290 Mr. Ernest Ewart Unwin on the 
which seems to communicate with the exterior through a 
small aperture (fig. 5). A number of fine branched hairs 
form a rosette around the spiracle tip. They remind one 
of the hairs on the extremity of the respiratory tube of 
the Eristalis larva,* and have probably a similar function. 
The power of extending or retracting the spiracles, 
although not considerable, is of great service to such a 
larva as Drosophila. When the larva is burrowing deeply 
in its food, and is completely submerged, the process is 
retracted, and the spiracles become flush with the surface 
of the body. When the larva is 
near the surface the process can 
be extended, causing the spiracles 
to reach the air, or even to be 
protruded into it. At such times 
the head and the rest of the body 
may remain buried in the food, 
so that feeding and breathing take 
place simultaneously. If the 
larve are put into shallow water 
the spiracle-bearing process is 
protruded, thus carrying the 
spiracles above the surface of the 
water. If more water is added 
so that the larve are entirely 
submerged, the spiracles cannot 
reach the surface, and in time 
Fie. 5. the larvee drown. Several experi- 
Posterior spiray) ev ments were tried, using water, 
vinegar, fruit-juice, etc., but in 
every case the submerged larvae ‘perished. 
THE PUPA. 
The larva, when full grown, shows considerable anxiety 
to leave the food in which it has been living. It seeks 
out some dry spot on the sides of the vessel, and there 
pupates. The natural place of pupation is no doubt the 
ground, as with many other Muscide. The fall of the 
fruit to the ground would make it easy for the larva to 
enter the soil. When Drosophila larvee, about to pupate, 
are placed on loose earth, they quickly burrow into it, and 
if uncovered show great eagerness to get below the surface 
* Miall (1895), fig. 72. 
