no Sinnott.— Foliar Gaps in the Osmundaceae . 
family. The second is to examine the structure of those portions of the 
living forms which are known to be retentive of ancestral features. 
Without going into the vexed question of the origin of the Osmun¬ 
daceae, and leaving aside all such protostelic members of the group 
as Zalesskya and Thamnopteris , in which, obviously, leaf-gaps cannot occur 
and which therefore have little bearing on the present question, we find 
in the genus Osmundites , which possesses a medullated cylinder with 
a parenchymatous pith, a close and suggestive parallel to the conditions 
in living forms. There are five species at present included in this genus, 
the structure of all of which is now well known. Of these, perhaps 
O. Skidegatensis is the most striking. This fossil was first described from 
the Lower Cretaceous of Canada by Penhallow ( 7 ), and has also been 
investigated by Kidston and Gwynne-Vaughan ( 3 ). It is very large, 
possessing a wide pith surrounded by a central cylinder which has internal 
as well as external phloem. The method of departure of the leaf-trace 
from the cylinder is particularly noteworthy. It differs from the condition 
found in any of the other Osmundaceae in that the continuity of the whole 
central cylinder is interrupted, and the tissue of the pith becomes directly 
continuous with that of the cortex, and not, as in the other forms, with the 
pericycle-parenchyma only. The leaf-trace thus goes off in a fashion very 
similar to that found in Adiantum pedatum or any other of the ordinary 
siphonostelic ferns. 
Osmundites Gibbiana of Kidston and Gwynne-Vaughan, O. Dowkeri of 
Carruthers, and O. chemnitziensis of Unger closely resemble one another 
in stelar structure. The general type of cylinder is very similar to that 
found in O. regalis or T. hymenophylloides. Foliar gaps are always present, 
though they are usually rather narrow, and the stem-bundles are con¬ 
sequently close together. 
In O. Dunlopii of Kidston and Gwynne-Vaughan, however, a nearly 
continuous and unbroken ring of xylem appears at first sight to be present 
around the pith. The figures presented by the authors show this ring 
to be deeply constricted at intervals, and it seems entirely possible that 
very narrow ‘ rays ’, such as we have described in several species of Osmunda 
and Todea , might have originally occurred at these constrictions but might 
be undemonstrable in the present indifferent state of preservation. There 
are, moreover, a number of actual breaks in the ring which, according to these 
investigators, are probably due to accident. Some of these may also possibly 
represent true foliar gaps, or * rays ’. Several figures are presented show¬ 
ing the departure of the leaf-trace from the cylinder. They strikingly 
suggest the condition in some of the forms which we have above described, 
where the gap does not appear for some time after the trace has broken away 
from the stele. It is noteworthy that where the trace is figured as just 
departing from the cylinder there are but slight indications of a gap, but that 
