THE LIVING CYCADS AND THE PHYLOGENY OF SEED PLANTS 1 47 
objection to putting the entire Cycadophyte phylum into one family. But 
when one turns to the reproductive structures, it is evident that the ancestral 
stock, the Cycadofilicales, has either differentiated into two lines, or has 
given rise to the Bennetti tales, which, very soon, gave rise to the Cycadales. 
The spore-bearing structures of the Cycadofilicales may be represented 
diagrammatically : In the center, a crown of much reduced leaves, bearing 
seeds; just outside these, a crown of reduced leaves — but not so much 
reduced — bearing microsporangia. But in none of the Paleozoic forms, the 
Cycadofilicales, is either of these two crowns of reduced spore-bearing leaves 
compacted into cones. This feature marks the Cycadofilicales, for in the 
succeeding forms one or both of these crowns of reduced leaves become 
compacted into cones (Plate VI). 
The Bennettitales and Cycadales are best separated from each other 
by the fact that, in the former, the microsporophylls have not yet been 
compacted into cones; while in the Cycadales the microsporophylls form 
closely compacted cones. In both groups, the ovulate structures form cones, 
except in the genus Cycas. 
The microsporophyll is easily traced, not only from the Paleozoic 
Cycadofilicales, but even from the ferns, up to the living cycads. It was 
the close resemblance of this microsporophyll to the spore-bearing leaves of 
Marattiaceous ferns, as well as the close resemblance in vegetative leaves, 
that led to the earlier geologists to call the Carboniferous "The Age of 
Ferns." 
Throughout the series, the microsporangia are borne on the margin or 
on the under (abaxial) side of more or less reduced leaves. In the Bennet- 
titales the microsporophylls, while much smaller than the foliage leaves, 
still show the pinnate character, with no tendency toward becoming com- 
pacted into cones. In the Cycadales, the pinnate character has been lost 
entirely and, in every genus, the compact cone stage has been reached. 
But the resemblance to a leaf is still seen in the prevailing distribution of 
the sori into two groups, representing the two series of pinnae, one on each 
side of a midrib (Plate VI). 
The structure of the individual microsporangium has changed very 
little since the phylum was differentiated from the ferns. It would be 
interesting to compare the contents of pollen grains of Carboniferous, 
Mesozoic, and living forms; but no satisfactory fossil material has been 
sectioned. It seems safe to say that there were no pollen tubes in the 
carboniferous forms. Engler's term Siphonogamia would not include these 
early seed plants. The small size of the pollen grains, together with the 
absence of the pollen-tube habit, would indicate that the sperms were very 
small and that germination and the development of sperms took place 
very rapidly, as in our living heterosporous ferns. 
The immense size of the sperms in the living cycads is an example of 
giantism which — so paleozoologists tell us — indicates that the phylum has 
reached its limit and is ready for extinction. 
