The shoot apex of
Rhoeo discolor
, a spirodistichous species, is described in detail. The mean divergence angle between successive leaf centres is 152°-6. Each new leaf covers only a small arc when first visible, and its flanks then extend round the apex until they meet in the later part of the plastochron. The new leaf is asymmetric, its half which is anodic in the direction of the genetic spiral being the shorter transversely and covering a mean arc of 166°-5 when the flanks meet. Thus the centre of a new leaf does not lie directly over the meeting-point of the flanks of the previous leaf, but above its anodic half. Experiments were made in order to discover the causes on which the divergence angle depends. As a preliminary some stem apices of
Rhoeo
were exposed and marked with transverse streaks of albumen and charcoal. But after several days the streaks had not been bent by the growth of the apex in the way which would be expected on Hirmer’s theory of spiral growth. In the main experiments the central part of
P
1
, the youngest leaf, was cut out completely. It was necessary first to cut down
P
2
and the older leaves of the bud nearly to their bases. The main results were that
I
1
, the next leaf to arise, was displaced towards the missing centre of
P
1
and that the angle
I
1
—
I
2
was larger than the normal, and sometimes exceeded 180° (see figure 6). In some other experiments the central part of
P
2
was removed as well as that of
P
1
In these also the angle
I
1
—
I
2
increased, and in one of them the genetic spiral reversed permanently, but the displacement of
I
1
towards
P
1
was usually less. It is concluded that the position of the centre of each new leaf depends on a balance between inhibiting influences exerted on the stem apex by existing leaf centres. The position of a new leaf
n
depends mainly on the inhibitions coming from the centres of the two youngest leaves,
n
-1 and
n
—2, and to a lesser degree on those from the centres of
n
—3 and possibly
n
-4. The strength of the inhibition thus decreases with the age of the leaf from which it comes and also with the distance from the inhibiting centre. The asymmetries of the leaves that arose after the operations are reported and discussed in comparison with those of the normal leaves. Many of the facts can be explained if it is supposed that the same influences from existing leaves which tend to inhibit the formation of leaf centres promote the extension of leaf flanks round the apex. But the asymmetry of
I
3
in most of the experiments is at present not explained on this hypothesis. The localization and determination of leaves in
Rhoeo
is further discussed and compared with the corresponding processes in
Lupinus albus
and other dicotyledons. It is concluded that in spirodistichous monocotyledons the localization of leaves, depending on physiological inhibitions, is different from what it is in dicotyledons, in which it was previously concluded to be a space-filling process. The process of determination is also shown to be different in
Rhoeo
, since a rather small central part of a leaf is determined first, and from it the determination extends by induction round the apex. In both these respects the experiments on
Rhoeo
, unlike those on dicotyledons, give good support to Richards’s theory of phyllotaxis, which, however, was intended to apply to all groups.
The role of epicormic shoot production in maintaining foliage in old Pseudotsuga menziesii (Douglas-fir) trees
