fuam trahat originem formamque propriam? Nam vix credibile, quod pollen flofculorum mafculorum plantæ fertilis fuas femellas impregnans, poffit alias plantas producere, quam matribus fuis fimillimas. Ergo ftatuendum, aut quod pollen ex planta fterili, quandoque fecundet ovaria plantæ fertilis, unde proles a matre diverfa ; aut quod pollen plantæ fertilis producere poffit fœtus, in totum forma et deftinatione diverfos: id quod autem miraculo foret proximum, quia hoc in cafu flofculi mafculi plantæ fterilis præter omnem fcopum atque neceffitatem a natura fuiffent producti. Habentur quidem exempla variarum plantarum fexu et forma diverfarum, et tamen ab eodem polline productarum; fed exemplum plantæ perfeca et ex natura inflituto fterilis, iftoque modo gignite, nullum, quod fciam, habetur. '

The male plant in this cafe (by no means perfect, nor probably ex natura inflituto fterilis,' being furnished with the rudiments of the female flowers) appears to be morbidly, though habitually, fteri! by abortion; a phenomenon obfervable in almost all the polygamous, and many of the dioecious plants. Nor is it unfrequent among plants that at first fight appear to be really hermaphrodite; fuch as feveral of the Sapindi, in which the ftamina of the flowers of one ftem will increafe at the expence of the ovary, which ceases to grow; while thofe of the female or fertile plant are found to contain incomplete anthers without any pollen. In the fame manner, what appear complete ftamina in the male flowers of the fertile plant of Arctopus, may poffibly be imperfect, fo that the influence of the other plant is requifite in order to effect fecundation,

Coprofma is one of thofe genera with whofe fructification we are but little acquainted, even after the defcription and figure here given of C. lucida, which, it may be remembered, does not appear to be that defcribed by Forfter under this name. We perfectly agree with Dr Gertner, that this genus belongs to the Rubiacea, and not to the Umbellifera; nor, indeed, do we know by whofe ingenuity it has been referred to the latter natural order. As Mr Brown of this city has moft probably met with feveral of its fpecies, in his interefting botanical expedition to New Holland, we may expect that this very acute naturalift will foon throw more light upon the history of the genus.

The two following plates (tab. 83. 84.) appear to us, more than the reft of this number, in the ftyle and fpirit of those of the preceding volumes; and, indeed, the fpecimens from which the figures are made (except Hydropityon, on which we shall remark by and by) were fo complete as to admit of that copiousness in the defcriptions, which conftitutes the chief merit of that work. Though feveral of the plants they reprefent are far from being uncommon, fuch as Cyclamen europæum, Soldanella alpina, Illecebrum, verticillatum, Coris, Gleux, &c. yet their fruits and feeds were

by

1

by no means fo well understood as it might be expected. With regard to Cyclamen, for inftance, we are even now uncertain whether its embryo be of one or two cotyledons. Dr Gærtner, the father, in defcribing Cycl. europæum, makes no mention of these organs, though fig. K. in the plate reprefents fomething like a dicotyledonous embryo; upon which his fon has made the following remark.

Embryo in hac ftirpe admodum fallax eft; in plurimis enim exemplis quæ pater defunctus fcrutinio fubjecit, ad, unum omnes invenit totos folidos; quamvis omnes oculorum intenderit nervos, optimafque adhibuerit lentes. Ex figura autem K, poft defcriptionem factam adjecta, elucet, ipfum tandem denique unicum ex centenis reperiiffe embryonem dicotyledoneum, de quo autem in defcriptione ne verbulum quidem fecit. Embryo itaque revera dicotyledoneus eft, cotyledonibus autem, in plurimis exemplis, breviffimis et vix ac nevix quidem difcernibilibus.'

But should we not rather suspect, that the solitary observation alluded to was founded in an optical or accidental deception, and that the embryo of Cyclamen is really of one lobe only? The general habit of the genus (though furnished both with calyx and corolla) would appear to be in favour of this fuppofition.

