Packard.] 
350 
[May 2, 
tlie view tliat the Siphoiiaptera are not a direct offshoot of the 
dipterous order. 
The stigmata. — Altlioiigli Wagner maintains that there are 
tliree pairs of thoracic stigmata, I do not find more tlian two pairs 
of thoracic and eiglit pairs of abdominal stigmata. Eacli pro- 
tlioracic spiracle is situated on the lower and liinder edge of the 
prothoracic tergitc near the 
base of and behind the epime- 
rum. The mesothoracic stig- 
mata are easier to detect, and 
are situated on the hinder edge 
of the mesothoracic segment at 
tlie base of the epimerum. What 
Wagner designates as the meta- 
thoracic stigmata I regard as the 
first abdominal pair. Wagner 
states that they differ in struc- 
ture from those in front, re- 
marking: "The stigma of the 
wing-like scale is constructed 
entirelj^ as in the abdominal stig- 
mata." 
On general principles we 
should adopt the view that there 
can be but two j)airs of thora- 
cic stigmata, as no existing in- 
sects are known to have a pair 
in each thoracic segment,^ 
The structure of the thorax. — 
A charactei'istic almost diagnostic of the fleas is the unusual 
equalit}' in develoi)ment of the segments, i.e., their horaonomy, 
soe.i 
Fig. 34. Longitudinal section 
tlirougli the liead of Brcmla coeca ; 
cl, clypeus ; Ibr, labrnni ; hp, hypo- 
pharynx; sd, salivary duct; oc, 
oesophagus ; v, vesicle ; br, brain ; 
soe.g, snboesophageal ganglion. — • 
After Mwjyenherg. 
' Among dipterous larvae those of Bibio and Dilopiius have ten pairs of spiracles 
but as Osten Sacken says, "this is a ver}' extraordinary nuniljcr, almost unique 
among insects. Nine pairs is tlie maximum number of spiracles for all the other 
peripneustic larvae of Diptera, and the occurrence of a supernumerary pair on the 
thoracic segments is a very exceptional character." He adds a parallel case referred 
to by Erichson of two larvae of Lampyridae from Java recorded in Westwood's Intro- 
duction, V. 1, p. 254, tig. 1, and j>. 259, fig. 1, in which there is, besides the normal 
mesothoracic pair, a supernumerary one in the corresponding place of the metathoracic 
segment. "If this latter pair should prove to be spiracles then those larvae would 
offer a remarkable anomaly in having ten pairs of spiracles." Berlin, ent. zfitsclir., 
1892, p. 450. 
