[ 147 ] XI. A Commentary on the Fourth Part of the Hortus Malabaricus. By (the late) Francis Hamilton, M.D., F.R.S. and L.S. Read February 21st, and November 7th, 1826. Mao, seu Mau, p. 1. tab. 1 et 2. X HE word Mange, which, the author says, is the name of this tree among the Indians, is of Malay origin, and was introduced by Garcias ab Horto, Acosta, and other early writers. These absurdly applied the Mangka, or Manga, of the Malays to the fruit, and called the tree Mangifera, which has been copied by modern botanists, although Rumphius properly called the genus Manga. His specific name domestica has been changed with equal want of propriety ; for the name indica is equally applicable to every species of this genus. The Sanscrita name Amra, corrupted in the vulgar dialects of Gangetic India into Am, is the source of the word Ambo, used by the Brah- mans of Malabar. For one circumstance in Rheede's description I cannot account ; and, as there can be no doubt that he knew the tree perfectly, and meant to describe it, this circumstance must be attributed to one of those errors into which even the most accurate are liable to fall. He says, "folia bina, terna, aut quatema simul ex eodem pediculo ramulis inhaerent" This, converted into Lin- nsean language, would imply that they are folia composita ; but this is per- fectly erroneous. Another error, respecting the stamina, induced Linnaeus to place this tree in the class Pentandria. Rheede says, flores — quinque intiis alhicantibusjibris,jlavescentibus apicibus dotatis — prcediti. Now in ninety-nine flowers out of a hundred only one filament has an anthera, and I have never observed one flower in which all the five stamina were complete. Ada maram, p. 5. tab. 3 et 4. Maram annexed to Ada signifies tree ; the Malabar name therefore is Ada, VOL. XVII. X 148 Dr. Francis Hamilton's Commentary ♦ or Saros. Rheede says that it grows in the woods of Malabar ; but so far as I have observed, it seemed to me to have been always planted, and reared with care in the neighbourhood of villages or in gardens ; and I suspect that it has been introduced from the great Oceanic Archipelago, where it would seem to be a spontaneous production, being, I suppose, the Catappa silvestris of Ruin- phius (see my Commentary on Herb. Amh. i. 175-)- Both Ada and Saros, however, may be Malabar words peculiar to this plant, which would seem to imply its being indigenous ; but Jibe, the name given to it by the Brahmans in Malabar, is also peculiar to that country ; nor does there seem to be any Sanscrita name for this plant, which would imply its being an exotic lately introduced. At any rate, that it is so in the North of India I have no doubt, because in the vulgar dialects spoken there it is called Budam, or the Almond- tree, on account of its kernels being like those of the almond. This, although a very slight affinity, seems to have at first satisfied Nieuhof, Ray and Pluke- net, who called the tree Amygdakis indlca (^Alm. 28.). Afterwards, indeed, on account of an absurd resemblance which he imagined to exist between its fruit and that of his Prunifera Fago similis arbor Gummi Elemi fundens,Jigura et magnitudine Olivce ex Insula Barbadensi {Aim. 306; Phyt. t. 217./! 4.), the last-mentioned author considered the Ada maram as nearly allied to this plant {Mant. 156.), which, although by no means the Amyris Elemifera of modern botanists, is certainly not the Ada maram ; nor, if it produces Gum Elemi, is it likely to be even of the same natural order, none of the Combre- tacece producing odorous resins. The elder Burman probably mentioned this tree under the following name. Arbor indica, amara, nucleis Amygdali facie, Katappas Lusitanis, Samandara zeylonensibiis, as I shall endeavour to show when I treat of the Hagam {Hort. Malab. vi. 37.). llumphius {Herb. Amb. i. 175.) described two kindred species, the Catappa domestica and silvestris ; and in the Appendix ( 1 76.) he notices the strong affinity which these have to the Ada maram, without mentioning to which of his kinds it is nearest. I have already stated that I think it is his C. silvestris. It was not introduced into the modern system until Linnaeus published the Mantissa, in which he improperly called it Terminalia Catappa (see my Commentary on the Herb. Amb. i. 175.), a name retained by more modern botanists {E71C. (m the Hortus Malabaricus, Part IF.- 149 Mdth. i. 348.; Willd. Sp. PI. iv. 967.; Hort. Kew. v. 441.). I must here cau- tion the young botanist against relying on the specific character given by these authors, however respectable. The leaves of the Ada maram, as well as of the Catappa dnmestka, have in general edges quite entire ; and the real difference between them consists in the former being pubescent, and the latter smooth. Panem Palka, seu Panam Palca, p. 9. tab. 3. This tree, according to Commeline, was well kno^vn to John Bauhin, al- though it is alleged that his brother mistook its fruit for that of a Palm. Plukenet called it Nux Myrlstica spuria {Aim. 265.) ; and the elder Bur- man, copying Herman, called it Nux Myristlca, ohlonga, Malabar tea {Thes. Zeyl. 172.). Under the name of Myristlca fnictu inodoro, Linnaeus (Fl. Zeyl. 588.) placed it among his Annihilutce, the explanation of which (" sunt planta- rum zeylonensium nomina, quae soni prsetereaque nihil,") seems very little applicable to a plant, the female of which has been described and figured excellently by Rheede. As, however, this author did not mention the male, Linnaeus, vnth the sexual system, was no doubt at a loss. Among the more recent botanists this tree was first taken up by Thunberg (anno 1782), who called it Myristlca tomentosa. M. Lamarck, overlooking this, or uncertain of what plant Thunberg meant, called it Myristlca malaba- rica {Enc. Mdth. iv. 388.), and distinguished it from the Nux Myristlca Mas of Rumphius, with which Burman in his observation {Herb. Amb. ii. 25.) had confounded it. Rumphius himself, although he admitted a great similarity, pointed out several differences, which should have prevented Burman's mis- take, especially as the latter had probably mentioned the Nux Myristlca Mas of Rumphius under Herman's name, Nux Zeylanlca, Moschatte rofundce simllls, oblonga {Thes. Zeyl. 172.), which is probably the M. Phlllppensls of M. Lamarck. Whether or not Gaertner could have seen M. Lamarck's account of this tree, first published in the Memoirs of the French Academy, I know not ; but in the same year (1788), overlooking also the account of Thunberg, he described the fruit of the Panem Palca by the name of .1/. dacfyloldes {De Sem. i. 195. t. 41./. 