MUSCLES AND VISCERA OF THE IMAGO. 225 



integument, and it is not possible that any muscular contrac- 

 tion can bring about this result. As the large spiracles are 

 thoracic, it might be supposed that the expansion of the thorax 

 acts as an inspiratory agent ; but no such expansion occurs. 

 The only possible way in which such an expansion could occur 

 is by the cavity being rendered more spherical by the contrac- 

 tion of the dorsales muscles, without any alteration of the extent 

 of its surface, as it is well known that a spherical envelope 

 encloses a maximum space ; but the contraction of these 

 muscles closes the syndesmoses, and renders the surface 

 smaller, so that there is no reason to believe that the size of 

 the cavity can be increased by their contraction. I shall here- 

 after show that the mechanism of inspiration is entirely inde- 

 pendent of either thoracic or abdominal movements, and that 

 the air is pumped into the tracheae by a special mechanism. 



The general arrangement of the tracheae in the larva has 

 been already described, and this appears to be the ground-form 

 of the tracheal system. In the Blow-fly imago this ground- 

 form still persists, but is masked by the complexity of the 

 branches, which are often larger than the longitudinal trunks. 



The vessels which represent these lie one on either side of 

 the thoracic blood-sinus (Fig. 37) these I term the para- 

 gastric trunks. The paragastric tracheae enter the head as 

 narrow vessels, one on either side of the neck. Each imme- 

 diately gives off a branch which passes below the jugum and 

 tentorium, and dilates to form the air-sacs of the proboscis ; the 

 main trunks then unite and form a single median vessel, which 

 ascends on the posterior surface of the brain ; its walls are 

 greatly thickened, it gives off lateral branches to each side, 

 and terminates above the brain by dividing into two again. 

 The branches of this vessel dilate in front of the brain into the 

 air-sacs of the supra-tentorial region of the head. 



The paragastric trunks are connected with the anterior and 

 posterior spiracles by transverse or spiracular trunks, and im- 

 mediately behind those from the posterior spiracles they 

 become very narrow, pass beneath the mesophagma, and 

 enter the great abdominal pulmonary sacs, which form the 



