^ MO E P H O L O G Y .
rise to an appearance often described as crenate. A truly crenale margin (P. ColkUii) is
exceedingly rare.
The disc is not enlarged at all in many species, but in the larger moiety it is
slightly enlarged anteriorly. This enlargement is not as a rale conspicuons, though in
one Indian species (P. ra) it projects as an almost vertical process from two.thirds
to qnite as long as the ovary, and protrudes in front between the margins of
the completely cleft calyx. °
The corolla consists of a tube, sometimes of uniform calibre but usually expanded
upwards, whereon the stamens are inserted. The limbns is divided into a patent 3-fid
lip (laUum) below and a compressed hood (¡alta) enclosing the style and stamens and
prolonged or not into a beak {roslrum) above. Its aistivation is that characteristic of the
EOPHRASIEJ!. The midlobc of the lip is outermost in the bud, the hood is inmost, and
where the lip is large enough to cover the hood completely, the left lateral lobe 'overlaps
the right. The beak, when present, is somewhat variably disposed. The tube, sometimes
even shorter than the calyx, usually equals or somewhat exceeds it, and at times
{as in most SlPHOSiHTlM and some EHYNCHOLOPHJ!) it is long and slender and much
exceeds the calyx. Even when ultimately longest it is late of development, the hood
and lip, as well as the anthers and ovary, attaining almost their full size before it begins
to elongate. In young buds of all species the filaments are inserted close to the base,
and where no great elongation takes place the staminal insertion is not carried any great
distance up the tube, very frequently only to a point opposite the top of the ovary"and
in a few cases (P. JSngii) not quite so far. The same condition is met with in those
cases where the elongation of the tube takes place chiefly in the portion between the
staminal insertion and the limbns (P. corpiibma, feclitahi, gmim). Where, on the other
hand, the elongation is uniform throughout the tube we obtain the condition most usual
among OETHOERHYNCILE (but also very common in other sections) of staminal insertion in
the middle of the tube, while in those cases where the elongation takes place only
between the staminal insertion and the base of the tube we obtain the staminal insertion
at the top of the tube—A condition that characterises the SIPHONANTH/E.
The amount of elongation varies considerably, as the subjoined table shows, and
does not therefore afford a reliable diagnostic character, but the portion that elongates
is always the same, and therefore supplies a character of the greatest value.
COROLLA.
TABLE 11.—FIRIAIION in hngth of Corolla tuie in I«iim sfeeiea of § SIPHOKAHTHA!.
I. II. III. IV. 1
s 11 n o s AST Haï.
LIMITS or TABIATION IN THE
SPECIES—
" A L L
LIMITS OF T
IKDITINUAL Localities of the specis
mens in column IV.
Of corolla tube Of calyx.
Of corolla tube. Of calys.
1. EusipUonanthoe.
P, bicornuta
elephantoides
loDgiflora
Degalantlia
rhiimntho'des
siphonantha
bella
1 Przewalskii
18—26 mm.
18 mm.
33—74 mm.
21 - 94 mm.
1 4 - 2 7 nun.
23—60 mill.
2 6 — m m .
28—34 mm.
16—! 9 mm.
14'5 mm.
10—11mm.
15—20 mm.
1 1 - 1 3 mm.
9 - 10 mm.
15 mm.
10—12 mm.
1
3 4
3 3
2 1
6S
3
21—26 mm.
1 8 mm.
42—55 mm.
4 0 - 5 4 mm.
4 4 - 5 2 mm.
1 4 - 2 1 mm.
3 0 - 8S mm.
2 8 - 3 0 mm.
16 mm.
14-5 mm.
10 mm.
10-5 mm.
16 mm.
12'5 mm.
9-6 mm.
16 mm.
10 mm,
PauKt.
Kashmir. ,
Alpine Sikkim.
Chutubi.
Buck-checm, Chumbi. ,
Pan^i.
Lebong Pass in Ka- i
Cliuni-k-ree.
Phari.
3. Maorantlioe.
Daltoni
Elwesii
Garckeana
macrantha
Î 7 - 2 3 mm.
13—16 mm.
1 6 - 2 8 mm.
10—13 mm.
12 mm.
11—17 mm.
14—17 mm.
13 mm.
7
10
7
8
1 8 - 2 0 mm.
18 mm.
1 6 - 1 9 mm.
11—13 mm.
13 mm.
11mm.
16 mm.
13 mm.
BoEkia-la. 1
Phari, 1
Tan-ka-la.' 1
Nampa Gadh in W. •
Nepal. •
Kogting io E. Nepal.
Alpine Sikkim.
Jongri. 1
nej^alonsis
ScuOjana
21—2« mm.
1 1 - 1 2 mm.
3 0 - 2 2 mm.
14 mm.
1 1 - 1 2 mm.
18—19 mm.
1
3
5
^^ nX' .
2 0 - 2 2 mm.
14 mm.
11 mm.
18-5 mm.
3. Gyrorrliyncha.
integrifolia 17—24 mm. 12—13 mm. 4 17—21 mm. 13 mm. Phar;.
The figures in coliinin III indicate the number of gatherings from -n-hich the measurements given in column II are
taken; to iiive the number of specimens measured in every case seems unnecessary. Measurements of over 200 specimens
of P. siphonantha have been made, and measurements of 23 specimens of P. nepalen^s. The only species in the
list whereof the measurements must be considered incoaclusive is P. e/ephantoides, of which but one gathering containing
only two specimens has been available. The localities in column V are those of tiie indiTidual specimena whose
measurements are given in column IV.
This table has been limited to § SlPHONANxnAi because this is the section of which an
elongated tube is most characteristic, and is the one in which most use has been made of
its morphological index—the relative length of calyx and corolla tubes. The table shows
that the character is too variable to be capable of other than a very general application.
The morphological index of the portion of tube that elongates—the point of insertion
of the stamens—affords, on the other hand, a character of high diagnostic value.
When the staminal insertion is carried to the apex of the tnhe, the calibre is uniform
throughout; when not carried so far, the tube, as a rule, enlarges slightly upwards from
this point. This enlargement takes place even among SiPHONANTH.iE in those species
where the staminal insertion is some distance below the apex (P. maeraniha, Scullyana,
Daltoni). But there are many speeies, more particularly among ORTHoERHYNCHAi, where
the staminal insertion is at the middle of the tube (P. flamsa, ehumhiea), and some
both amongst ORTnoRRHYNCHAi (P. peciinata) and amongst RllYNCHOLoPHAi (P. ffruiva),
where the staminal insertion is opposite the top of the ovary, yet with the tube cylindric
throughout or only very slightly enlarged at the apex. The contrary condilion,