10 
Siphonophora Rosarmn 
the floor of the front of the head. The lower part of each arm—that is 
from the joint downwards—has attached to it two curved rods of chitin, 
placed one above the other, and projecting one forwards and one back¬ 
wards into the head. The more dorsally placed rod from each side is 
continuous with the expanded base of the anterior seta of that side, and 
the lower rod from each side is attached to the base of the posterior 
seta. In this way, the transverse bar, the arms of the A, and the setae 
themselves are all in direct continuity with one another. This point 
will be dealt with again in connection with the protrusion of the setae. 
In the description of the pharynx it will be shown that the posterior 
wall of the pharynx is strengthened by a strong layer of chitin which 
extends from the mouth aperture to the point where the oesophagus 
makes its abrupt curve over the transverse bar, and in preparations to 
show the chitinous framework, this strip of chitin always remains 
behind and hence it appears in Fig. 1 (Plate I), though it does not 
actually contribute to the support of the mouth parts. The chitinous 
skeleton of the front of the head is raised into an ovoid dome, and on 
the inner surface of the dome the muscles which dilate the pharynx 
have their origin. 
The setae. With regard to these organs, S. rosarum agrees, so far 
as I am able to judge from the literature at my disposal, with most of 
the Aphides. The setae are really four in number, but, as Buckton and 
others have observed, the proboscis or beak only contains three stylets. 
Witlaczil (1882), figures (Plate II, Fig. 5) Aphis platanoides with 
four setae in the beak, but I feel sure that this must be an error. 
Krassilstschik (1893) states that in Phylloxera vestatrix, fusion of two 
setae has taken place, but he does not make it quite clear which pair of 
setae have fused. 
In S. rosarum the setae are arranged in two pairs—an anterior and 
a posterior pair—and it is the posterior pair which fuse to form a 
single median seta-like structure. In Fig. 2 (Plate I) will be seen a 
transverse section across the upper lip or labrum, just where the setae 
are leaving the head to pass into the groove in the dorsal side of the 
beak. The three setae are all cut across transversely, and lie in a 
straight transverse line. On closer examination it will be seen that 
each of the two external setae has a single clear spot in the centre, 
which represents a fine tube traversing the whole length of the seta. 
On looking at the section of the median seta, it will be noticed that 
there are two clear spots, showing that there are two tubes running 
down it. This in itself is very good evidence that the median structure 
