1256 
the imaginal colour-pattern arises rather suddenly in the last days 
of the chrysalid period, but this pattern is preceded by a different 
one, which passes only partially into the definite pattern, and for 
the rest is obliterated by it. 
This preliminary, original or primary pattern is distinguished from 
the imaginal one by its greater regularity, simplicity and at the 
same time completeness. Moreover it is identical in different nearly 
related species. 
2. On the wing sheaths of those pupae which are not simply of 
a brown or black bue, but show an elaborate colour-pattern, these 
markings form a design, which corresponds to the above mentioned 
primitive pattern, and consequently is similar in a number of related 
species. 
53. A comparison of the imaginal forms of the said related species 
shows, that the principal features of similarity from which this 
relationship is deduced, are especially furnished by those elements 
of the primitive pattern, which have been retained in the definite 
pattern. 
The second conclusion is, that the above mentioned simplicity, regu- 
larity and completeness of the primitive pattern depend on its con- 
nection with the course of the wing-veins, the markings either fol- 
lowing these veins, or being arranged in the interspaces between 
them, without transgressing their boundaries. As a proof of this 
connection the fact may be mentioned, that in many cases the course 
of veins which become obliterated during the development of the 
wings, still remains visible in the colour-pattern. For it must be 
noted that the system of wing-veins likewise undergoes a process 
of metamorphosis: the definite wing-skeleton arises from the modi- 
fication of a provisional and more primitive one, which shows 
smaller differences between fore- and hindwings, and greater similarity 
to a general ground-plan, holding good for all different groups of 
Lepidoptera. In those families, which for different reasons are con- 
sidered the most primitive, the imaginal system of wing-veins shows 
the least degree of deviation from this general plan, and for the 
same reason the greatest similarity to the distribution of wing-veins 
in other insect-orders nearly related to the Lepidoptera, such as the 
Trichoptera. 
The third conclusion is connected with the fact that in the bulk 
of imaginal patterns the markings on the upper side differ from 
those on the underside, and those of the forewing from those of 
the hind one (especially on the upper surface). This fact may be 
contrasted with the similarity existing, as remarked before, between 