2.). Willdenow {Sp. PI. iv. 870.) restored Thunberg's name, M. tomentosa ; but falls into Burman's error in considering the Nux X 2 150 Dr. Francis Hamilton's Commentary Myrlstica Mas of Rumphius as the same. As he quotes botli, I cannot take upon myself to determine which he really meant. If Thunberg did the same, the name tomentosa, being uncertain, should be altogether abandoned, as both M. Lamarck and Gaertner seem to have properly enough done. Samstravadi, seu Samstravari, seu Caipa Tsjambu, p.W. fab. 6. The second name, which is that on the plate, is evidently an error of the engraver. The third implies the plant to be a species of Tsjatnbu or Eugenia, an opinion adopted by Commeline on no other authority than that of the na- tives, and these not the men of science ; for the Brahmans call it Sada Pali, which Rheede says implies frugifera arbor. The vulgar Malabar generic name is not Fadi, as Burman would have it {Fl. Ind. 1 15.) by printing Samstra vadi. Samstravadi is evidently one word, and the prototype of a genus, as the following plant is called by the same name, with the specific term Tsjeria prefixed. Jussieu was therefore scarcely justifiable in calling {Gen. Plant. 361 .) this genus Stravadium, which consists only of half a word. Plukenet (Mant. 137.) suspected, but without being certain, that the Sams- travadi might be his Nuciprunifera Arbor, foliis densioribus, subtus argenteis Jloribus in prcelongam spicam dispositis,fructu tetragono; but, although nearly allied, the plants are no doubt different, as he might have concluded from Rheede's description, " folia superne colore atro-viridi splendentia, inferne viridi dilutiore." Linnaeus in the Flora Zeylanica (191.), still following the Hindu arrange- ment, called the plant of Rheede Eugetiia foliis crenatis, pomis ovatis, racemo longissi?no, which in the first edition of the Species Plantaium, and in Burman's Flora Indica (115.), became the Eugenia racemosa ; but now the Butonica syl- vestris alba of Rumphius {Herb. Amb. iii. 181. t. 116.) was added as synony- mous. Although in the explanation of the plate Burman says that it repre- sents the Butonica sylvestris alba, yet Rumphius himself called no plant by this name, but in the places quoted describes and figures the Butonica terres- tris alba, a species totally diflFerent from the Samstravadi. Willdenow, how- ever, {Sp. PI. ii. 966.) leaves the synonyms just as he found them. M. Lamarck (Enc. Meth. iii. 197) continues to call this plant Eugenia race- mosa, but notices its affinity to the Barringtonia or Butonica ; and although on the Hortus Malabaricus, Part IV. 151 he properly rejects the Butonica terrestris alba as synonymous, he falls into an error equally great in calling it the Butonica sylvestris {terrestris) rubra {Herb. Amb. iii. 181. t. 115.) of Rumphius ; for European botanists seem to have thought it necessary, as Rheede had described two Samstravadis, that these should be the same with the two Butonicas of Rumphius ; whereas the latter does not describe the Samstravadi, nor mention any plant by the name of Butonica sylvestris; nor does Rheede notice the Butonica tei-restris rubra. M. Lamarck saw specimens of his plant ; and from the account which he gives of the calyx, it was evidently the Samstravadi of Rheede. Willdenow, on the contrary, says nothing to enable us to judge whether his specimens belonged to the Samstravadi or to the Butonica terrestris alba. Jussieu was the first, as far as I know, to point out a tolerably correct arrangement of the Samstravadi, by separating it from the Eugenia and placing it {Gen. Plant. 361.) in the same genus with the Butonica of Rum- phius and Lamarck, the Barringtonia of Forster and the younger Linnjeus, and the Commersonia of Sonnerat, which the elder Linnaeus had placed among the Guttiferue in the genus Mammea. Perhaps M. Jussieu should have taken the genus of Rumphius as it stood, and included in it not only his three Buto- nicas, but the two Samstravadis of Rheede ; but Jussieu considered the Tsjeria Samstravadi and the Butonicce terrestres as forming a distinct genus from the Butonica, and called this genus Stravadium {Gen. Plant. 361.). Dr. Roxburgh however {Hart. Beng. 38.), as I have above proposed, includes in the same genus both the Butonicas of Rumphius and the Samstravadis of Rheede, calling the plant, of which I am now treating, Barringtonia racemosa ; but he does not quote Rheede, deterred probably by the following words in the description, "Arbor est vastae magnitudinis caudice crasso," while, I must confess, that the plant which Dr. Roxburgh and I knew, is only a small tree ; but I cannot on this account call it a different species. When I returned from Ava, I sent to England both specimens and a draw- ing of the Samstravadi, which were given to Sir Joseph Banks. A copy of the drawing has been lodged in the Library of the India House, where I have also placed specimens from India Proper. In deference to M. Jussieu I have classed it in the Catalogue with his second division of the order of Myrti ; but I sus- pect that it might with more propriety be arranged with the second division of 152 Dr. Francis Hamilton's Commentary the Guaiacance, as will appear from the following description. The natives of Ava call it Kiin gri, the first word being the generic term, and gri signifying great. Arbuscuia pulchra. Folia sparsa, apices versus ramulorum congesta, basi obtusa obovata, acuta, ultra pedem longa, costata, venis reticulata, nuda, serrata, petiolata. Racemi longissimi, penduli. Flores ex albido rubicundi, magni, speciosi, caly- cibus coloratis, striatis. Calyx foliolis concavis obtusis 2- sen 3-partitus, persistens, intus disco integro mellifero ad basin vestitus. Petula 4 sen 5 patentia, obtusa, concava, obliqua. Filamenta plurima filiformia, petalis longiora, basi coalita in annulum discum calycis cingens. Anthera; parvae. Germen inferum tur- binatum. Stylus longitudine staminum filiformis. Stigma simplex. Bacca molliuscula, tetragono-ovata, calyce coronata, obsolete quadrisulca, unilocularis. Semen unicum, oblongum, magnum. Perispermum forma seminis magnum. Embryo centralis, ovalis, dum non germinaverit abs- que partium distinctione indivisus. TsjERiA seu Sjeria Samstravadi, p. \5. tab. 7- In the preceding commentary I have already made several remarks appli- cable to this plant, which the Brahmans call Gove-sada-pali, using the last two words as a compound generic name, and the words, therefore, should have been printed Gove Sada-pali. Notwithstanding the utmost affinity between this and the preceding, Com- meline could trace scarcely any resemblance to the Eugenia, in which, not having been misled by the native nomenclature, he judged properly. Ray, liowever, more consistently witli his admitting the Samstravadi to be an Eu- genia or Jambos, allowed the Tsjeria Samstravadi to belong to this genus : but Plukenet more cautiously called it Nuci pomifera Arbor Orientalis Castanete equince foliis, fructu longo corticoso crasso, tetragono, summo apice (Pomi in modum) umbilicato, nucleum nudum angulosum includente (Aim. 266.), in which he entirely overlooked the leaves of this being simple, and those of the Horse Chestnut being compounded. cm the Hortus Malabaricus, Part IF. 153 Although neither Rumphius, nor his editor Burman, considered either spe- cies of Butonka terrestris as the same with the Tsjeria Samstravadi ; and although Linnaeus in the Fbra Zeylamca (190.) quoted the latter alone, with the synonynia of Ray and Plukenet, for his Eugenia folUs coronafts, peduncidis fermimntihus, pomls oblongls acutanguUs ; yet in the Species Plantarum, copied by the younger Burman {Fl. Ind. 114.), he introduced, as synonymous with the Tsjeria Samstravadi, the Baton ica terrestris rubra, adding to Eugenia the specific name acufangula. This arrangement was of course followed by Will- denow (Sp. PL n. 996.). M. Lamarck, however, observing, I presume, that the fruit of the Butonica terrestris rubra, as represented by Rumphius {Herb. Amb. iii. /. 