One of the most highly modified groups of teleosts, the Plectognathi or Tetraodontiformes, is represented by a tiny (5-16 mm) species of balistoids from latest Paleocene marine "shales", silicified, thin beds in dark, laminated clay (Lower Ølst Fm.) in N-Jutland. The clay grades upwards into diatomite, the socalled Mo-clay (Fur Fm.), perhaps crossing the Paleocene/Eocene boundary. The Mo-clay contains a rich fish fauna somewhat different from the one in the "shales" below. The probable sistergroup of the plectognaths, namely the Zeiformes, is found in the Mo-clay: two very small species (10-12 mm) of true zeoids (Bonde & Tyler MS in prep), and another tiny (8-15 mm) species of antigoniid-like fish (caproids, boarfish). The latter also occurs in the "shales" below and is described here in some detail. Like the mentioned zeoids, it is the oldest known representative of its supposed group. It is either the sistergroup of all later antigoniids or it could be a member of the "Squamipinnes"/Acanthuroids, a more primitive, "percoid-like" assemblage often supposed to be related to the plectognath (+zeiform) group. There is character conflict, and the tiny Antigonia-like fossil furthermore has an unusual basal caudal skeleton (Bonde in prep).

Surprisingly, the most advanced of the groups mentioned, the plectognaths, is the only one identified with certainty in the Late Cretaceous. The group is represented by a monophyletic assemblage of three families, each one monotypic, of small (less than 2½ cm), but highly derived fishes, toothless with strong body armour and some strong fin spines.

The phylogenetic analyses of these fishes indicate that much convergent evolution and a number of reversals must have occurred, both in this very advanced group of "higher" teleosteans and within the acanthopts, the spiny- rayed fishes, in general. And it certainly indicates that stratigraphic occurrence or the "fossil record" is a poor guide to evolutionary history.