DISTRIBUTION AND RELATIONSHIPS OF THE CYCADEOIDS 1 63 
The singular Dolerophyllum is another type very difficult to place, but 
with a seed fern antecedent in Pecopteris (fig. 3). The linear spirally 
inserted leaves are not unlike those of the Cordaites. The large pollen 
believed to pertain to these leaves is borne on a large, fleshy, peltate disc 
Fig. 3. Ancient staminate disks: A, Potoniea adiantiformis of the French Carbonifer- 
ous, a Neuropterid bearing staminate cups (natural size) ; B, Linopteris antiqua, also of the 
French Carboniferous, showing under side of disk (several times enlarged) ; C, Dolero- 
phyllum, a fleshy staminate cup (with pollen enlarged at D) ; £, Codonotheca caduca, show- 
ing the microspores (enlarged), and the toothed and symmetric campanula (natural size) 
from the Carboniferous of Mazon Creek, Illinois. A and B from Bertrand, C and D from 
Renault (Seward), E from Sellards. 
Note. The microspores of Ceratozamia are 40 microns long, those of Cycadeoidea 50 
to 100, of Stephanospermum 120, of Codonotheca 300, and those of Dolerophyllum 400 
microns long. All the evidence thus far tends to indicate that ancient microspores were 
large. 
6 by 5 cm. in a series of very elongate pockets more or less regularly radiating 
from th6 eccentric insertion. Whether these pockets are rows of more or 
less confluent sori or synangia is not clear, but possible, since vascular strands 
run between them. If the disc were symmetrical, or could it be shown to 
arise from fusion in a whorl of fertile leaflets, affinity to the staminate discs 
of Gnetum and the cycadeoids would be foreshadowed. Somewhat similar 
discs are seen in the Neuropterids called Potoniea and Linopteris, also in 
