supplied by a very curious tendency often visible in cases of intense
and extreme specialization. When an animal form becomes highly
specialized, the type at first, because of its specialization,
triumphs over its allied rivals and its enemies, and attains a great
development; until in many cases the specialization becomes so
extreme that from some cause unknown to us, or at which we merely
guess, it disappears. The new species which mark a new era commonly
come from the less specialized types, the less distinctive, dominant,
and striking types, of the preceding era.

When dealing with the changes, cataclysmic or gradual, which divide
one period of palaeontological history from another, we can sometimes
assign causes, and again we cannot even guess at them. In the case of
single species, or of faunas of very restricted localities, the
explanation is often self-evident. A comparatively slight change in
the amount of moisture in the climate, with the attendant change in
vegetation, might readily mean the destruction of a group of huge
herbivores with a bodily size such that they needed a vast quantity of
food, and with teeth so weak or so peculiar that but one or two kinds
of plants could furnish this food. Again, we now know that the most
deadly foes of the higher forms of life are various lower forms of
life, such as insects, or microscopic creatures conveyed into the
blood by insects. There are districts in South America where many
large animals, wild and domestic, cannot live because of the presence
either of certain ticks or of certain baleful flies. In Africa there
is a terrible genus of poison fly, each species acting as the host of
microscopic creatures which are deadly to certain of the higher
vertebrates. One of these species, though harmless to man, is fatal to
all domestic animals, and this although harmless to the
closely-related wild kinsfolk of these animals. Another is fatal to
man himself, being the cause of the "sleeping sickness" which in many
large districts has killed out the entire population. Of course the
development or the extension of the range of any such insects, and any
one of many other causes which we see actually at work around us,
would readily account for the destruction of some given species or
even for the destruction of several species in a limited area of
country.

When whole faunal groups die out over large areas, the question is
different, and may or may not be susceptible of expl