176 LIVERPOOL BIOLOGICAL SOCIETY. 
out of, and they must therefore have been in an exceed- 
ingly plastic condition towards the end of larval life, or 
else they must have been very much folded. Hither con- 
dition would fully explain why the larva gave up eating. 
The antennee of the larva as seen from without are a 
pair of rather short rod-like organs with a thick cuticle. 
This cuticle prevents any apparent growth, but a very 
extensive growth nevertheless occurs, only as there 1s no 
room to grow forwards, growth occurs backwards and the 
basal portion is thrust into the head and telescoped and 
much folded. 
At the time of escape of the pupa, the mouth-parts, 
antenne, wings, halteres and legs become straightened 
out and laid beneath the thorax and upon its sides, the 
legs only bemg folded. The fins and pupal siphons take 
on their final form and at once come into use. The 
mouth-parts which are cemented together are much longer 
than the thorax and their distal ends are turned up- 
wards under the anterior part of the abdomen, leaving a 
space between them and the ventral surface of the body. 
The wings are also much longer than the thorax and they 
extend down at each side of the mouth-parts, forming side 
walls to the cavity of which the mouth-parts form the 
floor. This cavity is partially occupied by the legs, the 
remainder of it is filled with air and serves as a float. 
All the mouth-parts of the imago are present in the 
youngest pupa and are indeed much larger than in the 
adult. The pulpy tissue with which they are filled rapidly 
contracts and the new cuticle is then formed. The hypo- 
pharynx is at first continuous with the labium but rapidly 
separates in the female (not in the male) and in both sexes 
the salivary groove is formed upon its anterior surface, and 
at the point of opening of the salivary duct (S, fig. 7) a 
thickening of the epidermis appears, becomes hollow, ac- 
