1264 
Though the most primitive type, that which is preserved to us 
in forms like Charagia mirabilis ©, Phassus giganteus and Dalaca 
assa, is not met with in the genus Pielus, yet stages’ corresponding 
to Ch. ramsayi y and Ch. splendens, can be recognized among the 
different degrees of modification belonging to it, which culminate in 
a form like Pielus hyalinatus var. barcas, where the above mentioned 
white bar stands out against an almost unicolourous dark back- 
ground. And even this characteristic bar is seen to occur as well 
in some forms of the New-Zealand genus Porina. 
The result, to which this comparison of the Hepialid genera leads 
us, is the independence of the colour-pattern in its original form 
as well as in its various modifications, of generic differences. 
For this fact no other explanation can be found. in my opinion, 
than the assumption, that phylogenetically the colour-pattern in all 
its manifestations is older than the division of the Hepialid family 
into genera, or, otherwise expressed, that the whole complex of 
modifications of the wing-design belonged to the inherent properties 
of the ancestry of the present Hepialids. 
Supposing this conclusion to be right, we are forced to assume 
that in all species of the different genera the hereditary factors for 
all these colour-modifications are present. Otherwise we should be 
obliged to believe, that in each genus, independently of the remain- 
ing ones, one and the same series of modifications of the common 
original pattern had developed, which means that as many cases of 
parallel evolution had occurred as genera of Hepialids exist. 
The latter supposition cannot be said to appear very probable, 
even for the genera of one family, but it becomes decidedly absurd 
as soon as we realise, that also in other families of Lepidoptera, 
akin to the Hepialids, the same types of modification, derivable 
from an identical groundform, are met with. 
If, however, our first assumption should prove justified, we cannot 
avoid the conclusion that also those specific forms in which the 
original pattern demonstrates itself with the least amount of secon- 
dary changes (Charagia mirabilis ©, Phassus giganteus, Dalaca assa, 
Porina characterifera, Oxyeanus fusconebulosus, Phassodes nausori) 
cannot for that reason be considered really to possess a more ori- 
ginal constitution than the farthest advanced modifications. The 
predisposition for these modifications must be equally present in 
their hereditary outfit, as in that of the other forms, whose outer 
aspect attains the highest degree of specialisation, or at least this 
hereditary tendency must originally have belonged to it. In cases of 
