397 



special development in the female, ramifies among the muscles 

 of the ovipositor (text fig. H). 



In the pupa of the fourth day a transverse ''commissural" 

 trachea is seen uniting the main longitudinal air sacs; I have 

 not observed the manner or time of its formation. 



The old ''spiral" intima of the larva can still be seen 

 lying within tho main air sacs ; probably it disappears by 

 being drawn out through the thoracic spiracles at the last 

 ecdysis; it is certainly not present in the adult wasp. 



The adult respiratory system (text fig. H) consists, then, 

 of a pair of main tracheal vessels, dil'ated in the thoracic 

 region into the ventral air sacs, and connected by three trans- 

 verse vessels — one anterior, another posterior, the third a 

 broad channel joining th© twO' air sacs in their mid-region. 

 Passing from the anterior to the posterior end of each ventral 

 air sac is a great dorso-lateral air sac. From near the pos- 

 terior end of the tracheae a pair of vessels pass downwards and 

 forwards and supply the abdomen. The head is aerated by 

 a pair of great tracheae which pass forwards from the main 

 vessels and divide into three great branches in the head. 

 From all these great vessels tracheoles are given off. Three 

 pairs of stigmatic trunks, two in the alitrunk, the other in ^ 

 the abdomen, connect these tubes with the exterior. 



The tracheoles of the adult insect, though essentially 

 intracellular structures, are not such remarkable structures 

 as we have seen in the larva. Certain cells of the developing 

 tracheal tubes do not flatten out when these form an intima; 

 on the contrary, they seem to grow in thickness, and then 

 migrating from the epithelium grow outwards in various 

 directions and ramify among the organs. They are the 

 tracheoloblasts, but they never assume abnormally large ' 

 dimensions. As they grow out from the tracheae (fig. 82) 

 they leave tubes of varying width behind them (fig. 95) ; in 

 the case of the smaller tracheoles the whole structure may 

 remain unicellular. The larger tracheoles, however, such as 

 those of the brain or of the appendages, are multicellular 

 structures; their formation can be clearly observed in the 

 legs. The nuclei of the small tracheoloblasts occasionally 

 divide, and the cytoplasm between the two nuclei thus formed 

 becomes drawn apart. Tliis process is repeated several times 

 and eventually the tracheole is seen as a narrow tube with a 

 number of oval thickenings, the nuclei, along its path (fig. 

 103). The tracheoles are essentially protoplasmic structures; 

 I could not detect, with any certainty, a chitinous intima. 

 Frequently chromatic ( ? ) granules may be observed within 

 the walls of the tracheae between the nuclei. 



