179 



sertioii and then bends in a loop and is brought back over the dorsum 

 and around the head and then back again on the ventral side, hugging 

 the legs, its tip reaching far beyond the tip of the abdomen. In this it 

 differs from the European Rhysfia ptrsuasoria^ in which the ovipositor of 

 the pupa, according to Snellen's figures, previously mentioned, reaches 

 ouly some two-thirds the length of the body behind the back. 



It remains only to state in reference to the habits and transformations 

 of Tremex (see Fig. 39) that, from the facts already mentioned, it would 

 seem that the imagines mature, as a rule, somewhat later than Thalessa, 

 and that the larvai are found of various sizes on the approach of winter. 

 We have also found, on one occasion, at Emporia, Kans., on December 

 16, 1874, in Celtis oecidentalis, a, Tremex imago somewhat torpid and eaten 

 half-way out of the trunk. There are no positive records in this coun- 

 try to show the length of duration of the larva state in either of these 

 genera, but we should expect the Thalessa larva to develop most rapidly 

 when once it finds its food, but to possess also great power of enduring 

 without food in early life. There is doubtless much irregularity in de- 

 velopment in both genera, especially after the pupa state is assumed, 

 while the period of oviposition, as we know, may cover several of the 

 summer months. 



The larva (Fig. 39, a) has the normal form of the horn tails, being 

 blunter at both ends than the Thalessa, with an anal thorn, short thor- 

 acic legs and strong gnawing jaws (Fig. 39, b). Our figure {a) shows 

 a young Thalessa larva attached about the middle, just as it has re- 

 mained since 1872 in our alcoholic specimens. The ovipositor in the 

 female pupa, as shown in the figure (c) is not bent. 



We have the authority of Kollar that the larva of Sirex gigas attains 

 full growth in seven weeks after the laying of the egg, and that in the 

 pupa state it may remain in the tree for several years. Normally both 

 Thalessa and Tremex probably go through their transformations within 

 a year. Tremex columba is at times abundant enough to materially 

 injure trees, and Mr. Jonathan Periam, the present editor of the Prairie 

 Farmer, sent us an account (November 28, 1873) of a hickory tree which 

 he believed was killed by it. Our figures will convey a very good idea 

 of the adolescent states of both. They were drawn by Dr. Marx, with 

 our assistance, from our Missouri material, and also from a pupa kindly 

 loaned for the purpose by Dr. H. A. Hagen, our own examples of the 

 pupa being too mature to permit of a good figure being made. 



EXPLANATION TO PLATE I. 



Thalessa lunaior: a, larva, side view ; h, head of larva from side ; c, pupa, side view; 

 d, tip of ovipositor of pupa, ventral view greatly enlarged to show five parts (includ- 

 ing sheaths) of which it consists; e, adult female;/, abdomen of adult female from side, 

 showing gap between joints 6 and 7, from which the membrane distends when ovi- 

 positor is in action ; g, adult male; h, anal extremity of abdomen of male enlarced. 

 (Original.) 



