Hemsley, Biologia Centrali-Americana

bd l. pte 1 1-96 pls. i-vii Sept. 1879 2 97-18h viii-xiii Nov. 1879

3 185-280 xiv, xv Feb, 1880 h 281-376 xvi, xvii Apr. 1880

5 377-h72 xviii-xxx June 1880

6 73-568 soo -xxxv Aug. 1880 2. pt. 7 1-96 xxxvi-xliii Apr. 1881 8 97-192 . xliv-xlix June 1881 (vol. 1: 569-619 (Oct. 1881 7 (vol. 2: 193-232 1 (Oct. 1881 10 233-328 Lli-lvi Dec. 1881 11 329-))2h) lvii-lxi Feb. 1882 12 26-620 lxii-lxviii Apr. 1882 13 521-621 lxix June 1882 3. pt. 1h 1-96 ixx-lxxvi Oct. 1882 | 1S 97-18 laxvii-lxxx Jan. 1883 16 185-280 Ixxxil-lxxxiv Nov. 1883 17 281-376 lxxxv-xciv Feb. 188) 18 377-72 xXCV=c Feb. 1885 19 73-568 Nov. 1885 -20 569-66 ci-eviii Dec. 1885 91 (3? 665-711 cix (Feb. 1886 (4: 1-8 . (Feb. 1886 he pt. 22 o-1hk Mar. 1887 23 15-298 Aug. 1887 2h 299-10 cx Dec. 1887 25 11-98 Oct. 1888

List of plates - Preface - Introduction, etc.

BIOLOGIA CENTRALI-AMERICAN A;

CONTRIBUTIONS TO THE KNOWLEDGE OF THE FAUNA AND FLORA OF

MEXICO AND CENTRAL AMERICA

EDITED BY F, DUCANE GODMAN AND OSBERT SALVIN,

BOTANY, VOL. I.

Wer BY W. BOTTING HEMSLEY, A.LS.,

HON. MEM. NAT. HIST. 80C. MEX.; ASSISTANT FOR INDIA AT THE HERBARIUM OF THE ROYAL GARDENS, KEW; AUTHOR OF THE “‘ BOTANY OF THE ‘ CHALLENGER’ EXPEDITION,” &c.

AND

A COMMENTARY ON THE INTRODUCTION AND APPENDIX.

BY Sir J. D. HOOKER, Late Drrecror or tHE Roya GarpEns, Kew. oo. it hash ; f ee a \ 744400 } ; SS A LONDON:

PUBLISHED FOR THE EDITORS BY R. H. PORTER, 10 CHANDOS STREET, CAVENDISH SQUARE, W.,

AND

DULAU & CO., SOHO SQUARE, W. 1879-1888.

FLAMMAM.

PRINTED BY TAYLOR AND FRANCIS, REP LION COURT, FLEET STREET,

PREFACE.

WueEN this work was commenced, a little more than twelve years ago, the main object in view was the collection of the widely scattered data bearing upon the phyto- geography of the region, in order to ascertain to what extent the phenomena agreed with, or deviated from, those obtaining in the Animal Kingdom, and also to supply as complete a synopsis of the flora as possible, so that its general character, relationships, and connections might be critically elaborated. Nevertheless, it was hoped that it would at the same time prove a substantial and useful contribution to Systematic Botany ; and as the work proceeded greater attention was bestowed upon this branch of the subject, so that ultimately it grew far beyond the dimensions originally laid down. As this great task was undertaken by one person, it was necessary to keep it within limits as narrow as were consistent with the aim in view, to ensure a reasonable prospect of its being completed. A critical determination of the vast amount of material in the Kew Herbarium alone was out of the question, to say nothing of the supplemental collections in other establishments; yet it was difficult to decide where to draw the line. At first it was thought practicable to include the named materials at Kew, the British Museum, and Paris—the first forming a wide and trustworthy basis for genera, and largely also for species, resulting from the labours of Bentham and Hooker, in connection with their now happily completed invaluable ‘Genera Plantarum ’; but this plan had to be abandoned in consequence of the risk of confusion arising from diverse determinations in the various herbaria; and it was decided not to attempt doing more than could be accomplished at Kew. ‘This course has been adversely criticised, but having intentionally and purposely thus curtailed the scope

of the work, such criticisms call for no further reply. Considering that, for obvious | a 2

iv PREFACE.

reasons, anything very closely approaching completeness could not have been attained had there been half a dozen workers in the field, instead of only one, the omission of some small historical collections is of the very slightest importance. Notwithstanding the fact that the work has been substantially restricted to the Kew Herbarium and Library, it has been, as already mentioned, considerably extended in a direction not originally contemplated. For the purposes in view it was thought that we might very largely rely on the names as we found them in the Kew Herbarium, and describe only such very evident novelties as did not involve too great an expenditure of time. In this manner the whole of the Polypetale was written out, and in the hands of the Editors, when an offer was received from Drs. Parry and Palmer to present the first set of a large collection of dried plants, chiefly from the State of San Luis Potosi, on the condition that we named the whole of them and embodied them in the “ Biologia.” To do this properly involved an enormous additional expenditure of time and money; yet the offer was accepted, and the writer, aided by his colleagues at Kew, spent nearly seven months on this collection, and the investigations it entailed, thereby greatly enhancing the value of the Enumeration, without considerably increasing its

bulk.

This critical examination of a large portion of the Mexican plants led to much fuller synonymy and references to the existing literature in the second and succeeding volumes than is given in the first, and we are fully justified in saying that the quality.

of the work has improved in consequence.

Although an immense amount of time has been spent in looking up published species not represented in the Kew Herbarium by authenticated specimens, some, doubtless, have been overlooked; but a few omissions are of little consequence. Had every name referring to a Mexican plant been taken up, the result would have been literary completeness, it is true, yet nothing more. Moreover, it would be mere affectation to apologize for shortcomings of this kind in a work professedly little more than a skeleton from the standpoint of a systematist. As explained in the Appendix, upwards of a thousand nominal species are left out of consideration in the geographical tables, because it is believed that their retention would swell the proportion of the

endemic element beyond what it really is.

PREFACE. v

The writer now has the great pleasure of recording his obligations and thanks to Mr. W. H. Fitch, the artist, and to his colleagues at Kew, especially to Professor D. Oliver, whose unrivalled knowledge of flowering plants is always at the service of others. Mr. J. G. Baker is almost wholly responsible for the nomenclature and limitation of the species of vascular cryptogams; but this matter is more fully

explained in the remarks under the various orders.

The writer also feels that he would very much like to be permitted to. mention that he has experienced the most liberal treatment from the Editors, who have expended a much larger sum of money (to say nothing of time) on the work than they can possibly see returned. Under other circumstances much more might be written on

this point.

The coloured Plates were taken from sketches made from the fresh plants by Mrs. Salvin in Central America in 1873-74.

WILLIAM BOTTING HEMSLEY. Chiswick, October 1888. |

CONTENTS OF VOLUME L

Page

PREFACE. © 6 6 ee ee ili 6

INTRODUCTION. © 1. ee ee 1X

Commentary on the Introduction and Appendix . . . © - - e/ + + ee ee lxii

Enumeration of the PoryreraLa, with Descriptions of New Species . . . . . . . - 1-576

INTRODUCTION.

AS stated in the Preface, this work was undertaken mainly in the interests of Geogra- phical Botany; and the distribution of the plants enumerated therein has been tabulated and discussed in considerable detail in the ‘‘ Appendix ” contained in the fourth volume.

The completion, or approaching completion, of several important works on systematic botany, dealing with the vegetation of large areas, such as Boissier’s ‘ Flora Orientalis,’ Hooker’s ‘Flora of British India,’ Gray’s ‘Synoptical Flora of North America,’ and monographs of large Natural Orders, together with recent botanical explorations in China, Madagascar, the mountains of Tropical Africa, and elsewhere, affords materials for a wider survey of the distribution of plants than has hitherto been attempted, and a closer comparison of the primary botanical and zoological regions of the world. To do this exhaustively would, of course, occupy much time and fill a large book; therefore only an exposition and rapid review of the principal facts will be attempted here *, and the inquiry will be limited to flowering plants.

Before approaching the examination of the botanical regions themselves it seems desirable to produce some further statistics and then endeavour to estimate their relative value yf.

Throughout this work the classification and generic limits of Bentham and Hooker's ‘Genera Plantarum’ have been followed, and all comparisons are made on the same, or practically the same, basis. Since the appearance of the first part of the ‘Genera Plantarum’ in 1862, very numerous new plants have been discovered, including some extremely singular and anomalous ones, though none probably to which the authors would have assigned the rank of a new natural order. The number of distinct genera and species has, however, been largely augmented. Elsewhere { some particulars have been given of the subsequent additions to the Composite, but it is unnecessary to enter into similar details respecting all the natural orders. Nevertheless, for purposes of comparison, it will be convenient to give here some of the statistics and rough approxi- mations arrived at in the ‘Genera Plantarum’ §.

* The questions discussed in the following pages might more appropriately have been incorporated in the “ Appendix ;” but this is a further development of the subject suggested by Sir Joseph Hooker after perusing the analysis of the Flora of Mexico.

+ Following the most approved nomenclature, the primary geographical divisions of the vegetation of the world are designated “ regions” and the secondary divisions “ subregions.”

+ Vol. IV. p. 249.

§ From a summary by N. E. Brown in the ‘ Gardener’s Chronicle,’ n. s. xix. p. 733.

BIOL. CENTR.-AMER., Bot. Vol. I., October 1888. b

x INTRODUCTION.

STATISTICS OF THE PHANEROGAMIC FLORA OF THE WoRLD.

Orders. Genera. Species,

Dicotyledones . . . . . . « 165 6052 77311 Gymnospermee. . . ... ; 3 4A 415 Monocotyledones . . . . . . 34 1489 17894: Totals . . . . . « « ~ 202% 7585 95620 +

At Kew it is the practice to post up all proposed new genera as they are published, and from a cursory examination of their claims to this rank (in Bentham and Hooker’s sense) the number of distinct genera now known is about 8000; and, allowing a proportionate increment of new species, the total may be placed at 100,000. Judging from the exceedingly large number of new forms in the latest collections from the Malayan Peninsula, Borneo, New Guinea, and Central China, future explorations will doubtless considerably increase these totals. Absolutely nothing is known botanically of immense tracts of the interior of Africa; whilst such comparatively well-explored countries as Mexico and Central America still continue to yield as much as ten per cent. of new species in collections made out of the beaten tracks, and in a North- Mexican collection of about 270 species made by C. G. Pringle in 1887, 20 per cent. are indicated as new in a catalogue by A. Gray and S. Watson.

The general distribution of the natural orders is given in our fourth volume, pages 201 to 207. Below are enumerated those natural orders of plants estimated by Bentham and Hooker to contain 1000 species and upwards :—

Genera. Species. Genera. Species. Composite . . . . 782 9800 Asclepiadaceew. . . . 147 1300 Leguminose. . . . 403 6500 Umbellifere . . . . 158 1800 Orchideet . . . . 3884 5000 Solanacee . . . . . 67 1250 Rubiacee. . . . . d4l 4100 Crucifere . . . . . 178 1200 Graminee . . . . 298 3200 . Boraginee. . . . . 68 1200 Euphorbiacee . . . 197 3000 Palme . ... . . I182 1100 Labiate . . . . . 186 2600 Campanulacee . . . 54 1000 Cyperacee . ... 61 2200 -Ericacee . . . . . 52 1000 Liliacee . . . . . 187 2100 Cactacee . . .. . 14 1000 Scrophularineer . . . 158 1900 Rosacee ..... 71 1000 Myrtacee. . . . . 78 1800 Piperacee$ . . . .. 8 1000 el a ‘og 500 Totals . . 25 . 4276 59200 Acanthacee . . . . 120 1350

* In this work the Fumariacesw are counted as a distinct natural order, which brings the total to 203.

+ In some instances Bentham and Hooker indicate the number of species of an order as ranging between two sums, the lower of which was taken by Mr. Brown. From an independent calculation, based partly on the higher, and partly on the mean of the two sums given, a total of 96680 species is obtained.

+ 4500 to 5000 in the ‘ Genera Plantarum.’ § Probably overestimated.

INTRODUCTION. xl

From the foregoing it will be seen that twenty-five, or one eighth, of the natural orders include much more than half the genera, as well as of the species; and all of them are very widely dispersed ; sixteen of them extending to the arctic regions.

Twenty natural orders have only from one to three genera and less than ten species,

namely :— Genera. Species. Genera. Species. Salvadoracee. . . . « 38 9 Balanopsee. . . . . . Il 8 Sarraceniez 3 8 Mayacez dl 7 Cyrilleze 3 8 Platanacez . 1 6 Roxburghiacee . 3 8 Coriaries 1 3 Flagellarieze 3 8 Moringese 1 3 Datiscaceze 3 4 Columelliacez . 1 2 Lennoacez 3 4 ‘Leitnerier . 1 2 Empetracese . 3 4, Ceratophyllee . 1 2 Philydracee . 3 4, Batidez . 1 1 Calycanthaceze 2 . 4 —_ — Canellaces owe ee 4, Totals . 40 99

Some of the foregoing orders are quite local, while others have a wider range, particulars of which are given in the fourth volume, pp. 171-205. There are six other orders of only one genus each, namely, the American Lacistemacee, com- prising sixteen described species, and the Australasian Stackhousiacee * of about twenty species; the widely spread Frankeniacee and Myricacee, and the Old-World Nepenthacez and Casuarines:: all rather numerous in species.

As another illustration of the Flora of the World, a list of genera numbering 300 species and upwards each is appended, together with their general distribution f.

Genera. . Species. Distribution. Senecio . . . . - 900 Nearly all over the world. Solanum. . . . . 700 Generally dispersed in warm regions, rarer in temperate. Piper. . . . . . 600 Tropics, and extending to Japan and New Zealand. Euphorbia . . . . 600 Almost everywhere, except the coldest regions. Ficus. . . . ... 600 Universal in the tropics, rare in temperate regions. Astragalus . . . . 500 North temperate and cold regions; rare in the south, and wanting in South Africa and in Australasia. Eugenia. . . . . 500 Tropical and subtropical regions. Psychotria . . . . 500 All over the tropics. Croton . . . . . 500 Tropics and extratropical America. Carex. . . . ~ . 500 In nearly all temperate and cold regions. Phyllanthus . 450 Very widely dispersed in warm regions, rare in temperate.

Salvia. . . . . +. 450 North temperate and subtropical regions, rare in the south, except the Andes and South Africa.

* The somewhat anomalous monotypic Australian genus Macgregoria, Muell., is referred to this order, a fact overlooked in the general table, vol. iv. p. 177. + See Bentham in ‘Gardener’s Chronicle,’ n. s. xix, p. 371.

b 2

xii INTRODUCTION.

Genera. Species. Distribution. Acacia . . . . . 482 Generally in warm regions; numerous in Australia.

Eupatorium. . . . 400 Temperate and subtropical regions except Africa and Australia, but most numerous in South America.

Erica. . . . . . 400 Europe, Mediterranean region, and very numerous in South Africa.

Peperomia . . . . 400 All over the tropics with few extensions into temperate regions.

Epidendrum . . . 400 Tropical and subtropical America; the Galapagos Islands.

Cyperus. . . . . 400 All warm regions; rarer in temperate.

Vernonia. . . . . 380 ‘Tropics generally; rare in temperate regions.

Cassia. . . . . . 338 All tropical countries; rare in temperate regions.

Loranthus . . . . 330 Warm regions; rarer in temperate.

Centaurea . . . . 820 Europe, Africa, and Asia; six in America; one in Australia.

Myrcia . . . . . 800 Tropical and subtropical America.

Miconia. . . . . 800 ‘Tropical and Andine America.

Mamillaria . . . . 800 Mexico to Bolivia.

Mesembryanthemum. 3800 Mediterranean region to Cape and Australasia; chiefly Cape.

Ipomea. . . . . 800 All warm and most temperate countries.

Quercus. . . . . 800 North temperate and subtropical regions and extending to New — Guinea.

Totals. . 28 . 12400

These twenty-eight genera contribute 13 per cent. of Bentham and Hooker's estimate of the total number of species of flowering plants. With regard to the actual numbers given, some of them, at least, are considerably below what the authors would them- selves probably distinguish as species. Thus, for instance, Bunge, in a monograph of the Old-World species of Astragalus, published since the corresponding part of the ‘Genera Plantarum,’ describes 971 species, and he has subsequently increased the number to upwards of 1000; and there are at least 200 American species. The estimated number of species of Piper and Peperomia is perhaps in excess of that actually existing. .

There is one more feature in the world distribution of flowering plants demanding attention, and that is the wide or peculiar range of certain species, which possess no special means of dispersion by animals or birds or the elements, and which are most unlikely to have been aided, intentionally or unintentionally, by man. Such species are not few in number, but a sample of them will be sufficient as an illustration of distribution, and a selection may be made from those inhabiting Britain. Sir Joseph Hooker has discussed this phenomenon with greater fulness than we have space to do in this place *.

.Radiola millegrana . . . 1. 1. 1 we Europe, North Africa, and mountains of Tropical Africa. Alchemilla vulgaris. . . 2. 2... us Western Europe to N.E. India, Labrador, Green-

land, and Alps of Victoria, Australia.

* ¢ Flora Tasmanie,’ i. Introductory Essay, p. xciv; and also “ On the Plants of the Cameroons Mountains,” in the Journal of the Linnean Society, Bot. vii. p. 178.

INTRODUCTION. xili

Cotyledon umbilicus Europe, N. Asia, and mountains of Tropical Africa. .

Europe, N. Asia, N. America, and Australia.

Europe, Asia, N. Africa, N. and 8. America, and Australasia.

Western Europe, mountains of Tropical Africa.

Europe, Asia, N. and 8. America, Australasia.

Europe, N. Africa, N. Asia, and N. America.

North and South temperate and Arctic regions and mountains of Tropical Africa.

Temperate and cold regions throughout the

World.

Lythrum salicaria Calystegia sepium

Sibthorpia europea . Brunella vulgaris Lycopus europzeus Deschampsia ceespitosa .

Luzula campestris

It is not intended to discuss the various means by which the above-named plants may have been thus dispersed—whether by migrations or natural agencies, inasmuch as this has already been done as exhaustively, perhaps, as the data permit.

The following table, extracted from an abstract of a paper by Mr. T. Comber on the world distribution of British flowering plants *, is a further illustration of the relatively wide distribution of what Sir Joseph Hooker has designated the Scandinavian Flora. It is a summary of the distribution of the vascular plants regarded as indigenous to Britain, divided into four climatal classes, according to the latitude or altitude they inhabit. The first column contains the names of these classes, which are sufficiently descriptive to be intelligible; the sixth column the total number of each class, and the intermediate columns the number of species extending to the countries or areas named.

General Distribution of British Plants. (After Comber.)

Europe Europe Europe. and and Universal. Total. Per cent.

Asia. America. Southern. ....... cee eee eee 149 123 6 16 294 26 Temperate ........ eee eee 61 299 12 264 636 57 Northern ........---e eee 12 13 5 94 124 11 Arctic... 2 cee ee eee mm) 3 8 53 69 6 Total .......eeeeeee 227 438 31 427 1123 100

Per cent. .......--55- 20 39 3 38 100

Jt will be observed from the totals in the sixth column that none of the species

* ‘Journal of Botany, 1874, p. 88. The original appeared in the ‘ Transactions of the Historic Society of

Lancashire and Cheshire.’

xiv INTRODUCTION.

appear in two categories; and it should be explained that a considerable number of the plants indicated in the second column as having only European extensions actually reach North Africa. The term “ universal” is applied to those plants which spread into all three of the northern continents ; and it also includes the cosmopolitan species. With regard to the “ southern universal ” the majority extend into the tropics; and of the “temperate universal ” sixteen species recur (in the southern hemisphere) in America only, ten in Africa only, and thirteen in Australia and New Zealand only; while fifty-six are more widely spread in south temperate regions. Though the foregoing numbers can only be accepted as approximate, they convey a good idea of the wide distribution of the British Phanerogamic Flora, which does not contain a single well- marked endemic species. |

STATISTICAL COMPARISON OF THE FLORAS OF LARGE AND WIDELY SEPARATED AREAS.

