On the Seasonal Dimorphism of Butterflies. 5 7 



larged until, finally, in the perfect Levana it unites 

 with another black triangle proceeding from the 

 front of the band, and thus becomes fused into a 

 black bar. The white band of Prorsa and the black 

 band of Levana by no means correspond in position; 

 in Prorsa quite a new pattern appears, which does 

 not originate by a simple colour replacement of the 

 Levana marking. In the present case, therefore, 

 there is no doubt that the new form is not produced 

 simply because a certain pigment (black) is formed 

 in larger quantities, but because its mode of dis- 

 tribution is at the same time different, white appear- 

 ing in some instances where black formerly existed, 

 whilst in other cases the black remains. Whoever 

 compares Prorsa with Levana will not fail to be 

 struck with the remarkable change of marking pro- 

 duced by the direct action of external conditions. 

 The numerous intermediate forms which can be 

 produced artificially appear to me to furnish a fur- 

 ther proof of the gradual character of the trans- 

 formation. Ancestral intermediate forms can only 

 occur where they have once had a former exis- 

 tence in the phyletic series. Reversion may only 

 take place completely in some particular characters, 

 whilst in others the new form remains constant 

 this is in fact the ordinary form of reversion, and 

 in this manner a mixture of characters might ap- 

 pear which never existed as a phyletic stage ; but 

 particular characters could certainly never appear 

 unless they were normal to the species at some 



