Stele in Two Species of Dip ter is, 7 67 
then gradually the internal endodermal and the sclerenchymatous cells 
disappear. 
Thus the behaviour of the petiolar stele was in this case quite different 
from the one described before. In the first one the internal sclerenchyma 
was at first in connexion with the cortex, and the internal endodermis was 
continuous with the endodermis surrounding the stele ; in the second petiole 
the internal sclerenchyma and endodermis are formed in the stele itself, and 
have no connexion with cortical tissue and outer endodermis. In the last- 
mentioned petiole the stele was never horseshoe-shaped, for the endodermis 
is never invaginated on the adaxial side. 
In a third petiole the different tissues behaved more like those in the 
petiole described first, but here the loop of endodermal cells surrounding 
the sclerenchymatous cells is never separated from the external endodermis. 
The two limbs of the horseshoe-shaped stele approach each other, separate, 
and approach each other again. The inner sclerenchymatous cells, or the 
ones in the curve or in the loop, have often thicker walls and are smaller 
than the cells of the cortex near the stele. 
In many petioles of approximately the same age as the ones described, 
no internal sclerenchyma or endodermis was found. The changes described 
take place rather suddenly ; it may therefore be that in some the structure 
was present in a small part, but not found. Seward and Dale 1 have found 
similar internal fibres in the stele of the main ribs of the leaf of mature plants. 
As already mentioned, Matonia sarmentosa has in its mature petioles 
internal sclerenchyma at the base, while on a higher level a horseshoe¬ 
shaped stele is present. Perhaps some may think that the resemblance of 
the structure in some young plants of D. conjugata with the mature M. sar¬ 
mentosa, and with some species of Gleichenia , must be regarded as an ancestral 
character, and that it adds support to the view that Dipteris , Matonia , and 
Gleichenia are related forms. But the irregularity of the behaviour of the 
internal tissues, the different ways in which they are formed, their rare 
occurrence and their presence in old leaves, are all facts which suggest to 
the writer that too much stress should not be laid on this character. 
It seems to me possible to explain the facts as follows. When the 
petiole is getting larger and the limbs of the horseshoe become more diver¬ 
gent, the central tissue gets a tendency to form sclerenchyma. In connexion 
with this, endodermal cells are formed separating the sclerenchymatous cells 
from the other tissues. According as different parts of the stele are affected 
the different structures occur: i.e. the internal sclerenchyma is surrounded 
by an endodermis in the centre of the stele or is attached to the outer endo¬ 
dermis or directly to the horseshoe-shaped strand. In older petioles, in which 
the limbs of the horseshoe are more divergent, the stele has always the 
horseshoe shape. We find a similar case in different species of Gleichenia , 
1 Seward and Dale, loc. cit., p. 498, Fig. 38. 
3 E 2 
