1562 
should have been derived from Trichoptera, as he cannot imagine 
that a group of insects possessing polypode land-living larvae could 
possibly originate from forms with hexapode aquatile larvae. 
He therefore inclines to the view, tliat we may trace their origin 
from the Panorpata, the larvae of which are caterpillarlike, and in 
fact only differ from those of Lepidoptera by the coalescence of 
two preanal segments. 
In any case, the order of Panorpata is regarded as nearly related 
to both the above mentioned ones. And among this order we find 
the genus Bittacus, in which the wings likewise bear internervural 
striae of coloured matter, but where moreover each one is provided 
with a transverse nervure, or in other words, in which in compa- 
rison with the two other orders, the number of the transverse 
nervural connections being still relatively large, each of these con- 
necting veins is provided with smoke-coloured margins. 
Supposing these veins to disappear and the coloured seams to 
remain, a pattern of transverse striae at regular intervals would be 
the result, i.e. the primitive pattern of Lepidoptera and Trichoptera. 
And this is in fact the real course of the evolution in Panorpata, 
for there in many eases we find transverse veins reduced to vestiges 
while the coloured bands have been preserved in undiminished 
intensity. Oceasionally on a Dittacus-wing I found the nervural 
connection present on one side, while on the other the correspond- 
ing vein had totally disappeared. 
So the three above mentioned orders provide us with a succession of 
stages through which the primitive colour-pattern of Lepidoptera has pro- 
bably passed in its evolution, and which may be formulated as follows: 
1. The membranous wings are provided with several transverse 
veins, lying between smoke-coloured margins. 
2. The transverse connecting veins disappear, the margins persist 
and so form transverse colour-markings, the hairs on the wing- 
membranes acquire corresponding markings. 
3. The hairs become transformed into scales, in which the colouring 
matter becomes accumulated, while the membrane loses it, being 
shut off from the light by the broadening of the scales. 
Supposing this to have been the real course of the evolution, we 
are led to ascribe the ornamentation of the anterior margin in the 
same Yhanner to a greater or less number of transverse veins. 
In fact, although the majority of Panorpata only possess a 
small number of these veins, there are genera to be found amongst 
them in which the number of transverse veins is more considerable 
e.g. Merope (Comp. Hanpuirscn, pl. V, fig. 18). 