115.), has no great resemblance to that of the Tsjeria Samstravadi, being too much attenuated at the ends, rejected this quotation, and considered the Butonica terrestris alba {Herb. Amb. iii. t. 116.) as the Tsjeria Samstravadi, the form of the fruit in the figures of these plants, by Rheede and Rumphius, having a great resemblance. I must, however, observe, that Rheede says of the Tsjeria Samstravadi, " Flores purpurei ; " and he represents the flowers as disposed in racemes; while of the Butonica terrestris alba Rumphius says, " petiolis (pedunculis communibu.s) insident capitula viridia sese in bina ter- nave crassa petala (calycis lacinias) aperientia, in quorum centre quatuor alia alba et extensa conspiciuntur petala, restans floris pars in medio repleta est albis staminibus ad basin rubescentibus, antheras fuscas gerentibus." Further, he not only represents the flowers and fruit as disposed in spikes, but says, " pomula sessilia, quum priora (id est, fructus Butonicce terrestris rubrie) ex pedunculo (pedicello) dependeant." We may safely, I think therefore, infer that, notwithstanding the similarity of the fruits, the Tsjeria Samstravadi and Butonica terrestris alba are not the same plant. In fact, neither species of the Butonica terrestris seems to have been described by Rheede, nor either species of Samstravadi to have been described by Rumphius ; as we may infer not only from the circumstances above mentioned, but also from the form of the leaves as represented by the two authors. The variations of opinion on the subject, among the best botanists, seem to have deterred Dr. Roxburgh from quoting either author for his Barringtonia acutangula {Hort. Beng. 52.), although I have no doubt that it is the Tsjeria Samstravadi. From Ava, where it is called Kiin ngceh (little), I sent speci- 154 Dr. Francis Hamilton's Commentary mens to Sir Joseph Banks under the name adopted by Dr. Roxburgh ; and I have since given specimens to the library at the India House under Jussieu's name of Stravadium acntangulum ; for, although I cannot approve of so violent a corruption, I must yield to his superior authority. In the dialects spoken in Gangetic India, where it is one of the most common trees, it is called Ijjal or Hijjal. Arbor magnitudine mediocris. Rami petiolorum cicatricibus exasperati. Folia sparsa, ramulorum apices versus approxiinata, obovata, apice nunc obtusa tunc acuta, basi cuneata, nitida, nuda, costata, venis reticulata, utrinque viridia. Petiolus brevissimus, supra planus, glaber, non stipulaceus. Racemus terminalis, simplicissimus, pendulus, foliis longior, nudus, glaber. Flores sparsi, parvi, filamentis coccineis rubentes. Calyx superus, laciniis erectis obtusis eequalibus 4- seu 5-partitus. Petala ssepius quatuor revoluta, oblonga, basi cohaerentia, ad staminum colum- nam adnata. Filamenta plurima, longissima, filiformia, basi coalita. Anthene parvae, subrotundae. Germen inferum, tetragonum. Stylus lon- gitudine et figura staminum simplex. Stigma indivisum. Bacca sicca, oblonga, tetragona, calyce coronata. Semen unicum, maximum, oblongum, circinatum. I have not noticed the structure of the seed, as the description was taken in Ava, before I had seen the work of Gsertner. Malla Katou Tsjambou, seu M. Catu Tsjambu, p. 17- tab, 8. Commeline joins the vulgar, Hindus, Portuguese and Dutch, in consider- ing this as a Jambu, or Eugenia, very nearly allied to the plants now called E.Jambos and E.malaccensis; while the Brahmans seem to err as much in calling it Mai Ambetti {montana Mangifera foemina) . It must be admitted that the figure represents the plant less like the Eugenia than it ought, be- cause the leaves have been drawn as if alternate ; but from the description we learn that this is erroneous (" Folia geminata brevibus petiolis decussato ra- mulis inhaerent"). So far, therefore, as to its leaves, it might be an Eugenia ; but then the flower is divided into five or six parts, the latter seeming to be the natural number, as the style is divided into three; and besides, some on the Hortiis Malabaricus, Part IF. 165 individuals would appear to be entirely female, as that described by Rheede, who does not mention any stamina. Both circumstances are incompatible with its being an Eugenia. Plukenet was as unfortunate as Commeline in comparing this plant to his Arbor Ind'ica Pi/ri densioribus et subrotimdli folds, fructu Nucis Moschatae magnitudine summo vertice coronato (Mant. 23. pi. 3. t. 336.), which is pretty evidently a Gardenia, and quite different from the Malta Katou Tsjambou. The elder Burman, in his observations on Rumphius {Herb. Amb. i. 128.), thinks that this is the Jambosa xilvestris alba, which again he considers as a variety, or rather as the female plant, of the Malacca Schambu, that is, of the Eugenia Jambos. In both opinions he is probably wrong ; for the E. Jambos has no flowers merely female, nor is the Jambosa silvestris alba the same with the Malacca Schambu, as I have endeavoured to show {Linn. Trans, xiii. 482.). It is, however, very possible that the Malta Katou Tsjambou, as the same Burman in another place alleges {Thes. Zeyl. 125), may be his lambos sylves- tris et montana fructu Cerasi magnitudine, which is the Maharatambola of the Ceylonese ; but it cannot be the Jambosa silvestris parvifolia of Rumphius {Herb. Amb. i. 129.; ii. t. 40.), with which Burman there joins it, because that is a real Eugenia with hermaphrodite flowers ; and the Malta Katou Tsjambou, or Maharatambola, on account of its dioecious flowers, terminal panicles, and trifid style, notwithstanding the authority of Linnaeus {Ft. Zeyl. 501.), I cannot consider as belonging to this genus. It seems, indeed, to have a greater resemblance to the genus Scopolia of Forster, as described in the En- cyclopedic M^thodique (vii. 14.; ///. Gen. t. 860.). Katou Tsjeroe, seu Cattu Tsjeru, seu C. Cheru, p. 19. tab. 9. Katou and Rana, the specific names used by the vulgar and learned of Malabar, have the same meaning, that is, signify anything wild or unculti- vated ; while a species that is planted round the corn-fields, and described in page 20, is considered the prototype of the genus called Tsjeroe or Cheru by the vulgar, and Bibo by the learned. It seems to be from a very considerable aflinitv between this tree and the Anacardium occidentale that the natives of India, according to Clusius {Enc. Me'th. Suppl. i. 753.), gave to the latter the name of Bybo, evidently the same with Bibo, used by the Brahmans of Malabar. VOL. XVII. V 156 Dr. Francis Hamilton's Commentary Commeline, however, does not venture to compare this with any plant then known ; and it was with uncertainty that Plukenet quoted it for his Prunifera arbor seu Nucipruntfera folio dodrantali longltudiue, Icevi mollitie prcedito {Aim. 306.; Plii/t. t. 218. y. 1.), a West Indian plant that I cannot trace in modern authors, unless it be the Achras Sapota, which, according to the Hor- tiis Kewemis (ii. 312.), is called the Bully-tree, if that be the same with the Bully-Bay used in Barbadoes according to Phikenet. Sliould this be the case, the West Indian plant can have no affinity with the Tsjeroe. M. Lamarck tliought that tlie Tsjeroe might be a ]\Iangifera, and it is ac- cordingly mentioned {Enc. Meth. Suppl. iii. 584.) under the name of Mangifera} racemosa, M. Poiret justly doubting of its being a real Mangifera. This is the only notice, so far as I know, that was taken of this tree by modern botanists, until I visited Chatigang in 1797, and Mysore in 1800. On my return from the former, I gave young plants to Dr. Roxburgh ; and on my return from the latter, I showed him a drawing and specimens, which were afterwards given to Sir J. E. Smith, under the name of Holigarna Vernix ; but Dr. Roxburgh called it Holigarna longifoUa {Hort. Beng. 22.). The plant, which I saw, seems to be that which Rheede calls Tsjeroe, or Biho, without prefixing a spe- cific name, and differs from the Cattu Tsjeru, or Rana Bibo, of which he gives a figure, in having much shorter racemes, and these not at the end of the branches, but from their sides, and also in a singular small tooth-like process on each side of the pefiolus. Dr. Roxburgh describes another species from Silhet, of which I have given specimens to the library at the India House. This genus, remarkable for the caustic nature of its juice, which is used as a varnish, I cannot reduce to any of Jussieu's natural orders. It comes nearer the Rhus than to any Linnsean genus ; but has the germen inferam ; on this account, as well as its caustic juice, it seems nearly allied to the Rak of Japan {Kcempf. Amcen. Exot. 793.), and to the Arbor f'ernicis of Rumphius {Herb. Ainb. ii. 259. t. 86.), which M. Lamarck {Enc. Meth. i. 350.) calls Terminalia Vernix. I should, indeed, have no doubt of their belonging to the same genus, did not Rumphius say, " flores plurimis staminibus rubris referti," which, if accurate, would show an essential difference between his plant and both the Bibo and lerminalia. In fact, the two latter have no sort of affinity, while the number of styles and the position of the germen distinguish the Bibo most on the Hortus Malabaricus, Part IF. 157 cleaily from the Mangifera. I shall now give the description, which I took in Mysore. HOLIGARNA LONGIFOLIA. Hoit. Beng. 22. Tsjiero sen Bibo. Hort. Mulab. iv. 20. Cheru Taulavse. Biba Concanse. Buchanans Mysore, iii. 186. Holigarna Carnatse. Habitat in Indiae sylvis montosis, humidis. Arbor verniciflua, succo caustico, venenato, recente albo seu hyalino, exsiccato nigricante scufens. Rami cicatricibus obovatis exasperati. Folia alterna, apices versus ramulorum conferta, oblonga, cuneata, acuminata, margine revoluto integerrima, costata, venis reticulata, glabra, junioribus tamen subtus pubescentibus. Petiolus semiteres, brevissimus, denticulo sub- ulato patente utrinque apicem versus instructus, non stipulaceus. Racemi infrafoliacei, sparsi, simplicissimi, adscendentes, folio breviores, undi- que pilis ferrugineis tecti. Flores dioeci, pedicellati, parvi, sparsi, vel ali- quando fasciculati, albi. Squamce in racemo et pedicellis vagoe. Masculini Jloris calyx minimus, quinquelobus. Petala quinque, ungue lato fere coalita, intus barbata, calyci inserta. Filamenta quinque petalis alterna et longiora, patentia. Antherce cordatse. Foeminei Jloris calyx brevissimus, cyathiformis, fundo setosus, ore obsolete quinquangularis. Petala quinque, linearia, intus villosa, ungue lato sub- coalita, calyci inserta. Filamenta quinque, subulata, brevissima, peri- gyna, petalis alterna. Antherce simplices, nescio an fertiles ? Germen magnum, inferum, compressum. Styli tres, erecti. Stigmata crassa. Drupa compressa, monosperma. Tani, p. 23. tab. 10. In the Hindwi dialect I cannot trace the name Gottinga, said to be used by the Brahmans of Malabar for this tree. According to Rheede, the vulgar in- habitants of Malabar reckon this the prototype of the genus Tani, which, however, is very unnatural, as this species has no affinity to the following plant, which is also called 7V/m/, with a specific name prefixed. As I under- V 2 158 Dr. Francis Hamilton's Commentary stood the natives of Malabar, it is the fruit which is called Tani ; for they called the tree Tani Cai Maram {Tani fructus arbor'), Buchanan's Mysore, ii. 342. The plant of C. Bauhin {Fructus in insula S. Marice, pyra majora referens intus muciilentuni), with which Commeline compares this, can scarcely be the same, on account of the size and mucilaginous quality of its fruit, and is pro- bably rather a Mabolo or Diospyros than a Myrobalamis, although Plukenet rather thinks it a Syalita {Dillenia), which, however, he confounds with the {Artocarpus) Bread fruit {Mant. 124.). In his Index he mentions the Tani, but without a reference to the part of his work where it is to be found, nor have I been able to discover the place. Commeline afterwards called the Tani a Primus, in which gross error he was followed by Ray and the elder Burman {Thes. Zeyl. 197.) ; the latter, indeed, was still further in the wrong, because he confounded it with the Dematha of the Ceylonese, which is the Gmelina asiatica, as Linnaeus, in rejecting Burman's synonyma, rightly observes {Fl. Zeyl. 230.). Gaertner considered the Tani as the same with his Myrobalamis Bellirica {De Sem. ii. 90. t. 9". ubi errore BelUrina dicitur), and certainly the fruits of the two plants are extremely similar ; but the form of the seed and locula- mentum is different, in that of Rheede being circular, and in that of Gaertner being angular. Whether or not the latter was right, in considering his plant as the Myrobalanus Bellirica of Blackwell and Breynius, I cannot say, not having it in my power to consult these authors ; but he says that Blackwell's figure is bad, or, in other words, does not entirely resemble his plant. M. Poiret {Enc. M^th. vii. 5/6.) seems doubtful whether Gaertner was right in quoting the