172 



upon is often at fault. In our own experience we have never found her 

 boring in uninfested trees, as others have done, and in cases where she 

 fails to reach a Tremex larva and to fasten her egg upon or near it she 

 must either reach a Tremex burrow or a Tremex larva must come in con- 

 tact with such egg or the larva issuing therefrom to insure peri^etuation. 

 The Thalessa larva no doubt actively searches for its victim within the 

 burrow, but, from the nature of its mouth-parts, is incapable of boring 

 wood as Mr. Harrington and Mr. Clarkson suppose. 



METHOD OF OVIPOSITION IN THALESSA. 



The method of ovij)osition in a creature with such an enormously long 



ovipositor as Thalessa 

 possesses must be of 

 particular interest. 

 We have had good op- 

 portunities of observ- 

 ing it. In preparing 

 for the act the position 

 is generally longitud- 

 inal or in a line with 

 the axis of trunk or 

 branch, the head ei- 

 therupordown .With 

 the abdomen raised in 

 the air the ovipositor 

 is taken and managed 

 with the hind legs, and 

 the tip guided by the 

 front tarsi. The two 

 outer sheaths are used 

 as props and do not 

 enter the wood with 

 the ovipositor proper. 

 They are generally 

 crossed — a position 

 which gives additional 

 strength and security 

 to them. Now, by a 

 movementfrom side to 

 side, and by arching 

 the abdomen and bear- 

 ing upon the oviposi tor 

 she gradually forces 

 this back through the tip of the abdomen into a membrane which issues 



Fig. 30. — Thalessa lunator. (a) Female in act of ovipositing ; (b) 

 abdomen showing outer sheaths in slightly diiierent position ; (f) ab. 

 doraen stretched to its utmo.**!, as when first inserting or finally 

 withdrawing the ovipositor, and showing the coil of outer sheaths (/ ), 

 the distended membrane (d), and the ovipositor coiled around in- 

 side it at periphery (original). 



