944 On the Genealogy of the Insects. [September, 
Though the earliest beetle known is a Carboniferous weevil, 
yet we imagine the Coleopterous type became established in De- 
vonian or Silurian times, when there may have existed the proto- 
types of the earwigs and beetles; for the two types may have 
branched off from some Thysanuran form. On the other hand, 
the primitive Coleopterous larva may have sprung from some 
metabolous Neuropterous form. The larva of Gyrinus has a 
striking resemblance to that of Corydalis and other Sialidæ, so 
much so that a terrestrial Carabidous form most probably was of 
Neuropterous origin, as indicated in our diagram. A 
Origin of the Diptera, Lepidoptera and Hymenoptera —The Eu 
glossata probably had a common origin in the first place from 
the metabolic Neuroptera. The Lepidoptera probably originated 
from the same group from which the Panorpidz and Trichopt 
branched off, and we agree with the opinion of H. Müller, who 
maintains that the Lepidoptera and Trichoptera “ proceed tome | 
common stock,” though we should suppose that the Panorpid@ 
in their larval stage represented forms like the ancestral cater 
pillar. a 
The adult structure and larval forms of the Diptera show that 
they originated from nearly the same stock as the moths. pa 
most perfectly developed Dipterous larvæ are those of the Culi- 
cidæ and Tipulidæ ; these were probably the primitive forms; th 
other Dipterous larvæ, notably the larval Muscidæ or magg 
are degradational forms, and the lower Diptera appear to = 
been degraded or degenerate forms. a 
The case is different with the Hymenoptera. The saw-fly a 23 
represent apparently the primitive larval form; and from their 
semblance to caterpillars and Panorpid larve, show that the Hy- 
menoptera and Lepidoptera may have had a common ' 
The footless larvæ of the parasitic Hymenoptera are cai BH 
with their parasitic mode of life, and the similar forms nae 
larval wasps and bees show that from disuse their nowt 
and legs became aborted, and the immobile larve became * r 
and thick-bodied. Hence such larve should be regaran. 
secondary, adaptive larval types. The high degree of 
tion of the bees’ mouth-parts, their concentrated bodies and a 
mented thorax, with other characters, show that they eo 
highest, most specialized and modern of all insects. 
Note. —It should be borne in mind that the embryo 
a 
