8 8 ATSNALS OP EOTi t BOTiSIO GAEDES, CALCUTTA.
ot which become. i«voM ta the inleredt;.! growth of the totu., so th.l the .ngle, of the letter
hecomo wmsed- in the S i t t m ipocimeM thi. wiiiK-lii. appond.fc ii, however, wonting, lo that it
mnnot be made ».e of .s . goncrio ohoi.ot.r; it i, hardly ot .nfhoieat importano. to found a
. e p » . t . specic. on. 1! C.m, be upheld a> a g««» the definition, of CM,a, CaMhoi.., and
J'rol.ius would te aa follows:—
C.iMc, Lim.ilcaT.e entire; »eotariferon. .iaminode. none; lorn, b.lwoen th. i».,tion of Ibo
c a l j i and th. gjnt«oinm not elongating after JowBring; seeds nsn.lly inserted aU along the ventral
snture; chalazal part of seeds more or less spongy. . .
Cafcifcio H f ct T.; leaves palmat.ly divided; nectariierons staminode. none; torns elongating
after flowering' between the insertion o! the calys and the bead of follicles; seeds inserted below th.
middle, chalazal part not spongy. ^ i i- v
Trcmu. Linn.; leaves palo. t . l y divid.d; neetariferou. st.m.nodes present; toms not elongating hj
iatersiitial growth; so«!, inserted .11 along th. ventral snture of the follicles, c h a W part not
spongYya. rieties of Cnhthoda piilmata—
AsABrni«.!, F.B.-. ovaries esappendioul.te, (Flowers soUtary.) Silkim, 10,000'(iioofar
«PraniooiAT» P if.; ovaries and foniolcs dorsally with a wing-shaped appendage
' neat' the ha..', Inflorcso.nc. racemo«, or panienlate. Hiipeb (Hmrj,, 1886-86, JVo.
6977 !) See also Hooker's Icones, plate 1935.
, , K f TK 1 flowermii plant from Sitkim; 3, bead of follicles of rsrietj 3,
Yar.
Tar.
.dg, • rip« MM" i»' belo"
T r o l l i u s PÜMILU3, Don.
B i . genu. IroU,,. is eomposed of two form-gronps-one group wbieb is ohar.ot^od b, the
n , . t ^ ror"t.mii.ode. being usuahy long-el.wed, and more espooially by the„ b . . g «neüy
r 1 nrf at the npoex edge ; whilst the staminode. of the second group have a very short daw .nd
t i ir ^ t « m l j i . tSl k . the margin of an ordmary petal. The former comprises the
the r T P " ™ " M the difflcuUly ..parable speoie. T. ««•oyae,.., TZZTtI L » 1 " . - t b tb.« - » » p - " "' i '™- ^ r t
f n g T the diserimioatio. of th. North.ru Asiatic and European spec... weuld be
ftccordiag to tke following ecbeme:—
A, Flores lUacini
B. i'loreä fiari vel auranliaci—
a. Styli discolores
b. Styli concolores—
a. Pétala stamina superaiitia
ß. Pétala etaminibus aubcequilonga
vel ijs gubbreviora
f . Pétala staminibia brevlora
... T. ¡¡¡acinus.
... T. ultaicui.
... T. aiiaticus.
... T. Inxui.
... T. euTopacus.
DKSCEIPIIONS OP NEW AND RARE INDIAN PLANTS. 8 r
wliicli the etftmiiijJes ato only slightly longer than the st'imons. la T. cwopaeus the staminodes
appear always to bo a little shorter than the stamens ; thoy are also moro distinctly obovate.
After all it raay be better to fall back on Lhinams' way of distinguisiiing T. (uropneus from
T. asiatious by the sepils bding coiinivent in the firmer aud patulous or spreading in the latter,
although it must be coiifesse.l that this method has its difBculties. liegel's T. i/a.'iwigaricus will
then liave to be separated agaia from T. evropacus. Its affinities are perhaps more with T. /«xk«.