In the ‘ Appendix’ (vol. iv. pp. 202-207) comparisons are made between the Floras of Mexico and Australia, and the relative positions of the natural orders shown, according to their predominance in species. No very special meaning or importance is _ attached to such comparisons ; yet they are exceedingly interesting, and teach something beyond mere numbers of species, especially to persons possessing a practical knowledge of the nature, size, duration, &c. of the plants constituting the various natural orders. It is thought, therefore, that an extension of such comparisons to a third area, that of British India, may be welcome. There are many similarities as well as diversities in the American and Asiatic areas: similarities in the altitudinal and latitudinal ranges of the two countries, and in‘a less degree in their climatal conditions; diversities in the direction of the principal mountain-chains, and consequently of the aspects of the slopes or exposures, and diversities in the land-connections, drainage, and coast-line. It would unduly lengthen this sketch to enter into further particulars on these -points, therefore a tabular view of the composition of the Indian Flora compared with those of Australia and Mexico* may follow here. This table was drawn up, with the assistance of Sir Joseph Hooker, from his ‘ Flora of British India’ as far as published, and from the Kew Herbarium and recent monographs, such as Engler’s ‘ Aracee’ and Baker’s various papers on petaloid monocotyledons. ‘The writer, however, is entirely responsible for the numbers of species of the genera of Orchidee, Scitaminee, Gramineae, and some smaller orders, whilst Mr. C. B. Clarke obligingly furnished the numbers of the Cyperacee f. At Mr. Salvin’s suggestion a column showing the position of the natural orders in the Flora of the World has since been added.

* For shortness Mexico is employed here and elsewhere instead of Mexico and Central America.

+ Owing to some mistakes in the figures, chiefly in the number of species of Quercus, discovered after the table was compiled, the percentages were calculated from a total of 13,647 instead of 13,700. The orders affected have been transferred to their proper positions, but it was not thought desirable to make any other alterations.

INTRODUCTION. XV

Sequence of the Natural Orders of the Phanerogamic Flora of British India according to their predominance in Species, with their relative percentages, and their positions in the Floras of the World, of Australia, and of Mexico.

Orders. , Species. Genera. Position Position Position Percentage in the in the in the Position in the Total in Number Total Number Flora of the | Australian Mexican Indian Flora. number. | Phanero- | endemic. | number. | endemic. World*. Flora. Flora. gamia. 3 7 3 1. Orchidew.............. 1060 17 969 106 20 2 1 2 2. Leguminosee .......... 831 6:09 482, 182 18 5 6 5 3. Gramines ..........6. 800 - §:86 476 134 11 6 9 8 4, Euphorbiacee .......... 624 4:57 473 79 12 4 14 7 5. Rubiacew.............. 611 4-48 446 89 21 1 4 1 6. Composite ............ 598 4:38 381 122 11 13 51 16 7. Acanthacee.....-...... 503 3°69 433 49 11 ‘8 5 11 8. Cyperacem ............ 385 2°82 171 24 7 15 9 9. Labiate .............. 331 2°43 226 55 8 12 33 20 10. Urticacese ............ 305 2°23 196 45 6 14 35 17 11. Asclepiadee............ 249 1°82 200 53 15 18 73 26 12. Rosacee ...........--- 218 1-60 144 26 2 10 30 14 13. Scrophularinee ........ 215 1:58 99 55 7 20 44 65 14, Laurines .............. . 205 1:50 159 16 1 32 87 76 15. Scitamines ............ 204 1:49 162 20 5 34 73 76 16. Anonaceee ............ 192 1-41 165 25 7 9 13 25 17. Liliacee .............. 178 1:30 93 36 3 11 103 21 18. Melastomacee .......... 166 1:22 120 21 2 22 97 53 19. Geraniacee ............ 165 1-21 137 9 11 2 46 20. Myrtacew.............. 157 1:15 117 11 1 14 19 34 21. Umbellifere .......... 154 1:13 114 37 3 21 32 19 22. Convolvulacez.......... 152 Lil 75 15 1 20 89 24 23. Aroides ........ cece ee 142 1:04 107 32 11 16 38 22 24. Boragineew ............ 140 1:03 82 32 3 16 37 37 25. Crucifere...........05. 137 1:00 55 43 4 28 62 42 26. Gentianee ............ 132 0:97 105 15 2 20 40 36 27, Apocynacer..........6. 131 0:96 94 39 3 23 121 18 28. Gesneracee ............ 129 0:95 107 25 10 23 29 33 29. Verbenacesw............ 128 0:94 78 23 17 55 23 30. Palme.............05. 126 0:92 94 30 1 25 7 53 831. Ranunculacee .......... 115 0°84 73 19 1 32 89 61 32. Anacardiaces® .......... 112 0-82 86 22 8 38 36 57 33. Tiliacesee ......... eee 110 0°81 56 13 3 34 79 59 34. Celastrinee ............ 105 0-77 85 13 2 19 60 52 35. Caryophyllacee ........ 104 0:76 57 19 1 34 112 31 36. Cupulifere ............ 99 0-72 95 6 45 71 85 37. Ampelidew ............ 94 0-69 71 3 25 57 44 38. Polygonacee .......... 93 |. 0:68 38 7 61 . . 39. Dipterocarpee.......... 92 0°67 80 9 2 23 18 13 40. Malvacesz............-. 91 0:67 40 22 6 28 15 55 | 41. Sterculiacee ) ........ 88 0°64 60 17 2 29 86 56 42. Myrsineze \ Leeeee “e 88 0°64 61 11 3

* Orders represented by the same number of species in the Flora of the World, in the Australian Flora, or in the Mexican Flora are indicated numerically as occupying the same position in the sequence, otherwise some of them would be removed a considerable distance from their true positions ; and in comparisons the same course should be taken with the Indian orders of equal representation, here bracketed.

XVi

INTRODUCTION.

Phanerogamic Flora of British India, &c. (continued).

Orders. Species. Genera. Position Position Position Percentage in the in the in the Position in the Total in Number Total Number _Flora of the | Australian Mexican Indian Flora. number. | Phanero- | endemic. | number. | endemic. World. Flora. Flora. gamla. 43 44 81 43, Meliacew ) ............ 84 0°62 71 19 3 42 66 96 44, Oleacese } beet e ee eees 84 0-62 67 10 1 45 97 134 45. Primulacew ............ 81 0°59 57 9 1 27 46 90 46. Saxifragacee .......... 80 0:59 56 14 24 12 80 47. Rutacee )........ 77 0-56 52 23 3 41 68 49 A8, Commnelinacer | voce eee 77 0-56 44 7 1 29 55 40 49. Loranthaces ..,....... 74 0-54 58 5 31 63 29 50. Cucurbitaceee .......... 71 0-52 33 29 7 23 20 29 51. Sapindacee ] .......... 70 0°51 41 23 3 48 125 101 52. Styraces } beeen rene 70 0-51 60 2 45 82 121 53. Ebenacew............6- 68 0-50 53 2 53 81 121 54, Olacines ...........26. 66 0-48 51 23 3 36 * 38 55. Begoniacew ............ 64 0-47 61 1 18 46 Al 56. Campanulacer.......... 64 0°47 49 13 3 18 108 46 57. Ericaceee ...........-4. 62 0°45 52 9 1 47 125 81 58. Guttifere.............. 61 0°45 51 6 3 18 89 12 59. Piperacesa ............ 56 0-41 40 3 37 65 7 60. Araliacew.............. 55 0-40 A2 18 3 41 60 59 61. Capparidee )...... 53 0-39 33 8 44 135 65 62. Ternstroemiacese } wees 53 0-39 36 14 1 38 68 96 63. Sapotacee ............ 52 0-38 44 8 33 25 58 64, Rhamnacee............ 51 0-37 30 - 12 1 36 39 65 | 65. Loganiacee ............ 50 0°37 33 8 51 125 69 66. Caprifoliacee ) ........ 49 0°36 36 8 2 40 135+ 73 67. Vacciniacese Leen eee 49 0-36 44 4 2 28 17 78 68. Chenopodiacee |} ........ 49 0°36 5 20 1 30 21 39 69. Amarantaces .......... 48 0°35 19 17 2 45 73 35 70. Lythracee ............ 45 0°33 22 11 46 54 88 71. Combretacew ) ........ 44 0°32 26 8 1 15 27 10 72. Solanacese } beens 44 0°32 14 10 34 103 42 73. Crassulacew............ 40 0°30 26 8 1 56 125 71 74. Burseracee ............ 39 0:29 35 10 8 57 125 112 75. Connaracee............ 35 0:26 25 7 1 * 52 22 94 76. Dilleniacee =) ...... 34 a 26 6 2 63 78 115 77. Meizer | beeeae 34 18 19 4 34 49 44 78. Polygalacee | ...... 34 eee 16 5 63 Le 153 79. Fumariacee............ 31 eae 24 4 65 135 143 80. Myristicacee } ........ 30 wean 20 1 51 79 94 81. Juncacesz } eee eee 30 wees 18 2 51 - 107 82. Salicineo.............. 29 Lee 25 2 68 112 128 83. Magnoliacee)} ..... .. 27 Lees 23 8 32 97 51 84. Bignoniacese } Levene 27 eee 16 11 1 50 185 88 85. Hypericinee |} ........ 26 eae 17 3 55 112 91 86. Samydacee +} ........ 26 18 3 39 71 128 87. Eriocaulee | ........ 26 Lee 20 1 54 96 98 88. Bixinee ) ..... cece ae 25 Lee 16 9 41 52 65 89. Conifere { ............ 25 15 13

* Leaves of what may prove to be a Begonia, have been collected in North-western Australia. t The somewhat anomalous Wittsteina is referred to the Ericacem by Mueller.

INTRODUCTION. Xvi Phanerogamic Flora of British India, &c. (continued). Orders. Species. Genera. Position Position Position in the in the in the Position in the Total | Number | Total | Number Flora of the | Australian Mexican Indian Flora. number. | endemic. | number. | endemic. World. Flora. Flora, 46 85 62 90. Violariee $j} ...... 24 13 3 535) 125 121 91. Ilicinee }$ ...... 24 21 1 55 135 115 92. Orobanchacee J} ...... 24 16 5 52 57 91 93. Lentibulariee) ...... 23 14 2 24 26 14 94, Amaryllidew }$ ...... 23 15 5 54 112 81 95. Dioscoreacew J ...... 23 16 2 1 58 112 86 96. Linacee ............ 22 14 7 2 75 .. 134 97. Sabiacew | ...... 21 14 2 1 71 97 150 98. Rhizophoree | ...... 21 8 10 2 41 108 28 99. Onagrariee [( ...... 21 3 5 60 185 105 100. Podostemacee } ...... 21 21 3 2 60 48 103 101. Naiadew.............. 20 6 8 7 135 107 102. Cornacese .........-4. 19 13 7 1 35 31 140 103. Thymeleacee ........ 18 8 11 63 98 104. Berberidee |} ........ 17 14 6 3 41 82 105. Valerianew } ........ 17 12 4 60 a .. 106. Dipsacee J ........ 17 15 4 1 61 97 78 107. Simarubee =) ...... 16 8 9 32 52 128 108. Ficoides | oe... 16 1 7 51 108 73 109. Aristolochiacee >...... 16 13 4 48 42 153 110. Santalaceze [ os... 16 12. 8 1 60 33* 153 111. Hemodoraceer J ...... 16 14 4 26 125 26 | 112. Malpighiaces.......... 13 11 3 69 1385 115 113. Papaveraceew | ........ 12 8 4 1 65 34 115 114. Halorages } eee eees 12 4 5 57 185 128 115. Ochnacee =—s) ........ 11 7 4 72 94 153 116. Hydrocharidee [| ...... 11 3 10 1 70 125 153 117. Burmanniacee {...... 1] 9 2 23 57 62 118. Iridee f...... 11 8 2 45 103 48 119. Passifloree | ........ 10 8 3 51 112 112 120. Plantagineew | .......- 10 i 1 49 108 49 121. Nyctaginee [........ 10 3 3 19 3 134 122. Proteacen | ........ 10 8 1 64 43 - 123. Pittosporee | ......-. 9 7 1 dl 112 140 124. Plumbagines | .......- 9 3 6 76 124 .. 125. Nepenthacee (.......- 9 3 1 69 103 103 126. Alismaceee |] ........ 9 1 6 1 74 112 115 127. Nympheacee )\...... 8 1 5 1 72 .. 143 128. Tamariscinee |.....- 8 4 2 63 66 105 129. Zygophyllacee >.....- 8 .. 4 77 135 143 130. Hamamelidee |....-- 8 5 8 2 65 93 131. Pandanacee )....-- 8 5 2 71 95 .. 182. Xyridee ]........-- 7 4 1 82 103 112 133. Lemnaceze } bene ee eee 7 1 2 59 49 98 134. Portulaces \ oe 6 3 2 82 112 153 135. Elatines lees 6 1 2 73 7 153 | 136. Chailletiacom \ vee 6 5 1 33 135 . 137. Eleagnaces (.eceee 6 2 2 74 135 143 188. Balanophoree | .....- 6 5 2 72 143 139. Gnetaceze Jovanes 6 1 2

* Sometimes united with the Amaryllidex, as in Bentham’s ‘ Flora Australiensis.’

BIOL. CENTR.-AMER., Bot. Vol. I., Octoder 1888.

Cc

INTRODUCTION.

XViil Phanerogamic Flora of British India, &c. (continued). Orders. Species. Genera. Position Position Position in the in the in the Position in the Total Number Total Number Flora of the | Australian Mexican Indian Flora. number. | endemic. | number. | endemic, World. Flora. Flora. 86 140. Salvadoracese ) ...... 5 .- 3 57 .. 141. Selaginew | ...... 5 3 1 85 135 .. 142. Taccacee [...... 5 2 1 83 125 143 143, Typhacee = j...... 5 . 2 77 we 143 144, Resedacee )........ 4 .. 3 62 41 e 145. Droseraceee |........ 4 2 2 72 121 128 146. Pedalinee |........ 4. 2 2 77 . 128 147. Juglandee >........ 4 2 2 74 oe 150 148. Myricacee |........ 4 2 1 87 135 .. 149. Flagellaries | ........ 4 2 2 83 .. 150. Triuridee J) ........ 4 4 1 63 23 .. 151. Stylideze Nee eee 3 2 1. 84 . 143 152. Monotropese | Leeeee 3 1 3 1 73 a 143 158. Chloranthacee \...... 3 2 2 55 82 107 |154, Monimiaces ...... 3 1 3 1 67 83 86 155. Cycadaceze J .eseee 3 2 1 87 135 we 156. Roxburghiacew ) ...... 3 2 2 1 92 a . 157. Moringes Vee eee 2 1 1 91 .. 153 158. Datiscacese J wae ee 2 2 51 10 153 159. Goodenovieew > ...... 2 1 64 *112 121 160. Illecebracese | ...... 2 2 74 185 121 161. Pontederiacee J Leaeee 2 1 84 97 153 162. Frankeniacee ) ...... 1 .. 1 90 Le 153 163. Coriariee = |...... 1 a 1 18 .- 6 164, Cactaceee $$|...... 1 . 1 40 8 + 165. Epacridee §$|...... 1 .. 1 87 .. .. 166. Diapensiacew |...... 1 1 1 55 .. 71 167. Polemoniacee | ...... 1 ve 1 55 125 70 168. Hydrophyllacee >...... 1 1 69 87 107 169. Phytolaccacee | ...... 1 .. 1 81 - 150 170. Cytinacee |...... 1 1 1 1 89 .- 134 171. Platanacee | ......]. 1 1 . 80 62 es 172. Casuarinee |...... 1 1 93 135 153 173. Ceratopbyllee |...... 1 1 91 121 .. 174. Philydracen J). 1 1

Summary of the Indian element in the foregoing Table.

Polypetale........ Gamopetale Incomplete ......

eevee ane

Dicotyledones Gymnospermeze

Monocotyledones

Grand totals

eeneese

Genera. A

‘a oN

Orders. Total. Endemic. 72 844 113 41 714 108 30 251 27 143 1809 248 3 16 see 28 446 56 174 2971 304

‘Species. Cc ~ oN Total. § Endemic. 4489 2965 4233 2982 1693 1159 10415 7106 34 18 3198 2246 13647 9370

* Associated with the Caryophyllaces in Bentham’s ‘ Flora Australiensis.’

INTRODUCTION. xix

Relative numbers of Orders, Genera, and Species in India, Mexico, North America, and Australia.

Orders. Genera. Species. India ........ 02.2008. 174* 2271 13647 Mexico .............. 162* 1794 11626 North America ........ 158* . 1513 9403 Australia ............ 154 1335 8575

Percentages of Dicotyledones (including Gymnospermex) and Monocotyledones in the four areas and in Europe.

Dicotyledones. Monocotyledones. India ......... ee eee. 76°57 23°43 =100-00 Mexico .............. 78°50 21:50 =100-00 North AmericaT ...... 80°62 19°38 =100-00 Australia ............ 81:50 18°50 =100-00 Europe $.........0.06- 82°70 17°30 =100-00

The mean proportions of the five countries are 79:97 dicotyledons and 20-03 per cent. monocotyledons, against 81°29 and 18°71 for the whole world, showing that the nume- rical proportions do not greatly vary for large areas, no matter how distant they are, nor how dissimilar is their vegetation as a whole; yet it is hardly necessary to add that very different proportions exist in smaller areas. Generally speaking, the drier the region the smaller the proportion of monocotyledons, and the greater the proportion of them bulbous plants. Maximowicz § in an analysis of the vegetation of different parts of Central and Eastern Asia gives the percentage of monocotyledons in the Phanerogamia as 14:1 in Tangout, or Northern Tibet, and 26-1 in Japan. But, as already hinted, figures of this kind convey no very definite information, inasmuch as they embody no idea of individual development, as a species of palm, bamboo, or banana of the tropics counts no more than a snowdrop, daffodil, or small grass of the temperate regions. Alphonse De Candolle ||, who presents very numerous statistics of the proportions of monocotyledons and dicotyledons in small areas, strongly insists on this point. It should be borne in mind that all statistical analyses of floras are to a great extent illusory, and can only be properly appreciated after a careful consideration of the composition of the elements. This fact is illustrated in the following somewhat detailed examination of the preceding table. —

One thing brought into great prominence by this table is the large number of

* Including the Fumariaces, treated as a suborder of the Papaveracee in the ‘Genera Plantarum.’

+ Calculated from the second edition of Oyster’s ‘ Catalogue of North American Plants,’ after deducting the introduced species.

+ Calculated from the numbers given in Nyman’s ‘Conspectus Flore Europes,’ p. 848, excluding the “* subspecies.”

§ Bulletin du Congrés International de Botanique et @Hortcultiure 4 St. Petersbourg, 1884, p. 158.

| Géographie Botanique Raisonnée, ii. p. 1166.

c2

&

xx INTRODUCTION.

natural orders represented in each of the areas under consideration, especially in India, where there are 86 per cent. of the orders retained by Bentham and Hooker. This is not put forward as something new, though it will be new to most people, but it comes out more strongly than might have been anticipated. It should be mentioned, too, that British India covers only a portion of one of the primary botanical regions, though, on the other hand, the upper belt of vegetation of the Himalayas belongs to the northern region. Let us pursue the ordinal distribution a little further, premising that the reader will remember that important particulars not given here may be found in the Appendix.

1. Natural Orders not known to be represented in British India *.

Calycanthacee. Bruniacez. Leitneriez. Sarraceniacez. Loasaceze. Lacistemacez. Cistacee. Turneracez ft. Empetracee. . Canellaceze. Calycereze. Bromeliacez. Tremandrez. Lennoacez. Mayacez. Vochysiacez. Columelliacez. Rapateacez. Chlenacez. Myoporinee. Cyclanthacee. Humiriaceee. Batidese. Centrolepidez. Cyrillacee. Penzeacee. Restiacee. Stackhousiacee. Balanopsez. .

2. Natural Orders not known to be represented in any part of America t.

Pittosporeze. Moringesze §. Balanopsez. Tremandree. Dipsaceze. Casuarinez §. Dipterocarpez. Salvadoracez. Philydracez. Chleenaceze. Nepenthacez. Flagellariez. Stackhousiacez. Penzacee. Pandanacez §.

3. Natural Orders not known to be represented in Australia.

Calycanthacez. Canellacez. Chailletiacez. Berberidacez. Vochysiacee. Cyrillaces. | Sarraceniacez. Tamariscines. Sabiaceze. Fumariaces. Dipterocarpeze. Coriariez. Resedacez. Chlenacez. Moringee. Cistacez. Humiriacez. Bruniacez.

* Those orders printed in italics are represented in some part of Asia. Forty-one natural orders are not known to be represented in Mexico: see the table in vol. iv. pp. 171-200. A Drosera, received at Kew from British Honduras, as this was going to press, reduces this number to forty.

Tt Turnera ulmifolia is extensively colonized in India.

