GENEALOGY OF INSECTS. 
297 
V. Hemiptera . — The only clew to the origin of this well circumscribed 
order is the fact that in the Physapoda (Thrips) and the Mallophaga 
the mandibles are free and adapted for biting. This would indicate that 
the entire group was derived from ancestors allied possibly to the Phy- 
loptera. The Mallophaga are by different authors referred to the Or- 
thoptera and Neuroptera, but the development of the bird-lice as worked 
out by Melnikow fully proves that in the form of the egg, the mode of de- 
velopment, and general form of the embryo, the Parasita and Mallophaga 
travel along the same developmental path until just before hatching, 
when in Mallophaga the jaws remain free, while in the Parasita they be- 
come farther modified and form a sucking beak. 
There is a possibility that the Hemiptera may have descended from in- 
sects remotely allied to the Pseudoueuroptera: perhaps forms resembling 
the Psocid;e; at least this family, the wingless forms of which superfi- 
cially resemble the Mallophaga, gives hints which may throw light on 
the origin of the Hemiptera. They are evidently the offshoot of a stock 
which had an incomplete metamorphosis, or they may have descended 
directly from a modified Oampodea-like ancestral form. 
VI. Neuroptera . — The members of this order are, excepting perhaps 
the Hemiptei’a, the most modern and least composite or synthetic forms 
that we have yet met with in our ascent up the insect series from the 
Thysanura. Moreover, in them for the first time do we meet with worm- 
like, cylindrical-bodied larvae, or what we have called eruciform larvae .' 45 
These larvae are secondary forms, derived, as Fritz Muller has in a gen- 
eral way suggested, from those larvae which have an incomplete metamor- 
phosis. By what line of descent, however, the lowest group of Neuro- 
ptera, viz., the Sialidae, arose, it would be difficult to say. The earliest 
winged insects were probably terrestrial; the aquatic larval forms of the 
Sialidae are evidently derivations from Gampodea like terrestrial larvae. 
But how the perfect metamorphosis with the quiescent pupa of the Neu- 
roptera was brought about, is indeed a problem. It is evident, however, 
that the eruciform larva is a derivation from a Thysanuriform 148 type, 
first stated by Fritz Mtiller. 
It seems to us that a consideration of the diverse larval forms which 
occur in the present order, throws some light on the origin of a com- 
plete metamorphosis in insects in general. In the Sialidae, as the larva 
of Corydalus, or Semblis, we have a Campodea-form provided with gills, 
and with the mouth-parts adapted for seizing and biting its prey. The 
terrestrial larvae of the Hemerobiidae are evidently modifications of the 
Sialid larval form ; the differences of structure in them, such as the long, 
l4fi See "Our Common Insects,” p. 175, 1873. Also the American Naturalist, vol. V, Sept. 1871. 
146 Wo liavo in the writings just quoted tailed the second class of larvte Leptifonn, but tlio term 
Thysanuriform, or Braner’s expression Campodea-foriu, is preferable. The Campodoa or primitive 
Hexnpodous form is evidently a derivative form, which points back to a common six-footed ancestor of 
all Trachosita, to which the term Lepti/oi'm may be applied. 
