32 
A. G. Tansley. 
of the leaf-trace leaves no gap in the stele because in relation to 
the breadth of the xylem ring the trace is too small to cause 
an interruption in the tracheal cylinder. In Tmesipteris however it 
appears, from the recent investigation of Miss Sykes 1 , that the 
sporophyll-traces, or as she would consider them “ branch-traces,” 
do make a gap in the xylem on leaving the stele, while the traces of 
the sterile foliage leaves do not. Professor Jeffrey might consider 
this fact conclusive evidence that the bifurcate “ sporophylls ” are 
really branches, but Miss Sykes shows that the traces of the two 
kinds of organs leave the stele in essentially the same way and 
that the difference between them in the matter of gap-formation 
is of one degree only. Furthermore Professor Bower ('08) has 
recorded a case in the upper part of the sterile region, in which 
the traces of sterile leaves are associated with gaps in the xylem 
ring. 
Again, the affinities of the very isolated Ophioglossales, which 
on Jeffrey’s test would fall into the Pteropsida, are quite doubtful, 
but the balance of evidence tends to connect them with the 
sporangiophoric, i.e., “ Lycopsid,” types rather than with the Ferns. 
Thus the most that can be maintained is that in the vast 
majority of the forms belonging to certain groups which are 
typically protostelic and whose modern representatives at least are 
typically microphyllous, the leaf-trace is attached only to the 
external layers of the xylem and its departure makes no gap, while 
in the majority of certain other, typically megaphyllous, groups a 
broad tubular type of stele—not found in the primitive types of the 
same groups—has been evolved, and the large leaf-trace is attached 
to the whole breadth of the xylem ring so that the continuity of the 
hollow cylinder is interrupted at the departure of the trace. Once 
such gaps have been formed they play an essential part in the general 
construction of the stele, so that in the higher types—the higher 
Ferns and the great seed-bearing series —the constituent strands 
of the leaf-trace appear as springing from the sides of the long and 
wide gap, instead of seeming to “cause” the gap, as in the middle 
(solenostelic) types. An analysis of the factors involved in gap- 
formation has already been attempted in the fifth and ninth 
lectures. 
The whole question resolves itself into the actual size- 
relations of leaf-trace and stele in any given form, as modified by 
the ancestrally determined construction of the stele itself. Thus a 
1 M. G. Sykes. The Anatomy and Morphology of Tmesipteris , 
Ann. of Bot., Vol. XXII., 1908, 
