—15— 



ules, upon each of which are eighth ambulator}' setce, one of the 

 luiinher being somewhat isolated and near the middle of the segment. 

 ( )n all twelve segments are four, short seti^, placed in pairs each side of 

 tlie middle ; those cm the thoracic segments being between the isolated 

 setce, just indicated. 



The pupiv average i6 mm. in length. In the female, the rostrum 

 reaches considerably beyond the base of anterior femora, while in the 

 male it extends very little beyond. On the rostrum, head and thorax 

 are a few, sparsely j)Iaced bristles. 



Beneath, the nine segments are smooth, distinctly separated, and 

 with a few short bristles. Above, the first segment is smooth, with ex- 

 ception of the scutellar lurbucle; the second is smooth, but on the 

 third, fourth and fifth, each, is a transverse row of fleshy spines, ending 

 with chitinous points, and placed near the posterior margins of the seg- 

 ments. On the sixth, the s[)ines are larger and almost dactyliform, 

 being placed on a transverse, fleshy ridge. On the seventh segment, 

 this ridge assumes a comb-like appearance, and is divided at middle, 

 and on each of the parts thus constituted are five fleshy spines, all much 

 larger and longer than any on the preceeding segments. The eighth 

 segment is nearly vertical, unarmed, but slightly wrinkled and con- 

 spicuously setigerous. The ninth segment is nearly horizontal btlow 

 and opixisite the eighth, or curled under, as it may be called. This is 

 armed with two, widely se])arated proleg-like appendages, each ter- 

 minating in a stout, chitinous hook, with the point turned directly 

 backward. Abdomen, slightly bioadest at third and fourth segments. 

 The spines above describetl are doubtless to enable the pupa to fuish 

 itself into a positicjn favorable to the escape of the adult, which, within 

 the stems of the plants, must be well nigh helpless. 



When full grown, the larvx" girdle the stems from within, at ir- 

 regular intervals, much as does the larva:; of Elaphidion within the 

 twigs of oak, hickoi}- and other trees, except that only the thin outer 

 bark is left intact, to be broken ofl" by the winds ; the object being to 

 give the larva time to witlidraw into its burrow and plug up the open- 

 ing. In proof of this the ends of the stems are filled with fibrous 

 matter, taken from within the stem. Doubtless this girdling is done 

 during a calm time, else the wind would surprise the larva with its work 

 but half finished, but how it obtains a knowledge of the state of the 

 weather outside, is a pr(»blem. 



Notwithstanding all their precautions in filling the cavities at the 

 ends of the stems, they are far from safe in their woody abodes, as, 

 while j)rotecting them from enemies of their own kind, even the thick 



