Also, I find the following overall patterns suggestive. Simplified, the
sequence goes; 1). a class appears, reaches maximum size, then disappears.
If the disappearance is not catastrophic, they dwindle in size. 2). New
class appears, reaches maximum size that is _less than the preceding class
maximum_, then dwindles gradually in size. This exact pattern is the rule
(since the mid-Jurassic, anyway), not the exception, for terrestrial
animals. Dwindling giantism, generally, seems to be the rule in the
terrestrial record (insects, amphibians, reptiles...).
------Interesting. The very largest pterosaurs were present at the very end
of the Cretaceous, so apparently the rule doesn't hold for them.