74 
Psyche 
[June 
the presence of the supposed gill-vestiges just mentioned 
suggests that at least some of them had aquatic nymphs. 
The Megasecoptera were probably no better fliers than 
the Palaeodictyoptera, and their long cerci must have 
handicapped them in their attempts to escape from ene- 
mies. Perhaps this had something to do with their abrupt 
disappearance at the close of the Permian, for no sign of 
them has been found in later strata. 
We next come to the insect dinosaurs- — the Protodonata. 
These include the largest insects known, living or extinct. 
Although all species were large, as insects go, not all were 
giants, as is usually stated; some had a wing expanse of 
five inches, which is well within the limits of many living 
insects. Three very large species have been found, all 
belonging to the family Meganeuridae. One, from the 
Carboniferous of France, was about twenty-six inches 
across the wings ; the other two, from the Lower Permian 
of Kansas and Oklahoma, were somewhat larger, with a 
wing expanse of about thirty inches. The distribution 
of these species, both in space and time, indicates that the 
giant meganeurids inhabited an extensive area of the 
earth for some fifty million years, though the whole order 
became extinct shortly after the close of the Permian 
Period. The protodonates resembled dragon-flies in gen- 
eral appearance, and the earlier forms were probably 
directly ancestral to the true Odonata. They had large, 
toothed mandibles and spiny legs, and were undoubtedly 
predaceous. What they fed on, we can only guess. The 
contemporary slow moving Palaeodictyoptera and Mega- 
secoptera, which, because of their wing structure, were 
unable to hide easily among plants or under rocks, were 
probably their chief source of food. Protodonate nymphs 
are unknown. They were probably aquatic, although Dr. 
August Krogh has asserted that nymphs of the giant 
meganeurids could not have breathed through caudal or 
rectal tracheal gills, as odonate nymphs do, since in order 
to convey the necessary quantity of oxygen to the head, 
their tracheae would have required a cross-section greater 
than that of the body itself. However, passage of oxy- 
gen in the tracheae might well have been sufficiently aided 
