pedeenn | 
LV. 
THE GEOGRAPHICAL DISTRIBUTION OF DRAGONFLIES. 
By GEORGE H. CARPENTER, B. Sc., Lond., F.E.S., 
Assistant Naturalist in the Science and Art Museum, Dublin. 
(Pirate XVII.) 
[Read DecEeMBER 16, 1896; Received for Publication DEcrmBer 18 ; 
Published Aprin 3, 1897.] 
Tue dragonflies are insects specially appropriate for study from a 
distributional point of view. They are an isolated group, so dis- 
tinct in structure and development from other insects which 
resemble them superficially, that most modern entomologists 
regard them as worthy to be ranked as a separate order (Odonata). 
Then they can be traced back to a somewhat remote geological 
period. Wemains of insects referred without doubt to some of the 
existing sub-families have been found in the Upper Lias of South 
‘England, and are numerous in the (Oolitic) lithographic stone of 
Bavaria ; while wings from the Devonian and Carboniferous for- 
“mations are believed by Brongniart to have belonged to insects 
nearly related to dragonflies, to which he applies the name of 
Protodonata. In studying the distribution of the dragonflies, 
therefore, we have to deal with a group of animals which had 
become differentiated into their principal existing types at a period 
when the dominant vertebrates were the ichthyosaurs, plesio- 
saurs, and dinosaurs, and when neither the placental mammals 
nor the higher birds had yet been developed. It is of considerable 
interest to find how well the regions, into which the earth’s surface 
has been divided to indicate the distribution of those modern 
groups, suit that of this ancient order of insects. 
Students of the dragonflies from any standpoint owe adebt of 
SCIEN. PROC. R.D.S., VOL. VIII., PART V. 21 
