496 
Agnes Arber .— The Phyllode Theory of 
The position of the Farinosae is somewhat doubtful, but the Family 
Pontederiaceae—in which, as we have shown (pp. 488-91), phyllodic 
characters are conspicuous in the leaf anatomy—exhibits considerable 
resemblance to the Liliiflorae. At present no other instances of phyllodic 
anatomy are known with certainty from the Farinosae. It is, however, not 
unlikely that the awl-like leaves of certain of the Restiaceae—unless their 
anatomy is too reduced—might reveal this type of structure on examination, 
and it is also probable that the vertical leaves of Anarthria 1 belonging to 
the same Family, and those of the Xyridaceae and Philydraceae, would prove 
to conform to the Iris type. The writer hopes in the future to test these 
suppositions. 
Phyllodic leaves of the isobilateral equitant type occur, though rarely, 
in the Orchidaceae (Microspermae). 
At present there appears to be no record of the type of structure here 
called phyllodic, from any other Cohort of Monocotyledons, but, until the 
Class has been exhaustively examined from this point of view, it is obvious 
that this result cannot be accepted as final. It may be useful, however, to 
take stock of the present position, while recognizing the inevitable incom¬ 
pleteness of the data. 
The Cohorts in which phyllodic anatomy is at present unknown are 
the Pandanales, Glumiflorae, Principes, Synanthae, Scitamineae, and Triuri- 
dales. On the theory that the Monocotyledons are descended from 
Ranalean ancestors, these Cohorts all seem to represent groups which have, 
on the whole, departed widely in vegetative characters from the original 
Monocotyledonous stock. 
We are thus led to the general conclusion that, so far as our present 
knowledge goes, phyllodic leaf anatomy is most common in those Cohorts 
of the Monocotyledons which, in other respects, seem to retain primitive 
characters. This conclusion appears to the present writer to afford some 
slight indirect support to the phyllode theory. If the ancestral Mono¬ 
cotyledon possessed a phyllode which performed the same functions as those 
of a Dicotyledonous blade, it is conceivable that the leaves of its more 
advanced and modified descendants might eventually lose those vestigial, 
anatomical characters which originally branded them as petiolar, and that 
they might ultimately approximate by homoplastic convergence to the 
structure of a Dicotyledonous leaf. There is also a second phyletic course 
which may in some cases have led to the loss of phyllodic characters. This 
is the occurrence of a further degree of reduction in the leaf, involving the 
disappearance of the petiolar portion and the retention of the leaf-base alone; 
in the latter region inverted bundles are characteristically absent. 
1 Brown, R. (1810). 
