46 



. 



serves to protrude, and to press firmly, the inner surfaces of the labellae, upon 

 any surface from which the insect wishes to scratch or dissolve substances for 

 food. That it is air, at least for the most part, and not water, which expands 

 the inner surfaces of the labellae, is easily proved by pressing carefully the head 

 of a fly between the fingers until the proboscis is fully extended, and the labellae 

 fully inflated, then, putting the fly under water, and pricking with a needle 

 the inner surfaces of the labellae ; they will at once collapse, bubbles of air 

 escaping, at the same moment, from the opening made in their surfaces by 

 the needle. 



The pseudotracheae on the inner surfaces of the labellae of Musca are 

 cylindrical channels, sunk, more or less deeply, into the surface of the labellae, 

 according to the amount that that surface is inflated, and they open on the surface 

 in zigzag slits. These channels are held open by partial rings, more strongly chitin- 

 ized than the rest of the membrane of the cylinder. As seen from above in Musca 

 domestica, the pseudotracheae (pi. 4, fig. 6, 6) appear to be supported by partial 

 rings, one end of each of which is forked. Such a ring, if isolated, would appear 

 as in fig. 6, d. These rings are apparently arranged so that one has its fork 

 on one side of the opening of the channel, the next ring the fork on the opposite 

 side of the channel, and so on, in alternation. This, I say, is the apparent 

 structure, for if one expands the membrane of the inner surface of the labellae, 

 to a sufficient extent, the channels, or pseudotracheae, are flattened out and their 

 true structure is revealed. PI. 4, fig. 7, represents a portion of such a flattened 

 out pseudotrachea of Musca domestica, the structure of which is immediately 

 evident; at the right-hand side of the figure is represented an irregularity, such 

 as now and then occurs in pseudotracheae. If such a piece of flattened pseudo- 

 tracheae as is seen at the left, in fig. 7, be formed into a cylindrical channel, 

 its appearance will be as in pi. 4, fig. 6, b- In Musca vomitoria the pseudo- 

 tracheae have a structure somewhat different in detail, but not in principle, from 

 those already described, as can be seen from fig. 5 without further description. 

 The pseudotracheae of Eristalis horticola are so nearly like those of Musca 

 vomitoria, that I have not figured those of the former. PI. 4, fig. 8 represents a 

 section through a portion of the inner surface of a labella of Musca vomitoria, 

 cut at right angles to the direction of the pseudotracheae which pass through it; 

 in this section the intrapseudotracheal spaces are represented as protruded beyond 

 the edges of the pseudotracheae themselves ; this is the natural position of the 

 pseudotracheae in repose, but when the fly inflates his labellae in the process of 

 eating, the soft intrapseudotracheal membrane is stretched, and the margins of 

 the pseudotracheae are protruded beyond the general level of the surface of the 