Tani for his Myrobalanus Bellirica, and in the Supplement (iii. 707-) to the Encyclopedic states this doubt more fully. Dr. Roxburgh does not quote {Hort. Beng. 33.) the Tani for the Terminalia Bellirica, which is a name not mentioned by Willdenow, although I suspect that Dr. Roxburgh's plant is what Willdenow calls T. Chebula, because he says, " foliis obovato-oblongis," while the Chebula of Dr. Roxburgh, the same with that of Retzius, has folia ovata. The Tani has folia obovata, and may therefore be the T. Chebula of Willdenow. In this case the Tani cannot be either the M. Chebula or M. Bel- lirica of Gaertner ; the former on account of the difference in the form of their on the Hortus Malabaricus, Part IF. 159 fruits, and the latter for the reasons I have already stated : and besides, the flowers of the T. Bellirica of Retzins, which in the Hindwi dialect is called Sahara, have an abominable stercoraceous smell, while Rheede says of his plant " flores suaveolentes." In the woods of Southern India [Buchanans Mysore, i. 183.) I found a tree called Tan in the dialect of Carnata, and Tani Cai Maram by those of Mala- bar, as already stated, which therefore, I have little or no doubt, is the Tani of Rheede, although I have not noted the smell of its flowers, by which chiefly it is distinguished from the Terminalia Bellirica. Specimens were given to Sir J. E. Smith under the name of Terminalia or Myrobalanus Taria, and I shall here annex a description. Jrbor magna, ligno firmo, albido, non resinoso, durabili. Ramuli sulco e petiolo utrinque decurrente angulati, surculis novis pubescentibus nudi. Folia decidua, subopposita, apices versus ramulorum conferta, obovata, aliquando acuta, ssepius cum acumine obsoleto obtusa, margine cartila- gineo integerrima, costata, venosissima, coriacea, eglandulosa; juniora pubescentia, adulta utrinque glabra. Petiolus compressiusculus, margi- natus, glaber, supra medium glandula, setate ssepe evanida, utrinque in- structus, brevis, non stipulaceus. SpiccB infrafoliacese vel axillares, petiolo longiores, pubescentes, laxse, nudae, solitarise. Flores sparsi : superiores masculini ; inferiores in eadem spica hermaphroditi. Drupa subcarnosa, angulis quinque obsoletis obovata. Nux semine esculent© monosperma. In the collection of specimens which I have given to the library at the India House, are those of several varieties of the Terminalia Bellirica, which, as I have said, I can scarcely distinguish from the Tani by any mark, except the smell of the flowers ; for I found a very considerable difference in the form and pubescence of the leaves, in the shape of the nut and seed, and in the presence or absence of glands, in the difi"erent trees that were admitted by all to be the Bahara, the name by which the plant with fetid flowers is known in the Hindwi dialect. In some places the Bahara was distinguished into two kinds, the great and the small, on account of a difference in the size of the fruit. The 160 Dr. Francis Hamilton's Commentary flowers of that with the small fruit are not so offensive as those of that with the large drupe, and therefore, in this respect, it approaches to the Tani -, but then the fruit of the Tani is as large as that of the Great Sahara, or as Gsert- ner's Badamia, while the fruit of the Small Bahara is like that which, l)y an error of the engraver, is called Bellirina in the 97th table of Gsertner. On the Mhole, these plants require still further examination. I shall, however, de- scribe the fruit of the large and small kinds of Bahara, the first taken at Domdoho, and the latter at Duriapur, both differing somewhat from the figure given by Gsertner ; but I think, as I have said, that these fruits are subject to verv considerable variations in the same individual tree. Bahara major. Drupa Juglandis integree magnitudine obovata, sessilis, umbilicata, junior pubescens, carnosa, obsolete pentagona, subsequilatera. Caro crassa, succo flavo scatens. Nux crassa, dura, circinata, cavitate quoque circi- nata. Semen forma cavitatis. Integiimentum crassum, membranaceum. Perispermum nullum. Cotyledones crasste, conduplicatse, una alteram amplectante, lit in gemma obvoluta, et minime circumactae ut in Termi- nalium pluribus. Baharcc majoris fructus. Sectio horizon talis. Bahara minor. pnmum. Bahar. 211. /. 3. referens, quae ultima tamen forte est Garcinia Mangostana, Horto Malabarico perperam citato. Bis (virosa) Moyen Bengalensium. Habitat in Indiee Gangeticse dumetis. Arbuscula Vanguerice facie. Rami rigidi, non pubescentes. Ramuli brevis- simi, ex anni prseteriti foliorum axillis (foliis deciduis nudati), subquadri- phylli. Rami nunc inermes ; tunc spinis oppositis supra ramulorum axil- las enatis, rectis, ramulos longitudine aequantibus armati. Folia opposite, approximate, oblongo-obovata vel cuneata, acuta, integerrima, glabra, subcostata, venosa. Petiolus brevissimus, marginatus. Stipulce petioli longitudine interfoliacese, ovatse, acutae, diaphanae. Pedunculi terminales 1 — 3, uniflori, petiolo vix longiores. Bractece vix ullae. Flores mediocres, lutei, inodori. Calyx glaber basi longitudine tubi corollae cylindraceo ; limbo quinquepartito laciniis patentibus, linearibus, acutis, corolla vix brevioribus. Corollte hypocrateriformis tubus crassus, brevis, teres, ad medium intus pilis duc- tus ; limbus glaber, aestivatione imbricata obliquus, quinquepartitus laci- niis obovatis, acutiusculis. At^therce quinque ad corollae incisures adnatae, oblongae, acutae, basi emarginatae. Germen inferum, globosum, glabrum. Stylus longitudine tubi teres. Stigma exsertum, ovatum, sulcatum, bi- partibile. z 2 166 Dr. Francis Hamilton's Commentary Pomum magnitudine fructus Juglandis subrotundnm, calyce truncate umbili- catum, parietibus crassis intus in putamen tenue induratis biloculaie. Receptacula e medio septi utrinque enata, membranacea, bifida. Semina plura horizontalia, bifariam in singulis pomi loculis nidulantia, pulpo carnoso tecta. It must be observed, that the Gardenia uliginosa {Hort. Beng. 13.; Hort. Kew. i. 370.; ^TUld. Sp. PI. i. 1228.) differs in no essential generic character from the preceding, and therefore I entirely approve of M. Poiret having called it Randia uliginosa {Enc. MMh. Suppl. ii. 829.), under which name I have pre- sented specimens to the library at the India House. That the Genipa {Gcertn. De Sem. t. 190.) is to be considered as a different genus seems very doubtful. I did not examine the position of the embryo in the seeds of the Randia uligi- nosa, and therefore cannot say whether it is similar to that in the Genipa ; but Gsertner's figure of the fruit of the latter is, on the whole, a good representa- tion of that of the Randia uliginosa ; and I must protest against such minute differences in structure, as Gsertner here relies on, being held as a sufficient ground for tearing asunder natural genera, a practice, I am sorry to say, now too common among botanists. TsjEROu Katou Naregam, seu Tsjeru Catu Narejam, ;>. 