394: SYSTEM OF INSECTS. 
to steer clear of being led by any system, but, with the 
exception of the Suborder, which I assume, to resolve it 
into natural groups gradually decreasing in value, or 
tending to the lowest term, which appear all of them to 
have been considered as such by preceding Entomolo- 
gists. The four Zribes into which the two subsections 
Geodephagena and Hydradephagena appear resolvable, 
are not only distinguished by the characters of the per- 
fect insect, but likewise by those of their larvae, which 
are constructed on four distinct types; those of the Gy- 
ronechina being the most perfectly Chilopodimorphous 
of the whole, and those of the Hunechina the least so*. 
The former appear rather to form an osculant tribe, or 
one without the circle, than one within it; and to be go- 
ing off towards another section, including Hydrophilus, 
Spheridium, &c. I must observe, that between Dytiscus 
and Hydrophilus there is a striking agreement both in 
their form and habits in their larvee?, and even in seve- 
ral characters in the perfect insect; so as in many re- 
spects to generate a doubt whether they ought not to 
enter the same circle and to follow each other. Yet the 
change of habits in the latter, which from a carnivorous 
larva becomes a herbivorous beetle; the consequent 
change of structure in their oral organs, their antennee, 
and other striking differences; and the evident interven- 
tion of the Gyronechina and some other osculant tribes 
between the two, forbid their union in one and the same 
circle. 
* Vor. HI. p. 167—. I formerly hinted (Ibid. p. 164.) that the 
larva of Cicindela L. may be regarded as Araneidiform: this is fur- 
ther confirmed by its having eight eyes, (and not six,) as 1 have since 
discovered, and by the aspect of its large head and prothorax. The 
other larvae of the Adephagana have twelve eyes. 