{ In the table referred to the Selaginew are indicated as not American; but one or two Asiatic species of Gymnandra (Lagotis) recur in the extreme north-west of America.

§ These orders, and perhaps some others, are represented by colonized species.

INTRODUCTION. Xxl Loasacee. Salvadoraceze. Myricacez. Turneracee. -Polemoniacez. Salicineze. Begoniacee. Columelliacez. Lacistemacez. Datiscacez. Selaginez *. Empetracez. Cactacee. Batideze. Gnetaceze. Valerianacez. Cytinez. Bromeliacez. Dipsaceze. Chloranthacez. Mayacez. Calycerez. Penzacee. Rapateacez. Monotropee. Platanaceze. Cyclanthacez. Diapensiacee. Leitneriez. Triuridez. Lennoacee. Juglandaceze.

4. Natural Orders represented in Mexico or Australia, but not known in the Indian Flora.

In Mexico. In AUSTRALIA. Cistacez. Tremandrez. Vochysiacez. Stackhousiacez. Loasacez. Myoporinez. Turneracez. Balanopsez. Lennoacez. Centrolepidee. Lacistemacez. Restiacez. Bromeliacez.

Cyclanthacee.

5. Natural Orders not known to occur in India, Mexico, or Australia.

Calycanthacez. Cyrillacez. Penzeacez. Sarraceniacez. Bruniacez. Leitneriez. Canellacez. Calycerez. Empetracez. Chlenacez. Columelliacecze. Mayacez. Humiriacee. Batidez. Rapateaceze.

6. Indian Natural Orders not known to occur either in Mexico or Australia.

Dipterocarpee. Selagineze. Triuridez. Dipsacez. Moringez. Salvadoracez. Diapensiacez.

The phenomena of ordinal distribution as set forth in the six preceding tables demand a few words of explanation. Respecting the first table, it is remarkable how little it contains ; how few important orders, considered either as to their extent or

structural peculiarities, in relation to the land-area of the rest of the world. The

* Specimens of the South-African Dischisma capitatum, Choisy, were collected by Drummond in West Australia, but Bentham regarded the species as ‘ most likely introduced,’ and Mueller treats it as an

undoubted alien.

Xxii INTRODUCTION.

succeeding tables afford further information on the distribution of most of these orders. Those printed in italics are represented in Asia; some of them by solitary outliers; others, the Calycanthaceze for example, as fully as elsewhere, and these may yet be found in the mountains of Northern India.

Not less remarkable is the small number of orders in the second table, especially in relation to the vegetation as a whole, of the two hemispheres*. With regard to the third, future explorations may probably reduce the total by about half a dozen orders, though not more. The fourth and fifth tables deal in different ways with orders included in the first; the fifth bringing also into great prominence the fact that there is exceedingly little ordinal peculiarity in the rest of the world, including the vast African region. The sixth table contains only one important natural order, namely the Dipterocarpee, which are known to extend to New Guinea, and some may possibly exist in tropical Australia. Altogether, twenty-six of the Indian orders are not found in Australia, and nineteen of them are not found in Mexico.

Natural Orders not known to extend beyond America.

Sarraceniacez. Batidez. Canellacez. Leitneriez. Vochysiacez. Lacistemacee. Cyrillaceze. Bromeliacez. Calyceracez. Mayacez. Lennoacee. Rapateacecze. Columelliacez. Cyclanthacez.

The Galapagos islands are regarded as belonging to the American region. Besides the foregoing orders, which appear to be absolutely confined to America, there are the Humiriacez, which are American with one exception in W. Africa; the Turneraces are represented by a few species in the African region ; the monotypic genus Kissenia is the only member of the Loasacez found out of America; and, with the exception of the genus Rhipsalis, the Cactacee are wholly American. Hence it will be seen that between what is wanting and what is peculiar to America in natural orders, the balance is somewhat in favour of the latter. This completes the review of the leading features in the distribution of the natural orders.

Something might be added here on the relative visible and spacious position occupied by the dominating orders in the different Floras; but as it is proposed dealing briefly with this part of the subject further on, some illustrations of generic and specific distribution are given first.

GENERIC AND SPEcIFIC COMPOSITION OF THE FLORAS OF DIFFERENT AREAS.

At page xix are given the relative numbers of orders, genera, and species in the Floras of India, Mexico, North America, and Australia ; and although no special significance * See vol. iv. pp. 202-207,

INTRODUCTION. Xxill

is to be attached to the fact, it is remarkable and noteworthy, as a matter of figures, what a close approach there is to uniformity in the proportions in each of these four dissimilar and distant areas. Thus :—

Average number of Genera Average number of Species to an Order. to a Genus. India. . . . . . . . 18:0 6:0 Mexico. . .... . 110 «64 N. America .. . ... 96 6°2 Australia . . . .. . 87 6°4

The proportions for the whole world, calculated from the numbers of the ‘ Genera Plantarum,’ are 37:50 genera to an order, and 12°65 species to a genus; from which it appears that there is, approximately, half of the average number of species of a genus, and a third to less than a quarter of the average number of the genera of an order in each of these large areas. Taking a portion of the Cape Flora, the average number of species to a genus is 6°6*; therefore between six and seven to one is probably the highest, or nearly the highest, average in large areas. Turning to other areas, the proportion of species is found to be much lower, and in certain insular Floras the genera are nearly as numerous as the species. In China the Polypetalous orders f yield about three species to a genus; and the proportions are nearly the same in the whole vascular plants of New Zealand, as well as in the Sandwich Islands. In Japan, the proportions are as 2°6 to 1; and of the probably endemic plants of St. Helena they are less than 1°4 to 1. .

Such are the averages, which, as has been shown (vol. iv. pp. 212-217), are made up to a great extent by genera numerous in species and genera of one species each. In the Mexican Flora, for example, eighty-five genera contribute 4760 species, or 39 per cent. of the total, whilst other 660 genera are represented by only one species each, upwards of one third of them being absolutely monotypic. The composition of all large Floras, in which there is a high percentage of species to a genus, is similar ; whereas in the Chinese and Japanese Floras there are exceedingly few very large genera, and at — the same time a smaller proportion of monotypic genera. We have not counted the monotypic genera of the Indian Flora, but the proportions are probably very nearly the same as in the Mexican Flora. As mentioned elsewhere, Mueller { states that there are 550 genera in Australia represented by only one species. With regard to large genera in the Indian Flora there are only seven of 100 species and upwards each, against ten in Mexico and four § in Australia ; and there is nothing in either India or

* Harvey and Sonder’s ‘Flora Capensis,’ vol. i, Ranunculacee to Connaracem, as estimated by Mr. N. E. Brown, in manuscript, in the Kew library.

+ Forbes and Hemsley, “ Index Flore Sinensis,” Journ. Linn. Soc., Bot. vol. xxiii.

+ Lecture on the Flora of Australia, 1883, p. 11.

§ Six according to Mueller in the place cited, but he unites some genera retained by Bentham and Hooker.

XXi1V

INTRODUCTION.

Mexico approaching the 320 species of Acacia and 120 of Eucalyptus in Australia, and the 300 species of Mesembryanthemum and about 500 of Erica in South Africa, that is to say in point of numbers or as a feature in the whole vegetation, except the

genus Micus.

Dendrobium . Strobilanthes Eugenia Carex . Impatiens Ficus . Habenaria Panicum .. Quercus . Cyperus Crotalaria Vitis

Eria . . Astragalus Polygonum .

Bulbophyllum .

Litsea Begonia Symplocos Senecio Diospyros Glochidion Loranthus Hedyotis . Ipomeea Phyllanthus . Fimbristylis . Psychotria Lasianthus Euphorbia Justicia Calamus . Andropogon . Desmodium . Ceelogyne. Vernonia .

British Indian Genera of twenty-five Species and upwards.

. 158 . 146 . 181 . 131 . 124 . 112 . 100 92 82 80 77 75 73 70 70 70 . 65 . 64 . 64 63 59 59 58 57 57 56 53 52 52 52 50 50 50 AQ AY AB

Ardisia Piper . Mallotus . Sonerila .

Rhododendron .

Primula . Jasminum Rubus. Elzocarpus . Indigofera Memecylon .

Didymocarpus .

Potentilla Saussurea Hoya . Oberonia . Leucas Bauhinia . Ixora . Gentiana . Pedicularis Liparis Smilax Grewia Blumea Ceropegia Calanthe . Saxifraga . Amomum

“Premna

Hibiscus . Nepeta Pollinia Plectranthus. Antidesma Sterculia .

45 45 45 43 43 43 43 41 40 40 40 40 39 39 39 39 38 37 37 37 37 37 37 36 36 36 36 35 35 34: 33 33 33 32 32 31

Ophiorrhiza . Capparis . Garcinia . Swertia Myristica . Saccolabium . Hedychium . Scirpus Baccaurea Ischemum. . Ebermaiera . Aporosa Salix Aneilema . Euonymus Anaphalis Artemisia. Solanum . Cleistanthus . Croton Macaranga

Cirrhopetalum .

Osbeckia . Barleria Cinnamomum Juncus Allium. Eriocaulon Polyalthia Argyrela . Pogostemon .

_Globba

Ariseema . Arundinella . Eragrostis Bambusa .

Totals: genera 108; species 5041.

31 30 30 30 30 30 30 30 29 29 28 28 28 28 27 27 27° 27 27 27 27 27 26 26 26 26 25 26 25 25 25 25 25 25 25 25

From these figures it appears that about 4°8 per cent. of the total number of genera

INTRODUCTION. XXV

of flowering plants in the Flora of India yield nearly 37 per cent. of the total number of species. In Mexico 4:6 per cent. of the genera comprise 39 per cent. of the species; and in Australia the amounts are about 4°64 and 37 per cent. Itis unnecessary to add that most of the above genera are widely spread, and many of them common to the New World.

Genera common to the widely separated areas of India and Mesico.

From the relations of the average number of species to a genus in the areas under consideration to the average for the whole world, it follows that the average area of a genus must be at least double one of these areas. But there is no necessary relation between the area a genus covers and the number of species it contains, though, speaking generally, monotypes have a restricted area. Indeed, if we exclude aquatic and sea- coast plants, and such as have probably been dispersed through human agency, it is difficult to multiply instances of monotypes with a large area. The curious Cressa cretica may be cited as a possible exception *. On the other hand, the species of some large genera are concentrated in one region, as Eucalyptus in Australia, and Miconia in America. |

Disregarding exceptions, it may be stated that the genera and species of the northern Floras have the widest range; those of the tropics an intermediate one, and those characteristic of the southern Floras, excluding that of the coldest zone, the most restricted range. Probably not less than 75 per cent. of the genera of the Flora of Eastern America, north of Mexico, are represented in the Old World, for in some statistics on the vegetation of the north-eastern part of the United States, drawn up by the late Dr. A. Gray more than thirty years ago}, it is shown that 63 per cent. of the then known genera were common to America and Europe, or America and eastern temperate Asia. Since that date many others have proved to be common to America and the Old World, and the rich collections made by Dr. A. Henry within the last three years in Hupeh, one of the central provinces of China, have added several conspicuous genera to the number.

A careful comparison of the generic composition of the Mexican and Indian Floras reveals the fact that 581, or 25°58 per cent., of the Indian genera are likewise repre- sented in Mexico. Our own table (vol. iv. pp. 207, 208) shows that more than a third of the Mexican genera are widely dispersed, that is to say, they occur as well in two or more of the large divisions of the Old World; and only 11 per cent. are endemic. Engler} finds that only about an eighth of the tropical dicotyledonous genera inhabit both America and some part or parts of the Old World. On the other hand, 30-5 per cent. of the Australian genera and 35:5 per cent. of the South-African are endemic.

* Grisebach (Symb. ad Fl. Argent. p. 266) refers a second species to this genus with an extended diagnosis. + Silliman’s ‘Journal of Science and Art,’ 2nd series, xxii. (1856) pp. 204-231. + Versuch einer Entwicklungsgeschichte der Florengebiete, ii. p. 174.

BIOL. CENTR.-AMER., Bot. Vol. I., October 1888. d

XXvi . INTRODUCTION.

Comparing the number of endemic species of the Indian Flora with that of other parts of the world, it is somewhat surprising to discover that it is less than 2 per cent. below the Mexican, and not far below that of the Australian; being no less than 68°67 per cent. But such statistics embrace only one class of facts, the value of which will be briefly discussed in connection with the delimitation of the primary botanical regions.

ON THE DISTRIBUTION OF SOME OF THE LARGEST NATURAL ORDERS.

Attention has already been directed to the illusory nature of mere statistical com- parisons of the vegetation of different regions. In many Floras the orders richest in species are not those which give character to the scenery, or constitute the bulk of the vegetation ; and the genera are sometimes less numerous and less diversified than those of many concomitant orders less numerous in species. This is pre-eminently the case with the orchids in India. Many of the species are exceedingly rare and local ; many are small plants with inconspicuous flowers; and the comparatively small number of common species having large conspicuous flowers, lend colour to the vegetation rather than give character to it, and this only in certain districts. Here and there in the mountains, as we learn from travellers and residents, the rocks are clothed with orchids, and when they are in full flower form a conspicuous feature in the land- scape. In South Mexico and Central America orchids, though third in order of predominance, pervade the whole country to a greater extent; yet here, as elsewhere, the existence of the preponderating epiphytic species depends largely on the arboreous vegetation. In Northern and Eastern India and Malaya, as in Mexico, orchids are specially abundant in the intermediate or oak region, and the numerous species of oak constitute one of the principal features in the vegetation. Lower down in the more tropical parts of India arboreous Leguminose, Euphorbiacee, figs, laurels, Dipterocarpee, Anonacee, together with Rubiacee, Acanthacee, Composite, Scita- mine, Aroidee, &c., constitute the bulk of the vegetation—some of these orders, their genera, or even species predominating in one district, some in another; but each or any of them entering more fully into the composition of the vegetation than orchids. Palms form a conspicuous feature on the coast and plains of India, where the more or less cultivated coco-nut, the palmyra (Borassus), and Phenix sylvestris abound and cover large tracts; but there is little variety in the prominent species; and the order occupies the thirtieth position only by reason of the large number of species of rattan (Calamus) that inhabit the dense forests, and render them almost impenetrable. Nowhere is there anything resembling the highly diversified palm-groves of the Amazon region.

In Mexico the Composite outnumber the order next in sequence by upwards of 100 genera and 500 species, and although few of them exceed the stature of shrubs, they grow in almost every variety of situation; are often excessively numerous individually ;

INTRODUCTION. XXVil

constitute a prominent part of the vegetation, and specially characterize the dry districts. Such highly characteristic Mexican orders as the Cactacee, Labiate (Salvia), Amaryl- lidaceee (Agave and Furcrwa), Aroidee (Anthurium, Philodendron), and Liliacez (Yucca and Dasylirion) are as numerous in species as they are prominent in the scenery

Turning to the Australian flora we find almost complete agreement between the numerical strength of the natural orders in species, and their degrees of domination in the composition of the vegetation. The Leguminose stand at the head with upwards of a thousand species, including the highly characteristic phyllodineous Acacias, numbering three hundred species, and prominently pervading the whole country. Next come the Myrtacee, to which belong Eucalyptus (120 species), Melaleuca (100 species), Verticordia, Calycothriz,and Darwinia, with thirty-five species each; followed by the Proteacee (Grevillea, Hakea, Banksia), Composite (Olearia, Helichrysum), Cyperacee, Graminez, Orchidew, Epacridee, Euphorbiacez, Goodeniacee, and Rutacez. .

Similar conditions obtain in the Cape Flora, and a person possessing a fair knowledge of plants can, from statistics alone, form some conception of the nature and composition of the vegetation.

Before leaving this part of the subject it may not be amiss to mention the fact that the vegetation of different districts of a country may present more striking diversities and much more abrupt transitions than does the whole flora of one country as compared with that of another. Within a few yards the whole character of the vegetation often changes, owing to differences in the substratum of the soil and other causes. Mr. Salvin was particularly impressed by this fact in Guatemala, where, in travelling, you often literally step from an oak-forest into a pine-forest, with the carpet and canopy (epiphytes, &c.) of vegetation equally as different in character as are the pines and oaks themselves.

Tre Primary BoranicaL REGIONS OF THE WORLD CONSIDERED IN THEIR RELATIONS TO THE: ZOOLOGICAL REGIONS.

Very various are the divisions and subdivisions of the world proposed by different botanists and zoologists who have written on the geographical distribution of plants and animals; but the comparisons instituted here will be with the zoological regions originally defined by Dr. Sclater *, and subsequently adopted, with slight modifications, by other eminent zoologists, notably by Mr. Wallace in his very elaborate treatise on the present distribution of animals, more especially of the mammals. For convenience, his table of regions t+ and subregions is reproduced here, as it is more intelligible than

* Journ. Linn. Soc., Zool. vol. ii. + The Geographical Distribution of Animals, 1876, i. p. 81. d 2

XXVlil

INTRODUCTION.

a brief description of them would be; and where it is necessary the exact boundaries will be indicated in any comparisons made.

Regions.

IT, Pan#arcric .

II. Ersior1an

TIT. Orrentar

IV. AvstRaian .

V. NrEorroricau

VI. Nearcric

Wallace’s Table of Zoological Regions and Subregions.

m © Ww —

Poo PO ee we 09 2 es

be oO wo eH

~ OF Oo

mo we

Subregions.

. North Europe.

. Mediterranean (or S. Europe). . Siberia.

. Manchuria (or Japan).

. East Africa.

West Africa.

. South Africa. . Madagascar.

. Hindostan (or Central India).

Ceylon.

. Indo-China (or Himalayas). . Indo-Malaya.

. Austro-Malaya. . Australia. . Polynesia.

New Zealand.

Chili (or S. Temp. America).

. Brazil. . Mexico (or Trop. N. America). . Antilles.

. California.

- Rocky Mountains.

. Alleghanies (or East U.S.). . Canada.

Remarks. Transition to Ethiopian.

Transition to Nearctic. Transition to Oriental.

Transition to Palearctic.

Transition to Ethiopian.

Transition to Palearctic. Transition to Australian.

Transition to Oriental.

Transition to Neotropical.

Transition to Australian.

Transition to Nearctic.

Transition to Neotropical.

Transition to Palearctic.

Before explaining my own ideas on the primary phyto-geographical regions of the world, it may be of interest to give in outline the two latest attempts to define them, especially as one of these has been published in the form of a very elaborate atlas for

educational purposes.

In 1882 Dr. Engler * grouped the botanical regions of the earth,

on the assumption that there existed in the Tertiary period four fundamental elements of the present vegetation, namely—the “ Arctic Tertiary ” element, the “ Paleotropical ” element, the “Neotropical” element, and the ‘Old Oceanic” element, which he

* Versuch einer Entwicklungsgeschichte der extratropischen Florengebiete der siidlichen Hemisphiire und der tropischen Gebiete, pp. 326-347.

INTRODUCTION. XX1X

briefly defines, and then proceeds to divide the vegetation of the earth into “ kingdoms, regions, provinces, zones, and districts.” To give the whole of his subdivisions would occupy more space than can be afforded, and carry us beyond what is necessary in this discussion. His primary divisions are four, corresponding to his four Tertiary elements. These are :—the northern extratropical floral kingdom, the paleotropical floral kingdom, the South-American floral kingdom, and the old oceanic floral kingdom. In conformity with English usage we will call his primary divisions regions and the secondary ones subregions ; and where we have occasion to refer to the smaller divisions, the names given above will be employed. Engler’s secondary divisions are as follows :—

1. NoRTHERN EXTRATROPICAL REGION.

SuBREGIONS: 1. Arctic; 2. Subarctic or Conifer; 3. Central Asiatic; 4. Mediter- ranean; 5. Mandshurian and Japanese; 6. North-American Pacific; 7. North-Ame- rican Atlantic.

2. PaLZoTRoPIC REGION.

SuBREGIONS: 1. West-African forest ; 2. African and Arabian desert; 3. Malagassy ; 4. Western Indian; 5. Tropical Himalayan; 6. East-Asiatic; 7. Malayan; 8. Arau- caria; 9. Polynesian ; 10. Sandwich Island.

3. SouTtH-AMERICAN REGION.

Susreeions: 1. Mexican Highlands; 2. Tropical American; 3. Andine; 4. Gala- pagos ; 5. Juan Fernandez.

4, Otp Ocranic REGION.

Suprecions: 1. Antarctic forest of South America ; 2. New-Zealand; 3. Australian; | 4. Kerguelen; 5. Amsterdam Island; 6. Cape; 7. Tristan d’Acunha; 8. St. Helena.

