Illustrations of Indian Botany, Vol. 1/Pittosporeae

by Robert Wight

LII.-PITTOSPOREAE.

A small order of plants, widely distributed, but principally confined to the tropics or the warmer portions of the temperate zones bordering on them, consisting of trees or shrubs with alternate, simple, entire, petioled, exstipulate leaves, and axillary and terminal, solitary, aggregated or racemose flowers.

" Sepals 5, deciduous, distinct, or partially cohering: aestivation imbricated. Petals 5, hypogynous, sometimes slightly cohering: aestivation imbricatpd. Stamens hypogynous, 5 , distinct, alternate with the petals. Torus not discoid, but often forming a stalk to the ovary of about the same thickness. Ovary solitary, usually imperfectly 2-5 celled; the dissepiments not uniting at the axis, and therefore apparently 1-celled: style 1: stigma 2-5 lobed, the lobes

equal in number to the placentas. Fruit baccate, or capsular, and loculicide ; cells polysperm- ous, usually incomplete. Seed* covered with a glutinous or resinous pulp, or arillale. Embryo minute, contained in a fleshy albumen near the hilum : radicle long : cotyledons very short. — Leaves simple, alternate, exstipulate. Flowers sometimes polygamous."

Affinities. The affinities of this order are still undetermined, and no two Botanists, who have given their attention to the subject, seem to be agreed on the station they should orctipy in the natural arrangement. " Brown in establishing this as an order remarks that it is widely different from Rhamnece and Celaslrinece, but without pointing out its real affinity" (Lindley). DeCandolle places it next to Poly galeae, but without assigning any reason for so doing. Achille Richard places it between Rutaceae and Geraniaceae, and remarks regarding it that " he genera which compose this family were formerly placed among the Rhamneae, but (heir hypo- gynous insertion removes them to a great distance from that family. M. DeCandolle places the Pittosporeae between Polygaleae and Franhen iaceae, but it appears to us that (bis family should be placed near the Rutaceae which it resembles in a great many characters." J.indley on the other hand refers the order to his group Albumenosae and places it between Ampeledeae and Olacineae, with both of which it associates in the structure of the seed, though as it appears to me it differs greatly in other respects. Notwithstanding the opinion of Dr. Brown, quoted above, we have, in our Prodromus, placed it next to Ce/astrineae, forming the last order of DeCandolle's sub-class of exogenous plants Di< hamydfjE Thalamiflor^. 1 his arrange- ment has been adopted, with a doubt, however by Meisner, but upon the whole seems more nearly correct than any of those which have gone before, and seems strongly supported by the fact of Dr. Roxburgh having referred one species of Piltosporum to the genus Celastrus, his C. verticellalus being in truth a Piltosporum. The variable character of the fruit in both, the loculicidal dehiscence of the capsular forms, and the usually albuminous seed of both, are all in favour of this station for the order, add to which, Bartling places it between Rhamneae and Celastrineae in his class Tiicoceae.

Geographical Distribution. New Holland is unquestionably the head quarters of this order, all the genera, except Piltosporum and Senaceae, being confined to that country ; species of Piltosporum are however very extensively distributed over the globe, being found not only widely diffused in Australia, but also in the Moluccas, China, Japan, and India, from the southern provinces of Ceylon up to the foot of the Himalaya mountains, and even in Madeira.

Properties and Uses. Nothing of any importance is known on this head. The wood of a species of Senacia, a native of the Mauritius, is handsomely veined, and the berries of a species of Bellardiera are eatable. The seeds of the Indian species are covered with a fragrant resinous fluid, which however soon dries on exposure to the air and loses its smell.

Remarks on Genera and Species. In addition to the three species of this order des- cribed in our Prodromus I have one undescribed from Ceylon, differing from all the others in having long obovate cuniate leaves and compound cory mbose inflorescence : that is, the terminal shoot and each of the lateral branches of the corymb constitute so many small corymbs. The leaves blacken in drying, which is not the case with the other species so far as I have seen. It forms a large shrub growing on the banks of water courses on the more elevated parts of the Island.

