Sinnott.—Foliar Gaps in the Osmundaceae . 109 
apparent absence of a gap was observed. Gwynne-Vaughan, however, 
claims to have found instances in this species where the leaf-trace departed 
in what he designates as a ‘ protostelic ’, or gapless, manner. The 
interruptions are all short, and the lateral continuity of the xylem is there¬ 
fore relatively great, but no instances of such a gapless condition were 
observed by the present writer. 
In T. superba the leaf-gap in almost every case is narrow, and there 
were numerous instances noted where the trace was well into the cortex 
before a break opposite to it appeared in the central cylinder. Fig. 5 
shows a trace apparently departing in a so-called 1 protostelic * manner. 
Fig. 6 exhibits the condition of affairs a little higher up, where a well- 
marked gap has been formed. Long series of sections were cut through 
the stem of this species and in no case was the absence of a leaf-gap 
observed. In their paper on the anatomy of Todea , Seward and Ford ( 5 ) 
figure a section of the stem of this species showing a leaf-trace departing 
apparently without leaving a gap. This figure is referred to by Kidston 
and Gwynne-Vaughan and by Tansley (6) as a probable example of 
the ‘ protostelic ’ exit of a leaf-strand. In view of the facts above 
described, however, it seems highly probable that a condition is present 
similar to that shown in Fig. 5, and that a gap in the xylem would 
subsequently appear. Seward and Ford do not call attention to the 
absence of leaf-gaps in this species in any other way than by the above- 
mentioned figure. 
In T. hymenophylloides, also, the gaps, or ‘rays’, are narrow in pro¬ 
portion to the size of the bundles. The great majority of traces depart in 
the usual manner. A number of instances were observed, however, where 
the gap was apparently absent for a while, but later appeared. Figs. 7 and 
8, two stages in the passing off of a leaf-trace from the cylinder, show such 
a condition. In this form, also, Kidston and Gwynne-Vaughan mention 
finding cases where gaps were absent. None of these were observed by 
the present writer. 
It would appear, therefore, that as far as the living Osmundaceae are 
concerned, foliar gaps are characteristically present, for in all the species 
where the whole course of a trace could be followed till it was far out in 
the cortex a break in the xylem ring opposite to it invariably appeared. 
The question now arises as to whether this family is primitively phyllo- 
siphonic—whether, as Jeffrey maintains, their ancestors belonged to the 
phylum Pteropsida and had foliar gaps, or whether, in the ancient members 
of the group which possessed siphonostelic central cylinders, the leaf-trace 
departed without causing a break in the continuity of the xylem ring. 
There are two general ways in which the answer to this question may 
be sought. The first, and by far the most satisfactory of the two if it can 
be successfully pursued, is to study the actual fossil progenitors of the 