The ternary subdivisions or provinces of the Northern Extratropical Region of the above scheme are about forty; many of them are divided into several zones, and some of them again into numerous districts. Subdivision is, for obvious reasons, not carried so far in the three other regions.

I have already put on record * my objections to that part of Dr. Engler’s scheme relating to oceanic islands and the “antarctic” Flora; but I have now to deal with it as a whole. Theoretically there may be much to justify his regions; yet it seems to me that any attempt at division based partly upon assumed, or even proved, anterior conditions, and partly on present conditions, is unsatisfactory and confusing. On this point Wallace says + :—‘ Our object is to represent as nearly as possible the

* Botany of the ‘ Challenger’ Expedition, i. Introduction, p, 50. ft The Geographical Distribution of Animals, i. p. 55.

XXX INTRODUCTION.

main features of the distribution of existing animals, not those of any or all past geological epochs. Should we ever obtain sufficient information as to the geography and biology of the earth in past epochs, we might indeed determine approximately what were the Pliocene, Miocene, or Eocene zoological regions; but any attempt to exhibit all these in combination with those of our own period must lead to confusion.” This objection applies with equal force to any botanical division; and as a critical exami- nation of Engler’s scheme would involve the discussion of questions beyond the scope of the present inquiry, it is sufficient for the purpose to have brought it under notice, though it should be added that it is full of valuable matter, and has been largely utilized in the present work. |

The other scheme referred to above is by Dr. Drude. It was originally published in 1884 *, and again in 1887 as an independent work}. As explained in the title given below, this is a representation of the present conditions of the distribution of plants; hence it has a greater demand on our attention. Drude divides the world into fourteen “ floral kingdoms,” namely—(1) Northern, (2) Central Asian, (3) Mediterranean, (4) East Asian, (5) Middle North American, (6) Tropical African, (7) East African Islands, (8) Indian, (9) Tropical American, (10) Cape, (11) Australian, (12) New Zealand, (13) Andine, (14) Antarctic. Most of these regions are subdivided, and the overlapping of the elements of different regions is indicated by lines and dots.

As Drude himself remarks, we are all striving and devising with the same aim in view, and we arrive by different ways to much the same conclusions. He claims that he obtains practically the same results as Engler, but by different methods; that Engler’s scheme is, after all, based essentially on present conditions; and says that there is far more difference in the arguments of the writers on phytogeography than there is in their deductions and cartographical illustrations. This is doubtless true to a certain extent, because there are certain facts which no student can overlook or disregard ; yet it is none the less true that one begins with four and the other with fourteen regions, and therefore there must be a wide difference in their value and extent.

Engler’s scheme, based upon a small number of primary regions, commends itself because these are much more nearly of equal importance than are Drude’s; but his old oceanic region is altogether inadmissible from the standpoint here taken, involving, as It does, the relegation of the Northern Island of New Zealand to one of his primary divisions and the Southern to another. |

A small number of primary divisions undoubtedly offers the least difficult basis for further division. Equivalent regions and subregions it is impossible to define, because -

* “Die Florenreiche der Erde. Darstellung der gegenwiirtigen Verbreitungsverhiltnisse der Pflanzen, mit 3 Karten.” LErginzungsheft no. 74 zu Petermann’s Mittheilungen. + Atlas der Pflanzenverbreitung.

INTRODUCTION. XXxi

they no more exist in nature than do absolute differential characters between many natural orders or genera of plants; but Drude has too many and unnecessarily unequal regions. Several of them correspond, or very nearly so, to the secondary divisions of other writers, and are not unnatural in this sense; yet we do not agree with the author in raising them to primary rank. On the other hand, his Indian region embraces tropical Asia, the whole of the Pacific Islands, from New Caledonia to the Sandwich Islands, and a large area of North Australia. It is unnecessary to dwell upon the great disparity of this region as compared with his East-African Islands region, or his New Zealand region of the same rank.

With regard to the Flora of the Sandwich Islands, it is so highly specialized, and its affinities so complex, that it cannot be included in any primary region without the question arising whether it might not with equal propriety have been included in another, as will hereafter be shown.

The Flora of North Australia undoubtedly contains a large tropical element consisting of species, many endemic, of Asiatic genera, or genera of wider range; but the elimi- nation of such species as are common sea-shore plants throughout the eastern tropical region would considerably reduce this element. It is equally true that some of the orders and tribes specially characteristic of the Australian Flora are almost entirely wanting, such as the Epacridee, Rhamnaceew, Myoporinee, Boroniee, the Podalyries, and some others; but are these two conditions sufficiently developed to justify separation in a primary division and annexation to the eastern tropical region? Drude appears to have separated it because it is tropical. Wherever the boundaries are drawn there will be overlapping of different elements to some extent, and a more natural boundary in this region is further north. Even if in the north-eastern coast district the com-— position of the vegetation is more Asiatic in character, it is not so in the north-west. Whatever the amount of infusion of Asiatic types may be in North Australia, and whatever groups are wanting or rare, the highly characteristic Australian gum-trees (Eucalyptus) and the phyllodineous Acacias are represented respectively by twenty-five and sixty-seven species ; Proteacez by about thirty-five species; Stylidiew and Goode- niaceee combined by upwards of fifty species; Amarantaceee by nearly sixty species; capsular Myrtacee, exclusive of Eucalyptus, by about thirty species; and many characteristic Australian genera, such as Dodonea and Stackhousia, are also present, though numerically few.

There seems even less reason for including New Caledonia in the Indian region, for although the Rubiacez and sarcocarpous Myrtacez appear to be the dominating groups, yet the vegetation generally is more Australian than tropical Asiatic in character. As Baron Mueller observes *, New Caledonia is the only country outside of Australia where capsular Myrtacee are largely developed, though they include

* A Lecture on the Flora of Australia, 1882, p. 16.

xxxii INTRODUCTION.

no species of Eucalyptus. Only fragments of the New Caledonian flora have been published, but from a rough manuscript list of New Caledonian plants contained in the Paris Herbarium, compiled by Sir Joseph Hooker twenty-five years ago, the Australasian character of the vegetation is evident. Examples are offered by such genera as Cordyline *, Dianella, Araucaria, Frenela, Dacrydium, Hedycarya, Casuarina (five species), Exocarpus, Grevillea, Cenarrhenes, Knightia, Stenocarpus, Myoporum (four species), Leucopogon (twelve species), Dracophyllum (ten species), Hibbertia (fifteen species), Pittosporum (twenty-five species), and Boronia (eighteen species). It is noteworthy that the phanerogamic flora of New Caledonia, unlike that of New Zealand, is exceedingly rich in species, yet, as in New Zealand, many characteristic Australian groups of plants are wholly wanting.

The flora of New Zealand has been so exhaustively discussed by Hooker, Wallace, Engler, and others, that it is inexplicable why Drude should have raised it to the rank of a primary region. It is remarkable for its poverty and the total absence of many of the most characteristic Australian types; yet, apart from the extraordinary development of such widely spread genera as Ranunculus, Epilobium, and Veronica, the flora is so essentially Australian that it is difficult to understand why there should be any hesi- tation in treating it as a subregion of the Australasian flora; especially by an author who would include the Sandwich Islands in the Indian region.

As far as the Flora of Madagascar and adjacent islands is concerned, the reasons for regarding it as a primary region are more intelligible; and some zoologists have pro- posed the same thing ; but recent explorations prove that it should rank asa subregion of Africa. Although rich in endemic genera and species, it is not relatively more so probably than the Malay Archipelago or Ceylon in Asia, It is true that the small order Chlenaces is apparently peculiar to the island, for there is little doubt that the two members of this order recorded from Mozambique were collected by Forbes on the Madagascar side of the channel of that name. Briefly, the forest flora of Mada- gascar.is. closely related to that of tropical Africa, while South-African forms reappear in the hill flora; and it exhibits still closer affinities with Mauritius, Bourbon, and the

neighbouring islands.

Sufficient has been advanced perhaps to show that Drude’s divisions are not always the best that could be devised, and that his primary divisions are too numerous, OY, from another standpoint, not numerous enough to attain the nearest approach to equality.

Ten years ago Mr, Thiselton Dyer { drew up a concise and pregnant sketch of the

* It must not be assumed that the generic identifications hastily made are in all instances absolutely correct.

t+ See Baker in ‘ Journal of Botany,’ 1881.

t “A Lecture on Plant Distribution as a field for Geographical Researches,” Proceedings of the Royal Geographical Society, xxii. 1878.

INTRODUCTION. XXxlil

geographical distribution of plants, in which he deals with both present and past con- ditions, more especially in relation to the theory of a general southward migration. He groups the Floras of the world into northern, tropical, and southern, and enters into particulars of the characters, relationships, connections, and dissimilarities of these groups. Drude groups his primary divisions in the same manner *; and this is perhaps the most philosophical method of dealing with them, though, on account of the greater differentiation the southern Floras present, it is preferable to consider them separately—that is to say to give the Australian, African, and South-American regions the same rank as the great northern region, and treat their tropical and temperate parts as subregions of so many regions, rather than subregions of a south temperate and a tropical zone respectively. Too close an adherence to climatal primary regions leads to unnatural combinations, as has been pointed out in regard to Drude’s Indian region. But before entering more fully into the limits of the primary botanical regions, Wallace’s zoological regions will be briefly discussed in relation to the distribution of plants.

CoMPARISON OF THE ZOOLOGICAL WITH THE BOTANICAL REGIONS.

Wallace based his zoological divisions (as tabulated, page xxviii) on the present distribution of mammals, having, after years of study and research, arrived at the con- viction that this class furnished the best foundation for the purpose. He also found that the distribution of birds and other groups harmonized sufficiently with such a division, and any anomalies or divergences in their distribution were capable of explanation by a study of the exceptional means of dispersal and conditions of existence. Still he is careful to emphasize the fact that any system of division must necessarily be more or less arbitrary and artificial, and not equally applicable to all classes of animals. Such mammals as the bats, which fly, and the oceanic mammals, which swim, possess exceptional means of dispersal, and therefore they are not taken into consideration. These exceptions and many other phenomena of distribution in the animal kingdom are paralleled in the vegetable kingdom, but no intelligible system of botanical division could be based on the distribution of any one group of plants less comprehensive than the Phanerogamia, because the large orders, such as the Com- posites or Leguminose, comprise plants of every size, habit, and duration, inhabiting every kind of situation. As an illustration of the wide dispersal of a very large pro- portion of the natural orders of plants, the reader may be referred back to the analysis of the flora of British India, pp. xv-xxii. Generally speaking, the smaller a Flora the larger the proportions of orders and genera to the total number of species t.

* Petermann’s ‘ Geographische Mittheilungen,’ Ergiinzungsheft lxxiv. p. 43. + The indigenous Phanerogamic Flora of the Bermudas, for example, is estimated at 120 species belonging to ninety-eight genera and forty-eight orders. See Botany of the ‘ Challenger’ Expedition, 1. p. 8.

BIOL. CENTR.-AMER., Bot. Vol. I., October 1888. e

a

XxXiv INTRODUCTION.

Notwithstanding the different and often greater means of dispersal possessed by plants, it is surprising how very similar are the broad features of the distribution of plants and animals. Doubtless this is owing in part to interdependence; and extensions of area of members of the two kingdoms have probably often been contemporaneous. Still, there are important divergences, and the primary regions of plants and animals cannot always be held as conterminous; assuming, of course, that Wallace has adopted the most natural divisions that could be found. This is most strikingly exemplified in the northern floras. Wallace was able to keep separate the eastern and western hemispheres, even in the north; and his palearctic and nearctic regions he defends against the opinion of Huxley *, the endemic element being nearly equal in the two. On the merits of the question of one or two primary northern zoological regions it is not proposed to enter ; but such a division cannot well be sustained in the vegetable kingdom, the alternative being more than two. Dr. Asa Gray long ago f pointed out the intimate relationships existing between the floras of Japan and North America, especially eastern North America: and the rich collections from Central China received at Kew within the last two years have added considerably to the number of genera, and almost identical species, common to Eastern Asia and Eastern America. Many of these extend to the mountains of North India, and a very few farther westward; but the affinities of the Floras of Eastern Asia and Eastern America are vastly greater than either exhibits with that of Europe. It is only in the higher latitude of North Corea and Mandshuria and northward that the vegetation bears a strong likeness to the European; but even there the relative proportion of woody plants is much higher than in Europe ~. How- ever, it seems clear that the whole north temperate and arctic flora is better considered as forming one primary botanical region, with extensions, or remains of extensions, through the mountain-chains to the Australasian Alps, Tierra del Fuego, and the. mountains of Tropical Africa, with only very faint traces in South Africa §.

The alternative of more than two primary northern botanical regions seems quite inadmissible ; and this is tbe opinion of Engler, who has also specially examined the paleontological evidence, which proves that many of the genera of E. Asia and Eastern N. America formerly existed in Europe. If more than one primary northern region be admitted, we must, like Drude, recognize five or six; and, after all, there are no

* See ‘ Proceedings of the Zoological Soviety of London,’ 1868, pp. 313-319 : “ the Geographical Distribution of the Alectoromorphe,” where the author suggests the propriety of two primary zoological regions, namely, a northern and a southern. As an alternative he proposes four primary regions, namely: 1. Arctogea (practi- cally Europe, Asia, North America, and Africa); 2. Austro-Columbia (South and Central America); 3. Australasia (Australia and New Guinea to Celebes and the Philippines); and 4. New Zealand.

+ ‘Memoirs of the American Academy of Arts and Sciences,’ n. s. vi. 1858-59; and more fully elaborated by Engler, ‘ Versuch,’ i. pp. 22-43 (1879).

~ Maximowicz in Bull. Congr. Intern. Bot. et Hort. St. Petersb. 1884, p. 152.

§ Sir Joseph Hooker enters fully into the distribution of “Scandinavian Forms,” Transactions of the Linnean Society, xxiii. p. 251.

INTRODUCTION. XXXV

sharper contrasts in the vegetation of different areas of this large region than there are in the Indian, the Australian, or anv of the southern regions. ‘The poverty in genera and species of the woody element of the European and West-American Floras and its extreme richness in the Floras of China and Japan and Eastern N. America are well known; but with very few exceptions the genera of the former Floras are represented in the latter. As Gray remarks:—“ All round the world in our zone the woods contain Pines and Firs and Larches, Cypresses and Junipers, Oaks and Birches, Willows and Poplars, Maples and Ashes, and the like” *. The distribution of many of the genera of herbaceous plants of the poorer Floras is even wider. Thus, in California, where there is an enormous development of peculiar genera of some orders, the Ranunculacee are represented by thirteen genera, whereof eleven are also British, and one other is European, and the solitary one remaining is anomalous and endemic, and has been referred to various natural orders. The Orchidacee, again, are represented by ten genera, eight of which are British. Calypso borealis is a native of Northern Europe and Siberia, and the monotypic Aplectrum ranges across the continent to the Atlantic, being the only one restricted to America.

The subregions of Wallace’s palearctic and nearctic zoological regions are perhaps less in harmony with botanical subregions than those of any other part of the world. His Mediterranean and Mandshurian subregions in the Old World, and his Californian, Rocky Mountains, and Eastern United States in the New World, are, however, substan- tially identical with botanical subregions. It does not come within the scope of this sketch to examine and discuss all these subregions, but a few remarks follow on the Chinese and North-Mexican Floras, which have recently been more fully investigated. In a lecture delivered before the Royal Institution of Great Britain in 1878, and after- wards published, Sir Joseph Hooker explains the main features of the distribution of the Flora of North America, and Professor C. S. Sargent has exhaustively described and elaborately illustrated cartographically the distribution of the arboreous element f; and Mr. C.J. Maximowicz has made a valuable contribution to the botanical geography of Central Asia in a paper which has already been referred tof. In these and various other more generally known publications much new light is thrown on the nature, composition, and distribution of the vegetation of the countries under consideration.

With regard to Wallace’s Mandshurian or Japanese subregion, as already hinted, a corresponding botanical subregion would be differently bounded; its northern limit barely reaching Peking and its western limit not extending so far, if at all, into

* « Forest Geography and Archeology,” Am. Journ. Sc. xvi. (1878) p. 183.

+ ‘Report on the Forests of North America (exclusive of Mexico).’? Department of the Interior: Census Office, 1884.

+ “Sur les Collections botaniques de la Mongolie et du Tibet septentrional (Tangout) recueillies récemment par des voyageurs Russes et conservées & St. Pétersbourg,” Bulletin du Congrés International de Botanique et

d’Horticulture & St. Pétersbourg, 1884, pp. 135-196. e 2

XXxvi INTRODUCTION.

Mongolia. In the present comparatively early stage of an enumeration of Chinese

plants * the data are not forthcoming to describe exactly where this subregion should

be bounded, but it is warm temperate in climate, and characterized by having a very

large woody element, with intimate Himalayan and Eastern North-American connec-

tions. Rarity of annual plants is also strongly marked. Collections received since the

publication of the parts of the ‘Index’ issued have very largely augmented the num-

bers of genera and species, especially of the earlier orders, which were elaborated before

even the first collection came to hand, Therefore data taken from it alone are im- perfect beyond even what is known of the flora. To give some idea of the large number of species concentrated in a small area, it may be mentioned that Dr. A. Henry has collected upwards of a thousand species of flowering plants within a short distance

of Ichang, and at a very moderate estimate ten per cent. of them were previously unknown.

A rough analysis of the Polypetale enumerated in the ‘Index’ gives a total, exclud- ing cultivated and doubtful ones, of 1514 species belonging to 476 genera and sixty- six (out of a total of eighty-five) natural orders. Fourteen of the genera and 626 (or 41:3 per cent.) of the species are apparently endemic, but these figures do not represent the true proportions, because all the common plants of the coast districts are included, and it is very imperfect as far as regards the interior.’ The small number of species to a genus (about three), half the number found in the larger areas examined, has already been commented upon. The distribution of the non-endemic species in the above total, so far as it is known, is as follows :—140 extend to Japan only; 90 to India only; 27 to India and Japan only ; 273 others are restricted to Asia, inhabiting some other part or parts than Japan or India, or besides Japan and India; and the remaining 397 (nearly a quarter of the non-endemic) are of wider range.

Other remarkable features in the Chinese Flora are the great latitudinal range of some of the species, and the high northern localities of some subtropical types, such as Nelumbo, Euryale, and Cedrela. ‘The same phenomenon is exemplified in the animal kingdom.

Returning to the New World, it will be seen that Wallace’s Rocky Mountains subregion embraces the North-Mexican province, as roughly defined in vol. iv. pp. 139 and 306, and the country northward to about 55° Jat., and from about 97° long. west- ward to the coast range ; thus covering the whole prairie and treeless regions, the central mountain-range, and Lower California. The corresponding botanical subregion occu- pies a similar but rather more restricted area, and it is essentially a dry one, falling into several provinces, one of which, the North-Mexican, is approximately defined and characterized in the place cited. Dr. Asa Gray and Sir Joseph Hooker seem to have had some such division in view, though they do not go the length of actually

* « Index Flor Sinensis,” Forbes and Hemsley in Journ. Linn. Soc., Bot. vol. xxiii.

INTRODUCTION. XXXVI1

proposing subdivision beyond the three great floras, namely, the Atlantic, Central, and Pacific *.

The Californian subregion, which is practically the Sierra-Nevada range and the narrow strip of country westward of it, is more highly specialized botanically than it would appear to be zoologically.

The only other of Wallace’s primary zoological regions which differs materially from the botanical region of the same part of the world is the Australian. This he divides into four subregions}, all of which, except the ‘ Austro-Malayan,’ are sufficiently indicated by their names. The Austro-Malayan subregion, of which New Guinea is the centre, includes the surrounding islands and groups of islands from the Louisiade Archipelago to Timor, Lombok, Celebes, Gilolo, Admiralty, New Britain, New Ireland, and the Solomon group. ‘There is no great change in the vegetation such as to justity the separation of Eastern Malaya and New Guinea from the India region, or Oriental region, as Wallace names it. Indeed, botanically, it is naturally a subregion of the Asiatic and not of the Australian region. It is true that a number of Australian types extend into this subregion, and a few even beyond; but they form a very small per- centage, and nowhere, so far as is known, do they constitute a feature in the vegetation. Including all the common sea-shore plants, Miquel’s statistics of the Malayan flora { show that less than 4:2 per cent. of the Malayan Phanerogamia are also found either in Australia or Polynesia. However, very little was then known of the vegetation of New Guinea; but from an examination of Dr. Beccari’s and Baron von Mueller’s consi- derable subsequent contributions to Papuan botany, it appears that although such specially characteristic Australian genera as Eucalyptus, Acacia (phyllodineous species), Styphelia, Stackhousia, Banksia, and Grevillea are represented by one or two species each, the bulk of the Papuan vegetation is more Asiatic than Australian in character, and exceedingly rich in peculiar forms.