We fhall now fay a word or two of fome of the rarer plants contained in these two plates, and begin with Hydropityon zeylanicum. Its defcription is deficient with regard to the interior structure of the fruit and the feeds. From the generic character, however, and the figure, we are enabled to say, that the fubjoined fynonyms, of which the leading one is Tsiunda-Tsiera Hort. Malab. 12. t. 36, are all completely wrong. It cannot be doubted, that the plant just mentioned of the Hortus Malabaricus, is the Hottonia indica L., and that both, again, are the fame with Gratiola trifida, an obfervation also made by the late Profeffor Vahl in his Enumeratio Plantarum. Though Gærtner could not know this, it is ftill furprising that a botanist of his accuracy fhould not have found out the diffimilarity between his plant and that of Rhede's work, efpecially as the latter is defcribed as having only two ftamina, and a corolla entirely different from that of Hydropityon. To enable our readers to judge for themfelves, we copy Gærtner's generic character.

• HYDROPITYON.-Cal. inferus, pentaphyllus: foliolis ovatis conca. yis magnis incumbentibus. Cor. pentapetala calyce paulo brevior: petalis ovato-rotundatis concavis. Stam, 10 filamentis craffis brevibus, apice dorfo antherarum et bafi receptaculo genitalium carnofo et molli ter longeque villofo infertus; antheris obefis cordatis bilocularibus. Ovar. fuperum oblongum definens in ftylum fimplicem, ftigmate orbicu lato præpilatum. Capf. monofperma, aut femen nudum compreffum fulcatum.

Both the figure and defcription of Derana japonica, Thunb.

are very instructive. The dry berry (angidium coriaceum) of this plant has the peculiarity of being crowned at the top with five fquamæ clinging clofely to the ftile, and conveying the idea of a calyx fuperus. Were it not that this fruit has been communicated (as are almost all the reft here defcribed and figured) by Profeffor Thunberg himself, we fhould have doubted its being the plant taken up under that name in the Flora Japonica, where we looked in vain for the defcription of the fcales now mentioned; though, according to Gartner, they conftitute almost exclufively the effential character of the genus! Another genus of the Flora Japonica is Deuzia, whofe place in the natural feries appears to be problematical. We are inclined to believe, in oppoition both to Thunberg and Gartner, that the calyx in this genus is really adherent; and, indeed, Kämpfer fpeaks of it as a Caliculus globofus carnofus, fructus futuri, ut apparet, rudimentum.' In the figure of Torenia afatica, the fruit is well enough reprefented; but the corolla, profeffedly drawn from fancy, is objectionable both in regard to fize and fhape.

The laft of those five plates, which we confider as the late Dr Gartner's own, throws light on some fruits which were not well known before, fuch as Halleria lucida, Ourifia of Commerfon, and Difandra proftrata. But with regard to Sarcodactylis, contained in the fame plate, our author labours under a strange mistake. Thofe who have opportunities of feeing Chinese drawings of ve getables, must have often met in them with a large, ftrange looking yellow fruit, fantaftically grown out at its widened top into flethy appendages, which bear a diftant refemblance to fingers, efpecially if they are only five in number, which is often the cafe. This is nothing elfe than the fruit of a fpecies of Citrus, and probably a monftrofity; it contains no feeds, but is generally furnished with fome irregular loculaments. The figure here given is undoubtedly that of fuch a lemon, of the fmaller fort; but Dr Gærtner miftakes it for the fruit of the famous hand-bearing plant, the Macpalxochi-quabuitl of the Mexicans, figured in Hernandez's work, from whence it is erroneously quoted by Linnæus as a fynonym of his Helicteres apetala. To reconcile the diflimilarity fubfifting between his figure and that of Hernandez, the author calls the latter miferrima; an epithet which, however applicable it may be to most of the figures of that infignificant work, is lefs fo in this cafe. It gives a tolerably good idea of what it is intended to represent.

The five laft plates prefent us with the first refults of the younger Dr Gærtner's carpological tour and fubfequent labours. They are feven genera, four of them entirely new, namely, Shorea, Roxb. MSS.; Dryobalanops, Dipterocarpus, and Lophira, Banks's

MSS.

