377 



Summing up the conclusions to which my researches have 

 led me, I should wish to mention not only those relating to the 

 topics just laid before you, but also some general laws, which, 

 according to mv view, may be relied on in ph^dogenetic 

 speculations. 



1. During metamorphosis, the colour-pattern on the butterfl}'- 

 wings passes through a series of stages, representing a phylogenetic 

 sequence, and therefore may be of use in judging of the ancestral 

 affinities of allied species. 



2. One of these stages is found depicted on the wing-sheath of 

 the chrysalis : another one, very similar to the first, but probably 

 still more primitive instead of more advanced, makes its appear- 

 ance during pupal life on the wings of the future imago. 



3. The components of this primitive pattern are strictly 

 confined to the internervural spaces, and show a great equality 

 among each other, thereby producing a similarity in design 

 between fore- and hindwings, upper- and underside. 



4. Some of the elements of this primitive pattern pass directly 

 into the definite one, forming some of those features of the 

 latter which are common to the species of one genus, to the 

 genera of one family, and probably even to families of one 

 order. 



5. These ontogenetic results give a fresh confirmation of 

 the principles derived from the comparative investigation of 

 imaginai forms ; telling us that similarity of fore- and hindwing, 

 of upper- and underside, and repetition of the same elements 

 of design in succeeding internervural spaces, may be considered 

 as being more primitive than the contrary, or at least as an 

 atavistic return to the ancestral condition. 



6. The special hues or shades of the colours of the pattern 

 are not of primary importance, but may be regarded as secondary 

 modifications in the constitution of one or a few fundamental 

 colouring substances, if they are not merely caused by the 

 physical structure of the scales. By these modifications the 

 primary pattern may be more or less disturbed, obscured, and 



-even rendered invisible. Monochromatism (self-colour), f.i., is 

 always to be regarded as a final specialisation ; the simplicity of 

 such a unicolorous form has nothing to do with primitiveness. 



7. In tlic dcvL'liijjnicnt of the (dift'iTcntiated) imaginai pattern 

 4S 