31. tab. 14. In the commentary on the Mai Naregam I have noticed the mistakes which have arisen from the carelessness of Rheede, or of his editors, in prefixing to the figure of that plant the specific names Tsjeru and Catu, which belong to this, with only the order reversed. The Brahmans of Malabar, as well as the vulgar, class this with the Citrus. With his usual want of care in the ortho- graphy of Indian words, Rheede in the plate not only spells the vulgar name differently from what he does in the text, but the name said to be given by the Brahmans in the plate is Naringi (Orange), while in the text it is Cit Rana Nimha {alba, f era Citrus). All these names, however, agree in classing it with the Citrus, while even Commeline condemns in some sort this arrange- ment, which was however adopted by Ray, who called it Mains Limonia Indica fructu pusillo (Hist. Plant. 1658.). Plukenet, who at first followed tlie same idea, and called it Malus Limonia Lentisci foliis Zeylanica, fructu minimo, twarum magnitudine ceviulo {Aim. 239.), afterwards {Mant. 125.) became sen- on the Hortus Malabaricus, Part IF. 167 sible of this error, and classed it with the Corn, of which, as I have said in treating of the Mai Naregam, tlie prototype is probably the Feronia Elephan- tum of modern botanists. In commenting on the Mai Naregam, I have already mentioned the error into which the elder Burman fell by quoting this plant for the mdhedi or Jahuawa of the Ceylonese, which, from the synonyma of Rumphius and Sloane, seems to be rather the small-fruited Citrus, called Lime by the English. Lin- nseus seems to have been aware of this, and therefore joined the Tsjerou Katou Naregam with the Diivul or Giwul, although by an error, probably typogra- phical, he quotes plate 12 in place of 14. On this subject I have in this com- mentary made already some remarks. The Tsjerou Katu Naregam, or Diwul, Linnseus in the Flora Zeylanica (17ft-) considered as a species of Schinm, thus placing it in the order of Terebinthacece ; but from his synonyma we must reject those of Burman and Sloane, which belong to the small-fruited Citrus. The younger Burman having become sensible that the Tsjerou Katou Nare- •ram could not be a Schinus, the fruit of which is a drupa, formed a new genus, which he called Limonia, and in this he included this plant and another, since called Triphasia, and thus returned to the old system of placing it among the Aurantia;, which shows how nearly the Aurantice and Terebinthacece are allied. The Tsjerou Katou Naregam may therefore be most justly considered as the real prototype of the genus Limonia, and is perhaps still the only species pro- perly belonging to it, several, at least, of those since annexed by Linnaeus and others having both a very different character and appearance. Burman, in- deed, added as synonymous the Anisifolium or Boa Balangan of Rumphius {Herb. Amb. ii. 133. t. 43.), which, however, that excellent botanist merely says has the same habit (foliatura) with the Tsjerou Katou Naregam ; and the elder Burman, in his explanation of the plate (43.), points out essential dif- ferences. We may infer, from Linnaeus quoting the plant of Rheede alone for his plant, that it was this he meant ; and as Burman's Limonia acidissima is the Schinus of Linnaeus, it cannot be the Anisifolium, although Willdenow continues to join them {Sp. PI. ii. 572.). Yet, that even he means the Katou Naregam alone, may be inferred from his describing the fruit " Bacca trilocu- laris, seminibus solitariis." The Anisifolium is now considered as forming a distinct genus, and is called Feronia Elephant urn {Enc. MM. Suppl. ii. 630. ; 168 Dr. Francis Hamilton's Commentary Hort. Be)ig. 33.). although the two plants have such a strong resemblance, that I retiu'n to the opinion of Plukenet, and doubt the propriety of separating them merely on account of some differences in their fruit ; at least, if a generic character exists in both their fructifications sufficient to distinguish them from the other plants of the natural orders of Aurantice and Terehinthacece ; for, except in habit, the Murraya comes very near them, and may not be easily distinguished by characters common to them both. Specimens of both have been presented to the library at the India House. Koenig somehow took the Anisifolium to be the true Limonia acidissima, and the Tsjerou Katou Naregam was therefore called the Limonia crenulata ; for he had discovered that the two plants were different ; and this nomencla- ture is followed in the Hortus Kewensis (iii. 43.), and even in the Hortus Ben- galensis (32.) and Encyclopddie {Suppl. iii. 44.) ; but in my opinion it is impos- sible to admit with propriety of such an innovation. Paenoe, seu Paenu, p. 33. tah. 15. The Brahmans of Malabar call this tree Doepoe, or Dupa, rightly translated Arvore Ensenza by the Portuguese, who probably used its fine resin as incense. The resin however, as Commeline observes, is very similar to the gum Aniine of America, and, in fact, is often sent to Europe as such ; for, as Commeline observes, a similar resin is produced by several different trees, having pro- bably little botanical affinity with each other, which is the case also with the resin now more commonly used as incense. The Paenoe is one of the most ornamental trees in India, and in the province of Canara, where alone I have seen it, is usually planted in rows by the sides of highways, making remarkably fine avenues (Buchanan's Mysore, iii. 89.). Ray, followed by Plukenet {Aim. 28.), was as usual very unfortunate in classing this tree, which he called Amygdake ajffims Indica fructu umhilicato, nucleo niido, cortice pulvinato trifido tecto {Hist. PI. 1482.). Linnaeus most justly considered it as a new genus, which he called Fateria {Fl. Zeyl. 204.), and in the Species Plantarum, he added the specific name indica {Burm. Fl. Ind. 122.)