594 MODERN CLASSIFICATION OF INSECTS. 
Lithosiide) ; but these stirpes, being established upon the larve, are 
evidently more numerous than proposed by Dr. Horsfield. Latreille, 
after separating the exotic genus Erebus (in which the wings are 
always extended and horizontal, and the last joint of the palpi long 
and naked, and which comprises some of the most gigantic insects in 
the order), proposes to divide the remainder into two extensive and 
parallel series. In the first, the larvae are geometrical in their mode 
of progression, some of which have 16 feet, but with the 2 or 4 
anterior ventral feet shorter than the others, and the others have only 
12 feet (fig. 108. 9. larva; 10. pupa of Plusia Gamma). The second 
series comprises the genera Calyptra, Xylina, Cucullia, &c., all of 
which have 16 feet, the anterior ventral ones being of the ordinary 
size, and their progression is rectigrade. Chrysoptera concha 
(Fischer, Ent. Russ. Lepid. i. iv.) in the former series, and Erastria 
in the latter, appear equally to lead to the Pyralides. Catocala, 
Ophiusa, and Brepha, on the other hand, appear most nearly allied to 
Erebus. 
One of the chief difficulties connected with the arrangement of the 
order already alluded to in p. 361., is especially evident in the present 
family. In the Sphingide and Geometride, for instance, we find the 
larva state affording the best characteristics of the families; but here 
the case is quite different. Plusia is in effect as truly a good type of 
the family as Polia, Miselia, Acronycta, or Agrotis, and yet the larve 
of all these genera are strikingly distinct, some being Geometrideous, 
others Arctiideous, and others Noctuideous, if we consider the latter 
to be characterised by a naked fleshy larva, without inequalities on 
the surface of the body, and 16 feet. Of these, many are radicivor- 
ous, but they are easily distinguished from the Hepialideous larve, 
although the resemblance between the latter and those of Gortyna is 
very close. Some of these naked larvze are external feeders, and 
have the body more coloured, and others have the eleventh segment 
of the body more cr less angulated above (Miselia, Phlogophora, 
Trachea, &c.; fig. 108.11. larva of Mamestra Persicariz ; 12. front 
of the head; 13. ocellar region; 14. mandible ; 15. labium and max- 
illa, with a thread issuing from the spinneret ; 16. antenna). Acro- 
nycta varies in its larve, being strongly hairy in A. Menyanthidis, 
and having an elevated horn near the extremity of the body in some 
of the other species. ‘That of Dipthera Orion nearly resembles that 
of an Arctia. Ophiusa has a naked larva, greatly attenuated at each 