MSS. Their appearance is fingularly beautiful, from the lucinia of the calyx growing out, after the flowering is over, into long wings, covering the pericarp, and of different fhape in the different genera. The conftruction of the embryo of the three former is remarkably complicated, but pretty well explained both by the descriptions and figures. With regard to his Dryobalanops, we notice two errors: first, this excellent timber-tree is not, as here stated, a native of Ceylon, but of Sumatra; and, fecondly, it does not yield cinnamon, but camphor, known by the name of the Sumatra camphor, and mentioned by Kæmpfer, and alfo by Mr Marsden in his account of that island. Befides thefe, we have Vateria indica, Genipa americana, and Tocoyena of Aublet. The firft of thefe genera being referred by Retzius and his followers to Elaocarpus, affords another proof how little attention botanists pay to the fruit; for even Rhede's figure fhews how little it has to do with that genus.

We cannot dismiss this work without expreffing our wishes that no obftacles may arife to its uninterrupted continuation; and, at the fame time, exhorting its author not to facrifice to expedition the proper selection of his materials, or the accuracy of his delineations and engravings. Dr Gætner cannot fail to know what anxious diligence his father beftowed upon the latter, and that he was even in the habit of fending proofs to London, for the infpection of an eminent botanical artift then refiding in that city.

ART. V. On an Artificial Subflance which poffeffes the principal Characteristic properties of Tannin. By Charles Hatchett, Efq.

F. R. S.

Additional Experiments and Remarks on an Artificial Subflance which poffeffes the principal Characteristic properties of Tannin. By the fame Author. From Phil. Tranf. for 1805. Part II.

WE E have generally contented ourselves with selecting from the memoirs of academies the papers moft interesting, either for novelty or merit, and have paffed over the reft, without attempting to give any account of their contents. In purfuance of this plan, we are now to direct the attention of our readers towards fome of the most curious and important fpeculations that are to be found in the late volumes of the Royal Society's Tranfactions. Much as Mr Hatchett has contributed to the advancement of chemical fcience on former occafions, we think the fervices which thefe recent inquiries render to that

branch

As

branch of knowledge, are of a different and a higher caft. they are contained in two papers, we have thought it right to take them together, and confider them as parts of the fame inveftigation.

Before Mr Hatchett entered upon this courfe of experiments, it had never been fuppofed that tannin could be produced artificially. Mr Chenevix had indeed found, that a decoction of coffee berries had not the quality of precipitating gelatine, unless they were previoufly roasted; and hence it might be concluded that the process of burning produced tannin. But this was only an indiftinct and imperfect inference, and we were ftill left to regard tannin as exclufively prepared by nature. The inquiries of Mr Hatchett, however, place this matter in a new light, and warrant the conclufion, that if not tannin itself, at least a body refembling it in its chief characteristic properties, of precipitating gelatine, and rendering fkins of animals infoluble in water and imputrefcible, may be obtained by a fimple procefs, both from vegetable, mineral, and animal matter.

When nitrous acid is digefted upon afphaltum, jet, or other bitumens, containing a portion of uncombined carbonaceous matter, a yellow vifcid fubftance is feparated, perfectly fimilar to that which we obtain by the fame procefs with refins; but the remainder of the folution is a dark brown colour, and has different properties. The former confifts of the effential portion of the bitumen; the latter of the uncombined carbonaceous matter. A fimilar product is obtained from the various kinds of mineral coal; but thofe which contain no bitumen, yield none of the yellow folution. Having by this process of digeftion with nitrous acid, obtained the dark brown folutions from bitumens, coals, and charcoal, they were evaporated to drynefs, and the refidua examined. They reddened vegetable infufions, were highly aftringent, and feparated glue or ifinglafs from their foJation in water, forming a precipitate quite infoluble, either in hot or cold water. A fimilar product was obtained from digefling ifinglafs itfelf in nitric acid; and the folution of ifinglafs in water being added to the fubftance procured from the digeftion of ifinglafs with acid, the ifinglafs was precipitated from the water in an infoluble form: uncharred wood, or even bovey coal, which had the appearance of being only half charred, yielded none of this fubftance by treatment with nitrous acid; but when charred, the fame bodies gave it in great a

bundance.

Our author having been engaged in an interefting course of experiments upen the production of coals in the humid way, he found, that by uniting the refults of this inquiry with the pro

cefs