*. * It is, however, probable that Linnaus mentions the same tree under a different name, Ktckvria ghaha {Fl. Zeyl. 630.), which is the Arbor Kcekuria ghaha odorata ex qua fluit Gumm. Elemi of the on the Hortus Malabaricus, Part IV. 169 Commeline, after stating the affinity of the gum-resin of the Paenoe to Gum Anime, had observed, " similis arboris meminit Reechus nomine Copalli mon- tana. Ad haec e Zeylan Insula simile adfertur gum mi, quapropter et hsec arbor non male forsan eo referri potest." On no stronger grounds, probably, Retzius considered this as the tree which produces Gum Copal, and called it EUBocarpus copalliferus, in which it is scarcely possible to say whether there is the greater want of care in tracing a substance used in the arts, or of skill in botanical arrangement, the Paenoe wanting every character by which the genus Ela;ocarpus is distinguished. Vahl, however, and Willdenow {Sp. PI. ii. 1 170.) adopt this name, but M. Poiret properly continues to call it Vater'ia indka {Enc. M6th. viii. 418.), as did Dr. Roxburgh {Hort. Beng. 42.). As Vahl says that his plant had all the generic characters of the Ekeocarpus in its calyx, corolla, antherse and fruit, we may safely conclude that it is totally dif- ferent from the Paenoe, especially if it has a germen inferum, as Retzius is said to assert. Dr. Roxburgh alleges that the resin of the Paenoe is called East India Copal, and perhaps it may have passed for such at an Indian cus- tom-house, where a skill in drugs is not very conspicuous ; but Mr. Turnbidl of Mirzapur informed me, that some he sent home for a trial would not sell for Copal, although it was allowed to be Anbne. The real Copal and Anime are, however, American productions. In 1806 I gave specimens and a drawing to Sir J. E. Smith ; and I shall here give a description taken in Canara, where the tree is called Dupada. In Car- nata it is called Cunglium, and in the Hindwi dialect its name is Gugulut. Arbor resinifera magnitudine Qiterci. Rami teretes. Turiones farina quasi aspersi. Folia alterna, magna, oblonga, utrinque obtusa, vel aliquando retusa, integerrima, glabra, costata, venosa. Petiolus teres, medio atte- nuatus, rugosus, nudus, brevissimus. Stipules geminae, laterales, caducae, sessiles, oblongae, integerrima;, obtusae, farina aspersse, brevissima;. Paniculce axillares, folio longiores, ramosissimse, laxae ramis alternis, teretibus, elder Bunnan (Thes. Zeyl. 28.), who properly quotes the Paenu (by error printed Paeru), but errone- ously joins it with an American tree that produces Gum Elemi, and is figured by Plukenet (Phyt. /. 217./. 4.). It must be also observed, that the quotation from Grimm respecting the G. Elemi pro- bably refers to quite another plant, the Kakuna of the Ceylonese, which Bunnan calls {Thes. Zeyl. 166.) Myrobalanus Zeylanica ex qua G. Elemi, fructu odore et sapore prastans. 170 Dr. Francis Hamilton's Commentary albidis, farinosis. Bractece stipulseformes, caducse, geminse ad singulas paniculse divisiones, et ad singulorum pedicellorum basin. Flores alterni, pedicellati, albi, odorati, calycibus extra farinosis. Calyx coloratus, persistens, patulus, laciniis oblongis obtusis ultra medium quinquefidus. Petala quinque, longitudine calycis sessilia, disci hypogyni lateribus inserta, calyce alterna, ovata, integerrima. Antherce plurimae, sessiles, disco insidentes, seta reourva terminatae. Gernioi superum, ova- tum, sulcatum, ovulis quinque foetum. Stylus subulatus, staminibus lon- gior. Stigma acutum. For a description of the fruit I may refer to Gsertner's account {De Sem. iii. 53. t. 189.), to which I have nothing to add. It would thus appear that the Paeroe does not belong to even the same natural order with the Elceocarpus, but is nearly allied to the Vatica, Shorea, Dipterocarpus, Hopea Roxburghii, Dryohalanops, and Lophira, which form a natural order, standing between the Gutt'iferce and Aurantioe ; while the Elceo- carpus, although placed by Jussieu among the latter, is, I think, more nearly allied to the Tiliacece. NvALEL, seu NiALEL, p. 57- tab. 16. With his too frequent want of care concerning native names, the author says that this tree by the Brahmans is called Lassa, which is usually applied to some species of Cordia ; but in the plate the name given by the Brahmans is said to be Rana Bori, and Rana signifying wilde, the generic name is Bori, to which it will be found that two other plants (t. 40, 41.), having little affinity to this, are also referred. Commeline compares the Nayalel to the Sambucus Indlca of Bontius, an author whom I have had no opportunity of consulting. Plukenet compared both (taking them, I presume, to be the same) with his Uvifera arbor Ameri- cana per funiculos e summis ramis ad terram usque demissis prolifera {Aim. 394. ; Phyt. t. 2^7. f. 5.); but I see no grounds for this comparison, for the leaves of Plukenet's tree are simple, and those of the Nayalel, like those of the Sam- bucus, are pinnated ; nor does Rheede hint at its branches sending down roots like a Ficus, to which genus the American plant perhaps belongs. on the Hortus Malabaricus, Part IF. 171 M. Jussieu {Gen. Plant. 297.) and M. Poiret (^Enc. Mdth. Suppl. iv. 93.) thought that the Malel perhaps belongs to the genus ntis ; but the habit is so different that, with all submission to such authorities, I cannot bring myself to this opinion, and rather think that it has a greater affinity to some of the Aurantice, such as the Coo/cla and Murraya ; and especially to the Lansium, as I have mentioned in a Commentary on Rumphius (Herb. Ami. i. 151. t. 54.). It is remarkable that in the island of Ternate the Lansium is called Lassa, one of the names by which the Brahmans of Malabar know the Nayalel. Angolam, seu Alangi, p. 39. tab. 17- Commeline does not venture any conjecture concerning this tree, and Plukenet {Aim. 31.), in quoting Ray's name, "Arbor Indica baccifera fructu umbilicato rotundo Cerasi magnitudine dtcocco," makes no advance beyond what is stated by Rheede. M. Lamarck was the first to introduce the Angolam into the modern system of botany, calling it Alangium decapetalum {Enc. M4th. i. 174.). He con- sidered it as belonging to the order of Myrti, and nearly allied to the Decu- maria ; but Jussieu doubts of the propriety of this arrangement, and rather thinks that it should be placed in his 4th division of the Onagrce, in which I entirely coincide. Willdenow {Sp. PI. ii. 1174.) and M. Poiret {Enc. MM. Suppl. i. 366.) allege, copying, perhaps, from Vahl, that the younger Linnaeus had previously de- scribed the Angolam under the name of Grewia salvifoUa ; but Linnaeus did not quote the Hortus Malabaricus, nor does his description agree with that of the Angolam either by Rheede or Vahl. What authority there may be for the allegation I do not know ; I suspect that it may be some specimen of the An- golam, marked by mistake Grewia salvifoUa, an accident very likely to happen, and therefore by no means a good test. Idou Moulli, seu Idu Mulli, p. 48. tab. 18. Moulli, or Mulli, signifying Thorn, is rather the name of a class than of a genus, and the word Idou, or Idu, must therefore be either considered as generic, or the two words considered as forming a compound, like our English words Buck-thorn, Haw-thorn, Black-thorn, all signifying different genera. VOL. XVII. 2 A 172 Dr. Francis Hamilton's Commetitary The word Elati-canto, used by the Brahmans of Malabar, is of asimilar nature, Canto signifying Thorn in the Hindwi dialect. Commeline made no attempt to class this plant. Plukenet, having thought that he had a plant nearly allied to the Wadouha (p. 97.) of this volume, con- ceived that it might be the Idu Mulli, and called it Wadouke Malaharicce haud multum dispar, Frutex aculeatus e Maderaspatan [Aim. 395.; Phyt. t. 69. f. 7-); but the figure which he gives seems to have little or no resemblance to either Idu Mulli or fVadouka. He afterwards {Mant. 133.) formed a more rational conjecture, and says, " Myrohalano Bellericce, ut nobis videtur Idu Mulli con- gener est, et nominari potest Myrobalanus Indica, Arbor spinis horrida, angus- tiore folio longo,fructu racemoso," Now, although from the number of stamina, as well as from the habit, this cannot be a Mi/rohalamis or TerminaUa, I have little or no doubt of its belonging to the same natural order. At one time I thought that it might possibly belong to the genus called Pyrularia by Mi- chaux [Enc. M^th. v. 745.), but which Willdenow has chosen, without any good reason, to call Hamiltonia {Sp. PL iv. 1114.). The appearance of the plants, however, differs so much, that I now think them likely to belong to different genera. PoERiNsii, seu PuRiNSii, seu Vercoepoelongi, j[;. 