2 ñ 6 STE-YOSTOJXA,
Tliis table shows that in 182S Dr, Wallicla issned two gatborings of G. panifloviim, one tolonginfj
to Hoyno's Herbfti-iiim and one obtained by. his Collector, DoSilra, in Silbet, The former, being
n. 2158/1 in tbo Cataloyae, is of necessity tbo true type of G. pnrtijiorum. Its exact locality, it
ought to be observed, is unknown. If it had, liko nearly all Hoyne's otbor speoimens, been oollooted
in Soutbem India, it is extremely sti'ange that it should never bavo been met witb by subsequent
colleotors. And Mr. Dontliam's opinion that it is fi'om Nepal is undoubtedly unfounded ;
"Wallicb does not state tliat Heyne's ploiit came from Nepal, indeed Heyne had no means of obtaining
Nepalese plants ; "Wallieh does not bimself appear to baye obtained any form of G. parviflonini
in Nepal, nor does lie anywhere say that he did; finally, if any fonn of G. pani/lonim is ultimately
found in Nepal, it is reasonable to espect that it will be vau. farhma (Bentbam's G. imiltijlorfm) tbat
will be met with there, since this is the only form that has hitherto been roported from SLkkim or tbo
Eastern Himalaya. So far as is tnowu, var. typica is confined to Indo-China—io the caH of the rirer
Brahmaputra; all the gatherings up to 1890 are from Lower Assam, Manipw, the Lnshai lulls, Chittagong,
the Cliindwin -valley. Lower Burma, and Siam; as Hejiio obtained plants from both the Indian
peninsulas, it is almost ea-tain that his plant came from some part of the area indicated above. Tlio
second gathering forms only part of Cat. n. 21S.'</2; it is the pait marked " Sillot " in tbo type Herbarium
and, though eonspeciiie witb n. 21Ô8/1, it constitutes a distinct variety and is the same as 2258/D
{^Leonurm farhmns Ham. Mss.) described above as G. paniflorum vak. farinosa. This is tbo plant tbat
was by Mr. Bentham (1835) aceoi-ded specific rank under the name G. muUifionm. The lemainder of
Cat. n. 2158/2, maa'ked in the type Herbarium " Gentea Mts.," agrees exactly witli the plant issued
as Cat. n. 2159/1 (1828).
I n the same yeai- (1828) Dr. 'Wallicb issued two gatherings of G. criintum, one collected by
himself in Penong in 1822, the other obtained by his Collector, Gomez, in Tavoy. These T\-ere publisbsd
as 2109/1 and 2159/2 respectively. But in 1829 Dr. WalHeh bad noticed that n. 2159/1 was the saaie
as tbe " Gentea Mts." portion of bis n. 2158/2. He therefore re-is.sued it in tbat yeax (as he explains in
nth. Cat. p. 92) as 6. parviPruni Cat. n. 21Ô8/C. So that in 1829 only 2159/2 was left as the typo
cf G. crinitum. But 2159/2 differs from 2159/1 (eonverfed into 2158/0) only in having condensed
whorls—a cbai'acter of no value—and in having glabrous calyces—an accident and not a character at
all—aud tbe two are ccrtainly conspeciflc. If, therefore. Dr. Wallicb rightly referred 2159/1 to (?. pwHjlnrnm,
there is no room for the species G. a-inUtm at all. But if there be a species G. crinit'-m—
and it must bo here repeated tbat 2159/1 and 2159/2 aro undoubtedly conapecifio—then there was no
room for the re-issue of pai-t o! that species as G. paroiflorum because it happens to agree exactly with
the "Gentea Mts." portion of n. 2158/2. For this very "Gentea Mts." plant agrees badly with
n. 2158/1, which is the type of Q. parciflorum, and agrees still less with the rest of 2158/2, along with
•n-bieb it w'as issued. At the time that Dr. Wallicb made this change in the designation and number
of 2159/1, he issued three more gatherings as G. pnrviji,„;„n. Two o! these appear as 2158/D, botli
•were collected by the same botanist (Buehanan-Hamiltnn), at the same place (Goalpara in Assam),
though at d.Serent seasons of the year. Dr. Buchanan-Hamilton named both "Lconurus farinosus, "
find as a matter of fact they are identical, and aro moreover the same as the "Sillet" portion of 2158/2.
They are specimens of G. pitriifioriun vab. farinosa, and therefore of true O. mulllfiorum. Beuth. Tho
remaining gatheiing belonged to Serb. Finlayson from Slam; it is tbe same as 2158/1, and therefore
is tj-pical G. paniflorum.
la. PI. -fifl''- ii (1831)—Table XIII, eolunm 3—Mr. Bentham followed Dr. "Wallich's disposition
of 1829, so tbat no remark is called for regarding this work, But in Lah. Gen. et Sp. (1835) and
ao-ain in DO. Prodr. xii (1848)—Table XIII, column 4—Mi'. Bentham notes tbo confusion introduced
by Dr. Wallicb into Cut. n, 2158/2, and to rectify it raises to specific rank the "Sillet" portion of
tliat gathering as G. mulHfloram, though rather strangely he does not mention the Goalpaia plant
2158/D, which exactly agrees with the fj-pe of his G. multiflorum. Having scpBiatod oiï G.
munifioruw, Bentham therefore confines to G. ¡.avtijionim three gatherings ; n. 2158/1 (wliieh ho supposed
to be Nepalese, bu£ which in i-eality earoe fi-om Heyne's HerbMium and the exact locality of wliich
we cannot now ascertain), n. 2158/0 (which was originally issued as n, 2159/1), aud tho " Gentea
Mts." portion of n. 2168/2, which is conspedfio with n. 2158/C. In these works therefore—and
Drs. WaberB and Mifiuel have both followed his airangement—Bentham has advanced to speciCc
•r
• 'î!'