Pitiasporum Ceylanicum, (R.W.) Shrubby, diffuse, ra- form, each division corvmbose, petals 5, linear, about mous, leaves obovale cuneate, bluntly acuminated, en- three times the length of the sepals: young fruit tire, glabrous, congested near the summits of the densely clothed with greyish tomeutum, and termina- branches : panicles axillary and terminal; several to- ted by the persistent style and stigma: fruit glabrous, gether on the apex of the branches, springing from the 2-celled.

axils of a whorl of 4 leaves: peduncles longish, fili- Ceylon on the banks of streams among thick jungle.

EXPLANATION OF PLATE 70.

1. Pittosporum Neelgherrense — natural size.

2. A flower partially dissected, showing the insertion of the stamens, the ovary, style and stigma.

3. Anthers back and front views.

4. Ovary cut transversely, imperfectly 2-celled.

5. vertically, ovules superposed.

6. A mature capsule after dehiscence.

7. A seed — natural size.

8. magnified.

9. cut transversely, all albuminous — the cen- tral circle a mere flourish of the draughtsman — with the exceptions mentioned, all more or less magnified.

DIV: II.— CALICIFLORAE, D.C.

The families we have been hitherto describing are all referred by DeCandolle to his first division Thalamifiorae : distinguished by the inferior (hypogynous) insertion of the parts of the flower with reference to the ovary, in other words, by the insertion of the petals and sta- mens into the receptacle, the Torus being usually very narrow and unattached to the calyx. This division or class, owing to the constantly inferior insertion of the petals and stamens, has received from Jussieu the name of Hypopetal/e.

In the division, on the consideration of which we are now about to enter, the Torus is broad, and lines the bottom of the calyx, hence, the petals and stamens springing from its dilated and adnate part are commonly said to be inserted into the calyx. The petals in this division are either distinct or united together : and, the ovary is either free ( superior J ; or em- braced by and united with the tube of the calyx by the torus, ( inferior.)

This division is sub-divided into the following five classes — Peripetalae — Epipetalae — Epicorollae Corisantherae — (anthers distinct) Epecorollae Synantherae — (anthers united) and Pericorollae. The characters of these classes will be given at the head of each.

CLASS PERIPETALAE.

Torus between the ovary and tube of the calyx, but not forming a disk on the summit of the ovary. Petals distinct, and stamens perigynous.

To this class is referable all those orders having a superior or free ovary.

CELASTRINE^E.

This is a large family of plants partly tropical, but the greater number of them natives of the warmer regions bordering on the tropics. They are all trees or shrubs, with round or square branches, alternate or opposite, simple, or rarely compound, entire or dentate, petiolate, leaves : small deciduous stipules : and regular hermaphrodite small flowers, the bottom of which is lined with a large flat disk.

" Sepals 4-5, united at the base; aestivation imbricated. Petals as many as the sepals and alternating with them with abroad base, inserted under the margin of the torus; very rarely wanting : aestivation imbricated. Stamens alternate with the petals, and as many, in- serted on the margin or disk of the torus : anthers 2-celled, dehiscing on the inner side. Torus a large thick fleshy flat disk, coveting the bottom of the calyx. Ovary more or less immersed in arid adhering to the torus, 2-5-celled, or rarely from abortion 1 -celled: ovules usually two (sometimes only one, sometimes several) in each cell, attached to the axis, and usually at its base, at first ascending (sometimes afterwards by the elongation of the axis peritropal, or at last resupinately pendulous): styles 2-5, usually combined into one, sometimes distinct: stig- mas combined or distinct. Fruit free from the calyx, 2-5-celled (often partly abortive) : either indehiscent, drupacious, or baccate, or samaroid ; or membranaceous and follicular ; or capsu- lar, 3-4-valved. and loculicidal. Seeds one or several in each cell, sometimes arillate. Albu- men usually fleshy, sometimes very thin or wanting. Embryo straight : radicle short, pointing to the hilum : cotyledons usually thick. — Shrubs or trees. Leaves alternate or opposite."