On OvutLyYInG AUSTRALIAN TYPES OF VEGETATION.

A few words respecting the wider extensions of Australian types apart from those belonging mainly to the cold temperate region, which have been pretty fully tabulated by Hooker and Engler, and more recently by the writer). Hucalyptus papwana and at least one other species inhabit New Guinea||; #. alba, and possibly one other species, is a native of Timor; and Blume records Eucalyptus deglupta, from the Celebes; but the genus of this tree is very uncertain, as neither flowers nor fruit were known to the

* «Vegetation of the Rocky Mountain Region,” Bulletin of the U.S. Geol. and Geogr. Survey, vi. no. 1, p. 62; and Sir Joseph Hooker’s Lecture previously referred to. —

+ See Table, anté, p. xxviii.

+ ‘Flora Indie Batave,’ iii. pp. 763-773.

§ Botany of the ‘ Challenger’ Expedition, i. Introduction, pp. 50-58.

ii Mueller ‘ Eucalyptographia, under Eucalyptus alba.

XXXVlli INTRODUCTION.

author. With regard to Eucalyptus multiflora, Rich*, from Mindanao, one of the Philippine Islands, there is also great doubt as to the genus; and no species of Euca- lyptus is in any of the older or the large recent collections at Kew from that archi- ‘pelago. Therefore the range of the genus Eucalyptus, so far as it is known with absolute certainty, is from Tasmania to New Guinea and Timor, and it is essentially a warm temperate type. Acacia, on the other hand, is almost wholly tropical in its distribution outside of Australia, and the phyllodineous section, which numbers hard upon 300 species in Australia, exhibits some much wider and very remarkable exten- sions. Two or three species of this section inhabit New Caledonia and other islands of the South Pacific. Acacia richit is a native of the Fiji Islands, and specimens in- distinguishable from it have been collected in Formosa, separated by forty degrees of latitude and sixty of longitude, though it may exist and yet be found in some inter- mediate stations. Another species (A. heterophylla) is indigenous in Mauritius and Bourbon, and possibly also in Madagascar, and the Sandwich-Island A. koa is so near it that the late Mr. Bentham was of opinion that it was a form of that speciesf.

The capsular Myrtacez, which are so specially Australian (New Zealand and New Caledonia) extend to China and the Malayan peninsula, where they are represented by one, or in some instances two or three species, of the genera Baeckea, Leptospermum, Tristania, and Melaleuca. Xanthostemon, another genus of the same group, is con-. fined to Australia and New Caledonia, with the exception of the Philippine Island A. verdugonianus. Stylidium is perhaps the only strictly Australian type extending into the heart of India, even to the foot of the Sikkim Himalaya. It is a genus numbering eighty-five Australian species and three Indian—one of the latter being also a native of North Australia, a second scarcely more than a variety of it, while the third is quite distinct. Leucopogon, an Australian Epacrideous genus of nearly 120 species, extends to the Philippines, yet there are only two species known from the whole Malay Archipelago. Helicia is apparently the only genus of the Proteacee extending north of New Guinea into Asia, and this, although represented in Australia, has its greatest concentration in Malaya and India—ten species being found within the limits of British India, three of which inhabit Ceylon and the Western Deccan peninsula. One species is a native of Japan, and one, or more, of South China and Formosa.

Extensions of Australian types into Eastern Polynesia are relatively more numerous than into Malaya, yet they do not predominate over the other elements of these small insular floras; and there is such an intermingling of American, Asiatic, and Australian types in the much more highly-developed Sandwich-Island Flora, that it cannot,

* A. Gray, Botany U.S. Exploring Expedition, i. p. 554.

+ Mr. J. G. Baker informs us, on the authority of Mr. Baron, that there is a possible chance of his lately- described Acacia atphoclada, from Madagascar, being an introduced Australian species.

t In addition to A. koa, two endemic species are described in Hillebrand’s lately-published ‘Flora of the Hawaiian Islands.’ .

INTRODUCTION. XXXIX

as already observed, justly be regarded as appertaining to any one of the great primary regions of vegetation. The characteristic endemic shrubby Composite* and Lobeliacee are most nearly related to American forms; Perrottetia is a Mexican and Colombian genus, and the pomaceous Osteomeles anthyllidifolia is a member of a genus all the other species of which are Andine. Nama is otherwise restricted to America; and Hillebrand regards the American Lythrum maritimum, Daucus pusillus, and Aster divaricatus (== A. evilis) as indigenous. Prominent among the Australian types and common in all the islands according to Hillebrand are :—Metrosideros polymorpha, an exceedingly variable tree or shrub scattered throughout Polynesia, eastward to Pitcairn Island, and Acacia koa, already alluded to. The Australian genera Cyathodes and Exocarpus are also represented. Cyrtandra, of which there are thirty-two species endemic in the Sandwich Islands, is more Malayan in character, and many other such relationships exist, besides other more remote ones in the highest mountain flora, which includes such forms as Luzula campestris, Rumex, Silene, Ranunculus, Drosera longifolia, Hydrocotyle interrupta, Fragaria chilensis, Vicia, Vaccinium, Aster, and Artemisia.

Added to the foregoing elements is a sea-shore element consisting almost exclusively of species having a very wide range in the Old World; many of them from the eastern coast of Africa or the Mascarene Islands and India to N. Australia, the Marquesas Islands and Easter Island. Further, the vegetation of the small and remote coral islands is entirely of this character. .

Besides the phyllodineous Acacia above alluded to, it has long been known that there were two or three other noteworthy outlying Australian types in Madagascar ; but no important addition to these has been made by recent explorations. One or two species of Hibdertia (Dilleniacez), and two or three of Rulingia (Sterculiacee-Lasio- petalee); and the genus Adansonia is represented by one endemic species in Mada- gascar, one in tropical Africa, and one in North Australia. These widely-sundered areas of distribution of closely-allied forms have given rise to much speculation, and it is very problematical whether satisfactory paleontological evidence will ever be forth- coming which will account for the existing distribution of plants.

BoraNIcaL DIVISION OF THE EARTH INTO PRimaRY REGIONS.

From the data adduced in the preceding paragraphs, and numerous more familiar facts which it is unnecessary to repeat, it is clear that a system of botanical geography should be based upon a small number of primary regions, similar in many respects to Sclater and Wallace’s zoological regions. It has been shown where the botanical regions do not even approximately coincide with the zoological regions, and to some

® See Bentham in Journ. Linn. Soc. xiii. p. 555.

+ The Madagascar plant, long supposed to be a species of Evocarpus, as only foliage was known, is Neobaronia, Baker, a new genus of Leguminose, of which flowering specimens of two species have been collected by Mr. Baron.

xl INTRODUCTION.

extent also this has been done for the subregions or secondary divisions; but it is obviously beyond the scope of the present sketch to pursue this to the end. Yet, without attempting to define secondary and further divisions of all the primary regions, it may be possible to give further particulars, which, with what has preceded, shall convey some general idea of the extensions of the characteristic types of the floras of these regions. .

Wallace states that nothing like a perfect zoological division of the earth is possible ; and a perfect botanical division is equally impossible. Therefore the aim is a division that conveys in itself some notion of the extent of the dispersion of the characteristic types of each region. As already pointed out, it is obvious that a botanical division cannot be based on one class of plants, equivalent, if such could be found, to the Mammalia ; but rather on the general composition of the vegetation, and largely, too, apart from latitudinal position. On this plan it is not difficult to distinguish five primary regions; but if this number be exceeded, it must be doubled or trebled with a less satisfactory result. These five primary regions are :—

1. Northern. 4, South American. 2. African. 5. Australasian. 3. Indian.

These regions are unequal in extent, and the southern divisions are meridional rather than latitudinal; but the greater degree of differentiation of the vegetation justifies such a division; and a natural subdivision of the northern region, excluding the arctic zone, and, perhaps, the coldest arboreal zone, is in the same direction. But these five _ regions do not include quite all, because, for reasons given, the Sandwich-Island Flora cannot consistently be included in any one of them; and it would convey a false impression to rank this comparatively small, though highly peculiar Flora, as a primary region. Further, there isa remnant of an antarctic Flora which is scattered all round the hemisphere, constituting the coldest southern zone of vegetation. ‘These exceptional phanerogamic Floras will be separately discussed.

An alternative primary division, which is more in accord with the writings of many botanists, and has some practical advantages over the one proposed, may be thus

stated :— 1. Northern Region.

2. Neotropical Region. 3. Paleeotropical Region. 4, Andine Region. 5. Cape Region. 6. Australasian Region.

The anomalous Sandwich-Islands Flora and the fragmentary Antarctic Flora would be unattached in this, as in the foregoing plan. It is unnecessary to define these alternative regions, as their denominations are sufficiently descriptive to be intel- ligible. One of the advantages this division possesses over the one advocated, or rather the one point in which it radically differs, is the separation of the tropical and temperate

INTRODUCTION. xli

Floras in the primary regions of the southern hemisphere. Elsewhere reasons are given for treating the Cape and Andine Floras as subregions of more extended areas. Another important difference is the retention of three south temperate regions against one north temperate region, which brings out more prominently the greater differen- tiation of the vegetation of the southern hemisphere. Practically the plan proposed in this sketch recognizes and deals with these facts, although it does not bring them into direct contrast.

There is yet another way of treating the subject, one that obviates the necessity for defined regions and subregions, and one that may be effectively employed for conveying an idea of the broad features of the distribution of plants: it is by grouping the Floras into northern, tropical, and southern, as was done by Thiselton Dyer in the lecture on the geography of plants, which has been referred to before.

Some further justification of the plan here adopted will now be attempted, but it is not intended to recapitulate well-known facts and traverse familiar ground. The inten- tion is rather to bring together some facts and data additional to those collected in the Appendix, or adduced in the preceding paragraphs of the present sketch.

The Northern Region.

This corresponds very closely to Wallace’s Arctic and Palearctic regions combined, and is adopted for reasons already sufficiently explained; the close relationship of the Floras of Eastern Asia and of North America, especially eastern, than of either of these and the European being the principal reason. A rough subdivision of this region gives eight subregions, namely :—an Arctic, a North and Mid-European and Central Asian, a Mediterranean, a Chino-Japanese, and in America a continuation of the Arctic, a boreal, an Atlantic, a Central, and a Pacific subregion. In the Old World we find that the vegetation of North and Central Europe extends to the extreme east of Asia, associated there with an endemic element, which increases so much in China, south of about the fortieth parallel, and in Japan, as to constitute of these countries a distinct subregion, exceedingly rich in trees and shrubs—richer even than the Atlantic subregion of North America. The transition from the Chinese Flora to that of a more European character is very sudden on entering Mandshuria, where probably in some districts fifty per cent. of the species are European. Maximowicz gives* some interesting figures illustrating the gradual diminution of the absolute numbers and percentages of species having a wide area—that is circumpolar, or common to both Europeand Asia. Thus in Baikal-Dahuria the figures are 747 species, =53°4 per cent. ; in Mongolia 599 species, 46-2 per cent.; in Mandshuria 533 species, =39°6; Peking district, 318 species, —31:9 per cent.; Japan, 442 species, =16°2 per cent. The percentages of endemic species in these five areas are respectively 9-4, 8-3, 8:7, 13°3, and 44-0. Maximowicz comments on the great fall in the percentage of the endemic element passing from

* Bull. Congr. Intern. Bot. Hort. Pétersb. 1884, p. 164. BIOL. CENTR.-AMER., Bot. Vol. I., October 1888. f

xhii INTRODUCTION.

Japan to the continent. This is so in the latitude of Peking; but the more southern province of Hupeh, for instance, would show a much higher percentage than Peking, and possibly even higher than Japan, inasmuch as the Polypetale of the whole of China proper, so far as known *, contain 41°34 per cent. of endemic species, and there is no doubt that the further exploration of the interior provinces will materially raise this percentage.

Maximowicz very elaborately analyses the affinities of the endemic species in the areas named, from which it appears that it is only in Japan that there is a considerable development of endemic species of what he terms the Chinese type. On the other hand, the “levantine type” is apparent in 13°9 per cent. of the endemic species of Mongolia. To be brief, Maximowicz sums up this question in words to the effect that the plants of the plains of Northern and Central Europe constitute the greater part of the flora eastward to the Pacific coast, if not in identical forms, at least by forms connected by intermediates with European species f.

It would carry us too far to attempt to give the exact eastern limits of the Mediter- ranean region in Asia, as it would involve a discussion of the vegetation of the various provinces of this subregion. The southern limits of the northern region in the Old World may be indicated approximately as the tropics, varying in different longitudes. ‘It includes the alpine and temperate Flora of the Himalayas, and in Africa it extends to the Great Atlas f.

The subregions of the northern region in the New World are five, namely :—Arctic, Boreal, Atlantic, Central, and Pacific, which may be roughly defined. Briefly the Arctic is a continuation of the same subregion in the Old World; the Boreal is a westward extension of the hardier elements of the Atlantic subregion, and perhaps better regarded as a province of it; the southern or Mexican province of the Central subregion is fully described in the Appendix; and further particulars of the characteristics of the whole Rocky-Mountain Flora will be found in the joint essay, by the late Dr. A. Gray and Sir Joseph Hooker, previously cited. The distribution of the North-American Flora generally is admirably summarized by the latter §, who distinguishes the Sink country between the Rocky Mountains and the Sierra Nevada as a separate Flora, while admitting that cacti and yucca attain their maximum development further south in the same meridian. Professor Sargent describes his northern forest-region as extending southward to the fiftieth parallel on the Atlantic coast and to the fifty-fourth at the hundredth meridian ||.

* « Index Flore Sinensis,”’ Journ. Linn. Soc. xxiii.

+ Ina recent collection of about 500 species, made in Mandshuria by Mr. H. E. M. James, nearly a third are British species (see ‘ Proceedings of the Royal Geographical Society,’ 1887, p. 548).

t For an account of the vegetation of these mountains consult Hooker and Ball’s ‘ Tour in Morocco,’ 1878.

§ Proceedings of the Royal Institution of Great Britain, 1878.

|| Report on the Forests of North America, p. 3.

INTRODUCTION. xiii

Allusion has been made to the southward extension of elements of the Flora of the northern region through the great mountain-ranges to the southern limits of vegetation, and indications where there are remains of it on the mountains within the tropics and in south temperate countries. Sir Joseph Hooker's reports on the vegetation of Clarence Peak, Fernando Po, of the Cameroons mountains, and of Kilima-njaro are among the most important of the later contributions to the literature of this subject *.

The African Region.

The phytogeographical essays last referred to afford some most interesting details of the relationships and apparent migrations of the components of the vegetation of the great African region, which is here understood to comprise the whole of Tropical and South Africa, Madagascar, Mauritius, Bourbon and the contiguous groups of islets, and the Cape Verde Islands—Madeira and the Canaries being regarded as a province of the Mediterranean subregion of the Northern region. Ascension, St. Helena, Tristan da Cunha, and Amsterdam and St. Paul Islands, though presenting some curious anomalies in their vegetation, may be regarded as appanages of the African region, or they might be left unattached to any primary region. Excepting the Composite, the affinities of the flora of St. Helena are distinctly African ; and Phylica nitida, the only tree, or even shrub more than a trailer, in the other two groups of islands, is Mascarene f. Instead of arctic and temperate climates there are in this region tropical and temperate climates; yet botanically, as well as geographically, this is one of the most compact of the primary regions, and naturally divides into three subregions, namely:—Tropical Africa, South Africa, and Madagascar and adjacent islands. Eastern and Western tropical Asia should only be regarded as provinces of one Flora, as will presently be demonstrated.

Besides remote connections with the Northern, American, and Australasian regions, there is a very intimate connection of the African with the Indian region, traceable from the Cape up the eastern side of the continent to Abyssinia and by way of Socotra, Southern Arabia, Persia, and Afghanistan to the Panjab and Gangetic plain, and less distinctly southward into the Deccan peninsula, with a few extensions into the Malayan peninsula and archipelago ; and notwithstanding the presence in the Madagascar sub- region of such eminently Asiatic types as Nepenthes, Lagerstremia, and a few others hitherto not found in continental Africa, it is evident that the interchange between Africa and Asia is far greater than between Madagascar and Asia, whether we con- sider tropical or temperate types. As might be expected, among the species common to India and Africa, those characteristic of a dry climate largely preponderate. On this

* See Journ. Linn. Soc., Bot. vi. p. 1, vii. p. 171, xiv. p. 141, and xxi. p. 392. + For a full account of the botany of these islands, see Botany of the ‘ Challenger’ Expedition, i. part 2,

and Introduction.

f 2

xliv INTRODUCTION.

point Dr. Schweinfurth and Dr. Bayley Balfour’s botanical investigation of the island of Socotra furnishes conclusive data, collected by the latter *.

About one third of the species of flowering plants of Socotra are endemic; and a third are species belonging to tropical Africa and tropical Asia. Exclusive of species having a wider area, about eighty-five of the species of dicotyledons are common to Africa and Asia, though comparatively few of them penetrate India eastward of Scindh. The affinities of the entire Flora of Socotra are essentially tropical African and tropical Asian, but the African element predominates, and is mainly composed of the features of the Flora of the mountainous region of Abyssinia, with an in- fusion of West-African, South-African, and Madagascar types. Among South-African types are Graderia, Babiana, Thamnosma, Lasiocorys, and Euryops, of which the first two are not known to be represented in the intervening country. In this connection it may be mentioned that Dr. Aitchison discovered in Afghanistan the very distinct Fingerhuthia, a genus of grasses previously known only from South Africa. It was one of the most abundant grasses between Thal and Shinak in the lower Kurram valley, and presents no obvious characters to separate it specifically from F. africana, though Boissier has described it as a different species f.

As in the Madagascar flora, so in the Socotran, there are a very few isolated types whose nearest allies are in the New World. Balfour specially notices his new mono- typic genera Dirachma (Geraniacese) and Celocarpus (Verbenaces) as belonging to this category. The three or four Turneracee in Madagascar, the arboreous Mathurina of the same order in Rodriguez, and Ravenala madagascariensis are other examples ; yet this element is by no means so prominent in the flora as it would appear to be in the fauna, judging from Wallace’s remarks thereon f.

To include the whole of tropical Africa in one subregion is unusual, but the facts seem to warrant this course, and the next division should be into several provinces. Interruptions in the continuity of the vegetation there are, and the forests of the eastern side of the continent are probably nowhere so rich as those of Guinea: but their composition is essentially the same. Taking the first volume of Oliver’s ‘Flora of Tropical Africa,’ which is, of course, exceedingly fragmentary, it would appear that about one fifth of the species there enumerated are common to both sides of the con- tinent ; but subsequent investigations leave no doubt that the proportion is really much higher. Engler) has analyzed the composition of the flora of tropical Africa as far as published in the work cited, namely the Polypetale and the Gamopetale to the end of the Ebenacee; but the results can only be used in respect to the general relationships. The Leguminose rank first and the Composite next in regard to number of species,

* Proceedings of the Royal Institution of Great Britain, 1883, and Transactions of the Royal Society of Edinburgh, vol. xxxi.

t ‘Flora Orientalis,’ v. p. 569. + ‘Island Life,’ p. 420. § Versuch, ii. p. 276.

INTRODUCTION. xlv

Very small areas have, however, been thoroughly investigated botanically; hence we are far from knowing the extent and degree of richness of the flora.

From Engler’s tabulation of Oliver’s enumeration it appears that there is in W. Africa a larger proportion of endemic species, and fewer widely dispersed genera, than in K. Africa, and that there are about as many genera otherwise only represented in Madagascar and the neighbouring islands as there are in E. Africa. Further, W. Africa has more genera in common with tropical Asia than has E. Africa, which might be held as a sufficient ground for following Engler and others in regarding the tropics of the Old World as a primary region. Among Asiatic species discovered in Angola by Dr. Welwitsch was the remarkable Naregamia alata (Meliaceee), previously only known from the Deccan peninsula of India. West Africa is relatively rich in genera otherwise restricted to America, though these are mostly represented by one or very few species. Many of them have seeds or fruits that float and bear long immersion in sea-water without injury, hence their presence in Africa may be due to oceanic currents ; others may have been introduced with ballast. But after eliminating all these, there remain many remarkable connections between the two floras which are not so easily explained *.

