tion, as described above, while in the latter these
organs are short, and, as already explained, apparently composed of only
three joints.

Within the divisions named--Orthorrhapha Nematocera, Orthorrhapha
Brachycera and Cyclorrhapha--the constituent families are usually
grouped into a series of "superfamilies," distinguished by features of
structure or habit. Certain extremely aberrant Diptera, which, in
consequence of the adoption of a parasitic mode of life, have undergone
great structural modification, are further remarkable for their peculiar
mode of reproduction, on account of which the families composing the
group are often termed Pupipara. In these forms the pregnant female,
instead of laying eggs, as Diptera usually do, or even producing a
number of minute living larvae, gives birth at one time but to a single
larva, which is retained within the oviduct of the mother until adult,
and assumes the pupal state immediately on extrusion. The Pupipara are
also termed Eproboscidea (although they actually possess a
well-developed and functional proboscis), and by some dipterists the
Eproboscidea are regarded as a suborder and contrasted as such with the
rest of the Diptera, which are styled the suborder Proboscidea. By other
writers Proboscidea and Eproboscidea are treated as primary divisions of
the Cyclorrhapha. In reality, however, the families designated
Eproboscidea (_Hippoboscidae_, _Braulidae_, _Nycteribiidae_ and
_Streblidae_), are not entitled to be considered as constituting either
a suborder, or even a main division of the Cyclorrhapha; they are simply
Cyclorrhapha much modified owing to parasitism, and in view of the
closely similar mode of reproduction in the tsetse-flies the special
designation Pupipara should be abandoned. Before leaving the subject of
classification it may be noted in passing that in 1906 Professor
Lameere, of Brussels, proposed a scheme for the classification of
Diptera which as regards both the limits of the families and their
grouping into higher categories differs considerably from that in
current use.

Little light on the relationship and evolution of the various families
of Diptera is afforded by fossil forms, since as a rule the latter are
readily referable to existing families. With the exception of a few
species from the Solenhofen lithographic Oolite, fossil Diptera belong
to the Tertiary Period, during which the members of this order attained
a high degree of development. I