128 BOTANICAL GAZETTE: [OCTOBER 
mens, and in some phyllodic specimens, such as that outlined in 
text fig. 7, where the carbonaceous matter is dissolved away in 
the center, the wide extent of the striated, possibly lignified, trans- 
fusion elements of the nerve endings is very clearly seen by the aid 
of the microscope. The nerve endings appear here to have wholly 
usurped the position of the central loculi (cf. fig. 6). The enlarged 
nerve ending is a familiar feature of the sterile pinna of Lyginopleris, 
and probably occurred in that of Heterangium also. The macerated 
condition of some of the specimens may be explained by the condi- 
tions to which they were subjected during fossilization. If we 
accept GorDON’s explanation (8) of the mode of formation of the» 
Pettycur plant deposits, we may attribute it partially to the action 
of thermal springs and warm pools in which the accumulation of 
the plant remains were deposited and gradually infiltrated. 
Text figs. 3 and 4, which illustrate the preceding statements, 
are both mere diagrams based on definite sections showing ripe 
spores. The positions in which the latter are clearly present have 
been indicated by S. They were probably present in other loculi, 
but were welded together into an impenetrable mass, owing prob- 
ably to the presence of some vestige of the spore mother cell, which, 
as in recent cycads, appears to have remained long enveloping the 
tetrad. In the specimen shown in figs. 11 and 12 the spores are 
' free, and there is no welding together of the contents of the loculus. 
In text figs. 3 and 4 sclerotic plates are indicated by groups of 
hexagonal cells. Two vertical plates can be seen in text fig. 3 
(sp), and where plates are incorporated with the wall of a loculus 
they are indicated by s‘p'. Larger horizontal plates are also seen 
in text fig. 4 in the central region. Vascular bundles, their delicate 
ramifications and large nerve endings, are indicated by hatching. 
The wall of a loculus has apparently broken down, as shown in 
the center of both text figs. 3 and 4, and spores are found among 
the swollen fibers. In both diagrams there is strong suggestion 
of segmentation of loculi, as Professor Bower kindly pointed out 
to me. 
Although the distribution of the vascular strands is mainly pe- 
ripheral, at least two may be found in the foci of the ellipse formed 
by a cross-section (text fig. 4, vb,, vb,), but the number cannot be 
