Polycycly. 237 
The leaf-trace (Fig. 80) is formed of an abaxial arc of from 
four to six strands (e) and of adaxial wings, sometimes, according to 
Mettinius, formed from two strong lateral strands, each of which 
immediately splits into several (a, b). The same investigator found 
that two strands (k, k,) from the second cylinder occasionally con¬ 
tribute to the leaf-trace, either directly, or by anastomosis with 
strands of the outer cylinder. These second-cylinder strands appear 
to occupy a position as the terminal strands of the adaxial wings of 
the leaf-trace (cf. Matonia). It should be mentioned that Miss 
Shove failed to find these strands in,the specimen she examined. The 
number of the leaf-trace strands rapidly increases by branching, and 
internal strands (i, i,) arise from those of the abaxial arc as in the 
other genera. 
Thus in the Marattiaceae we find the internal vascular strands 
of the stem playing exactly the same part as in the Leptosporangiate 
Ferns. In the simplest forms there is one internal strand only, 
which runs from leaf-gap to leaf-gap as in the simplest polycyclic 
Leptosporangiate types such as Dennstaedtia adiantoides. In the 
more complicated cases the internal system increases in complexity 
in relation to the increase in size and crowding of the leaves, till 
the highest grade of complexity is reached in Angiopteris. The 
gap-filling segments of the second cylinder are sometimes drawn 
into contributing to the margins of the leaf-trace, as in Matonia , but 
typically they do nothing more than fill the gaps in the outer 
cylinder caused by the departure of the leaf-traces. 
On a review of the phenomena of polycycly, which we have been 
surveying, it is apparent that the primary stimulus to the development 
of an internal vascular system is the need for providing additional 
conducting or water-storing tissue at the leaf-gap, when the 
departure of the trace largely diminishes the conducting capacity 
of the stele above that point. In some cases the accessory system 
is drawn into supplying the leaf-trace directly, owing apparently to 
the ease with which parts of a petiolar system undergoing 
progressive complication may become attached to the internal 
system at its point of junction with the edge of the gap. In other 
cases ( Pteris ) the main function of the internal system is to supply 
directly the complications of the leaf-trace, of which it then appears 
to be a decurrent development, but in such cases the internal 
system always seems to help fill the leaf-gap, though it may be in a 
very small degree, and the possibility is not excluded that that was 
