3°6 
Arber. — The Leaf Structure of the I rid ace ae , 
Gladiolus , to Acacia neitrophylla , W. V. Fitz, and A. uncinella , Benth. ; 
Figs. 45 and 46 show the close resemblance between Gladiolus ornatus , 
Klatt, and Acacia incurva , Benth., while Figs. 27 and 28, p. 311, indicate the 
identity in type of Sisyrinchium junceum , E. Mey., and Acacia teretifolia , 
Benth. The similarity of these Irid and Acacia leaves seems to me to 
supply a strong argument against the congenital concrescence view. There 
can obviously be no question of concrescence in Acacia, since here we have 
no sheathing base, and thus the limb cannot possibly be interpreted as 
arising through the fusion of the margins of such a sheath. Yet the 
parallelism between the very various typ'es met with among Acacia 
phyllodes, and many of the Irid leaf forms, is too exact in all essentials to 
be treated as fortuitous ; it appears far more likely that the same 
morphological interpretation applies to both. The evidence of the leaf 
succession in the seedling makes it certain that in many Acacias the leaf 
of the mature plant is a petiolar phyllode, and the balance of probability 
favours the idea that the Irid leaf should come under the same category. 
(iv) Semi-equitant leaves . 
In a previous paper 1 I drew attention, very briefly, to the case of the 
Liliaceous plant, Phormium tenax , Forst., the New Zealand Flax ; further 
work has inclined me to an alternative interpretation of the curious leaf 
structure of this plant — an interpretation which I think is more likely to 
prove valid than the view I tentatively suggested in 1918. In Phormium 
and the related genus Dianella , there is, in the mature leaves, a sheathing 
base (a in Figs. 15 and 22) and a normal bifacial limb (c in Figs. 15 and 22), 
but between these parts there is a short, vertically flattened region ( b in 
Figs. 15 and 22) whose structure and anatomy (Figs. 19 and 23) recall the 
limb — or rather the transition region between sheath and limb — in the 
ensiform species of Iris. Inconsequence of this peculiarity, Velenovsky 2 
describes the leaves of Phormium and Dianella as ‘ semi-equitant ’, and 
claims them as an obvious case of congenital concrescence. To test his 
view I have looked into the ontogeny of the leaf, which I find to be wholly 
different from that of Iris . The very young leaves are entirely open and 
sheathing (Figs. 16 and 17), and this bifacial stage, in which there is no 
vertical ensiform region, continues until the leaves have reached a con- 
siderable size. For instance, in a leaf a little more than 30 cm. long, the 
closed vertical region was found to be just beginning to differentiate itself 
from the sheathing base. In the seedling, also, the first plumular leaves 
are entirely open* and sheath-like (Figs. 21 A-C). The evidence from 
ontogeny and from seedling structure thus appears definitely to support 
Velenovsky’ s contention that we have here a genuine instance of congenital 
2 Velenovsky, J. (1907). 
1 Arber, A. (1918). 