Affinities. — Jussieu originally confounded this order with Rhamneae, from which it was afterwards separated by Brown on account of its imbricated aestivation, the stamens being alternate with, not opposed to, the petals as in that order, and its ovary always quite free. It is also nearly allied to Hippocratiaceae (which see) and to Euphorbiaceae, near which, in his alliance Euphorbiales Lindley places the order. Bartling takes a similar view and refers the order to his class Tn'coccae, the type of which is the Euphorbiaceae. Brogniart, who has so ably illustrated the Rhamneae , states it as his conviction that Celastrineae have more relation, to some orders with hypogynous stamens, especially Malpighiaceae, than to any with perigynous ones. Achille Richard places it between Rhamneae and Euphorbiaceae : Blume however continues to follow Jussieu in uniting them to Rhamneae. As all these eminent Botanists seem agreed on the owe point, that this order forms the transition from the hypogynous to the perigynous divisions of DeCandolle's arrangement, the one adopted in this work, it seems evi- dent, that it is most judiciously placed in its present situation as the connecting link between the two.

Geographical Distribution. This is a very widely distributed order, species being met with in the warmer regions of every quarter of the globe, but much more abundantly beyond than within the tropics. The Spindle tree, Euonemus Europaeus is a native of the north of Europe, and has several congeners in the Himalayas, and also in the alpine districts of this Peninsula. I have two species from the hills about Courtallum, and excluding one or two, which I am as yet unable to determine, two from the Neilgberries. Celastrus in like manner has a wide range in India, extending from the southern extremity of Ceylon up to the Himalayas, several species occurring at either end of these distant limits. Species of both of these genera are found in Java and China, as well as in America, and of the latter, in Africa. Celastrvs coreaceus of Senegambia, seems indeed almost identical with our C. montana. Elaeodendron has fewer species and a more limited range, being confined to the East and West Indies, and has one species in Norfolk Island. The tribe Ilicineae of which the Holly is the type has recently been removed from this, and placed as a separate order among the monopetalous orders, where it ranks near the Ebony's and Olives. Dr. Wallich's genus Kurrimia is found in Silhet and Ceylon, from the last of which stations I have two species. According to Wallich's list there are about 70 species, natives of India, referable to this order.

Properties and Uses. But little is known of these. The berries of Euonemus Europaeus are said to be violently emetic, and to prove poisonous to sheep. Of the Indian species, not one is mentioned by Ainslie as being used in medicine. The researches however of the late Dr. Herklots brought one to light well known to the natives as possessed of very peculiar powers, the seed namely of Celastrus nutans (Malkungnee of the natives;, from which he extracted, by a rather rude process an Emperumatic oil. This he administered with great success in Berriberri. In other hands it has not proved quite so successful. As however the subject is as yet imperfectly known and seems to merit further inquiry, I presume I need offer no apology for introducing a rather long extract from Mr. Malcolmson's essay on Berriberri, giving some account of its introduction into medical practice, and of its active constituents, as ascer- tained by chemical analysis, in the hope, that the very simple and easily made experiments of Mr. M. may induce others to extend the enquiry to the seed of other species, and thence perhaps, add another not less powerful remedial agent from a family which has not hitherto been sufficiently investigated.