PAIIVIFLOEA, 2 5 7
rank a plant that is only a variety of G. pmxiflorum, while ha hta described as G. parvijlorum a nuxturo
of that specics and G. criiiitum.
But the fact that n. 2158/1 and n. 2158/E ai'e, even if varietally seporable, still conspecifio
with n. 2158/D and with the "Sillet" portion of n. 2158/2 could not long pass undetected, so that
it is not suiin'ising to find that in tbo tj^te "Wallichiau herbarium—preserved by tbe Linnean Society
of London—the necessary reduction has been efíceted. There, however, in placo of G. muUiflorum
having been again merged in G. pardflorum, typical G. pardflonmi Wall,, Oat. n. 2168/1 (1828) has
been named G. imiltiflonm Bentb., a name tbat dates fi'om 1835; and tho name " G. parrífiorum" has
been restricted to n. 2158/C and the part of n, 2158/2 that is tho same as 2168/C—to the gatheiiags,
in fact, that do not belong to G. parvijioram at all. The species to which they do boloug-for theso
gathermgs are undoiibtedly conspecifio with n. 2159/2—is thus pronded witb two names, G. parviflorum
aud G. crinitum. This result is masked, so far as the Tyi)e-Herbarium is concerned, by the
manner in which the specimens have been there glued down, for n. 2158yC, tho transfer of which to
G. pavdjlorim may be said to have given rise to the confusion, occurs on sheet 1, while n. 2158/i
occupies sheet 4. This being the ease, and ñu'ther because true G. pmxiflonun haa become more
familiai'ly known as G. multljlorum, it almost seems advisable to allow the names indicated in column
5 to stand. But the practical objectiona are insuperable. For fii'st, there is no room for both
"G. parciflorum," as understood in tbe Linn. Soe. Herbarium, and G. crinitum. Again, since tho
Linn. Soc. "<?, parvifiorxon" is based on 2158/0, which only received the namo G. pan-ijforum in
1829, it is the name G. erinitum (which in any case dates from 1828) that must stand. This being
so, the suppression of the Wallichiau namo 0. paniflvrmn in favour of Bentham's name 6. mulnflwum
has nothing left to justify it. Eor even in 1828 Wallich's G. parvifiorum contained two
gatherings of that species a.s agaiust one gathering that should have been excluded, and even after
h á readjustment of 1S29 contained five gatherings of the species against only two of another. And,
fui-ther, though it is true tbat alt tbe specimens named G. muliiflonun in tbe Linn. Soc, Herbarium
are conspecifio, we know that Mr. Beutham. himself in the two works in which he desciibed G. niulti-
Jiorum only applied tbo name to una varuiy of that species. Sia' Joseph Hooker in Ilor. Brit. Ind., iv
(1885) is thus amply justified in suppressing Bentham's name G. multiflorum. And as his treatment
Table XIH, column 6—restores WalEch's aiTangement of 1S29 and Bentham's of 1831, thus removing
all the confusion subsequent to these dates, it is only necessary fuither to eliminate from G.
parvifionim those gatherings of G. crini/am that were unfortunately included in it by Wallicb aud
restore them to the latter species.
Specimens from Java ((?. (UahoCnmum Zoll. et Mor.) have xinfortunately not been a^-ailablo for study.
But tbe position of the Java plant hardly admits of a doubt. The original description of the species
by its authors is quite satisfactory, and (unconsciously) laj-s emphasis on the coroUine chaiacters tbat
incUcate its position in S Stenostuma " tuho eoroliino ^uOfiUfonni fuuccm versus tarJton^ lta..iu>»-\ dilalato
" ( n c c ! í ¿ w rcUíiiiis Gorphostimmutis ¡.upra medioin vflaic)." Bentham, writing of authentic specimens
in ncrb. DC., says " valde aj)ii\e G. pavvifloro; talycü dentón nñnut acuminati," while Miquel goes
further caiyci» dmtúus minus acuntiaatis G. parviiloro differt oui forsmi nmis affine." That is to
say, according to both Bentham and Miquel, G. dichotonnnn otihj differs from G. pardjlormn in
having less acuminate calyx teeth. Neither Bentham nor Miquel state precisely whether the calyx
teeth are or ai-e not shorter than the calyx tube, but the original description says " ca/ycibu, B-Jidii,"
which implies, unless the term is loosely applied, that the teeth ai-c shorter than the tube is. Both
Bontham and Miquel, however, mention a character omitted by Zollinger and Moritzi " bracleis
caly;es (equantibus." This makes it differ both from G. pareflo}-um and G. crimluni, but only to
the extent of almost assuring one that G. parv/lorum and G. crinilum are not specifically separable,
and that in G. dickotoinum exists the intermediate foi-m necessary to support the formal union of
' lb should be noted tliat Walpcra (Si^ert. vi, 684) corrects tUs misprint into "ianJem." But tiis alter,
at ion aliosetber annuls the forco ol the parenthetic remark wl.icb follows. And Miquol. who did «khorate
the penus-tho Eepirtoyium is, as its name implies, only a compilatory work-rccognises the true meaaing of
the sentence and makes (Fior. Ind. Hat. ii. 086) the ol>TÌously correct altcrution.
' i l i '
IfflHl
'Ì
i
li