BY R. J. TILLYARD. 811 



at first complete!}' hiding the liindwing. While the basal half 

 of the forewing expanded rapidly, the distal half remained bent 

 under it for some time. Meanwhile, the hindwings expanded 

 rapidly, and their apical portions appeared beneath the still 

 bent forewings, as seen in Text-fig.ll, d. Next, the distal por- 

 tions of the forewings became straightened out, though at first 

 they remained full of longitudinal rucks, as seen in Text-fig.ll, e. 

 The wings were held well away from the abdomen up to 5.49 

 p.m., by which time they were about three-fourths expanded. 

 Then the apical rucks became rapidly smoothed out, and the 

 wings were arched forwards, forming a steep roof over the body, 

 so that the costal margins of the forewings come almost into 

 contact with the blotting-paper (Text-fig.ll,/"). From the time 

 of leaving the cocoon to the time that the wings were fully ex- 

 panded, only thirty eight minutes elapsed. After resting for 

 some time in the position shown in Text-fig. 1 1,/j until its wings 

 were sufficiently hardened, the imago spread them out in the 

 broad, flattened, roof-like manner usually adopted in this family. 



The pupal skin left behind is a flimsy, wliite structure. It is 

 very difficult to detach from its position. But, when once this 

 has been done, the slightest breath of air causes it to move 

 readily, so light and delicate is its construction. The thorax 

 and base of the abdomen are flattened, and wide open dorsally. 

 From the prothoracic spiracles, two long threads stand up; these 

 are the intima of the tracheal trunks, withdrawn through these 

 spiracles when the imago emerges. The back of the head is split 

 open, and the large dark brown pupal mandibles are left gaping 

 wide apart. 



The Imago. (Text-fig. 12). 



A few details about the imago which have not been published 

 may suitably be given here. 



The mouth-parts are figured in Text-fig. 12. The lahrum re- 

 sembles that of the pupa, but is not definitely bifid, there being 

 only a very slight indication of the division into two distinct 

 lobes. It carries numerous hairs, not set upon raised bases. 



The mandibles ai-e smaller and weaker than in the pupa, and 

 show considerg^ble modification in shape. The ri^ht mandible 





