507 
Evolution of the Angiosperms. 
the nature of anomalies than typical of the Gnetales of the past. The 
embryology of the Gnetales, as a whole, is still imperfectly known, and 
it would not surprise us if further research should result in bringing the 
three genera into closer agreement as regards their embryo-sacs. 
Fossil Evidence. 
In the previous section we have shown that the most reasonable 
interpretation of the fructifications of the Gnetales appears to us to be 
that they are very reduced pro-anthostrobili, aggregated into dense clusters, 
which have often a spike-like habit. If this be true, it follows that at one 
time more primitive members existed, bearing typical pro-anthostrobili, 
which possessed a perianth and both male and female organs, the latter 
arranged above the male. On this view the Gnetales are the last survivals 
of a race, the strobili of which must have presented much variety in form 
and detail, and consequently the group itself must have been of some size 
and complexity. We ought, therefore, to find in the rocks evidence of such 
a stock. Yet such is not the case. So far as we are aware, no reliable 
evidence exists at present of fossil Gnetales. Solms has already expressed 
a similar conclusion in his Fossil Botany . 1 
It is true that some authors in describing fossil plants, especially 
of Tertiary age, have attributed a few remains to this group. In most 
cases these determinations have been founded on very obscure, badly pre- 
served, or fragmentary impressions, which are not worthy of consideration 
in this connexion. The fossils termed Ephedrites Sotzkianus by Unger , 2 
and E. antiquus'by Heer , 3 are such examples. The Ephedra Johniana 4 * 
and E. Mengeana 5 of Goeppert, fossils preserved in the amber of the 
Baltic, are now referred by Conwentz 6 to the genus Patzea of the Loran- 
thaceae, and are certainly not members of the Gnetales. The Palaeozoic 
seeds referred to the genera Gnetopsis , Samaropsis , &c., which Renault 7 
regarded as belonging to the Gnetales, are not now believed to have any 
connexion with that group. The former are with little doubt the seeds 
and cupules of a Pteridosperm. 
Although the Gnetales are unknown to us in the fossil state, we 
are by no means precluded ipso facto from the conclusion that the group 
may have once flourished much more vigorously than at the present day. 
The progress of Palaeobotanical research during the past forty years has 
taught us that many groups, such as the Equisetales and the Ginkgoales, of 
1 Solms (’91), p. 126. 2 Unger (’51), p. 159, PL XXVI, Figs. 1-1 r. 
3 Heer (76), p. 82, PI. XIV, Figs. 7, 24-32, and PI. XV, Figs. 1 « and 1 b. 
4 Goeppert (’83), p. 47, PI. XVI, Figs. 243-5, 2 47> 2 47 a - 
■ IbicL, p. 48, PI. XVI, Figs. 248-50. 
6 Conwentz (’86), pp. 136-9, PI. XIII, Figs. 8-20. 
7 Renault (’85), vol. iv, pp. 176-83. 
M m 
