Dec. i8 97 .] Grote : Classification of Lepidoptera. 153 
of the three primary veins. Its progress is not uniform, but is evi¬ 
denced in different ways. The comparative completion of this effort 
affords a particular gauge of the standing of the form. The second 
direction occurs sporadically in* very different groups. It consists 
in an absorption of the branches of the Radius, so that their normal 
number is diminished. It is probably reminiscent of what has taken 
place on the hind wings, as we see from Hepialus. When we apply 
our knowledge of these two tests of specialization to the Day-Butter¬ 
flies, we find that the second, or sporadic direction, occurs in the 
Parnassidae, Pieridae, Lycaenidae, thus independently in otherwise 
very different groups. It is thus a secondary character and we 
find it again in a group so dissimilar to the Day Butterflies as the 
Saturniades, while it is not indicated in the Hawk Moths. The first or 
general direction of specialization we find indicated by most Lepidoptera, 
in some of its stages. It is a fundamental movement and has probably 
a mechanical cause. The Pieridae unite the two directions in a palpa¬ 
ble manner, more strongly so than the Lycaenidae, which exhibit, in the 
Theclinae, the second direction very completely. In the Pierinae 
(. Mancipium , Pier is, etc.) the first direction is shown by the transfer of 
vein IVi, the upper branch of the Media, to the Radius. This state of 
affairs we find only again so strongly marked in Nemeobius. In the 
four-footed Butterflies the first direction, or suppression of the Media, 
asserts itself in the total degeneration of the crossvein; while the two 
upper branches of the Media are pulled towards the Radius, the cell 
opens completely. Thus the Media, as a system, ceases to exist. But 
in the Nymphalidae, the upper branch of the Media does not become 
completely absorbed by the Radius, as in the Pieridae, in which latter 
the cell is never so completely opened as in the former family. Again 
the second direction is not taken up at all by the brush-footed butter¬ 
flies, the Radius remaining generalized, five-branched. Judging 
from the condition of the hind wings especially, the Agapetinae 
and Limnadidae are less specialized than the Nymphalidae. The 
Libytheidae overlap the more generalized Meadow-Browns. The neuia- 
tion of the Libytheidae is almost repeated by the Nemeobiidae, which 
latter retain no essential wing characters of the Riodinidae (Erycinidae) 
or Lycaenidae. I tried to explain its position on the Lycaenid branch 
by the view that the evolution of the neuration has taken a parallel 
direction to that of the Pieridae and the four-footed Butterflies. On 
the neuration by itself we must, and I now do, exclude Nemeobius 
from the Lycaenid branch. Its junction with this branch must remain 
