526 MODERN CLASSIFICATION OF INSECTS. 
The larve and pupz of various species of Ctenophora and Tipule 
which agree with the above description, have been observed by Réau- 
mur (Meém. tom. v. pl. 1—3.); De Geer (tom. vi. pl. 18, 19. and 25.); 
Fischer ( Oryctographie Imper. Rossie); Bouché (Naturgesch. tab. 2. 
f. 24—29., tab. 3. f. 1—4. Tipula pratensis) ; Rdsel (Abhandl. vol. ii. 
Muse. et Culic. tab. i.). 
I have several times observed battles taking place between males 
of different species of Tipula. 
The larvee of Limnobia* xanthoptera and pilosa inhabit species 
of Agaricus (Stannius, Beztr. Entomol. Schles. vol. i. p. 202.); they 
offer no marked peculiarities of structure; the pupz also are of the 
ordinary form, without elongated appendages to the thorax (tab. 5. 
fig. 14.). 
M. Van Roser has noticed several larve of this and allied genera 
(Verz. Wurtemb. Dipt.); that of the curious genus Anisomera re- 
sembles that of Tipula, and is commonly found by him under the 
sand of the banks of the Neckar. 
All the larvae of this family are not, however, terrestrial, the pre- 
paratory states of several groups being undergone in water; of these 
the genus Ptychopteryx (fig. 126.6. Pt. paludosa}) exhibits a sin- 
gular departure from the rest of the family; the larva (fig. 126. 8.) 
being very long and worm-like, but mnuch narrowed at the posterior 
extremity, which is terminated by a very long and delicate tube 
serving to convey the air to two trachee, which extend through the 
entire length of the body. In the pupa (jig. 126. 9.) this peculiarity 
is reversed, the body being terminated by four small points, whilst the 
anterior extremity is furnished with a very long thread-like appendage 
serving, like the former, to supply air to the insect, its extremity 
being extended to the surface. Lyonnet has beautifully illustrated 
the transformations of P. paludosa (Rech. Posth. pl. 18. f. 1. 7.); 
* See Ann. Soc. Ent. France, 1836, p. xviii. on a fossil Limnobia found at Dal- 
marnock. 
+ The species of this genus possess a remarkable peculiarity in the imago state, 
which I have observed in no other insect of the family, and which does not appear 
to have been previously noticed, namely, the possession of two minute membranous 
and setose appendages between the base of the balancers and metathoracic spiracles. 
( Fig. 126.7. x base of the balancer ; 0, metathoracie spiracle ; +, the above-mentioned 
appendage.) The discovery of this appendage throws further difficulty in the way 
of the solution of the question, whether the halteres be the real representations of 
the hind wings, 
