804 STUDIES IN AUSTRALIAN NEUROPTERA, vii., 



sometimc'tj vields a larva. Veiy*soon the larva will begin to 

 move rapidly backwards, feeling all the time, with its mobile 

 anal papilla, for a suitable crevice in which to hide. Jf once it 

 succeeds in reaching one, it will not be easy to secure it. Hence 

 it should be picked up at once with the forceps, and transferred 

 to a glass-tube, whose open end should be stopped up with cotton- 

 wool. Not more than one larva must be put into one tube, 

 unless each is separated from the next by a plug of cotton-wool; 

 for one is sure to attack the other and suck him dry. 



The larva may be reared right through, in a single SRiall glass- 

 tube. The best food for the earlier stages is white-ants. These 

 should be given alive, but with their heads crushed; otherwise 

 there will be a fight between the larva and its intended victim, 

 in which the tables may be turned. In the second and third 

 instars, the larvae of the Oodlin Moth make excellent food; but 

 these also must be offered with their heads crushed, unless one 

 is willing to take the risk of damage to the Psychopsis larva from 

 the strong mandibles of the Codlin grub. Under natural con- 

 ditions, the Psychopsis larva guards himself from attacks of this 

 kind by retreating into a crevice, after he has seized his prey. 

 With his body thus covered, and only his large jaws projecting, 

 he is perfectly safe. But it is different in a glass-tube, where 

 the whole of the larva's soft body is exposed to attack from the 

 jaws of his writhing victim. 



The larvae feed but seldom; each meal, however, is a very sub- 

 stantial one. A larva supplied with a Codlin grub will usually 

 attack it at once, advancing cautiously with its jaws wide apart, 

 until they are well placed on each side of its victim. Then, with 

 a sudden vicious snap, the jaws are driven home, and the victim 

 is secured. If the victim struggles furiously, the Pyychopsis 

 larva will frequently let go its hold, and retreat into a corner, 

 with every sign of fear; nor will it be induced to attack the same 

 victim a second time. Generally, however, the larva is able to 

 hold on, until the loss of blood occasioned by the wound weakens 

 the victim sufficiently to prevent its struggling further. Then 

 the jaws are driven far in, and the play of the maxillae to and 

 fro beneath the grooved mandibles shows that the larva is suck- 



