744 CHARLES PauL ALEXANDER 
The type specimen should, of course, be selected only when there is 
absolute certainty of the identification, and in most cases this determination 
can be made only by rearing the species. After the species has been reared 
(this should be done many times, if possible, in order to check up the 
identity), a good representative specimen may be chosen as the typeof the 
stage. In the cases in which the species is known only from a single 
specimen, the nepionotype may be the larval skin, the neanotype the 
pupal skin. The remaining specimens of the original series become para- 
types. The types of the immature stages possess fully the value of the 
type of the adult and should be as carefully preserved. The types herein 
designated are in the collection of the writer. They are preserved in 
alcohol, but the larval heads of most species have been removed, treated 
with caustic potash, and mounted in balsam. 
EXTERNAL MORPHOLOGY 
The larvae and the pupae of crane-flies show considerable diversity 
in their general form. The fundamental plan of structure remains much 
the same thruout the group, but the details are widely different and furnish 
the characters in use for the separation of the various tribes and lesser 
divisions. 
The immature stages of crane-flies have evolved more rapidly than 
have the adult flies, and in many features they show a greater specialization. 
The head capsule of the larva seems to be the most constant feature, 
the same fundamental type of structure recurring in the generalized 
members of all the various groups, indicating a close phylogenetic relation- 
ship. On the other hand, the respiratory organs of both the larvae and 
the pupae vary greatly in the different species and are obviously molded 
by habitat. The often-repeated statement that the inside of an organism 
shows what it is, while the outside shows where it has been, is well illus- 
trated here. 
The larva 
General features 
The form of the larval body is, as a rule, moderately elongated and 
usually terete. The head is eucephalous and non-retractile in the three 
families Tanyderidae, Ptychopteridae, and Rhyphidae. It is incomplete 
and more or less retractile in all the species of Tipulidae. The body 
is shortest in the more generalized forms, becoming greatly elongated 
