The ammonoid cephalopods range from the early Devonian to the late Cretaceous,
 a period of some 320 Ma.  Because of their importance for biostratigraphic
 discrimination and their use in practical age dating for this period they have been
 intensively studied.  Major extinctions at the close of the Devonian, end Permian, end
 Triassic and end Cretaceous have long been recognized and linked with regressional
 palaeogeographical events.  The recognition of smaller-scale extinction events is
 relatively new and is especially well shown in the Palaeozoic, when there was a
 simpler distribution of land and sea pathways than in later periods when the influence
 of latitudinal distributions and local provinces was more severe.  Extinction events in
 the Devonian show the nature of the process.  Usually a gradual decline in diversity
 is followed by extinction; then there is a period of low diversity but often
 individual abundance.  Then novelty appears and is seen in new characters of the
 early stages; elaboration and diversification follow.  These fluctuations can often be
 correlated with changes in other groups and also with sedimentological and
 palaeogeographical changes.  Usually a regression-transgression couplet is involved
 with evidence of ocean turnover indicated by anoxic or low-oxygen events.  A new
 family, Sobolewiidae, is diagnosed.
    A new analysis of diversity, appearances and extinctions is made at the family level
 for 2 Ma time units throughout the history of the Ammonoidea.  This record is
 compared with modern attempts to portray sea-level fluctuations and onlap and
 offlap movements of marine seas.  The correlation, even in detail, is impressive and
 gives support for the species/area theory.  But it is argued that temperature, as well
 as sea-level factors, is important.
    The evidence, on both large and small scales, shows an association of evolutionary
 change with palaeogeographical change.  The new evidence does not suggest a role for
 periodicity above the Milankovitch Band level.  Whether or not periodicity is
 involved, such factors seem more readily explained in endogenic earth causations and
 for the present these provide the most parsimonious explanations.