The chief peculiarity of spirodistichous phyllotaxis, shown, for example, by
Rhoeo discolor
, is that it is spiral instead of distichous, although each leaf makes contact at its base with the next older leaf only. Previously (1951) the writer suggested as an explanation that in this species the position of each new leaf depends on a balance of physiological repulsions from the centres of the two or more next older leaves; for it did not seem possible to explain the spirodistichy on the basis of a spaceoccupying theory such as had been found to hold good for various dicotyledons. This suggestion was based on experiments in which the central part of
P
1
, the youngest visible leaf, was removed, and the angular positions of the subsequent leaves,
I
1
and
I
2
, were found to differ from the normal in such a way that these leaves appeared to be displaced towards the removed leaf centre. But later it was found that when the central part of
P
2
was removed,
I
1
was not appreciably displaced towards
P
2
, as it was expected to be, unless the part removed was very large. This made the previous explanation very doubtful, and another explanation was sought. It was found by measurements of the angles between older leaves that the growing-point of the apex shifted away from the wound after the operations on
P
1
, but only after the more severe of the operations on
P
2
. It is shown that on the basis of a space-filling theory of leaf determination the shift of the growing-point, acting in various ways, is enough to account for the displacements of
I
1
and
I
2
towards the wound after the previous operations on
P
1
and the displacement of
I
1
after the more severe operations on
P
2
. The question arises therefore whether the normal spirodistichy can be explained without postulating physiological repulsions. It is shown that in
Rhoeo
the flanks of a leaf, extending round the apex, are able to approach the summit of the apex in passing over an obstacle, and it is suggested that the phyllotaxis is spiral because each new leaf,
n
, is caused to deviate laterally by the flank of leaf
n
-1 beneath it, which rises towards the summit of the apex in passing over the centre and axillary bud of leaf
n
-2, a suggestion first made to the writer by Professor G. van Iterson. Also the results of other operations on the apex of
Rhoeo
are reported and are interpreted similarly on the basis of a space-filling theory, as due to the shift of the growing-point acting in various ways, and the peculiar powers of development of the leaf flanks. The development of the axillary buds and seedlings is described and interpreted on a similar basis. So it is not necessary to postulate physiological repulsions in order to explain leaf determination in
Rhoeo
, nor to suppose that the process is fundamentally different from what it is in dicotyledons. The base of each leaf in
Rhoeo
is normally asymmetric, its anodic half being shorter peripherally, but thicker radially. After various operations on the apex the shapes of the bases of the next three leaves that arise are often changed, and a study of these changes has made it possible to suggest an explanation of the asymmetries of the leaves both after the operations and in normal apices. Some changes in the eccentricities of axillary buds are also reported.
Occurrence of circular vessels above axillary buds in stems of woody plants