SEASONAL POLYMORPHISM. 



129 



above, each causing two parallel lines to exist, and one furthermore 

 bears in ruind that the tone of the dark pattern varies from pale grey 

 to black and that of the ground colour from white to yellow (the 

 black and the yellow being, on broad lines, the result of dampness), 

 one can classify very fairly the principal variations of the first genera- 

 tion of /-". »«/>/, which at first sight seern so hopelessly complex as to 

 defy any such attempt. To do this in connection with all the 

 individual variations would, however, be quite outside the object of this 

 paper. To be interesting and useful it must be done on a large scale 

 and systematically, not only in the entire species, but sorting out the 

 characters which are generic and comparing them with their equivalents 

 in the other J'ieridi and even with broader groups. What I want to 

 do here is to try and grasp the characteristics of the geographical 

 races on the whole, such as can only be done and clearly seen by 

 comparing adequate series from each locality. It becomes obvious, by 

 so domg, that the vast niajority of the innumerable individual varia- 

 tions run through all or most of the races and can be ignored from 

 this point of view, so that European races are reduced to quite a small 

 number. As in the cases of litniiicia pidaeas, L., Aricia medon, Hufn., 

 Leiitonia sinctjiis, L., and other species I have studied in this way in 

 these columns, we find that the features of the races of F. vapi 

 consist simply in a series of grades along one single line of variation ; 

 as in A. medon, the main line bifurcates at its two ends by producing 

 variations which are certainly not successive, but collateral to each 

 other. The order in which the races fall in the most natural way is 

 the one which begins by those whose nervural streaks are most highly 

 developed and leads down to their minimum extent. The latter meet 

 and partly overlap the features of the summer generations, which take 

 up variation at this point, gradually reduce and abolish the nervural 

 pattern and tend to develop the transverse pattern alone, or nearly so. 



Grades in the extent of the dark pattern, taken on the whole, and 

 races of P. napi, L., in Europe, detectable in the first generation : — 



Grade I. : amcolor, Ruber, in Seitz's (jroHs-Schmett., p. 49 (1907), 

 is the name which has been given to " individuals in which the 

 yellowish ground-colour, especially on the forewing, is almost entirely 

 suppressed by the greater extension of the dark scaling." I have 

 noticed that in some regions of the Alps and, to my knowledge, more 

 precisely in the Austrian ones, this extreme form of bnjoniae is 

 frequent, whereas m others, such as the Maritime Alps, it never 

 occurs, and the females are, in a general way, much less heavily scaled 

 with dark. This observation seems to make it necessary that the 

 darkest forms of hri/oniae should be introduced as a grade in geo- 

 graphical variation. One may anticipate that the name of conculcm 

 will be extended to a race, although I lack the necessary material to 

 establish this now. 



Grade II. : This may be described as having broad nervural 

 streaks over the entire wing, separated from each other by narrow 

 spaces of clear ground-colour. As individual forms they are very 

 definite and characteristic, including all those usually known in a 

 general way under the name of bnjoniae. As races it is very restricted, 

 because, as a rule, the forms just mentioned do not predominate, but 

 are found amongst individuals belonging to the following less heavily 

 marked grades, and the average extent of the dark pattern falls well 



