Fossils of
Tetraxylopteris have so far been discovered in two locations.
T. schmidtii was named from the Catskill Clastic Wedge, New York State, United States of America, in rocks of Middle to Upper
Devonian age (around ). They have been called "non-laminate proto-leaves" by
Beerling and Fleming, reflecting the belief that such structures were precursors to true leaves, which evolved by first '' – flattening to produce a two-dimensional branched structure – and then 'webbing' – tissue growing out between the flattened branches. One consistent difference between the two species is that
T. schmidtii had branches which very distinctly narrowed along their length whereas only the third order branches of
T. reposana showed any tapering. Another is that
T. reposana had swellings at the bases of first and second order branches. In both species, the primary
xylem strand was central to stems, branches and appendages. In the main stems and branches it was X-shaped in cross-section, corresponding to the four rows of branches. In the final appendages it became circular in cross-section. The
xylem development was 'mesarch', i.e. the first maturing protoxylem had later maturing metaxylem on either side. Protoxylem occurred both at the tips of the lobes of the xylem strand and in the centre. The general anatomy of the woody stem resembles that of
seed plants. The spore-forming organs or
sporangia of
Tetraxylopteris were born on a very complex 'fertile branching system'. Firstly the main axis of the system branched twice dichotomously. Then each of the four resulting branches was three times
pinnate. Each ultimate unit had an elongated sporangium at its end which split longitudinally to release the spores which were trilete, ranging from around 70 to 170 μm in diameter. The complex three-dimensional branching pattern implies that both species would have been open bushy plants. The exact height cannot be determined from the fossils, which consist of broken-off portions. The longest known section of
T. schmidtii is 50 cm – rather less for
T. reposana, which may nevertheless have been the taller species. An overall height of a few metres has been proposed. Hammond and Berry suggest that
T. reposana may have grown in dense thickets so that plants supported one another. ==Taxonomy==