162 DR. V- E. SHELFORD ON THE LIFE-HISTORIES AND 



Under fuA^ourable conditions this stage lasts about five weeks. The time is 

 doubtless modified by temperature and food conditions, as some individuals 

 require longer. The activities connected with the second larval moult are 

 the same as those connected with the first. 



Last Larval Stage. — The changes are essentially the same as in the first 

 moult ; the diameter of the prothorax is 3^ to 4 mm. (PI. 23. figs. 10 & 11). 



During the latter part of August and the month of September the larva? 

 disappear, closing the outer end of their burrows and going to the bottom to 

 remain until spring. (In captivity they were kept as nearly as possible under 

 normal conditions and the survivors were found with their holes opened in 

 the early part of April.) They feed until the middle of June, when they fill 

 the upper part of the burrows. Each then constructs an ovoid cavity (fig. 13) 

 near the bottom of its burrow. Having carefully packed and smoothed the 

 soil forming the wall of this cell with its mandibles and the ventral side of its 

 head, each larva remains at rest with the body leaning against one of the 

 sides of the pupal cavity. The head is uppermost and the animal gradually 

 becomes less and less active as time goes on. At the end of two or three 

 weeks it is almost incapable of moving the legs. The abdomen gradually 

 thickens and becomes a clear, characteristic, cream colour, due to internal 

 changes. The tubercles of the first five abdominal segments of the pupa 

 (fig. 12) can be seen through the larval skin. They are folded toward the 

 centre of the back. 



Pupal Ecdysis. — A few days after the conditions just mentioned have 

 become apparent, the pupa emerges from the larval skin. The cuticula splits 

 in the mid-dorsal line of the thoracic segments and the head ; the split on the 

 head bifurcates, following the suture at the posterior side of the clypeus. 



The process of shedding the skin requires only a few minutes and has been 

 seen in the variety limbalis. It is accomplished by contractions of the 

 ventral body-muscles which tend to elongate the dorsal side and bring the 

 old skin under tension. The thoracic portions of the body are the first to 

 emerge ; the skin meanwhile slips backward on the abdomen and the head 

 is gradually withdrawn, usually becoming free before the abdomen. The 

 abdomen is freed by its later movements. The pupa possesses very few 

 bristles that could assist in the removal of the exuvium. The only ones are a 

 few on the margins of the pronotum and several long ones on each abdominal 

 tubercle. The bristles of the tubercles together with the tubercles themselves 

 obviously assist in working the exuvium from the abdomen. An exuvial fluid 

 is apparently present, and this together with the great expansion of the wings 

 and appendages is sufficient to insure safe emergence. 



Tlie Pupa. — At the time of emergence the pupa is only a little shorter 

 than the larva, but it soon contracts and assumes the form shown in fig. 12. 

 As has been noted by former workers, the dorsal tubercles serve to hold the 

 body away from the substratum on which it rests. 



