IS94-1 353 [Packard. 
"The chitinous portion of the penis consists in general of two canals 
lying within each other, the inner one of which is formed of two 
channels, the npper and the lower. The lower channel forms with its 
end the point of the penis. The walls of the eflerejit canal of the penis 
are snpported by two chitinous bands." 
"From the vagina of Vermipsylla $ pass laterally two outgrowths 
which are attached to a special knob-like thickening of the eighth dorsal 
and ventral plates of the abdomen. Their function is unknown." 
To recapitulate the characters in which the Siphonaptera differ 
from the Diptera, and which seem to prevent our including them 
in one and the same order, we have in the adults many characters 
not wholly adaptational, viz., the absence of a clypeus andlabrum, 
of compound eyes, the head widening behind and articulated bj^ 
its broad base with the prothornx, not being capable of easily 
turning around as in Diptera ; the large hypopharynx forming a 
functional piercing organ ; a |)air of large well-developed lal)ial 
palpi, while the large, long, doubly serrated mandibles are also 
adapted for piercing; the free homonomous thoracic segments, 
without vestiges of wings ; the absence of sternites, the free large 
epimera and coxae, and the ten pairs of stigmata. 
As regards the internal structure, fleas differ fi-om flies in 
having no "sucking stomach," and the proventriculus is liner] 
within with a layer of long, slender, tooth-like projections. 
The footless eruciform or maggot-like larva has a more perfectly 
formed head than that of any known dipterous larva ; an egg- 
shell burster is present ; the brain is contained in the head ; there 
are no coecal appendages of the stomach. 
Tlie pupa is without any vestiges of wings or halteres. 
Hints as to the xikylogeny of the Siphonaptera. — As to the oiigin 
of the Diptera we are at present almost wholly in the dark. Thi§ 
seems due to the fact that they have so widely diverged,, owing 
to the high degree of specialization of certain organs, and the 
atrophy of others, from the primitive forms of the order. It is 
not so with the other orders of tlie higher metabolic insects. 
The Lepidopteraliave apparently descended from the Trichoptera, 
or to state it in another way, a common ancestor may have given 
rise to three orders, more or less affiliated, i. e., the Mecoptera 
(Panorjndae), the Trichoptera, and the Lepidoptera. Tiie 
eruciform larva of the Tentliredinidae suggests that the Hymenop- 
tera have descended from forms not far removed from the 
