was found to be extremely abrupt; indeed, in 1956 the border 

 between populations with the two kinds of spot distribution was 

 found to be a hedge which was not a barrier to the passage of indi- 

 vidual butterflies. There was no sign of a gradient between the two 

 types; if anything, the difference was greatest at the point of con- 

 tiguity. In 1957 the border between the two forms was found to be 

 3 miles east of its 1956 location, and the transition was more gradual. 

 The boundary itself was a strip some 150 yards wide which was 

 occupied by an intermediate population. Spot distributions also have 

 been studied extensively on a small archipelago, the Scilly Isles, off 

 the western tip of Cornwall. Numerous differences were found 

 among the islands, but each island tended to have a characteristic 

 population which remained the same from year to year. 



Spot distributions certainly are under strong selective control, 

 although it seems clear that the selective value is manifest in other 

 effects of the genes involved rather than in inconspicuous spot 

 changes. It is difficult to formulate any other explanation for the 

 sharp and mobile border between the Devon and Cornwall types 

 than the shift from one highly integrated gene complex to another. 

 The types seem highly successful, as they extend for a considerable 

 distance in either direction from the border (especially eastward) 

 and are relatively undisturbed by the heterogeneity of the habitats 

 they occupy. It does not seem likely that the environmental factors 

 responsible for the change in selection coefficients actually change 

 as abruptly as the spot frequencies; the reason for the sharp border 

 almost certainly lies in the genetics of Maniola. 



