406 BOTANICAL GAZETTE [DECEMBER 
Within this horseshoe-shaped cordon of bundles is inclosed a mass 
of cortical tissue, which to all intents is a pith. Fig. 4 represents 
a cross-section of the bundle system of the pinnular petiole at 
a higher level, where, as in Gleichenia cited above, the edges of 
the horseshoe have fused together. It is clear that as a result 
of this fusion we have a mass of included cortical tissue entirely 
surrounded by bundles, in other words a condition which the 
present writer has hypothetically considered to exist in the case 
of the central cylinder of the stem, where it contains a pith. On 
the right of the fused horseshoe-shaped complex of bundles two 
small strands are seen passing off to supply a pinnule of the second 
order, leaving a gap precisely like that found in connection with 
the departing leaf traces of the stem or main axis. It is impos- 
sible to disregard this obvious resemblance of all the structures 
in the petiole under discussion to those found in the stem. It 
is at once suggested that the interpretation of the two sets of facts 
must be the same, and that if, as is admitted, the petiolar strands 
are able to inclose tissues of the cortex, there is no reason to ques- 
‘ tion that the same process of the inclusion of fundamental tissues 
by the fibrovascular system may take place in the stem. This 
view of the matter gathers force from the hypothesis, which has 
won favor in England and on the continent, that the leaf of ferns 
represents only a modified branch. If this view of the morphology 
of filicinean leaves is adopted, it makes the study of that organ 
of great use in throwing light on the conditions found in the stem, 
especially as the leaf by its well-recognized conservatism will 
supply us with evolutionary stages which have become obsolete 
in the stem proper. 
Fig. 5 represents a cross-section of the entire rachis of M arattia 
alata at a point where the petiolar bundles have formed a closed 
system by the fusion of the originally free ends of the horseshoe- 
shaped complex. Here we see precisely the same conditions as 
are found in the case of Angiopteris evecta. Fig. 6 shows the same 
leaf axis at a higher point, where a vascular supply for a secondary 
axis is being given off. This plane of section is of special interest 
because the large medullary bundle formed from the fusion of the 
two margins of the original horseshoe-shaped complex is func- 