"Oleum Nigrum, (Herklots). It remains to communicate such observations as I have made, regarding another native remedy of great value, styled by Dr. Herklots, to whom we owe its first introduction to the notice of Europeans, the oleum nigrum. It is prepared in the following manner. Into an earthen pot whose bottom is perforated by a number of small holes are put, malkungnee seeds tbijss, benzoin, cloves, nutmegs, mace, of each 3ss ; the mouth is closed and the pot placed over another and luted to it. They are then placed in a pit three feet deep and nearly as wide, and surrounded by cakes of dry cow-dung which are set on fire, and when they are consumed about six ounces of the oil is found in the under vessel ready for use. It should be kept in well closed bottles. As we would expect from the nature of the process, the appearance and sensible qualities of the product resemble thoss of an emphreumatic oil, possessing the peculiar powers of the principal ingredient, the malkungnee seeds. These do not appear to be the produce of the Circars, or as far as I can learn, of the Deccan, but Mr. Royle has in- cluded them in a list of Indian materia medica, and states that the plant is the Celastrus nutans.™ It is stated by Hukeem Mahomed Hussein Khan, to be a native of Hindoostan, and the seeds are imported from Calcutta or brought to the Circars from Hyderabad, and are only found in the principal towns, where they are kept in bags containing the seeds, seed vessels and stalks, from which the genus can be ascertained, as was done by Dr. Wight some years before the publication of Mr. Royle's catalogue. The name used in Hindoostan has been introduced into Arabic, Persian, Teloogoo. and Tamul. A tree called by the Mahrattas " malkanee," the seeds of which afford an oil, grows in the forest of Dongatal, north of Nagpore, but I did not see the tree nor am I sure that it is the same. 151 'J he seeds differ in quality, sometimes from age, but I have seen very fresh ones of inferior quality, apparently from their being pulled too soon. When new, the seeds are partially surrounded by a yellow unctuous tasteless farina, and when they are reduced to powder they form a paste from which a yellow oil, having in some degree the qualities of the seeds, may be expressed. These have a hot biting taste, permanent in the palate, and if many of them are masticated, a sense of giddiness and a peculiar slight sensation extending over the face and brow are felt„ A very slight taste and no acrimony is given to water distilled from them, and the oil on the surface of the residuum is bland. The seeds retain their taste. It was evident from this, that their virtues did not reside either in a fixed or volatile oil. Some of the seeds were coarsely powdered, and alcohol rubbed up with them and afterwards filtered ; it bad acquired a light yellow tinge. On this being dropped into water, an immediate separation of the resin it held in solution took place, in a white flake, which had a strong biting acrid taste exactly like that of the seeds, but much more powerful, and in which it appears the virtues of the plnnt reside. The alcoholic solution evaporated, leaves a beautiful yellow resinous paste which also possesses the qualities of the seeds. The black oil itself is a thick deep brown fluid, burning with a white flame and not acted on by acids. Its specific gravity is, at ninety, 1097,5, which is higher than any of the fixed oils ; its taste is rough, bitterish and acrid, its smell empyreumatic and peculiar. Water distilled from it is limpid, but has a good deal of the taste of the seeds, and the oil floating on the water in the retort is tasteless. The oil when rubbed up with alcohol forms a deep olive brown partial solution, and there is only a thick oily matter left on the filter, which has much less taste than the oil itself or the oily tincture. This last, on being dropped into water, separates into a white flake sinking in the wafer and of the .same quality as that obtained from the alcoholic solution of the seeds, and a fixed black oil having a slight bitter oily taste. On the mixture of the oil and alcohol standing for some time, the fixed od partly separates from the tincture ; and seems to pass through the filter with, but not dissolved by it. From these observations it appears, that the resin is combined with the oils existing in the ingredients employed, which are partly converted into an empyreumatic compound, and thus acquire the property of partially dissolving the resin."