Gustav Mann’s botanical exploration of the temperate regions of the Cameroons Mountains in 1861 and 1862 resulted in some remarkable revelations published by Sir Joseph Hooker +. Nearly all the genera and half of the species are common to the mountains of Abyssinia, and one of the genera and many of the species are not found elsewhere. The number of European genera represented in this flora of fifty-six genera and 237 species found at elevations above 5000 feet is forty-five; thirty-eight of the genera and twenty-seven of the species are British. The South-African element is a small one, and consists almost wholly of species which also inhabit Abyssinia; it includes the genera Anthospermum, Bleria, Ericinella, Peddiea, and Geissorhiza.

Returning to Engler’s summary, the much greater development of the Mediterranean forms is one of the most striking characteristics of the Flora of Eastern Africa, where they meet and intermingle with South-African types. More recently the mountains of eastern tropical Africa have been explored and the results given to the world ¢. Mr. Thomson’s Kilima-njaro and other mountain collections, more particularly referred to here, consist of 140 species belonging to 107 genera, and add no fewer than nine northern genera to the equatorial-African Flora. Altogether they contain twenty- seven genera and thirty-seven species of a northern type; and the rest are almost exclusively South-African in character, some of the species being identical—Calodendron capense, Clematis thunbergiana, and Alepidea amatymbica, for example. In Angola

* See Engler, ‘ Versuch,’ ii. pp. 176-179. + Journal of the Linnean Society, Bot. vii. pp. 171-240. + See Journ. Linn. Soc., Bot. xxi. pp. 392-406, and Trans. Linn. Soe. 2nd ser., Bot. ii. pp. 327-355, tt. 60-63,

Sir Joseph Hooker and Professor Oliver.

xlvi INTRODUCTION.

such South-African types as Faurea (Proteacez), Selaginee, Cyrtandree, Hricacee, Mesembryanthemum, and Aloe were encountered by Welwitsch *.

The Cape subregion must be dismissed with a reference to the latest essay on the composition and subdivision of its vegetation f.

The relationships of the Madagascar subregion to tropical and South Africa have been discussed mainly from data extracted from Mr. Baker’s published contributions to this rich Flora}, and he has kindly permitted the use of some additional facts taken from his unpublished catalogue of the known flowering plants of the island. The number of species is about 3650, belonging to 1000 genera and 141 orders. This number represents probably less than half the flora of Madagascar ; but it may be regarded as a fair sample, consisting partly of the upland or Cape element and partly of the low- land or tropical forest element. The twelve largest orders are :—

Genera. Species. Genera. Species.

§Leguminose . .. . 8l 352 Acanthacee . . . . . 25 117 “§Composite. . . . . 59 261 Graminee. . . . . . 48 110 Euphorbiacee. . . . 39 207 Urticacee. . . - . . 20 73 Orchidee . . . . . 388 169 Tiliacee . 2. 2. . 6 72 Rubiacee . . . . . 56 138 Labiate . . .... I18 56 Cyperacee. . . . . 22 129 Sterculiacer . . . . . 10 55 Totals. . . . 419 1739

Ten of these orders are among the first twelve in the Indian Flora, though occupying relatively very different positions, and the other two, Tiliaceee and Sterculiacez (replacing Asclepiadez and Rosacee), are brought into this position by the very large number of species of Grewia and Dombeya. Tighteen orders are represented by only one species each, and thirty-five others by twenty species and upwards. Of the endemic order, Chlenaces, seven genera and twenty-two species have been defined. The Ternstre- miacee afford an example of a widely spread order poorly represented both in Africa and Madagascar, whence only one species is recorded. Among characteristic South- African genera in Madagascar are Hricinella, Philippia, Selago, Aloe, Aristea, Geissorhiza, Gladiolus, Faurea, Alectra, Harveya, Disa, Satyrium, Lasiosiphon, Phylica, and Antho- spermum. It is singular, too, that the solitary known Madagascar species of the genera Viola, Geranium, and Drosera occur in the mountains of tropical Africa, and the Drosera also in South Africa, though none is known to have a wider range. Sufficient evidence has perhaps been adduced to justify the course of treating the whole of tropical and South Africa and the Mascarene islands as a primary region, divisible into three subregions.

* « Sertum Angolense,” Trans, Linn. Soc. xxvii.

+ “A Sketch of the Flora of South Africa,” by H. Bolus. A reprint from the ‘ Official Handbook to the Cape of Good Hope,’ 1886.

+ See Journ. Linn. Soc. xv., xvi., xviii., xx., and xxi., and Journ. Bot. 1881 (phytogeographical), 1882, 1884.

§ The same position as in the whole of tropical Africa.

INTRODUCTION. xl vii

Conspicuous among orders absent from the African region are the Cupulifere *, the characteristic order of trees of the northern region, and hardly less so of the mountains of tropical Asia (extending southward to New Guinea, though not south of the Ganges in Western India), and of the mountains of Mexico and Central America, extending nearly to the equator (Quercus), reappearing both in the American and Australasian regions in the extreme south (Fagus). Other notable orders or tribes unrepresented in the African region are:—Magnoliacee, Acerinee (Maples), Pomacee, Hydrangez, Cornaceze, Caprifoliacee, Vacciniacese, Rhodoracee (Rhododendron), and Abietines. There is also a very much smaller development of such essentially tropical orders as the Myrtacez, Aroidex, and Palme than in either the Indian or the South-American region. On the other hand, the northern genus Erica, which covers thousands of square miles in Europe with very few species, is represented by hundreds of species in a comparatively small area in South Africa. Such anomalies occur in nearly all Floras: take the genera Ranunculus, Epilobium, and Veronica in New Zealand, for instance, where combined they constitute eight or nine per cent. of the flowering plants.

The Indian Region.

This, it is assumed, should include the whole of Wallace’s “ Oriental” zoological region and those portions of his Australasian region indicated in a previous paragraph (p. xxxvil), and Western Polynesia. Indeed the whole of Polynesia, except the Sand- wich Islands, might be included. It is not intended to discuss the subdivision of this region, as the collection and examination of the data would involve great labour. New Guinea and some of the adjacent islands to the west, and those eastward to the Fiji group, constitute a distinct subprovince. Whether the remainder of the Malayan Archipelago should, with the Malayan Peninsula and Cochin China, all be included in one subpro- vince is not quite so certain. Some parts are exceedingly rich in endemic species and proportionately in genera, while others, the Philippines for example, are remarkably poor in endemic generic types, for only six genera in upwards of 1000 are endemic. It is here, too, that the highest proportion exists of monocotyledons to dicotyledons in any Flora of considerable extent of. which there are available statistics, it being as 1 to 157+. Further materials will probably modify these figures, though not perhaps to any great extent.

Miquel records some statistics { of the Flora of the Malayan Archipelago, but as he took a much more restricted view of genera and species, especially of the latter, than the other authorities cited, they will only serve for approximate comparisons. The twelve natural orders most numerous in species are:—1, Leguminose, 676 ; 2, Orchidee, 616 ;

* Even in the wide sense of Bentham and Hooker, for although the European Alnus glutinosa is now widely spread in South Africa, it is perhaps beyond doubt that it was introduced by man.

+ BR. A. Rolfe in Journ. Linn. Soc. xxi. p. 292.

t ‘Flora Indie Batave,’ iii. p. 768.

xlviii INTRODUCTION.

3, Rubiaceew, 594; 4, Urticacee, 448; 5, Graminee, 430; 6, Euphorbiacee, 265 ; 7, Acanthacee, 257; 8, Composite, 250; 9, Laurinee, 234; 10, Palme, 234; 11, Melastomaceze, 224; 12, Myrtacee, 215. The total number of phanerogams is set down as 9118 species; and the monocotyledons and dicotyledons are as 1 to 3°5. It should be added that Miquel’s enumeration is very far from complete for many of the islands.

Sufficient particulars have been given of the botany of British India as a whole, and it is not intended attempting to characterize the subregions. It may be mentioned in passing, however, that Malayan types have penetrated to the flank of the Himalayas and traversed the Deccan peninsula to Ceylon. Thwaites, who enumerates upwards of 2600 phanerogams in Ceylon *, states that the hill Flora resembles very much that of the Neilgherries ; in the humid south it is more akin to that of the Malay Archipelago, and in the dry north it is very nearly identical with that of the Coromandel coast.

The supposed special relationships between the Flora of the Deccan peninsula of India and those of Madagascar and tropical Africa alluded to by many writers, are probably not greater than those existing between the African region and Malaya.

The South-American Region.

The data brought together in the Appendix relative to the composition and the distribution of the Flora of Central America and Mexico demonstrate very clearly that, apart from the peculiar Mexican element and the southward extensions of northern types, there are two other distinct elements, namely, the Andine and the Tropical, answering to the two subregions of the South-American region. Though only two subregions are recognizable, the development of the types characteristic of each of these subregions varies very much in different areas. Thus, Chili, considered as a province of the Andine subregion, has Californian connections, and wants some of the most characteristic and universal of South-American types, while others attain their maximum development in this province. These peculiarities are chiefly due to the varying amounts of heat and moisture in different districts. Similar conditions produce similar results in some districts east of the Andes. How far many of the characteristic types are generally spread in the South-American region, and within

* «Enumeratio Plantarum Zeylaniz.’

+ Polakowsky, H., “ Die Pflanzenwelt von Costa Rica” (16 Jahresb., Ver. Erdk. Dresden, 187 9, pp. 26-124, mit einer pflanzengeogr. Karte), Just. Bot. Jahresb. viii. (1880), 2, pp. 502-506. In this paper, previously overlooked, the author gives a sketch of the composition and physiognomies of the vegetation, and brings together all the data afforded by his own collections and professedly those of CErsted, Warscewicz, Wagner, Scherzer, Wendland, and Hoffmann. He tabulates the number of species of the natural orders, and his totals are :—Monocotyledons 209, Dicotyledons 748=957, or 129 less than our total (Vol. IV. p. 218). But the total is made up in a very different way. Thus, Polakowsky enumerates 127 Composite against our 100, and only 57 Orchides against our 210.

INTRODUCTION. xlix

what limits others are restricted, may be gathered from the various tables and para- graphs in the Appendix, more especially from those paragraphs on the distribution of the more prominent natural orders (vol. iv. pp. 235 to 282), and from the analysis of a sample of the mountain vegetation (p. 298); but it may be of interest to illus- trate this further by a few examples, selecting such groups as are spread over the tropical, or subtropical, and temperate parts of the country.

Foremost among the types not extending into Chili are the Melastomacee ; but their absence can hardly be explained by climatal conditions alone. The large genus Cuphea, which ranges from the south-eastern states of North America to Uruguay and Chili, is represented in the last country by only one species (C. spicata), which covers nearly the whole area of the genus. Juchsia extends from Mexico to Magellan’s Straits, is represented in San Domingo by two species, and by three or four species in Brazil, and it reappears in New Zealand. The Turneracee are essentially temperate and sub- tropical plants of the western side of the continent, yet there are a few species in Brazil and Buenos Ayres. The specially characteristic Cactacee are spread all over South America and the West Indies, though they do not inhabit all districts. Their northern extensions are given in detail in the fourth volume, page 247. The Brome- liaceee are spread all over South and Central America and the West Indies, and penetrate the south-eastern states of North America. The more tropical Cyclanthacee, Marcgraviacee, Vochysiacee, and Myrtacee-Lecythidee are mainly eastern South American, though they are also represented in Central America; and the distribution of such genera as Tropeolum, Baccharis, Coccoloba, Lacistema, Roupala, Lisianthus, Hyptis, Philodendron, Anthurium, Brassia, Dichea, Epidendrum, Oncidium, and many others proves that we have to deal with one large primary region. Epiphytal orchids generally are local, or occupy relatively small geographical areas, yet many of the American species have a considerable range. A list of thirty Mexican species with their distribution is given in the fourth volume of this work (p. 270), and it may be added here that about eighty of the Mexican species extend to the West Indies, and upwards of 100 to South America, many of the species being the same in the three areas, as is apparent from the fact that 800 out of 938 are endemic in Mexico. So far as the distribution*is known, a larger number of the Mexican species reach Brazil than Peru. Approximately the numbers are :—Colombia (New Granada and Ecuador) seventy-five species ; Peru twenty-two; Guiana forty-five; and Brazil thirty-six *.

Wallace divides his neotropical zoological region, which corresponds to our South- American region, into four subregions, namely a Mexican, an Antillean, a Brazilian,

* In the discussion on the northern limits of epiphytal orchids in Mexico (vol. iv. p. 269) the improbability of Meiracyllium gemma and Odontoglossum madrense being North Mexican was advanced, though in conse- quence of the habitat being given as Sierra Madre they were 80 recorded. It appears, however (Veitch, Odontoglossum, p. 51), that they were found in the neighbourhood of Colima, in about 19° N, lat., and far from the Sierra Madre, in Durango, where Seemann collected.

BIOL. CENTR.-AMER., Bot. Vol. I., October 1888. - g

] INTRODUCTION.

and a Chilian. From what has preceded it is evident that such a division would not satisfactorily indicate the leading features of the distribution of plants, as there are only two distinct elements, the andine and tropical, both represented at different altitudes in several of the provinces. A better botanical division is an Andean subregion (which merges into the North Mexican and Californian to the north and into the Antarctic in the south) and a Tropical subregion, both divisible into several provinces. In conclusion, reference may be made to a statistical account of the Andine Flora of Ecuador *, and to Mr. Ball’s recent work +, in which he suggests that the ancient mountains of Brazil formed a great centre of development of plant life.

The Australasian Region.

This includes the whole of Australia and the adjacent islands, which may be sub- divided into a south-western and a north-eastern subregion; and New Caledonia and New Zealand, which constitute other subregions. The remote connections with the American Andine flora are reviewed in the Appendix (vol. iv. p. 234), and in the description further on of the Antarctic Flora. Reasons for including New Caledonia and New Zealand in this region are given in the discussion of Drude’s botanical and Wallace’s zoological divisions of this part of the world}. It may be of interest to add here a few statistics of the vegetation of two or three definite areas of the Australasian region.

Tasmania. (After Mueller §.)

Orders. Genera. Species.

Dicotyledones . . . . . 72 257 662 Monocotyledones . . . . 15 99 272 Gymnospermee . . . . . 1 7 ll 88 363 945

Extratropical South Australia. (After R. Tate ||.)

Orders. Genera. Species.

Dicotyledones . . . . . 78 365 * 1244 Monocotyledones . .. . 16 113 322 Gymnospermee . . . . . 2 2 3 96 480 1569

* Dressel, L., Charakteristik des eucadorianischen Pflanzenschatzes: Natur und Offenbarung, xxvii., 1881. Abstract in Just’s Bot. Jahresb. x. 1882, pp. 435-441, including a table of the number of species of each natural order. Cultivated or introduced plants appear to be counted with the others, as he has one Resedacee.

+ ‘ Notes of a Naturalist in South America’ (1887), Chapter vi.

+ See pp. xxxi and xxxvii. § Census of the Plants of Tasmania.

|| Transactions of the Philosophical Society of Adelaide, 1880.

INTRODUCTION. li

New Zealand. (After Engler *.)

Orders. Genera. Species. Dicotyledones . . .. . 74: 207 697 Monocotyledones . . . . 12 94, 241 Gymnospermee . . . . . 1 By) 17 87 306 955

The smallness of the numbers of genera and species strikes one most, especially on comparison with those for the whole of Australia or with those of other areas. That this is not altogether attributable to insularity is clear from the richness of the Flora of New Caledonia, computed at 3000 species of phanerogamst. Japan, of similar extent, and lying in about the same latitudinal position in the north that New Zealand occupies in the south, shelters nearly a hundred more genera of flowering plants than there are species in New Zealand, and about three times as many species.

Flora of the Sandwich Islands.

Wallace treats this as a subregion of his Australian region; Drude regards it as a part of his Indian region; while Engler makes it a province of his ‘ Paleotropical Floral Kingdom.’ Considering the complexity of the affinities of the flora and its extent, and the fact that no element largely predominates over the others, it seems desirable to leave it unattached, without, however, giving it the rank of a primary region. Could Engler’s ‘Ancient Oceanic Floral Kingdom’ (which includes the Antarctic forest region of South America, the Southern Island of New Zealand and outlying islets, extratropical Australia, the Cape, Kerguelen, Amsterdam, Tristan da Cunha, St. Helena, and Ascension Islands) be regarded as a satisfactory solution of a difficult problem, the Sandwich Islands should be referred to this rather than to the Indian region; but the basis of such an arrangement is altogether too hypothetical from our standpoint, and it brings together the most diverse Floras. As Hillebrand remarks }, the Sandwich Islands are the only Polynesian group which contain a large number of indigenous plants of American affinities. Ina previous paragraph (p. xxxix) examples are given of the more striking genera or species of the different elements of this highly interesting Flora; and Engler’s tabular view of the Flora and its affinities affords much fuller information § on this point. Engler’s enumeration contains 669 species of vascular plants, of which he estimates 500, or 74°6 per cent., to be endemic. The recently published ‘ Flora,’ cited below, of the late Dr. W. Hillebrand, who spent twenty years in the islands, has increased this number by nearly 200 species, nearly all of them endemic. His summary is here reproduced.

* Versuch, &c. ii. p. 84. + Brongniart, in Ann. Se. Nat. 5° série, 1865, p. 187. t Flora of the Hawaiian Islands, Introduction, p. XxIX. § Versuch, il. pp. 104-131.

g 2

li INTRODUCTION.

Summary of the Flora of the Sandwich Islands. (After Hillebrand.)

Species. | Gtanecton | introduction, | Eedemic. |] Original Total. Dicotyledones .........+.: 13 92 500 584. 689 Monocotyledones ........ 11 23 74 121 155 Phanerogamia .......... ' 94 115 574 105 844 Cryptogamia Vasculares.... 0 0 79 155 155 Totals ..........00.0 000s 24 115 653 860 999

The 860 species of vascular plants regarded as indigenous belong to 265 genera, being as 3:25 tol, or about half that in large continental areas such as Mexico, Australia, and British India. Notwithstanding this low average number of species to a genus, the most striking feature in the flora is the large number of species of many of the endemic genera. Thus, Schiedea (Caryophyllacez), 17; Kadua (Rubiaceze), 13; Cler- montia (Campanulacee), 11; Cyanea (Campanulacee), 28 ; and Stenogyne (Labiatz), 17. Further the Polynesian genera Pelea (Rutaceze) and Phyllostegia (Labiate) are repre- sented by 20 and 16 species respectively, and Cyrtandra by 29 species*. Of the indigenous species 81:4 per cent. are endemic—a proportion exceeded in no other Flora of the same extent, perhaps, except that of West Australia, in which it is 85 per cent. Ninety-five orders of Phanerogamia are represented; and the proportion of Mono- cotyledones to Dicotyledones is as 1 to 4°8, or higher than might have been expected ; but the former consist largely of Glumifere, petaloid monocots being rare, and orchids | are represented by only three terrestrial genera, and one endemic species of each genus.

The question arises whether any of the other islands or groups of islands in the Pacific can appropriately be associated with the Sandwich Islands. 'The connection with the other Pacific islands beyond the sea-coast and Australian elements is so very slight that the Sandwich Islands stand almost completely isolated. Nadeaud f describes a plant from Tahiti which he refers to Phyllostegia, otherwise exclusively belonging to the Sandwich Islands; Sclerotheca arborea (Campanulacez), also a native of the Society Islands, is said to be more nearly related to the American genus Siphocampylus than it is to the endemic genera of the Sandwich Islands}; Litchia, the only genus of Compo-

* Thirty-two according to Clarke’s monograph, overlooked by Hillebrand. t Enumération des Plantes Indigéne de I’Ile de Tahiti, p. 56. t Bentham and Hooker, Gen. Pl. ii. p. 548.

INTRODUCTION. lili

sites peculiar to the Society Islands, is related to the Sandwich Islands genera only in the sense of belonging to the same tribe (Helianthoidew), which is eminently American ; Pelea is represented in the Navigators Islands. These are practically all the outside affinities with the peculiar element of the Sandwich Islands flora. The smaller islands possess no endemic plants; and even the Marquesas are almost destitute of an endemic element * ; and the vegetation of all the other groups of islands of Eastern Polynesia is much more largely Malayan in character. Nadeaud includes nineteen species of orchids in his enumeration of the plants of Tahiti, amongst them half a dozen epiphytal species of the genera Bolbophyllum and Dendrobium. Seventy-nine orders of Phanerogamia are represented by about 500 species collectively; but the fragmentary character of the Flora may be estimated from the fact that thirty-three of the orders are represented by only one species each. ‘Taking all the facts into consideration, the remainder of Polynesia may be included in the Indian region.

