1883.] On the Genealogy of the Insects. 937 
branches near the base, while there is a lateral row of slender fil- 
aments and a row of ventral verticillate hairs. It thus bears a 
resemblance to the larvae of some butterflies, as Vanessa antiopa, 
and especially the young Polyommatus (Heodes hypopleas) or the 
Bombycid larve of Anisota stigma or Platysamia, as. well as 
Selandria larvæ. Brauer’s figures show a pair of abdominal, two- 
jointed feet to each of the nine abdominal segments, while just as 
in Lepidopterous larve and in that of Panorpa there is a pair of 
prothoracic spiracles, none on the mesothoracic or metathoracic 
segments, and there are nine pairs of abdominal spiracles, accord- 
ing to Brauer’s figure, or one more pair than in Lepidopterous 
arvæ. 
The fact that there are in the larval Panorpidæ collectively a 
pair of feet tò each abdominal segment (the terminal segment in 
Panorpa bearing what are evidently homologues of the anal prop- 
legs of caterpillars) is of much significance when we bear in 
mind that while no caterpillars are known to have more than five 
pairs of abdominal or prop-legs, some of the segments bearing 
none, yet the embryos, as shown by Kowalevsky, have temporary 
embryonic indications of legs, a pair to each segment (uromere) ; 
it is a significant fact that the eruciform larve of the Panorpide 
actually have two-jointed legs to each abdominal segment, the 
penultimate segment in Bittacus bearing such legs, and the ter- 
minal segment bearing leg-like processes in Panorpa. The ori- 
gin of the Lepidoptera from the. same stem-form as the Panor- 
pide thus seems a reasonable hypothesis. 
In the metamorphosis of Mantispa, as Brauer has shown, there 
is a hypermetamorphosis, i. e., two larval stages. The first stage 
is Campodea-form but the second is suberuciform. The transfor- 
Mations of Mantispa. appear to give us the key to the mode in 
Which a metamorphosis was brought about. The larva, born a 
Campodea-like form, active, with large, long, four-jointed feet, 
living a sedentary life in the egg-sac of a spider, before the first 
molt loses the use of its feet, while the antennz are partly 
aborted., The fully -grown larva is round-bodied,. with small, 
caterpillar-like feet and a small round head. Its, external appen- 
dages retrograding and retarded, acceleration of growth goes on 
Within, and thus the pupal form is perfected while the larva is 
full-fed and quiescent; hence as a result the pupal stage became 
a quiescent one, and by inheritance it gradually became a perma- 
