MOEPHOLOGY.
however, nowhere prevails : the portion of tube betireeii the staminal insertion and the
base, whether short or long, is of uniform calibre. The staminal insertion iii J?, peotinata
and P. tenuirostris is at the base of a transverse indentation of the tube, but this arrangement
occurs nowhere else, and is absent from the very nearly allied P. pyramidata.
The tube in Sii'HONANTHIE and in OETHOEEHYNCH/E is straight, but in many UuYNcno-
LOPH/E and in most BIDENTAT^ it is cm-vedj the concavity being anterior. In ANODONT^
it is very frequently angularly bent either within the tube or opposite the limbus of
the calyx. In some cases this infraction is absent, and the tube may be straight (P.
Kingii) or incurved fP. pijcnmiha). The amount of infractioQ varies somewhat, but the
situation of the angle is constant for individual species.
The lip is always 3-lobed, and usually is rounded in front, very rarely narrowing
gradually from a wide base. In ANODOST^ and BIDENTAT^ the lip is stipitate, the base
being erect and 2-cristate above, and the lobes usually too small to envelope the hood in
the bud. Even when sufaciently large to do so. they are early deflected and permit the
hood to appear. In SIPHONANTH^ and ORTHORRHYNCH^E the lip is sessile, the lateral lobes
overlapping in the bud and quite covering the hood, while the base is narrow and
uncrested. In RHYNCHOLOPHJE the lip may be stipitate with a wide base and with distinct
crests (P. Collcttii, zeylanica), with a narrow base and then with only rudimentary folds
near the base of the lobes (P. Chrhei, Trainiana\ or without folds (P. laclmoglossa);
or it may be sessile (P. carnosa, gruina) and without folds. The lobes are usually
rounded, with the margin entire, or with the midlobe (P. siphonantha), or all three (P.
longijlora) truncate, but are sometimes ovate (P. corymhosa), or obovate (P. Perrofteiii),
Wlien rounded the midlobe is as a rule somewhat narrower than the lateral lobeswhen
ovate or obovate the lobes are nearly always equal in size. In a few cases (P.
Kingii, fragilis) the lateral lobes are narrower than the central; in these instances the
midlobe is rounded while the lateral lobes are ovate. The margin is often ciliate, a
character which is extremely constant in individual species.
The hood consists of two distinct portions—a basal continuous with the tube and an
upper which encloses the anthers. The basal part is always of the same consistence as
the tube and generally of the same width, is always erect and generally of the same
length as the upper portion, though it may be distinctly longer (P. schizorrhynoha), or
may, on the other hand, be almost absent (P. ezcelsa). The margin is usually distinctly
thickened, sometimes sinuate (P. longijlora, furfuracea), sometimes toothed (P. siphonantha,
odontophora), and sometimes replicate (P. Oederi, pycnaniha). The upper part or hood
proper is, except in ANODONTIE, of firmer consistence, and is generally a little wider than
the basal portion. In ANODONTJE it is erect and continuous with the base; in BIDENTXTIE,
most RHYNCUOLOPB^, and in SIPIIONAXTH.E it is curved; while in OETHORRSYNCH/E it is
angularly bent. The coargin in ANODONTJE and in some RNVNCHOLOPILIE is entire or is
toothed only at the lower angle (P. Regeliana) ; in BIDENTAT^ it is toothed immediately
below the apex; while in most RHYNCHOLOPH/E and in all save one species of the
LONGIROSTRES it is prolonged at the apex into a distinct beak. The beak, like the tube,
is late of development, and though discernible in the bud, it is impossible to say from its
condition there whether it will remain rudimentary (P. glohifera), become short and thick
(P. alaschanica), slender and falcate (P. gruina), straight {J?. fiezuosa), fiexuose (P. peciinata),
or circinnate (P. siphonantha). Neither the absolute length of beak nor the relative
lengths of beak and lip or of beak and hood afford reliable characters in diagnosis.
The apex of the beak may be entire, emarginate, or deeply bifid, and this condition, which
is evident in the very youngest bud, is of high diagnostic valuj The ^stivahon of the
beak appears to afford important evidence as to the structural affinities of species. Owing
to the lateness of development of the beak, young buds afford less information than might
be expected But the interpretation of the conditions is materially assisted by various
accidental circumstances. In the S:PHO.A.rH. the flowers in the axi s of the lower
bracts, if they open at all, open later than those higher up. 'Ihese lower unopened
flowers are not, however, necessarily undeveloped; on the contrary, the beak elongates
till it attains it^ full size and the tube also elongates and lies in zig-zag olds imprisoned
in the lower portion of the calyx-cup. Only careful study m the field ca^n_ determine
^.•hether we have here a mere case of centrifugal opening or a condition akin to cleisto
amy. The morphological aspect of the arrangement is, however, as interesting as the
biological, and, so far as the beak is concerned, admits of the following definition : -
of bcalc parallel to median line of lip wilh direction reversed, since the dorsal entire
aspect of the beak is apposed to the lip, and the apes is directed to the throat of the
corolla. It may be expressed by the notation ^ ^ ^ (the arrow R standing for the
beak, the arrow L representing the lip). In all OiiTHORRHYNcn.E an opposite condition
prevails. It is seen in its simplest form in species where beak and lip are of the same
length (P. gracilis, fiezuosa), but is equally evident in species with long flexuous beaks
(P. peeiinata, pyramidata), and ma y b e defined as follows :—ATiS of heak parallel to median
line of lip with direction the same, since the ventral divided aspect of tlie beak is apposed
to the lip, and the apox is directed to the margin of its middle lobe. The notation
R - • (R as before representing t here would be he beak and L the lip).
It is not easy to imagine a condition intermediate between two such oppositearrangements,
but it is interesting to find that in most of the beaked RHYNCHOLOPH^ a
condition exists of which the two arrangements already described might be equally well
derivates. It may be stated thus -.—Axis of beak -perpendicular to median line of lip with
direction at right angles, since the ventral divided aspect of the beak is directed towards
the throat of the corolla, and the apex is apposed to the middle of the lip. This condi-
, K ^ ^
tion may be symbolized as (where R is again the beak, L the lip).
The correlation of beak and tube deserves notice. In those species where the beak
is pronounced the corolla tube tends to become of uniform calibre, in those species where
the corolla tube is infundibuliform there is no beak. This is made more striking owing
to the fact that in the SIPHONANTH^E, of which a tube with uniform calibre is highly
characteristic, species which have a short beak (P. macrantlia, robusta) have also a slightly
widened tube, and in the highly natural groups, so common amongst RnYNcnOLOPHiE, of
which some members possess while others are without beaks, the beaked forms (P.
carnosa, gruina, pohjphylla) have almost uniform tubes, while the corresponding beakless
ones (P. zeylanica, Henryi, tenuisecta) have them infundibuliform. The lip also is involved
in this alteration; as the beak elongates and the tube constrict« its base shortens, its
crests disappear, and its lobes enlarge. Inside the tube in front it is usual to find in
short-beaked specics a number of long hairs; these may be confined to the lower part
of the tube (P. sgylanica), or extend to the base of the lip (P. rex). This condition also
tends to disappear when the corolla tube is contracted, as it is in long-beaked species.
ANN. ROY. BOT. GAUD. CAI.C