THE MUTATED GENE 249 



conditions within the wing as a unit. Other typical relations 

 of the pattern to definite parts of the wing, such as the veins, are 

 known, and these special structures also must be a part of what 

 have been called the systemic conditions for pattern formation. 

 Since a general plan of pattern is found in the different larger 

 groups, different systemic conditions are supposed to be present 

 in these groups. Finally, since similar or even identical patterns 

 may be combined with different chemical or physical structures 

 of an area — or, vice versa, the same chemical behavior over 

 different areas — pattern formation and chemical specificity, e.g., 

 guanin, melanin, carotin, as pigments, must be caused by inde- 

 pendent processes of determination. 



To these general points of view two more special ones of general 

 importance have been added. 



5. Forms are known in which different generations (alternate 

 spring-autumn or wet-dry seasons) exhibit different patterns, 

 which sometimes, in tropical forms, may be so different that they 

 can hardly be conceived of as simple quantitative shifts within a 

 pattern; they actually show a different plan of pattern. Exter- 

 nal conditions acting upon the same genotype may then produce 

 extraordinary changes of the type of pattern. (They can be pro- 

 duced also experimentally; see the classic case of Vanessa levana- 

 prorsa). In view of what precedes, it may be inferred that the 

 embryonic causation of such very different patterns may be due 

 to relatively simple differences in the causative agents. 



6. Cases are known of extreme polymorphism within a species, 

 e.g., the many different forms of females of Papilio dardanus, 

 mimicking the patterns of very different families. In such cases, 

 it is known that a few Mendelian genes are responsible for the 

 entire array of patterns. It follows that extraordinary differ- 

 ences of pattern may be produced by the simple differences of 

 reactions due to a mutant gene. In the same case, the different 

 patterns are further controlled by sex; they are limited to one 

 sex. The differences in development of the two sexes therefore 

 may permit or prevent the realization of different types of pat- 

 tern. This shows, again, that a relatively simple alternative 

 may lead to far-reaching consequences regarding the pattern. 



To this general review of the situation we may add a list— in 

 no way complete — of known effects of mutant genes upon pattern 

 in Lepidoptera: 



