3 
Vascular System of Pteris. 
completely separate, while if the development of the branch 
precedes that of the leaf and takes place higher on the petiole, then 
the stretch at the base, common to leaf and branch, has a structure 
like that of the main axis. We shall refer to these points later. 
De Bary in his “ Comparative Anatomy ” (’77) repeats the 
leading points of Hofmeister’s and Stenzel’s descriptions, placing 
Pteris aquilina among the ferns with accessory cortical bundles, but 
adds nothing to our knowledge of the subject. Conwentz (’75) also 
follows Hofmeister, and does not advance the subject. 
In 1890, Leclerc du Sablon (’90) in his well-known paper “ Sur 
la formation de la tige dans les Fougeres ” described the transition 
from the primary root to the stem, up to the point at which 
the vascular system of the latter has the form of two curved 
plates. Leclerc du Sablon did not follow the development of the 
young plant to a more adult stage, but his figure of the two plates 
recalls so strikingly the two inner strands of the adult rhizome 
as to lend great colour to Hofmeister’s erroneous description of 
the inner strands of the rhizome being directly continuous with 
these two primary ones, while the outer circle represented acces¬ 
sory cortical strands arising from the inner by branching. This 
conclusion was however, as already mentioned, entirely upset by 
Jeffrey’s account (’00) of the behaviour of the vascular system of 
the young plant above the level at which Leclerc du Sablon’s 
figure was taken. This shewed quite clearly that the ventral 
primary strand traced upwards, gives off two fresh strands in suc¬ 
cession from its internal face, and that it is these latter which are 
continuous with the internal strands of the rhizome, while the two 
primary strands gradually break up at a higher level to form the 
outer circle. 
Now while this discovery brings Pteris aquilina into line with 
the majority of ferns having a polycyclic vascular structure, in so 
far as it turns out to be the outer ring of strands which is the 
primary one, corresponding with the single ring of monocyclic ferns; 
yet the exceptional fact that about half (sometimes more, some¬ 
times less) of the leaf-trace arises from the inner (secondary or 
additional) strands, remains to be explained. It is clear that in the 
hypothetical ancestor with a single vascular cylinder, the whole of 
the leaf-trace must have arisen from this, and it is the business of 
the morphologist to try to discover the steps by which the supple¬ 
mentary supply to the leaf originated. 
The relation of the leaf-trace to the vascular system of the 
