32 I'llOCEIiUlXGS OF THE 



The fjicts set I'orfcli in the above table are best explained by 

 the strong endurance ot" a hereditary feature gradually weakened 

 as the Alri<'an invaders became further and further separated 

 fronj the original stock where the feature is prominent in all the 

 species 1 have examined except 2\ neptunia. I do not myself 

 doubt that the weakening has been brought about by the selection 

 in mercedoiiia and mon/eni of individuals which approached most 

 nearly to the pattern of the indigenous African Danaines of the 

 genus Ainaaris; but from tlie present standpoint I am only 

 concerned to ])oint out that the existence side by side in the same 

 locality of individuals with a minute trace of the mark and 

 individuals with no trace at all, and the comparison of these with 

 the individuals of allied forms in oti)er localities, cannot be 

 reasonably explained except on the supposition that the degree 

 of development of the feature is hereditary. And the different 

 degrees distinguished in the table on p- 31 are veri/ small 

 variations, 



b. Geographical Changes in Mimetic Species corresponding ivith a 

 Change of Model. — Under this head a large body of evidence is 

 available in support of the conclusion that mimetic patterns arose 

 by successive stages, and not suddenly as Professor Punnett 

 maintains — evidence which at the same time is available for the 

 general conclusion that small variations are hereditary. The 

 limits of time and space prevent anything more than the dis- 

 cussion of one good example from Africa and another from 

 S. America. 



c. The Mimetic Females of the African Charaxes etheocles. — 

 A wonderful example of geographical change in a single species 

 is seen in the mimetic females of Chara.ves etheocles, which in 

 various parts of Africa mimic the females and sometimes the 

 males of the larger species of the same genus. Among these local 

 changes are some which are quite as small as the steps by which 

 I believe that an elaborate mimetic pattern has been built up, 

 aud it is as certain that these small variations ai'e hereditary as 

 that any geographical race is permanent and fixed. It might 

 indeed be argued by one who has not compared tlie specimens, 

 that each mimetic female foi-m in each locality had developed 

 from the same ancestral jiattern. The following examples, to 

 which others might be added, aiford evidence that the female 

 forms are not thus evolved separately but by the modilication of 

 those in adjacent areas. 



The female form viola, of the West Coast, possesses a broad 

 orange bar beyond the middle of the fore wing and a still broader 

 white one shot with blue beyond the middle of the hind wing. 

 This form is believed by Dr. Karl Jordan to be a rough niimic of 

 Charaxes epijasias. In parts of British East Africa and of Uganda 

 the model is the male of Charaxes ansorgei and in correspondence 

 with its i)atteru the bars of both wings are, in the form kirlci. 



