PSEUDOTRACHEAL MEMBRANE OF MUSCA 



43 



is shaped like a squat flask with a short neck, is kept open by a series 

 of transversely arranged bars of chitin bent to the form of an incom- 

 plete loop, the shape of the channel (Plate VIII, figs. 2 and 3). Each 

 of these incomplete rings is bifid at one end and simply expanded at 

 the other; they are set closely side by side throughout the length of 

 the channel, and each ring is placed in the reverse direction to those 

 on either side of it. There is, therefore, a series of alternate bifid ends 

 and expanded ends on each side of the channel (Plate VIII, fig. 4), and 

 over each the membrane is tightly tacked down. When the labella 

 are in use the two ends of each ring are approximated to one another, 

 and there is little if any space between them through which the fluid 

 could pass ; the actual openings into the lumen are the shallow grooves 

 between the bifid ends of the rings, at right angles to the long axis of 

 the channel. 



Graham-Smith, from whose recent paper many of the above details 

 are taken, terms the space between the arms of the fork the ' inter- 

 bifid space ', and the groove which leads through it to the lumen of 

 the channel the ' interbifid ' groove. Now it is evident that, if this be 

 the normal path up which the fluid is sucked, the size of the groove 

 or space will limit the size of the particles which the fly can ingest, 

 a somewhat important point with regard to its disease carrying powers. 

 Graham-Smith gives the following measurements for some parts of the 

 system : 



PSEUDOTRACHEAE 

 DIAMETER AT 



INTERBIFID SPACES 



DIAMETER NEAR THE ENDS 

 OF THE PSEUDOTRACHEAE 



Calliphora erythrocephala 



Sarcophaga carnaria 



Lucilia caesar 



Fannia (Homolomyia) caniculans 



Ophyra anthrax 



Musca domestica 



The method by which the pseudotracheal channels terminate at the 

 prestomum differs a good deal in different species. In some the rings 

 become increased in depth, and arranged obliquely instead of perpendic- 

 ular to the surface, and then split in the middle line, the lateral halves 

 separating and coming to lie one in advance of the other. The terminal 

 portions are larger and longer than the rest, and become attached either 

 to the discal sclerite or to the membrane immediately distal to it. In 



