SEASONAL DIMORPHISM OF LEPIDOPTERA. 205 



be the case, that hotlt forms will have protective significance ; 

 but it is not so easy to prove, on account of the uncertainty in 

 estimating the biological value of many of the details of mark- 

 ings which here come into consideration, and at all events 

 observations are required on the spot. 



I have cited Vanessa prorsa-levana as a hypothetic example 

 of an adaptive seasonal dimorphism of the perfect insect, and I 

 have relied on the remarkable similarity which the upper side of 

 the prorsa-iovm, with its black provided with a white band, has 

 to Limenitis sihylla and Camilla. But I do not ignore the diffi- 

 culties which stand in the way of a proof that this is a case of 

 mimicry. We do not yet know whether these species of Limenitis 

 enjoy an immunity, or whether they are sought after by birds, 

 or w^ere sought after at an earlier period. But if it could be 

 proved that they are privileged, and that prorsa obtained pro- 

 tection by its similarity to them, it still has to be explained, how 

 the levana-iorm has an adaptive value, and, indeed, on its upper 

 side, which generally possesses no adaptive value among the 

 butterflies. To be sure, I was able to show years ago, that the 

 dark upper side of the female blues actually conferred protection, 

 as they lay their eggs with outspread wings, and are then con- 

 siderably less conspicuous than the blue males are, when they sit 

 with outspread wings. But we do not know the habits of the 

 levana-fonn. sufficiently accurately, and if we did know them, it 

 would still be always sufficiently uncertain whether we could 

 regard the undoubted similarity of the upper surface to the dead 

 leaves of the spring wood as protective. 



But, as it appears to me, we cannot well conceive that 

 seasonal dimorphism can arise, unless both seasonal forms have 

 an adaptive value. Granted the one alone is adaptive (here, for 

 example, the mimetic prorsa-fovm), it would therefore be con- 

 ceived as arising by selection, i.e. the determinants (anlagen)* 

 of their wing-colouring would have been changed little by little 

 from the Zera/m-determinants to the p?'orsa-determinants. Now 

 if, as my theory of heredity assumes, many primary constituents 

 of the entire individual (the **ids") are present close to one 

 another in each germ, then, although their selection need not 

 necessarily follow the same course, one cannot see what should 

 prevent the collective ids, little by httle in the course of gener- 

 ations, from containing only j9ror,<?6i-determinants, and from 

 the lev ana -deteTUiinsints being supplanted. Now generally, if 

 the ^rors<x-form has an advantage over the levana-ioi'm even 

 only in the summer, all individuals which are not true prorsa 



''' Anlagen is here used as synonymous with determmant. In reference 

 to the difficulty attending the proper rendering of anlagen in its various 

 meanings, cf. Prof. Parker's note to his translation of the 'Germ-Plasm' 

 (Cont. Science Series). He frequently renders it " primary constituent." 

 Later on Prof. Weismann uses it as synonymous with " id." — Translator. 