The Fragmentary Antarctic Flora.

We retain the designation ‘‘ Antarctic,” first employed by Forster and adopted by Hooker f, for the coldest southern Flora, although it is not geographically an admis- sible term. ‘There can be little doubt that the present scattered and isolated fragments, forming the fringe of the southern limits of phanerogamic vegetation, constitute the remnant of a Flora formerly covering a more extensive area, probably in a higher lati- tude. The surviving portion is poor, but from its uniform composition it is probably of remote origin {. Northern species are associated with such as now exist only in the south, and the same species occur in the scattered fragments of vegetation all round the hemisphere. It includes the Auckland, Campbell, and Macquarie Islands, Macdonald (Heard), Kerguelen, Crozets, and Prince Edward Islands, South Georgia and the Falkland Islands, Fuegia (Tierra del Fuego) and a small portion of South-western Patagonia. It is true that the vegetation of the Auckland group and Campbell Island is essentially New Zealand in character and relatively rich, while that of the Falkland Islands is almost wholly Fuegian ; yet there is a stronger common element than was apparent when Sir Joseph Hooker wrote his ‘Flora Antarctica.” In the fourth volume page 234, are tabulated the more striking connections between the Australasian and Antarctic Floras and that of the mountains of Mexico; and in the Botany of the ‘Challenger’ Expedition is summarized all that was known of the Botany of the Antarctic Region, and the relationships of the Australian and American Floras§. Since

* See Jardin and Jouan in Mém. Soc. Sc. Nat. Cherbourg, iv. 1856 and xi. 1865.

+ Flora Antarctica, 1844-47.

t In a recent communication from Dr. H. P. Guppy he suggests that the vegetation of these remote islands is due to the agency of birds. Admitting the probability of such a thing, it must have happened very long ago, and not altogether in the direction indicated by him, or how is the endemic element to be accounted for? See ‘ Nature,’ xxxviii. p. 40. § Introduction, pp. 50-65, and pt. 2, pp. 183-281.

liv INTRODUCTION.

this was done the Germans have botanically explored South Georgia *, where thirteen species of flowering plants were collected, nine of which are common to the eastern _ part of the Antarctic Flora, from Kerguelen to the islands south of New Zealand; four of them reach New Zealand itself, and one (Colobanthus subulatus) the Alps of Australia.

In Fuegia the beech-woods (consisting mainly of the deciduous-leaved Fagus antarctica and the evergreen F. betuloides) are a conspicuous feature; but all the islands are absolutely treeless, except the Auckland group, where there is an arboreous Myrtacea (Metrosideros lucida) and two or three other large shrubs or small trees. But the beech element in the southern hemisphere is one of the most interesting, and it is very fully described by Hooker +. In the northern hemisphere Fagus sylvatica inhabits Europe, Asia Minor, Northern Persia, and Japan, but is not known to occur in the intervening country; and the eastern North-American F. ferruginea is exceedingly near it, so near, Indeed, as to be regarded by some botanists as a variety. Besides these there is a Japanese species recently described by Maximowicz}, which strongly resembles F, sylvatica in foliage, though it is very different in the fruit. Japan, Northern Persia, North Italy, and Florida.are the southern limits of the genus Fagus in the northern hemisphere, where it is represented by at most three species. In the southern hemisphere, on the other hand, there are at least a dozen distinct species divided between South America, New Zealand, Tasmania, and the mountains of Victoria and N. S. Wales, with a maximum development in New Zealand and extra- tropical South-west America. In continental Australia the genus is represented by two isolated outlying endemic species, one occurring at the head of the Macleay river in about 31° of latitude, and the other on the Yarra-Yarra in about 37° 30’; and in America F. obliqua inhabits the Andes in as low a latitude as 33°. Between these stations and the northern ones indicated above there is no living trace of the genus 9. The foregoing particulars concerning these two widely separated northern and southern races of Fagus are given as another illustration of the intimate relationships existing between the northern and southern Floras, because the genus is so distinct and sharply defined that there can be no question about the generic identity of the two races, and because Fagus is the only genus of the characteristic northern Cupulifere that reaches high southern latitudes. Quercus reaches New Guinea in the east, and Popayan (about 2° 30’ N. lat.) in America. The allied Salicinese (Populus and Salix) also do not reach

* See Engler, Jahrbiicher, vii. p. 281, and ‘ Nature,’ xxxiv. p. 106.

+ Flora Antarctica, p. 345.

+ Mélanges Biologiques, xii. p. 542.

§ Fagus argentea and F. gavanica, enumerated in Steudel’s ‘Nomenclator Botanicus,’ attributed to Blume and recorded from Java, were probably manuscript names given by Blume to some sterile specimens of Castanopsis. He himself does not mention them in his ‘Cupulifere Javanica,’ nor does Miquel in his ‘Flora Indie Batave,’ and it is almost absolutely certain that no species of Fagus exists in Java.

INTRODUCTION. lv

cold southern latitudes, though one species of Salix is a native of South Africa, and another species extends about as far south as the northern limit of Fagus in the Andes, and further south in eastern 8. America.

Let us now briefly examine the composition of the very poor insular portion of the Antarctic Flora, excluding the Falkland, Campbell, and Auckland Islands. The islands or groups of islands more or less explored botanically are South Georgia, Marion (Prince Edward group), the Crozets, Kerguelen, Heard (Macdonald group), and Macquarie, lying between 38° W. and 160° E. longitude.

The accompanying table demonstrates the existence of an antarctic phanerogamic element all round the southern hemisphere. It will be perceived that the only island possessing endemic plants is Kerguelen, and these may possibly yet be discovered in the Crozets or Prince Edward Island, just as Pringlea antiscorbutica has been. ‘Two genera,

Flora of the Antarctic Islands from South Georgia in the west to Macquarie Island in the east.

Distribution in the Distribution beyond the islands. islands. g £ | Distribution of the Name. & . : . rd a genera. & _| & aI @)/sl/a|]e]4]4 : = 3 a aa ~ | ele) elel2i¢e/ 8/2) 2/2/28] 2 5 S/e/El Pl Ele E/P/SIE ES) 2/5 alse /ole |B alsa e/a, sa), | 44,3] 4 RANUNCULACER. Ranunculus acaulis ....| .. |... | -- | .. | .. | * .. | * * Wide. biternatus ........ * | * * * | * trullifolius........ * * * moseleyi ........ * CRUCIFERZ. Pringlea antiscorbutica ..| .. | * | * | * | * CARYOPHYLLEZ. Colobanthus muscoides ..) .. | .. | .. | -. | -. | ® | ee | ee | ee | YO ., |Andesand Australasia. kerguelensis ...... we | we foe | ® subulatus ........ we |... fo. | we | ee Lee | we | ® fee | ee | we] OY] OF erassifolius ...... x |. df... | ww | we fee Pee | #* LO Lyallia kerguelensis ....) .. | .. | -- | * PoRTULACER. . Montia fontana .......- * x |... | we |... dee | w® | me] ee | ee] OL O® x |Wide. RosacE.z. Acena buchanani ...... x |. | ue |e. | ee | ee | ® |e) | + jAndes, Polynesia, and Australasia. adscendens ...... * i * | * | #*# ].. | ® *x | * | * —— levigata.........- e |... fue |. foe fee | * |e] ® Carried forward ....| 6 4 | 2 8 1 4. 2 7 5 3; 3 | 5 2 1

lvi INTRODUCTION.

Flora of the Antarctic Islands, &c. (continued).

Distribution in the Distribution beyond the islands. islands. a g Distribution of the Name Po ; j 4 genera, 3 g ¢ $/sl/alEl a] #2 d/2#/8ilg]/5}a/]¢]/8 12/2} ¢8) 3] & S1S$i/8lelelsi4alea/S8/4/ 38 lalale2 Brought forward....| 6 | 4 | 2 | 8 1/4]/,2/]7;])5;38]38;,5 {24,1 CRASSULACEZ. Tillea moschata........ .. | * * * * | * | * .. |Wide. —— sinclairili ........ we foe doe | we fee | # Le . * HALORAGER. Callitriche verna........ x | «wf... | «law |... | &® | & | we | & | oe | & | c&® | hU& 6 (Wide. UMBELLIFERZ. Azorella selago ........ -- | *] *] *# | #] e] ee | RH]. Australia, Andes. lycopodioides...... wef ee foe | we | ee | w# Lee | we] ARALIACEA, Stilbocarpa polaris ...... we fee | we | we fee | ®t ee | we [ee | ow] we] RUBIACES. Coprosma repens ...... ee fae fee | ee foe | w]e. | ee | ee | we] wR | O® Polynesia, Australasia, Juan Fernandez. Galium antarcticum ....} .. | .. | # wx |... }.. | ee | we LO .. |Wide. CoMPosItz. Pleurophyllum criniferum| .. | .. | .. | .. |. |-# | ee | ee | ee | we | Cotula plumosa ........ re rr 2 2 Wide. ScROPHULARINES. Limosella aquatica......) .. | .. | .. | # |e. | ee | ® |] *® ] ® ] ee | ee | | ow | ® [Wide JUNCACER. Juncus scheuchzerioides..| .. | .. | .. | * | .. | .. | ¥® | * * | * | * | x : Wide. nove-zealandiz....| * | .. | .. | .. | we | ee Jee f we | ee | we de. * Luzula crinita.......... te | ee | we fee foe | fee | we fee | HL .. | .. |Wide. campestris........ we fee fee fee | ee | we ] ® | wR] ee Pee | we | we] we | Rostkovia magellanica ..| * | .. | .. |.) | -. | ee | *® | * | w]e | . |Fuegia, Auckland and Kermadec Islands. CYPERACER. Uncinia compacta ...... a a ee Andes to W. Indies, Sandwich Islands, GRAMINEZ. Australasia. Poa cookii ............ .. * | .. | * | * a Wide. foliosa .......... .. % oe * * * flabellata ........ oe ee rr * * Festuca erecta ........ * |... |... | ¥ .. * | * .. |e. | .. Wide. duriuscula........ rn rr sr re * * * * E kerguelensis ...... ~-- [oe | we | # Phleum alpinum ...... * i]... fe. fee fee | ee | ® |] OR Tee fee | ee] e. * |Wide. Aira antarctica ........ * | .. |... | #& | ee dee fee * * * Wide. Agrostis magellanica....) .. | .. | .. | *# Je. fee foe | YL ® * .. |Wide. Totals: 24 gen., 40 sp.}18/ 8 | 5 |21/ 4 115) 9 |22/16/12|)14)18| 7 | 6

INTRODUCTION. lvii

Pringlea and Lyallia, are confined to the islands under consideration ; two, Plewro- phyllum and Stilbocarpa, do not extend beyond the New-Zealand region; six are represented only in the American and the Australasian regions; fourteen are of wide, mostly of almost universal, distribution, and six of the species are of nearly equally wide range. The monotypic Pringlea antiscorbutica has no near ally in the southern hemisphere, but it is closely related to the northern genus Cochlearia, differing more in habit of growth than in floral structure. And Lyallia is of the same affinity as the Andine Pycnophyllum and the Mexican Cerdia.

As already stated, the Tristan da Cunha group and Amsterdam and St. Paul Islands can hardly be included in the antarctic region, unless we make it more extensive in New Zealand and in South America, because the bulk of the vegetation consists of Phylica nitida and Spartina arundinacea, types of a warmer region; the former, the only tree or even shrub larger than the trailing Empetrum, being a Mascarene species, and the latter a tall reed, whose nearest ally is a native of eastern temperate South America. Not one of the plants enumerated in the foregoing table is recorded from the Tristan da Cunha group, and only two, Ranunculus biternatus and Uneinia compacta, inhabit Amsterdam or St. Paul Island. On the other hand, several of the plants found in the Tristan and Amsterdam groups are common to New Zealand and South America. Of the twenty-nine flowering plants known to inhabit the Tristan group sixteen are apparently endemic; three are South-American and are not repre- sented eastward, while six extend eastward, three reaching New Zealand. Nineteen flowering plants are recorded from St. Paul and Amsterdam, eight of which have not been found elsewhere, and the distribution of the remainder is similar to that of the Tristan da Cunha non-endemic element. Numerically as to species, then, the compo- sition of the Tristan and Amsterdam Floras is that of the cold temperate region and very similar to that of the islands farther south; but several of these species are quite rare, and the conspicuous vegetation, apart from ferns, is almost wholly Phylica and Spartina, at least in the Tristan da Cunha group and Amsterdam Island.

CONCLUDING REMARKS.

The facts brought together in the preceding pages and in the ‘ Appendix ’* have an interest apart from any conclusions arrived at, and whether the views therein put forward on the botanical regions of the world meet with acceptance or not, it will be generally conceded that although the broad features of the distribution of plants and animals are essentially the same, they are by no means identical.

When it is considered how much more potent and diversified are the means of

* It may be well to repeat that this is a review of the results of comparatively recent investigations rather than an attempt at an exhaustive discussion of the subject.

BIOL. CENTR.-AMER., Bot. Vol. I., October 1888. h

Lviii INTRODUCTION.

dispersal of plants than are those of animals, the divergences in distribution are no greater than might be expected.

Beyond this, many natural orders most widely separated from each other in floral structure repeat the same modifications of their vegetative organs under similar circum- stances, and possess equal capabilities for adaptation to external conditions.

The present distribution of orders, genera, and species, and all that is known of the past, seems to point to one original centre of creation and development, and to such physical conditions at different periods as permitted, or even favoured, the general spread of all the principal types of plants and the wide migrations of many of the same ~ forms or species. Assuming this to have been so, any system of regional division is arbitrary, and only useful in proportion to its agreement with the present distribution of plants, inasmuch as it is merely a foundation on which to build a knowledge ‘of botanical geography, not a representation or classification of the facts. An exami- nation of the extensions and isolated remains of extensions of the types characteristic of each region throws more light on the subject than can otherwise be obtained.

Professor Huxley seems to have been so convinced of this in the Animal Kingdom that he even goes so far as to say that he thinks it would not be difficult to show that the whole surface of the globe should be primarily divided into a northern and a southern region in order to display best the geographical distribution of Birds and Mammals*. And he further remarks, in connection with his proposed four primary zoological regionsf (1, Arctogea; 2, Austro-Columbia; 3, Australasia; 4, New Zealand), that the three latter are in some respects less unlike one another than they are unlike the first. The same might be said of the plant-regions, taking tropical and South Africa and the Indian region out of Huxley’s Arctogea.

Even the most highly specialized Floras exhibit merely a local development of species, of genera, or groups of genera, belonging to orders of universal dispersion, often differing more strikingly in their vegetative organs (roots, stems, and leaves) than in their repro- ductive organs (flowers and fruits) from the usual character of the order. The nature of the medium in which plants grow, combined with the climatal conditions, determine the character the development assumes, and similar phenomena in development are repeated in widely-sundered areas, where the prevailing physical conditions are the same or similar. Familiar examples of this kind of parallelism are offered by the Cactacez of Mexico and certain African species of the genus Euphorbia, some of which so strongly resemble the branching Cerei, some the spheroidal Melocacti, as to deceive any but the most experienced eye. Such American genera as Yucca, Agave, and Dasylirion are replaced in Africa by Aloe, and in Australia by Xanthorrhwa. Some of the species of a genus develop tuberous roots, like many of the Australian Sundews, and some thick fleshy stems, as South-African species of Pelargonium and Vitis, thus adapting themselves to

* Proc. Zool. Soc. Lond. 1868, p. 313. T See ante, p. Xxxiv.

INTRODUCTION. lix

local conditions. Some of the species of a genus having formally flat leaves are remarkable for their great likeness to a cypress, a lycopod, or a Salicornia. Develop- ment in this direction is exemplified in some of the New-Zealand species of Veronica.

Another remarkable illustration of the almost universal dispersion of northern types is offered by the composition of the bulk of the vegetation of the central desert-region of Australia, where the Chenopodiacee number upwards of a hundred species, including thirty of Atriplex, fifteen of Kochia, twelve of Chenopodium, and seven of Salicornia, and associated with these are the European Sueda maritima and Salsola kali, and several endemic genera, mostly of one or very few species, except Rhagodia (12 species), which differs chiefly from Chenopodium in having a fleshy fruit.

It has been shown (pp. xix-xxii) how few natural orders are unrepresented in any one of the large areas under consideration; and ninety-five or nearly half of the natural orders are represented in the Sandwich Islands. So far as we know, the African region is poorer than either the South-American or the Indian region, but further explorations may reveal the existence of several of the natural orders hitherto undiscovered. Those of relatively restricted areas are almost all small, and consist of one or few genera and few species, as may be seen on referring to the table on page xi; and most of them are more definitely characterized than some of the larger, generally-dispersed orders. Indeed, it is the absence, through destruction, dying out, or some other cause, of connecting links, that gives some at least of these small groups the status of natural orders.

Turning to the groups intermediate in rank between orders and genera, a consider- able number of which are regarded by some botanists as distinct natural orders, we can better appreciate and estimate the amount of differentiation in development in different areas. As an illustration, a few of the more important in the thalamifloral Polypetale are noted. The Fouquieriese, Fremontiew, Limnanthee, Clusiee, Rhizobolez, Marc- graviee, Malpighiex, Gaudichaudiee, Cuspariee, and Luxembergiew are examples of distinct tribes or suborders restricted to America; and the Dilleniew, Dombeyee, Aurantiexw, and Phytocrenee are peculiar to the Old World. On the other hand, the Zanthoxylee is one out of many tribes that are generally spread in warm countries. The Lardizabalex are divided between Peru and Chili and Eastern Asia, from North India to Japan; the Hermanniee are African, with three or four representatives in Mexico, and the Colletieze are Andine and Australasian.

Similar illustrations of the distribution of plants might be indefinitely multiplied. It is clear, as Sir Joseph Hooker suggests, that a classification of plants by Linneus or Jussieu would have been essentially the same had it been based entirely upon Chinese, Australian, South-American, or Mexican instead of mainly upon European plants.

Whether the plants (and animals) of the earth had a common northern origin, as supposed by some writers, and the highly differentiated southern forms are descendants

of northern ancestors which have undergone their great differentiation in the south, are h 2

Ix INTRODUCTION,

problems difficult of solution. As regards a very large number of northern types, they can be distinctly traced southward. Many exist only in isolated localities, as they have been driven out by climatal conditions, while others present a nearly continuous chain; but, speaking generally, they gradually decrease in volume from north to south. The theory that the Proteacee, Eucalypti, and other southern types inhabited Europe in early Tertiary times is far from being established on satisfactory data, and all the indisputable evidence points to a northward migration of these types and a southern origin. If it can be proved that the prototypes of Eucalyptus, Proteaceee, &c. existed in Europe, and that the northern outliers of these types are survivals of stragglers of the southward migration, it follows that differentiation equal to the greatest differentiation in the whole world took place in the northern hemi- sphere, and that there has been comparatively little beyond specific differentiation in the southern. This may beso, and the extraordinary development of the genera Erica, Mesembryanthemum, and Pelargonium in South Africa and of Ranunculus, Epilodium, and Veronica in New Zealand might be cited in support of the argument.

In conclusion, a few additional remarks on the adoption of a few large primary regions. With regard to treating the north temperate and arctic countries as one primary region, it may be contended that, although there is a large number of genera common to Europe and western North America, to give an extreme illustration*, there is also a larger number not represented in the two countries, including genera numerous in species, such as Medicago, Cousinia, and Acantholimon on the one hand, and Dalea, Gilia, and Pentstemon on the other. This is true; but are the differences greater than between Ceylon and Borneo or between the latter and New Guinea, or between South- west Australia and South-east Australia? Among the numerous genera peculiar to West Australia are such prominent ones as Kingia, Dasypogon, Anigozanthos, Conostylis, Andersonia, Dryandra, Pileanthus, Verticordia, Tremandra, Platytheca, and Chorilena. Among those common to the two are Kucalyptus, Persoonia, Hakea, Grevillea, Dampiera, Leschenaultia, Myoporinee, and Xanthorrhea. Further, upwards of eighty per cent. of the species of vascular plants are endemic. ‘The absence of a large number of orders and suborders represented in Eastern Australia is also remarkable, and on a par with the poverty of the European Flora as contrasted with the Chino-Japanese or the North- American. Nevertheless few persons would refer these two areas to different primary botanical regions.

