12 
Siphonophora rosarum 
has been brought about by the presence of the habit of flight. As has 
been mentioned before, the third joint (Plate I, Fig. 4) has its entire sur¬ 
face covered with a number of small, clear, dome-shaped tubercles, which 
vary somewhat in size, and give a most remarkable appearance to the 
joint. Each tubercle consists of a blister-like excrescence on the chi tin 
of the joint, and the epidermal lining is continued into the cavity thus 
formed. The sides of the tubercle (see Plate I, Fig. 5) are pigmented 
like the surrounding surface of the joint, but the top is quite clear, and 
has the appearance of a membranous covering to a short cylinder 
of chitin. In nature, the top is almost hemispherical in shape, but in 
the section represented in Plate II, Fig. 5, considerable contraction has 
taken place so that the thin portion has become flattened and distorted. 
The cavity of the tubercle is filled up with a group of cells which 
seem to be modified epidermal cells which have taken on a sensory 
function. These cells are expanded at their base and taper off as they 
near the thin covering of the tubercle. 
It is of course almost impossible to say what kind of sensory organ 
these tubercles represent, but one thing is certain, that, as they are not 
present in the degenerate apterous form, they must meet some need 
which is experienced during the period when the insect possesses the 
power of flight. It does not follow, however, that these organs are 
directly connected with the actual process of flying; because flight, 
in giving to the insect the power of migrating from one host-plant to 
another, may create a need for sense organs to enable it to recognise 
the correct host-plant suited to its wants. Possibly these organs 
satisfy that need ; without some such sensory apparatus it is difficult 
to see how these winged individuals can find the correct plant upon 
which to carry on their life cycle, for it is not conceivable that such 
an important matter would be left entirely to chance. 
The Tracheal System. 
Owing to its having proved impracticable to obtain good pre¬ 
parations of the tracheal system of the winged stage, no detailed 
comparison of the two stages can be made. From the very imperfect 
preparations that it was possible to make, it would seem that in the 
abdomen, the system is a little more complicated than in the apterous 
stage, but chiefly in the larger number of fine branches given off from 
the larger tracheae. The thickness of the thoracic wall and the 
presence of dark pigment made it impossible to obtain any idea of the 
