Notes. 
5H 
in form, and that their breadth is greater than that of the rib, and equal to, or greater 
than, their length. 
No trace of the ligule has been found in the present specimen. This organ may 
or may not have occupied the groove between the successive leaf-bases. At present 
the evidence as to its exact position does not seem satisfactory. 
As in S. scutellata , the ribs of the present specimen, which are about 9-10 mm. 
broad, consist largely of phelloderm, of 2-3 mm. in thickness, probably developed 
on the inner side of a meristematic zone, although no actual evidence of cambial 
activity can be seen. The prismatic fibres of the phelloderm appear to be rather 
shorter and smaller than in A. scutellata , and we have not noticed any evidence of 
their being chambered. 
External to the phelloderm, a small zone of primary cortex is found in the 
longitudinal grooves, as in S. scutellata. The leaf-bases are likewise bracket-shaped 
structures, though, as has been already remarked, of a different shape. The 
external layers of the cortex appear to have been more highly suberized than in our 
previous specimens, three or four rows of cork-like cells being observed in some 
cases. 
The leaf-trace bundle, when passing through the leaf-bases, also possesses 
a double xylem strand , as we found to be the case in A. scutellata. The leaf-traces are 
usually not preserved very clearly, as is frequently the case in petrifactions of this 
genus, and it is only in a few instances that the xylem elements can be distinguished. 
There is some evidence, however, to indicate that the two xylem strands united as the 
leaf-trace passed inwards through the rib. 
The form of the parichnos appears to agree with that of S. scutellata. The 
various stages can be followed, leading up to the state where the two arms unite 
below the bundle ; but we have seen no example in which they are also united 
above, and completely surround the trace. It may be that the absence of tangential 
sections, cut deep through the ribs, will account for the lack of evidence of this 
stage. 
In conclusion we find that S. mamillaris confirms several of the main conclusions 
of our previous paper on S. scutellata , as regards the structure and relation of the ribs 
and the leaf-bases. Although the leaf-bases are entirely different in form, the relation- 
ships remain unaltered. The discovery of a second case among the Eusigillariae in 
which the leaf-trace, when passing through the leaf-base, possesses a double xylem 
strand, closely similar to that found in the leaf Stgillariopsis , confirms Dr. Kidston s 
conclusion as to the leaves of this species, and would seem to imply that this pecu- 
liarity may be of frequent occurrence, if indeed it be not the rule, among members of 
this group of the Sigillarias. 
E. A. NEWELL ARBER. 
HUGH HAMSHAW THOMAS. 
The Palaeobotanical Laboratory, 
Sedgwick Museum, Cambridge. 
