DEHISCENCE OF THE FEMALE PrPA OF FT-^rEA CASTA. 325 



abdomen from this position only for a brief period, some five minutes 

 altogether perhaps, during tlie visit of the male. There are two 

 reasons, however, that show that the keeping-in-touch idea is prob- 

 ably erroneous, certainly not the whole matter. The dehisced ex- 

 tremity of the pupa-case is a very definite and recognisable structure, 

 with a free aperture that the tactile arrangements of the ovipositor 

 ought easily to recognise, and, more convincing, if the female be 

 removed from her position she gets back to it without, apparently, 

 much difficulty. My own belief is that she sits so closely to prevent 

 the entrance of any parasites or enemies, the terminal wool forming a 

 good cJtcraiLr-(h'-fri>n' against anything from mites upwards. Fumea 

 does not mix much wool with her eggs, but accumulates a good deal 

 about the pupal opening during oviposition, and does a lot of work 

 after, chiefly, apparently, with the object of introducing as much as 

 possible of this on the top of the eggs and about the opening, as a 

 fence against marauders. These facilities for oviposition, and the 

 necessity of these special protective devices on the part of the moth, 

 are to be found in the method of dehiscence of the female pupa. The 

 pupa-case is not brought out of the sack, as in Solenobias, nor is it left 

 entirely within it, as in the Psyches, but comes forwards so far as to 

 bring the mesothorax level with the mouth of the sack. At this point 

 the 3rd thoracic and wing-cases remain together, and form an impedi- 

 ment to the further advance of the pupa, whilst the parts in front are 

 so disposed as to form a ring just outside the opening of the sack, and 

 so, as it were, rivet the pupa-case in that position. 



The mesothorax splits dorsally, and separates from the metathorax, 

 but remains firmly attached to the wing-cases and assumes a sloping 

 position, its dorsal end being just outside the sack, its other extremity 

 firmly fixed to the wings. These pieces, therefore, one on each side 

 (or rather more dorsally), are all of the exposed pupal parts that occupy 

 the narrow neck of the sack, and, by their elasticity, taking a firm basis 

 by their attachment to the wings, maintain the opening. Outside the 

 sack the prothoracic piece is attached to the anterior margin of the 

 dorsal end of the mesothoracic piece in such a way that it lies along 

 the edge of the opening of the sack, its inner surface sloping outwards 

 so as to make the opening funnel-shaped where it is, and form the 

 dorsal portion of the outer member of what I have compared with a 

 rivet. It carries at its anterior margin the dorsal head-piece, which 

 has no other function than to slightly extend the slopes of the funnel. 



Round the rest of the opening of the sack, i.e., opposite the venter 

 of the pupa, extend from the ventral ends of the prothoracic pieces, 

 two filmy but very tough shreds, united to the portion that includes 

 the head-piece. This portion consists of the head, antenna- and mouth- 

 parts, and the anterior legs. These filmy shreds are, in fact, the 

 pupal investment of that part of the prothorax that is covered by the 

 appendages in the pupa, as well as the basal portion of the leg-cases. 

 It takes its attachment ventrally (one is apt to fall into familiar, but 

 inaccurate, language and say in front, though front is opposed to back, 

 and is so far correct, but is also synonymous with anterior, which 

 would be wrong) to the points of the femoral pieces, the pieces that in 

 these pup:e look so like maxillie or labial palpi. The front piece thus 

 tethered leans outwards, and forms the slope of the funnel-shaped 

 aperture in front (as opposed to back). To the same point the second 