The character by whch T. aliaic,,^ is separated from 1'. asiaticm is hardly of sufficient weight to
prevent its being considered a subspecies of the latter.
Of how little importance the lelalivc length of ovary and style as a diagnostic character may be, is
exhibited by fig. 5 to k, all of ivhich are from Sibeiian and doubtlessly good specimens of T. asiaticusx
compare, for instance, h with k. This has an important bearing on T. acatdis as a species of the
first order. Ubre the p¡siil has very commouly the shape depicted in fig 5, c; but in a specimen,
for instance, irom PhQldlariJ in the Nlla valley tlie pistil is exactly lilce 5, 5, which represents
the pistil of T. a,„'i-iB-tius iliihlb. from North América. Altliough the majority of specimens of T.
acnulU are similar to those described and figui^ed by Liudley in tho Botauiual Eegister of IS42 and
1843, tlie name has become singulaily inappropriate, siuco leafy specimens as much as half a meter
in height have buen gathered in the North-West Himalaya. As a i-ule the Siberian forms of T. laxiu
have the blade of tlie siamiu.ides linear-ob'ocg (see fig. h), whilst iu T. araiilU they are usually
obovate or obovate-oblong isce S, d, e , f , y) ; but in the Anierioan varieties of T. I'xus, the nectariferous
staniinodes are like those of acaulis. As a matter of fact, T. aeaidis is connected with 2\ hxui by
a series of iuteimediate forms, especially frcm Gilgit, some of which ai-e hardly distinguishable from the
true T. lazu,. That T. patula, and T. c.ucaoica, are n.thi.,g but varieties of T. kt.us is pretty
generally admitted.
i l u c h importance should not ba attributed to the persistence or otherwise of the sepals; they
are very oumnoaly so in T. pumUu-^, but the character is not a constant one; the sepals of
T. acauHs are mostly deciduous, but specimens may he observel in which the head of ripe follicles
is accompanied by maxcescent sepals The relative lenj'h of ovary and style is freely made use of
by Ledebour, but the character U of little value, for we find the length of the style in terms of
the length of the ovary to be in specimens of T. i, f.
Remarkable Ls the close resemblance of T. Uocmu^ to the typical T. acanlü-, the colour of the
sepals of T. Idaciiw would hardly provumt its being unite! with T. laius, the more so as the
sepals of T. americanas are suffused with a pinkish hue on the outside, and in the variety alba from
the Rocky Mountains they are white. Tha only character by which it is possiblo to distinguish
T. clUacinm from all the other species of TrMiis is the abáeuce of glanlular papill» on the pislüs. a
character which, however, is hardly of sulhcient value to make it a subspecies as is done by Prantl.
VVe would thus arrive at the following grouping of the various forms of TroUiui:—
I . Staminodes manifestly thickened at their upper edge, with long cIa^vs.—y. pumiluB, Dou.
I I . Staminodes quite petaloid, flat at their upper margiu, with short claws.
A. Sepals patulous or spreading:—
a. Staminodes shoiter or somewhat longer than the stamens, obovate c
o. Carpels papillose. Sspals yellow or white. - T. laxus, Salisb.
Carpels glabrous. Sepals lilac.—T. Idaoinus, Bungo.
r obovate-oblong.
b. Staminodes mostly distinctly longer than the stamens, linear or linear oblanceolate
or style of a different colour to the ovary.—T. asiatiats, Lin.
P.. Sepals connicent, staminodes usaally somewhat shorter than the stamens.—T. ,
Linn.
. europiKUs,
We now pass on to a more detailed description of the forms whicli constitnte the spociea T. pumHus.
Subspecies.
^ Sobapccics I. AK-EMOKIFOUD«, P . ; slor>. spnrsoly folints; bkde of basal leavo,
.r.lobod to noar the base, nsnally 3 - 5 cm. long; the seg.iicl.ts sub-distant; the inijdle
ASN. EOV, Iot. Oasd. Cai.o., Toi. V.