the Biology of some British Neuroptera. 515 



Although all Neuropterous larvae have, to some extent, 

 similar eversible processes, these appear to be best developed 

 in Osmylus. After stabbing a large Chironomus larva, or 

 other prey, with its jaws, the anal processes are everted to 

 obtain a firm hold. 



The larva hves on the wet mossy margins of streams, 

 etc., and is amphibious in habits. It is generally sluggish, 

 but if the surface film of water in contact vvith it is disturbed, 

 even some millimetres away, attention is at once actively 

 directed to that spot. From various experiments, con- 

 ducted with larvae in small depths of water, I think the 

 sensitive parts are situated at the tips of the maxillae and 

 in the empodia, as in Chrysopa. Structure would tend to 

 confirm this. If a movement is detected in the moss, or 

 wet substratum on vv'hich the larva is walking, at once 

 there is a stabbing downward of the sucking spears, until 

 these strike a living object. A Chironomus larva, larger 

 than the Osmylus larva, is cjuickly paralysed and then 

 leisurely sucked of its juices. At first young larvae were 

 offered aphids, but although these were hampered by their 

 wet surroundings, the Osmylus larvae appeared to find 

 them difficult to manage. An aphid, about the same bulk 

 as a newly hatched Osmylus larva, continued to walk for 

 two minutes after being stabbed by the latter. With 

 Chironomus the poisonous action of the Osmylus' saliva 

 appears to be more rapid, and with a large Osmylus a 

 full-grown Chironomus rvparius larva dies often in ten 

 seconds. As mentioned, at first young Osmylus were fed 

 upon aphids, under the impression that a normal-sized 

 Chironomus larva would be too large for them to manage, 

 but such is not the case. If the Osmylus larvae be supplied 

 from the first, with mud containing Chironomus, they will 

 probe this effectively and quickly feed up. From the 

 fauna of places where I have taken Osmylus larvae, I am 

 quite sure that the natural food is Dipterous larvae. 



The length of each instar depends on food supply and 

 temperature. Larvae well supplied with food first moulted 

 at the end of fourteen days. Other larvae have been 

 retarded by lack of food. Winter is passed as a larva in 

 the second or third instar, generally the second. 



At the end of April or beginning of May a cocoon is 

 spun, still among the damp moss, in nature. This is of 

 thin, but closely woven yellowish-white silk, irregularly 

 oval in form, usually 1 cm. long by -8 cm. broad. Moss is 



TRANS. ENT. SOC. LOND. 1922. — PARTS III, IV. (FEB. '23) M M 



