392 ENTOMOLOGY 
far from primitive, and their ancestors have been obliterated. 
The general plan of wing structure, as Scudder finds, has 
remained unaltered from the earliest times, though the De- 
vonian specimens exhibit many peculiarities of venation, in 
which respect some of them are more specialized than their 
nearest living allies, while none of them have much special 
relation to Carboniferous forms. 
Carboniferous insects are more nearly related to recent 
forms than are the Devonian species, but present a number of 
significant generalized features. Generally speaking, the tho- 
racic segments were similar and unconsolidated, and the two 
pairs of diaphanous wings were alike in every respect—in 
groups that have since developed tegmina and dissimilar tho- 
racic segments. The Carboniferous precursors of our cock- 
roaches, phasmids and May flies have been mentioned. Pale- 
ozoic insects are grouped by Scudder into a single order, 
Paleeodictyoptera, on account of their synthetic organization, 
though other authors have tried to distribute them among the 
modern orders. This disagreement will continue until, with 
increasing knowledge, our classification becomes less arbitrary 
and more natural. 
Mesozoic insects are interesting chiefly as evolutionary links, 
notably so in the case of cockroaches—the only insects whose 
ancestry is continuously traceable. In this era the large fam- 
ilies became differentiated out. 
Most of the Tertiary species are referable to recent genera, 
peculiar families being highly exceptional, while all the Quater- 
nary species belong to recent genera. 
Hemiptera appear in the Silurian; Neuroptera (in the old 
sense) in the Devonian; Thysanura and Orthoptera, Carbonif- 
erous ; Coleoptera and Hymenoptera, Triassic; Diptera, Juras- 
sic; and Lepidoptera not until the Tertiary. 
