SHORT CUTS BY BIRDS TO NECTARIES. 395 
A quite sufficient reason for the upward-pointing direction of the flowers 
of these Grevilleas is probably to be found in the fact that were they to 
point otherwise the honey would spill. This factor is not so greatly present 
in Erythrina with its more or less closed “ honey-box," and is probably quite 
sufficient to account for the fact that Grevillea Banksii, unlike most of the 
other ornithophilous species of this paper, points its flower-openings away 
from the direction from which its birds arrive. Nevertheless, tilted well 
outwards, an open cup full to the brim and without intervening obstacles, 
the majority of them offer no greater difficulty than do those of the other 
species. 
Н. OBSERVATIONS ON Аменовга RHODESIANA, Rendle. 
This, oar common “red-hot poker plant," is exceedingly conspicuous 
when in flower, with its bare smooth stems and long conical flame-coloured 
tops of buds, the lower rows only of which are open. It sometimes occurs 
in hundreds on a single hillside, when it affords quite a striking spectacle. 
The stem is smooth, yet soft enough to yield slightly to pressure, so that it 
probably affords some slight “ grip” in itself to the birds which perch on it 
when the lowest flowers of the raceme * commence to open. They leave 
their elaw-marks on it. Triangular bracts are also present, however, scat- 
tered over the surface of the stem for а short distance immediately below 
the flower-head, and, as I have actually seen, these must also be quite a help 
to the birds in the raceme’s early stages. Later, the first flowers to open 
are hanging fairly flat against the stem, the perianths withered but still 
firmly attached, and they form a dense, firm, brown mat growing longer and 
longer as fresh rows of buds open and wither. It must afford a magnificent 
foothold to the birds. Immediately above this mat are the flowers, already 
trampled and bruised but still juicy, that will be the next to be added to 
it, and these in turn, packed densely, serve to support and so “ give апе” 
to their newly-open neighbours higher up. These once more are usually 
packed close, like cells in a honeycomb, each protecting the next and pro- 
tected from above by unopened buds. А dense mass of stamens projeets 
downwards from them all round, making it impossible for a bird to enter 
a flower without coming into contact with many pollen-filled anthers. The 
honey is very abundant. Slits, above, below, or at the sides, but more 
commonly, I think, below, are present in a larger or a smaller proportion 
of the perianths in nearly every head. Where the flowers are closely packed 
in their normal manner, these slits would certainly prove no hindrance to 
pollination, nor, without examination, would they often be seen, far less 
utilized, by the visiting birds. 
* “Spike” might seem at first sight the better term, but the tlowers are not sessile in 
this species although “ pedicellis brevissimis ” (Journ. Linn. Soc., Bot. xl. (1911) р, 214). 
