of the mass. ‘he mechanism for dehiscence has not been 
observed. It is noteworthy that the male gametophyte 
in the pollen-grain is multicellular, often with more than 
twenty cells. However, the great majority of pollen in 
the sacs have only five to seven cells visible. 
I should like to call attention to an anomalous feature 
of the young shoot illustrated in figure 3. Serial sections 
(fourteen) through this specimen revealed only one de- 
veloped strobilus on three sides of the inflorescence; the 
fourth side was destroyed when the coal ball was first 
sawn in two. However, there are the unmistakable struc- 
tures of a verticillate arrangement. The vascular strands 
occur in series of four at approximately 90 degree angles. 
Attention is called also to figure 6, which might well be 
from the same individual, although it was cut from an- 
other portion of the coal ball. Since these two specimens 
are quite immature, [ see no reason for placing them in a 
distinct species. It is possible that they do actually belong 
to another species, but only additional material could 
offer conclusive evidence. 
RELATIONSHIPS. 
When one attempts to compare Cordaianthus shuleri 
with other cordaitean fructifications one encounters cer- 
tain fundamental difficulties. The relatively abundant 
preserved material from central France which shows 
structure belongs presumably to the stems known as 
Poroxylon. Most of the older cordaitean stems (as found 
in England, Germany and Holland) are distinguished by 
significant differences in the nature of the protoxylem 
and leaf traces, and are referred to the form-genus Mes- 
oxylon. Strobili attributable to stems of the Mesovylon 
type are very imperfectly known. This uncertainty is un- 
fortunate because all of the cordaitean stems from Iowa 
coal balls, thus far studied, belong to Mesowylon. 
[7 | 