Remarks on Genera and Species. Five genera only of this order had been met with in this country at the time we published our Prodromus, since then I have added one (Icones Plant. No. 162) intermediate between Celastrus and Euonymus. Of these I find Meisner proposes to change the name of one, substituting Sckrebera for Eloeodendron. Our Eiceodendron being in truth Retz's Schrebera but not Linnaeus', I confess I am not prepared to coincide with him in this alteration, for so far as I can see, he has not afforded satisfactory reasons for doing so : the characters of his genera Elceodendron and Schrebera, with the exception of the seed, being in effect the same. It appears to me, he has been induced to change our name, partly by a remark of ours to the effect, that " if Geertner be correct in describing the fruit of El. orientate, the type of the genus, as a3-celled drupe, the seeds with a fleshy albumen, and remarkably thin membranous cotyledons, then the Indian species must be removed." This passage is guarded by an if — and not without reason, since it does not appear by any means certain, that Gaertner's E. indicum, is identical with Jacquant's E. orientate, on the contrary, he (Jacquant) particularly mentions the 2-celled nut, which renders their identity very doubtful. With reference to the only other distinction assigned, viz. the seed being exalbuminous with thick cotyledons in the one ; and sparingly fur- nished with albumen and having thin foliaceous cotyledons in the other ; I consider it a mark of inferior importance, and object to it the more, as being in this instance a very unnecessary refine- ment, and but little applicable to general practice, where we have so often to distinguish genera from flowering specimens only. For these reasons I cannot with my present information adopt the innovation. A curious blunder of Meisner's may be here noticed, as it seems not improbable it had some influence in leading him to make two genera, where I think one might have served — He says, " Obs. Sclirebera stamina apud Schreb. (gen. 446) et Jiiss. I. c. (gen. plant) decuntur basi int'us squamulis minutis ciliatis prcedita, quartern vero nulla facta est mentio in char-generico apitd. W aiad Arn." The error here is, in supposing the Sclirebera of Linnaeus and Retz the same— the Lknasan plant to which the squamulis minutis ciliatis belongs, is a species of Cusuita— and the Linnaean S. schinoides was Cuscuta Africana growing on a Myrica. In the Sclirebera of Retz no such character exists, and as already remarked, I think it an unnecessary genus, and regret to see it taken up anew, by so excellent a Botanist as M eisner, in a work so likely from its general accuracy to have an extensive circulation and to be looked up to as one of considerable authority.

The genus Lophopelalum is characterized by having a 5-lobed calyx : 5 petals, each fur- nished with a crest, inserted under the margin of the torus : 5 stamens inserted on the surface of the disk; a broad 5-lobed disk covering the whole hollow of the calyx, the lobes opposite, cover- ing, and adnate with the base of the petals, and a 3-celled superior ovary with two rows of ascend- ing collateral ovules in each cell.

This genus is exactly intermediate between Celastrus and Euonymus, having the 3-celled ovary and quinary petals of the former combined with the insertion of the stamens on the face of the disk the cymose inflorescence and opposite leaves of the latter. It differs from Celastrus in habit and in the insertion of its stamens ; and from Euonymus in the unequal number of the cells of its ovary and petals, and in the cells being polysporous. Adopting these last as the essential characters of the genus, the crested petals, from which it takes its name, and the lobed disk may be viewed as characters of secondary importance, and thus a thiid species, of which 1 have specimens, partaking of the essential characters but wanting the others, might be introduced. Dr. Arnott proposes to add as a fourth species Wallich's Euonymvs grandifforus, which I cannot adopt, owing to its having 4 petals, 4 stamens, and a 4-celied ovary, thus virtually destroying what I consider the essential character of the genus, namely, the inequality existing between the floral envelopes and cells of the ovary. That plant ought perhaps to form the type of a genus, distinct from Euonymus, depending on its numerous ovules and seed but ought not to be allowed to mar the uniformity of this, by combining under one name, plants having symmetrical and insymmetrical flowers.

The genus Euonymus though not very extensive will I suspect require revision, but this cannot be undertaken until we are better acquainted with the fruit. Some new species of it are found on the Neilgherries of which I have specimens communicated by Mr. Gough, an enter- prising young Botanist, who explored much of these elevated regions, but who is, unhappily for science, forced by bad health to leave the scene of his usefulness. Two of these can be readily defined, the rest cannot be satisfactorily made out and must be left for future investigation, since, it is worse than useless to name plants which we have not the means of describing with sufficient precision to admit of their being afterwards recognized from the description.

