88 Journal New York Entomological Society, t^oi. xxiv, 



surely occurs at this period. On June 19 when the larv?e were first 

 encountered, they were well on in the second stage and tunnelling 

 near the top, sometimes entering where the petiole sheath encircled 

 the stem at the second leaf downward. Their presence at this tender 

 portion causes the part above to die and furnishes a ready clue to 

 their position. The borings finally get down to the base though so 

 many changes from one plant to another occur that in no instance 

 under observation did this happen with an individual plant. During 

 the penultimate and last stage every two or three days sees a change. 

 It requires but a few moments for a larva to leave a plant and bore 

 its way in out of sight into another, as they move and operate with 

 a nervous haste quite out of the ordinary with such borers. Entry is 

 a little above the ground level and they work both upward and down 

 into the root, though rather avoiding the latter until quite mature. 

 Sanrurus has a horizontal rootstock which extends a long way for so 

 small a plant and affords a good chance for extended mining. Occa- 

 sionally a larva will mine 50 cm. in these roots and yet make the long 

 journey back to the ventilating orifice in the stem once in ten to fif- 

 teen minutes, for disposing the frass outside. Except for the first 

 work at the top, the plant shows no wilting or browning of the foliage. 



Pupation seems normally to occur in the gallery at the root crown 

 or in the root according to Mr. Jones, there being no further en- 

 largement of the orifice. Our experience had to do with diseased 

 larvae only and these left the plant but imperfectly transformed under 

 a covering of moss. Parasitism in the early stages (two and three), 

 from a small Ichneumon as yet undetermined, was very severe. The 

 orifice then made is too small for this wasp to enter and it may 

 pierce the stem with the ovipositor to reach the gallery or host, since 

 a similar species has been observed puncturing the enfolding leaf- 

 roll which sheltered a Pyralid larva, in order to reach its host. Ap- 

 parently the parasitic larva attains growth in a few days when it 

 spins a white cocoon nearby, and in fifteen days gives up the imago. 

 The host is exceedingly small at this time to support this species and 

 the larval period must needs be brief. 



A close relationship to Papaipema larva is shown in the species 

 under consideration, even the darkened girdle appearing, though it 

 is not so pronounced. There is little change up to maturity. 



Stage III. — Head brown, no side line, ocelli prominent; con- 



