birds from their common ancestors, that is from early Cainozoic times at
latest. On the other hand, the various kinds of such highly specialised
parasites as the warble-flies of the oxen and deer, must have become
differentiated during those later stages of the Cainozoic period which
witnessed the evolution of their respective mammalian hosts.

[13] The 'Little River' beds of St John, New Brunswick, Canada, by some
modern geologists however considered as Carboniferous.

The foregoing brief outline of our knowledge of the geological
succession of insects shows that the exopterygote preceded, in time, the
endopterygote type of life-history. We have already seen that those
insects undergoing little change in the life-cycle, and with visible,
external wing-rudiments, are on the whole less specialised in structure
than those which pass through a complete transformation. These two
considerations, taken together, suggest strongly that in the evolution
of the insect class, the simpler life-history preceded the more complex.
Such a conclusion seems reasonable and what might have been expected,
but we are confronted with the difficulty that if the most highly
organised insects pass through the most profound transformations, then
insects present a remarkable and puzzling exception to the general rules
of development among animals, as has already been pointed out in the
first chapter of this volume (p. 7). A few students of insect
transformation have indeed supposed that the crawling caterpillar or
maggot must be regarded as a larval stage which recalls the worm-like
nature of the supposed far-off ancestors of insects generally. Even in
Poulton's classical memoir (1891, p. 190), this view finds some support,
and it may be hard to give up the seductive idea that the worm-like
insect-larva has some phylogenetic meaning. But the weight of evidence,
when we take a comprehensive survey of the life-story of insects, must