Judging from a solitary not very perfect specimen in my herbarium of the Eu.fimbriatus Wall, a native of the Himalayas, communicated by the late Countess of Dalhousie, it appears, that that species belongs to the genus Pterocelastrus of Meisner, one hitherto only found at the Cape. My specimen is not in fruit, but the ovaries, shortly after the fall of the flowers, show the wings of the carpels already well formed. The specimen is from Masoori, and has much the habit of an Euonymus. This plant along E. japonicus. equally a native of Nepaul and Japan, adds another to the, already existing, numerous links, between the floras of these remote countries.

LOPHOPETALUM, (R. W.)

Calyx scutelliform, 5-lobed, lobes rounded, short. Petals 5, sessile, orbiculate, expanding, usually, fur- nished with a crest, and covered near the base with the projecting lobes of the disk. Torus discoid, 5-lobed, or angled, thick, fleshy, covering the whole cavity of the calyx, the lobes adnate to the base of the petals. Stamens 5, alternate with the petals, inserted on the disk, anthers versatile, ovate, 2-celled, dehiscing longitudinally, ovary free, 3-celled, ovules in a double series, 4-12, in each cell, ascending, style short, persistent, stigma obtuse, capsule — , seed — .

Glabrous trees, or shrubs, leaves opposite, petioled. Flowers numerous, on large spreading terminal cymes: Petals sometimes without a crest and the disk not lobed.

In the construction of this character, which is consi- derably altered from that published in the Icones 1 have availed myself both of a more extended acquaintance with the order, derived from recent study, and of Dr. Anion's character, which reached me as this article was passing through the press, to improve it to the ut- most, and at the same time, so to fix its limits as to prevent them interfeiiBg with those of Euonymus, which Dr. Arnott's character does not attain, and in so far weakens both. I give below Dr. Arnott's generic character.*

The genus naturally divides itself into two sections, one distinguished by its crested petals and projecting lobes of the disk, the other by the petals being without a crest, and having a 5-angled not 5-lobed disk.

§ I. Petals crested, disk 5-lobed, lobes adnate to the bases of the petals.

1. L. Wightianum, Arn. Leaves elliptic, oblong, slightly acuminated, quite entire, coriaceous, slightly villous beneath; crest of the petals a corrugated mem- brane.

Hab. — Malabar.

The villi on the under surface of the leaves, in this species is much more sensible to the touch th in evident to the eye, even when assisted with a high magnifier.

2. L. fimbriatitm. Leaves ovate, oblong, slightly acu- minated, quite entire, membranous, glabrous on both sides, crest of the petals fimbriated.

Hab. — Mergui — Griffith.

The much thinner and membranous leaves with the fimbriated crest of the petals at once distinguish this from, the former species. In addition to which the flowers are scarcely half the size.

§ 2. Petals not crested, disk 5-angled, not lobed, an- gles alternate with the petals.

This section ought perhaps to form a distinct genus.

3. L. floribundum, R.W. Leaves linear elliptic, acumi- nated, quite entire, coriaceous, glabrous on both sides ; panicles large, diffuse, many-flowered; flowers small, young capsules (?) acutely triangular, pointed.

Hab. — Mergui. — Griffith.

This is a most distinct species, and cannot be con- founded with either of the preceding, but it is still doubtful to me, whether it ought to remain in the genus.

EUONYMUS.

E. Goughii, (R. W.) Shrubby, glabrous, ramuli com- pressed : leaves somewhat tripli nerved shortly petioled, quite entire, oblong ovate, acute at both ends, acumi- nated: peduncles axillary-, short, 1-3 flowered: calyx cuellate, 5-lobed: petals 5, orbicular, fimbriated on

the margin : stamens5, inserted on the disk, connectivum of the anthers broad, cells placed transversely, dehis- cing lengthwise, ovary immersed in the disk, 5-celled, with two ovules in each.

Hab. — Neitgkerries. — Gough.

This species seems to approach E. grandiftorus, (Wall.) iii the size of its flowers, but differs in the quinary not quaternary number of parts, and in its petals being fimbriate on the margin.

