76 THE AMERICAN NATURALIST. [VoL. XXXV. 
have been described as Ferns, the best known being the genera 
Neuropteris and Alethopteris. 
The earliest of all true seed-bearing plants were undoubtedly the 
remarkable group, the Cordaitez, as to whose affinities there has 
been much discussion. Their remains occur abundantly from the 
Devonian through the Carboniferous. They present certain conif- 
erous features, especially in the character of the secondary wood, 
while, on the other hand, their structure recalls the Cycads, which they 
resemble in the structure of the leaves. Unlike most fossil plants, 
the flowers and fruits have been preserved with extraordinary per- 
fection, even to the pollen grains which are found within the pollen 
chamber, much as in the case of living Cycads. Most extraordinary 
of all, so perfectly are the pollen grains and ovules preserved, that the 
antheridia and archegonia are still recognizable! 
Whether the Cordaitee really represent a type intermediate 
between Cycads and Conifers, may perhaps be questioned, but they 
certainly are one of the most interesting of all the groups of 
fossil plants. 
The Cycads, although occurring sparingly in the later Palaeozoic 
formations, are especially characteristic of the Mesozoic, where, as is 
well known, they formed one of the principal plant types. It is 
evident that the Mesozoic cycadean forms were much more varied 
than the existing genera, which show comparatively little variety of 
structure. - 
While some of the fossil forms approach closely their living 
representatives, both in this character of the vegetative and repro- 
ductive parts, others are extremely different, this being especially 
true of the Bennettiteæ. These combined typical cycadean vegeta- 
tive characters with fructifications of a very different kind, and not 
readily comparable to that of the true Cycads. The seeds, which 
have been very perfectly preserved, show a large dicotyledonous 
embryo, nearly filling its cavity. These remarkable fossils are espe- 
cially abundant in our own Potomac formation, and from Jurassic 
and Cretaceous formations of the Black Hills, from which Professor 
Lester Ward has described many new species. 
Of the true Cycadaceæ, Cycas probably goes back at least to the 
Lias. 
The Coniferæ are but briefly treated. The earliest typical Conifers 
seem to have been allied to the Taxodieæ, to which the fossil genus 
Voltzia of the Upper Permian and Triassic seems to be allied. The 
Permian genus Walchia, which has been supposed to be allied to the 