43. tab. 19. The Portuguese and Dutch names arise from the saponaceous quality of the fruit ; but whether or not any of the Indian names allude to this quality I know not, all the Indian names for soap that I know being derived from the Portuguese, by whom, probably, this substance was introduced ; nor is it yet common, except among persons employed by Europeans. Commeline remarks, that the natives of hot climates {Indi) use various saponaceous fruits ; but that the Poerinsii was of a genus totally unknown to botanists. Ray, in arranging the plants of the Hortus Malabaricus, threw no further light on the subject by calling it Prunifera fructu racemoso parvo, nucleo saponario, although J. Bauhin had given the name Saponaria to some American plants nearly allied to this ; but the Nux Portoricensis ampUssimif: foliis venosis et Icete virentibus, with which Plukenet compares it {Aim. 265.; Phyt. t. 208./. 2.), having simple leaves, can have no affinity with the Poerinsii, nor with the Sphcerulce saponarice of J. Bauhin. on the Hortus Malabaricus, Part IV. 173 The elder Burman, on the authority of Comnieline's Flora Malaharica, joins the Poerimll with the Saponaria arbor Zeylanica trifoUa, semine Lupini of Herman ; but if Herman's specific character is not very bad, they must be different, the one having folia ternata, and the other folia pinnata; yet we can scarcely suppose Commeline to have been in such an error, and some of the leaves in the plate of Rheede no doubt are represented as ternate. If this circumstance, which is borrowed from an imperfect specimen, be admitted, and if Herman's specific character be amended, the Conghas of the Ceylonese may be the Poerimii ; but to this I shall again have occasion to revert. Bur- man, although with doubt, quotes also as synonymous the Arbor prunifera, sphcerulas saponarias ferens, tetraphylla, ex India Orientali of Plukenet {Aim. 47.; Phyt. t. 14. f. 6.), which, as well as the Poerimii, has pinnated leaves, but so different in form, that I cannot think them the same ; and I shall afterwards describe a plant, which perhaps is that of Plukenet, and totally different from the Poerimii. In the Flora Zeylanica (603.) the Conghas was left by Linnaeus among the Barbarce amiihilatce, which he could not attempt to arrange ; nor does he quote for it the Poerinsii, deterred, probably, by observing that the leaves, when perfect, were really pinnated. When, however, he published the Species Plantarum, he joined the Conghas, that is, the Saponaria arbor Indica trifolia of Herman, and the Saponaria arbor trifoUata semine Lupini of the elder Burman, with the Poerinsii ; and the name Saponaria having been given also to an herbaceous plant of the order of Caryophyllece, the Saponaria arbor of old botanists was now called Sapindus, and the Poerinsii became Sapindus trifoUata foliis ternatis {Burm. Ft. Ind. 91.), although its leaves, when perfect, as may be seen in the figure, are pinnated. " Folia bina et bina sibi invicem opposita tenerioribus surculis (petiolis) proveniunt." At the same time, Lin- naeus and Burman {Fl. Ind. 91.) constituted another species of Sapindus called Saponaria foliis impari — pinnatis, caule inermi, for which the only authority is the Saponaria of Rumphius [Herb. Amb. ii. 134.) ; for the other authorities quoted, Browne, Sloane, Commeline, and Plukenet, all refer to an American plant, no doubt different from that of India, as any one may see by looking at the figure in Plukenet (Phyt. t. 217- /■ 7)- Rumphius, in speaking of his Saponaria, says, " Similis Saponaria arbor descripta quoque occurrit in Hort. Malab. part. 4. fig. 19. nomine Poerinsii." This does not positively assert that 2 A 2 174 Dr. Francis Hamilton's Commentary Rumphius considered them as the same, but only alike. In the descriptions of the two authors, however, I can perceive no essential difference ; for although in the figure of Rheede some of the leaves are represented as ternate, or even as binate, yet others are represented as pinnate ; and although he says that the pinnse are opposite, yet in the figure some are represented as alternate. It must be observed, that in order to represent all the parts, Rheede's painter has selected the extremity of a branch containing flowers, young fruit, and leaves ; and in such cases, the extremity of the young flowering branches will be rarely found to have perfect leaves, especially where these are pinnated, because in this state the leaf has not arrived at full growth, and will be afterwards elon- gated by the extremity of the rachis communis pushing out new pinnae. Rum- phius has unfortunately given no figure ; but I am inclined to think that his Saponaria is the same species with the Poerinsii, and with the Sapindus trifo- liata of Linnaeus and Burman, although it may happen that these great bo- tanists had actually specimens of a Sapindus with ternate leaves, and did not entirely borrow their ideas from the figure of Rheede. If the latter was the case, the name trifoliata being absurd for a plant having pinnated leaves, Willdenow, copying Vahl, has done properly in calling this species Sapindus lauri/oliits (Sp. PL ii. 469.), and in rejecting altogether the S. Saponaria as an Indian plant, the plant so called by Burman being identically the same with the S. laurif alius. Of this I have given specimens to the library at the India House. It must be observed that both Willdenow and M. Poiret {Enc. Me'th. vi. 664.), copying Vahl probably, agree in quoting the Flora Zeylanica (603.) for the Sapindus trifoliata. This erroneous name was reserved for the Species Plantarum, and could not be given in the Flora Zeylanica, where no specific names are used. The Conghas is mentioned in the place alluded to ; and if that has really ternate leaves, it is neither the Poerinsii of Rheede nor the Saponaria of Rumphius. This can only be determined by inspecting the her- barium of Herman ; but in the mean time I must observe, that Dr. Roxburgh describes the Schleichera trijuga as the A'i