a carnivorous Tipulid larva. 249_ 
respiratory trumpet lies an expanded trachea, which 
fits loosely to the chitinous integument, except along 
the external, the superior, and part of the internal 
margin, where the two structures blend. Along the 
junction is a regular row of small oval apertures. These 
are so small, and it is so difficult to see them except 
through a layer of cuticle, that I have been unable to 
decide whether or not there is a transparent membrane 
across each aperture. However this may be, the 
openings no doubt serve for the admission of air to 
the trachea. Even if they are completely pervious, 
their minute size must constitute an effective provision 
against the entrance of water or dirt. 
The hinder end of the abdomen of the pupa is moulded 
upon the corresponding segments of the imago, and in- 
dicates the sexual modifications. The female pupa is 
distinguished at once by the long, triangular valves of 
the ovipositor. On the dorsal surface are a number of 
roughened plates. ‘These are armed with rather strong 
and dense spines. The first two abdominal segments 
have no such plates; the 3rd has a single plate in the 
fore part of the segment, the lateral extremities being 
most spiny; the 4th has two plates, anterior and pos- 
terior, most spiny laterally ; the 5th and 6th are similar, 
but here the median part of the plates is the most spiny ; 
the 7th segment bears only one plate, the lateral ex- 
tremities being the most spiny. The remaining segments 
have no spiny plates, being modified for reproductive 
functions. A similar provision of spines upon the pupal 
integument, usually taking, as in this case, the form of 
incomplete rings, is so common among insects as hardly 
to require exemplification. Every entomologist will 
recollect the spiny rings of the pupe of the Clear-wings, 
the Goat-moth, the Wood-leopard, and the Swifts. Very 
many dipterous pupe (Tipula, Limnobia, Bibio, &c.) 
offer examples of the same contrivance. The roughening 
of the mobile abdomen in such cases gives it a sufficient 
hold of the wood or earth in which the pupa lies to 
enable the insect to creep to the surface before the final 
change takes place, thus avoiding the damage to gauzy 
or plumed wings, which would inevitably result if the 
moth or fly emerged in the narrow gallery, or below the 
surface of the ground. 
Five pairs of fleshy pseudopods project from the ven- 
8 2 