2. E. acutangutus. Younger branches and ramuli acutely 4-angled, glabrous : leaves coriaceous, glabrous, quite entire, ovate, acuminated: cymes axillary, pedun- cles about half the length of the leaves, twice or thrice dichotomous: capsules conical, broad, truncated above, tapering downwards, 5-celled.

Ne ilgherries. — Gough.

For specimens of both these species I am indebted to Mr. Gough, who gathered them in the course of his excursions on the Hills.

3. E. revolufus, R.W. Young shoots, obtusely 4-an- gled, afterwards terate : leaves subsessile, coriaceous, revolute on the margin, quite entire, from elliptic to obovate-cunniate, with a short blunt acumination: peduncles axillary, paired, slender, 1-3 flowered, about one-fourth the length of the leaves; when one flowered, furnished with two minute bracts above the middle : calyx 5-lobed, lobes rounded: petals 5, orbicular : ovary immersed in the disk, 5-celled, with two ovules in each.

Hab. — Ceylon.

E. Walkerii, R. W. Branches slender, terate, extreme, shoots obscurely 4-sided : leaves subsessile, ovate, acu- minated at both ends, serrated towards the apex, quite entire below, somewhat coriaceous, glabrous: peduncles axillary, short, once or twice dichotomous ; flowers small, petals 5, orbicular, slightly undulated on the margin, capsule conical, 5-celled, large, 5-angled, tur- binato at the apex.

Hab. — Ceylon.

Flowering specimens of this plant were first commu- nicated to me bv Colonel Walker, I afterwards collected them in fruit, between the two there are some unimpor- tant differences, the leaves of the flowering specimensbe- insr larger, endingin a longer and more pointed acumen and more acutely serrated. These differences however could not be employed as specific characters to sepa- rate the two forms.

• Calyx scutelliformis 4-5-lobus, lobis rotundatis brevibus. Petala 4-5 sessilia orbiculata patentia, supra circa

basin enstis lobulisque carnosis instructa versus marginem nuda lajvia. Torus discoideus, 4-5-gonus crassus car- nosus calveis cavit item omnino implens. Stamina 4— 5, petalis alterna, supra discum inserta : filamenta persist- entia subulata. Antherce ovatffi, biloculares, longitudinaliter dehiseentes. Ovarium disco semi-immersum, 3-4- loculare. Ovula bisetialia, 8-12 in quoque loculo adscendentia. Stylus brevis persistens. Stigma obtusum. Capsula 3-4-angularis, 3-4-locularis, loculicide dehiscens, loculis sub-dispermis. Semina arillata.

EXPLANATION OF PLATE 71.

ElcBodendron Roxburghii — 1. Flowering branch — na- tural size.

2. An expanded flower seen from above.

3. The same, the petals removed to bring more clearly into view the disk and insertion of the stamens under its edge.

4. Stamens back and front views.

5. Ovary cut: vertically, showing the ovules attached to the base, erect.

6. Ovary cut transversely, showing it 2-celled, with two ovules in each.

7. A fruit nearly full grown.

8. Cut transversely, l-celled, I-seeded, the other ovules having aborted.

9. The same cut vertically, the seed still immature and not yet filling the cavity of the cell — all more or less magnified.

CELASTRINEÆ

ELAEODENDRON ROXBURGHII. (W.&A.)

CELASTRINEÆ

CELASTRUS PANICULATA. (Willd.)

EXPLANATION OF PLATE 72.

Celatrus paniculala—1. Flowering branch male plant—natural size.

2. An expanded flower seen from above.

3. The same, the petals removed, but the disk not clearly shown.

4. Stamens back and front views.

5. Ovary cut vertically, ovules erect.

6. ---transversely, 3-celled, with two ovules in each.

7. A raceme of fruit—natural size.

8. A seed—natural size.

9. The same—magnified.

10. A seed cut transversely, showing the embryo surrounded by copious albumen.

11. Cut vertically, showing the embryo in situ.

12. The seed, the testa removed.

13. The embryo detached, cotyledons foliaceous.

PITTOSPOREÆ

PITTOSPORUM NEELGHERRENSE. (W.&A.)