( 'ADDIS-FLIES. 73 



the hind-wings are pale brown, paler near the base. The two anterior pairs 

 of legs have the tarsi and tips of the tibia black. The posterior pair are 

 pale yellowish-brown throughout. The antennae are twice the length of the 

 anterior wings. The maxillary palpi are very hairy, and about £ the length 

 of the antennae. 



The transformations of this peculiar insect are very 

 interesting. The larva, which is found in clear, dashing, 

 forest streams, usually inhabits a piece of twig, through 

 the centre of which it has tunnelled out a cylindrical hole. 

 In this simple habitation it is enabled to live secure from 

 all enemies, as the stick is always long enough to allow 

 the insect to retreat a considerable distance from the 

 opening at either end. In fact, I have found small larva? 

 inhabiting twigs nearly five inches long. If menaced in 

 any way the larva always remains thus concealed in the 

 middle of the twig, consequently its habitation cannot be 

 distinguished from any other fragment of twig, of which 

 there are usually hundreds floating about in the streams, 

 where the insects abound. When the larva is not appre- 

 hensive of any danger it projects its head and thorax from 

 the opening, and proceeds to walk about dragging the 

 stick after it. The presence of a number of these ani- 

 mated twigs, of various shapes and sizes, in a basin of 

 water has a most remarkable appearance. The enclosed 

 insects are not readily seen, and it is thus often difficult 

 to account for the singularly rapid movements with which 

 the sticks are endowed. Although this larva usually in- 

 habits a plain stick, hollowed out in the middle, more 

 complicated dwellings are not infrequently constructed 

 by the insect. Fragments of the stems of fern fronds 

 (see Plate X., fig. 4), the green liverworts, which grow so 

 plentifully on the stony banks of many of these forest 

 streams (fig. 3) , as well as many other kinds of vegetable 

 remains, are built up by the larva into the most fantastic 

 cases imaginable. In addition, the insect frequently 

 employs ordinary fragments of wood from forest trees, 

 which it hollows out in a similar manner to the sticks 

 (fig. 5). This last-named habit proves, I think, that the 

 larva does not select twigs already hollowed out, as I at 

 first supposed, but bores them out itself to suit its own 

 requirements. 



The food of these larva? consists chiefly of fallen leaves 

 and sodden wood, both of which are present in great 

 quantities in all forest streams. I have actually seen 

 captive specimens feeding on the leaves of Melicytus 

 ramiflorus, and on the berries of Aristotelia racemosa. 