* As already urged, it is a comparison of the vegetation of Eastern Asia and Eastern North America that reveals the most striking similarities; the affinities between these two regions being much stronger than those existing between the vegetation of Europe and Eastern Asia.

t Sir Joseph Hooker sufficiently indicates in his essay on the Australian Flora the striking characteristics of, and the more striking diversities between, the vegetation of Eastern and Western Australia; but Dr. Engler’s tabular view and partial analysis of the Flora of the whole of Australia (Versuch, ii, pp. 12-54) brings out

‘these peculiarities much more prominently.

INTRODUCTION. Ixi

It is usual to rank the South-African Flora as a primary region, but here it is made a subregion of a region comprising tropical and South Africa and the Mascarene Islands, because the mountain flora of Madagascar and of tropical Africa, especially eastern, is largely composed of South-African types, a fact that is in a manner suppressed in treating them as separate regions. The same arguments apply with regard to South America.

In a cartographical representation of the floras of the World it seems at first easy enough to indicate the centres and extensions, or isolated mountain-areas, of each, but the practical difficulties are insurmountable. Yet it would not be difficult to show, on so many maps of the world, the areas and extensions into other regions of the five regions proposed. There would also be space for indicating the density and limits of some of the leading types. This should not be done, however, in the sense of denominating a vast area as the “ province of the tea-tree,” the “ province of the cedar,” and so on, which conveys an altogether false impression ; yet unfortunately this method — has been continued in the latest Atlas of Plant Distribution.

Sir Joseph Hooker, who has kindly read the proof-sheets of the foregoing pages and suggested many improvements, and who has consented to write a brief commentary on this ‘Introduction,’ does not share some of the opinions expressed by the writer, who himself might modify them for a primary botanical division of the world based on climatal conditions alone. To a great extent the facts speak for themselves, yet it is doubtful whether any two persons would deal with them in exactly the same manner.

It is perhaps unnecessary to apologize for the occasional repetition of a fact or an argument, or for slight discrepancies in the figures cited in different places, as all persons engaged or interested in similar work can appreciate the difficulties of the task, and will excuse imperfections of such a nature.

A COMMENTARY

ON

Mr. HEMSLEY’S INTRODUCTION AND APPENDIX

TO THE

BOTANICAL PART OF THE ‘BIOLOGIA CENTRALI-AMERICANA.’ |

By Sir J. D. HOOKER, Late Director oF THE RoyaL GARDENS, Kew.

Havine been deeply interested in the progress of Mr. Hemsley’s work on the Botanical portion of the ‘ Biologia Centrali-Americana,’ and especially in that author’s elaboration of the important Appendix relating to Botanical Geography, which is contained in the fourth volume, I have been requested by Messrs. Salvin and Godman to contribute to the pages of their ‘ Biologia’ some observations of my own relating to the scope and character of the work accomplished by Mr. Hemsley, especially as regards its value as advancing knowledge of the geographical distribution of plants.

At the same time, I have been asked to add, to any comments of my own upon the contents of the Appendix, my conceptions of the principles which should guide the _ botanical geographer in the limitation of the primary botanical regions of the globe— conceptions which have been tardily matured, during botanical visits to many countries and a long familiarity with the dominant features of their Floras, obtained both on the spot and from a study of large herbaria.

No country of equal area presents a richer or more varied vegetation than Mexico. Except perhaps the Javan and Indian, no tropical Flora of great extent is so well explored and so fully represented by collections; and none has been subjected to so searching an analysis, in respect of the correlation of its botanical features and the defi- nition of its botanical regions, as has this Flora under Mr. Hemsley’s judicious, pains- taking, and accurate methods of study. My view of the merit and completeness of this part of the work urged me to suggest to the authors of the ‘ Biologia’ the desira-

COMMENTARY ON THE INTRODUCTION AND APPENDIX. Ixiii

bility of Mr. Hemsley’s instituting a closer comparison between the Mexican Flora and that of some tropical region in the Old World presenting as rich and varied a vege- tation, and I indicated the British-Indian Flora as especially suitable, both on this account, and because the ‘Flora of British India’ was sufficiently advanced in respect of available published and unpublished materials to supply accurate data for such a comparison. And further it appeared to me that by availing himself, together with these materials, of the geographical data appended to every genus of phanerogamic plants contained in the recently-concluded ‘ Genera Plantarum,’ Mr, Hemsley might very greatly advance that most instructive branch of phytogeography which originated independently and coincidently in the minds of Humboldt and Brown, and to which the former gave the name of Arithmetice botanices.

Messrs. Godman and Salvin cordially responded to my suggestion, and I feel sure. that the results embodied in the “ Statistics of the Phanerogamic Flora of the World” (Introduction, pp. ix—lxi) will be received with gratitude by all botanists as a very valuable supplement to a work that owes its existence to those naturalists’ travels, collections, learned labours, and munificence.

The tables at p. xv and following of the Introduction are particularly valuable, and give information previously unattainable. ‘The areas compared are approximately within the same latitudes, 9° N. and 83°N., but separated by nearly 180° of longitude, the Asiatic in 70° to 95° E., the American 80° to 115° W. Each presents a hot, moist tropical, a temperate, and a frigid climate. It is impossible to find, in the Old and New Worlds respectively, two areas more similar as to physical features, or in which the vegetation of their respective continents is more fully represented; and yet the comparison of their Floras shows that, with an almost total diversity of species, genera, and of many natural orders, the proportion of monocotyledonous to dicotyledonous plants is nearly the same in each; that the number of natural orders is only 12 fewer in Mexico; that the number of species in each differs by only 2000 (11,626 in Mexico, 13,647 in India); that the average number of genera in each order is nearly the same in each (11 in Mexico and 13 in India); that the average number of species in each genus even more nearly coincides (6:4 in Mexico and 6-0 in India) ; and, more singular still, that the percentage of endemic species in each differs by only 2 per cent.

It is instructive to observe that these marked resemblances in proportions do not arise out of a resemblance in the elements from which they are derived; for, turning to the natural orders that contribute largest to the Flora of each area, they are very differently represented as to number of species in each. Composite, which take the first place in the Flora of the globe and of Mexico, are reduced to the sixth place in India. Leguminosee, which are second to Composite alone, are second in both Mexico and India; but Orchidew, which hold the third place in the world and in Mexico, are first in India; Rubiacew, the fourth in the world, are the seventh in Mexico and fifth in India; grasses are fifth in the world and in Mexico, but only third in India.

lxiv | COMMENTARY ON THE

Descending in the systematic scale to the lowest term of the series, the differences between the elements of the two Floras become greater and greater, until genera are reached; thus, as Mr. Hemsley shows, only 25 to 26 per cent. of these are common to the two regions. As yet data do not suffice to ascertain the exact number of species common to India and Mexico, but it may not exceed 600 of the 25,273 which is approximately the sum of the species of both Foras.

It is not my purpose to discuss the nature or origin of the unexpected likenesses and expected unlikenesses that Mr. Hemsley has shown to exist between the Floras of Mexico and India: of these the former are due to causes which have influenced vege- tation as a whole; the latter to more or less local causes. As an illustration of what I mean, I would say that the conditions which have resulted in monocotyledons main- taining their numerical position of one to four or thereabouts of dicotyledons in the globe and in all large areas thereof are, in the present state of science, inscrutable ; but that the comparatively low number of Compositze in India is explicable by the intrusion into India of the Malayan Flora, which is abnormally deficient in Composite. Not that this offers any real solution of this latter phenomenon, which lies much deeper. It must be shown whether the intrusive Malayan Flora found in India a Flora already deficient in Composite, or whether it prevailed over and displaced the pre-existing native Composite ; and it must also be shown why the Malayan Flora is deficient in this ubiquitously dominant element of all other floras, whether tropical, temperate, or frigid, insular or continental, humid or arid.

It would be very interesting to know whether any of the larger divisions of the animal kingdom present phenomena comparable with those derived from large remote botanical areas. It may be supposed that the great unconformity that exists between the geographical regions of plants and those of animals, as traced out by the most competent of zoologists, and which unconformity is so strongly, and as I think rightly, insisted upon by Mr. Hemsley, is opposed to such a parallelism existing ; but I do net see the force of this objection if, as I think, the problems presented by these “ Arith- metics ” are deeper than those of regional distribution.

Another point of resemblance between the Floras of Mexico and India is that each is botanically as well as geographically a n@ud. The lofty mountains in each are con- tinuations from more temperate latitudes in the north, favouring an immigration of temperate species which have retained their characters in the higher elevations and become modified or been extinguished in the lower. In each Rosacew, oaks, and Coniferee of northern forms descend into the tropics, even to 3000 feet elevation, whilst palms ascend to 8000; and in each epiphytic orchids abound, ascending in cool tem- perate regions to 8000 feet and upwards.

Perhaps the most striking case of parallelism in the Floras of tropical America and Asia, as prominently put forward by Mr. Hemsley, is that in each so many temperate types, especially oaks and Pinus proper, are continued far into the tropics, but cease at

INTRODUCTION AND APPENDIX. lxv

or near the equator. If, as has with much probability been surmised, the existing types of vegetation originated in northern latitudes and by migration southwards, and by differentiation and dispersion, peopled the southern hemisphere, there is no obvious obstacle to the prolongation of these two particular types (and many others) into South America, New Guinea, and Australia. But there is this difference between the oaks of the tropics of the Old World and those of the New, that the Asiatic in their prolongation southward to Malacca and thence eastward to Borneo in descending to the sea-level assume different types of structure from their northern allies, forming distinctly tropical sections of the genus, as Cyclobalanus, Chlamidobalanus, Lepidobalanus, and Castanopsis. On the other hand, the tropical-American oaks where they descend to the sea-level do not assume a character sectionally differing from temperate-American species. It is startling to have to regard the genus Quercus as tropical rather than temperate; but so it is, and especially in India, where about 70 species are purely tropical and only 12 purely temperate.

Turning now to the concluding pages of Mr. Hemsley’s Introduction (pp. xxxix-Ixi), which are devoted to an exposition of the primary botanical regions of the globe, I find that these regions are based upon far more complete and reliable data than had previously been available. In the last paragraph but one Mr. Hemsley remarks that I do not share some of the opinions which he has expressed in his previous pages. This remark must not be taken as conveying the impression that I dispute either his facts or methods. I shall now, in preference to discussing either, give in outline my own idea of the principal botanical regions of the globe, from which it will be seen how far I differ from him in the limitation of the primary Floras of the globe.

I limit the primary botanical regions of the globe to two, the Tropical and the Temperate ; these are distinguished by both climatic and botanical features—tropical- country plants will not, as a rule, bear a temperate climate, nor temperate-country plants a tropical climate. Botanically the two regions are distinguished by the restric- tion of certain natural orders to one or the other, and the prevalence of others in one or the other. The geographical limit between these Floras in either hemisphere varies with every few degrees of longitude, being affected by elevation of surface and local climatal conditions. I do not distinguish the northern and southern Frigid Floras as primary regions separate from the Temperate, for these regions contain no genera and very few species different from the Temperate, and the geographical limits of any group of species that may be selected as inhabiting the coldest region of the globe are unde- finable by latitude or by isothermal or isotheral lines.

If a distinctive name is desirable for the two primary regions, I would suggest that of Botanical Empires.

The regions next in importance to the two primary are in my view seven,—two north temperate, of the Eastern or Old and Western or New World respectively ; two tropical, corresponding to the above; and three south temperate (America, Africa, and

BIOL. CENTR.-AMER., Bot. Vol. I., October 1888. i

Ixvi COMMENTARY ON THE

Australia). Ifthe term Empire is applicable to the primary divisions, that of Kingdoms may be accepted for thesecondary. I have a few observations to offer upon these seven kingdoms :— | |

I. The north temperate Kingdom of the Old World.—This coincides with the Palearctic zoological region of Dr. Sclater, with the exception that it includes Greenland, the Flora of which, as I have elsewhere shown, is more European than American. In detail of circumscription,it, with this exception, follows exactly the sinuous course traced for it by Wallace in the third chapter of ‘Island Life.’

In attaching greater importance than do Mr. Hemsley and other botanists to the difference between north temperate Floras of the Old and New Worlds, I am influenced by a somewhat close study of the elements of each, together with impressions gathered from journeys in both continents. I am aware that such impressions are often quite untrustworthy, and must be so if not supported by adequate data derived from a com- parison and contrast of the plants constituting the Floras of the two areas.

The chief arguments for the uniting these kingdoms into one are the prevalence of coniferous and cupuliferous and other trees of the same genera in both, that a consi- derable number of other genera are common to both, and that the Floras of their higher latitudes are practically one. But it would not be difficult to give examples of similar tropical features which prevail over the continents of the Old and New Worlds, and which might with much reason be adduced in favour of uniting them into one region, reinforced as the argument would be by the great number of genera (more by far than is usually supposed) that are common to the tropics of the Old and New Worlds. Having regard to the composition of the Flora of the two northern temperate regions, if the Floras of their middle and southern temperate latitudes are taken into account, the botanical differences between them appear to me far to outweigh the resemblances. Of the forest-trees scarcely one is conspecific with an Old-World one. Of the 10,000 (or thereabouts) known temperate North-American flowering plants not more than 700 are common to Europe, and these include upwards of 150 water and marsh plants of more or less mundane distribution; and there are nearly fifty natural orders in tempe- rate America that are wanting in Europe. Again, taking Boissier’s ‘ Flora Orientalis’ as a standard of comparison with that of temperate North America, I find that of 1100 oriental genera less than 400 are American, and of about 9500 species less than 350 are American. In Gray’s ‘ Flora of the Eastern United States,’ of nearly 2300 species not 370 are European. In Coulter’s ‘ Flora of the Rocky-Mountain Region,’ out of about 1800 species about 300 alone are European, and of the Californian Flora of 2700 species there are not 230 European.

But these statistical data, important as they are, are far less so than what is supplied by the genera and species themselves. The generic, tribal, and specific differences between the Liliaces, Labiate, Scrophularinee, Umbellifere, Onagracee, Leguminose,

INTRODUCTION AND APPENDIX. xvii

and Malvacee of America and Europe are no less striking than is the comparative rarity in the former country of Cruciferze, Caryophyllee, Geraniacee, Carduacee, Campanulacee, Lobeliaceze, Primulaces, and Orchidex, and the prevalence of Vitacex, Anacardiacee, Crategi, Asclepiadeee, Polemoniacer, Nyctaginex, and Cyperacez.

Turning now to the Floras of the Asiatic and American continents, it is difficult to say which is the most striking phenomenon, the wonderful identity of certain isolated genera and species of the eastern shores and islands of the Old World with those of the eastern side of the New World (and which has been so admirably worked out and explained by Asa Gray), or the total dissimilarity of the Asiatic and American temperate Floras in other respects. I may select the Japan group in illustration of both phenomena, because geographically it is perfectly well suited for a comparison with the Pacific-coast Flora of America, and because it is the head-quarters in Asia of the representative genera and species of the Eastern American Flora. The Japanese Flora contains about 200 species common to North America, but nearly three fourths of these are species found all round the globe in the north temperate regions; the remainder are chiefly the Eastern American genera and species alluded to above. Of the North-American Flora proper there is not a trace in Japan; there is not one of its multitude (nearly 150) of peculiar genera to be found in Japan of Cruciferae, Capparidee, Papaveracee, Rosaceze, Saxifragee, Onagracee, Composite, Polemo- niaceee, Hydrophyllaceee, Scrophularinee, Nyctagineee, Polygonez, and Liliacex, nor are there any endemic representatives of them. Further, the numerical proportions of the Japanese orders are European and Asiatic, not American. Leguminose, of which there are only 19 genera in California, is represented by 41 in Japan; though. the total number of phanerogamous genera is 879 in California and only 839 in Japan. Of Orchidee there are only 10 genera and 22 species in California against 34 genera and 67 species in Japan*, and the contrast might be carried much further by taking many other natural families. In short, the differences between the Palearctic and Nearctic botanical areas are so many and various that I have no hesitation in regarding them as two botanical kingdoms.

Il. The tropical Kingdoms of the Old and New World.—<An analysis of the Floras of the Old and New World as a whole shows, as was to be anticipated, that the Old World Flora is by far the richest ; it contains probably 6000 known genera of Flowering Plants, and the New World nearly 4000, there being only about 1200 genera common to both. It is a singular fact that the ratios of Monocotyledons to Dicotyledons is the same for the endemic genera of the Old World (1°48), for the endemic genera of the New World (1:46), and for the genera common to both (1°47). How far this holds

* The Californian genera of Orchids are, with one exception, European, and most of them Asiatic, which

renders the absence of so many in Japan very anomalous.

Ixviii COMMENTARY ON THE INTRODUCTION AND APPENDIX.

good for the species I have not attempted to ascertain, nor have I attempted the sepa- ration of the tropical genera of either world for the totals above given.

I regard the tropical African Flora as a subdivision of that of the Old World, because I find no other essential difference between the Asiatic and African vegetations taken as

wholes than the poverty of the latter, and because the peculiar botanical features of large tracts of Asia are repeated in Africa. Thus the Punjab, Scinde, and S. Persian Flora is largely represented all over north tropical Africa, extending to the Cape de Verde Islands; the notable absence in the Deccan peninsula of India of Cupulifere and the extreme rarity of Conifere are conspicuous characters of all tropical Africa; and the Indo-Malayan Flora has- its representatives in Madagascar, and measurably on the coasts of the African mainland. If any part of the tropical Old World could be sepa- rated as a primary region that should rank as a kingdom, it would be New Guinea.

III. The three Southern temperate regions | (Extratropical America, Africa, and Australia).—I cannot accept the merging the South-African Flora into the tropical African. Of the six well-defined botanical provinces of South Africa so ably established and limited by Mr. Bolus, not one is represented anywhere in tropical Africa, where there is no region of heaths, of Composite, of Crassulaceew, of Campanulacee, of Prote- ace, and of Restiacese, and where such few representative species of these orders as do occur are either confined to mountain-regions or are isolated amongst the prevalent Indian types of vegetation. On the other hand, the genuinely tropical types of Africa are few and scattered in its south temperate regions, where the Anonacee, Menisper- macee, Guttiferee, Rubiacee, Acanthacee, &c. are hardly even represented, and only locally. It is true that Mr. Bolus designates the western seaboard of South Africa as the tropical region, mainly because a palm there extends to 33° 30! lat. south; but as in New Zealand, the Himalayas, and at Gibraltar palms enter the middle temperate zone, their presence does not necessarily imply a tropical heat; and as the plants of the so-called tropical South-African region require no greater heat than that of an English conservatory, I cannot regard them as typical of a tropical Flora.

In the above sketch I have taken no account of exceptional Floras like those of St. Helena and the Sandwich Islands, whose relationships must be determined by a study of the flowering plants they contain. Nor have I taken into account theoretical con- siderations of any kind. |

With regard to exact geographical limitations of any of these seven botanical areas, such are possible only where geographical features present insuperable obstacles to the further spread of the plants that characterize them. Where two are conterminous, there is always a neutral ground, often a very broad one, and this neutral ground may itself present a Flora which may be regarded as either tropical or temperate.

BIOLOGIA CENTRALI-AMERICANA.

BOTANICA.

PHANEROGAMIA.

DICOTYLEDONES.

POLYPETALZ. Series I. THALAMIFLORZ.

Order I. RANUNCULACE.

Ranunculacee, Benth. et Hook. Gen. Plant. i. p. 1.

Genera thirty, species about 540, dispersed over the whole world, but within the tropics almost confined to the mountains. With the exception of the genus Clematis, all the members of this order are herbaceous.

Tribe CLEMATIDEA. 1. CLEMATIS. Clematis, Linn. Gen. Plant. n. 696; Benth. et Hook. Gen. Plant. i. p. 3.

About 100 species, dispersed over nearly all temperate regions, a few occurring in the tropics. Chiefly woody climbers.

1. Clematis acapulcensis, Hook. et Arn. Bot. Beech. Voy. p. 410. Sour Mexico (Parkinson), Acapulco (Sinclair, Beechey). Hb. Kew.

9. Clematis americana, Mill. Dict. n. 14; DC. Prodr. i. p. 6. Sour Mexico, Campeche (Houston, Shakespear ).—EQUATORIAL AMERICA. Hb. Mus. Brit. Grisebach (Fl. Brit. W. Ind. p. 1) refers both of these to C. dioica, L.

3. Clematis caripensis, H. B. K. Nov. Gen. et Sp. v. p. 36. Clematis caracasana, DC.

Sour Mexico, in thickets near Jalapa and below San Salvador, on the road leading to Jalapa (Schiede), Yucatan and Tabasco (Johnson, 89); Nicaragua, Chontales (Tate) ; Panama, Volcan of Chiriqui, Veraguas (Seemann).—CouomBIA ; ‘TRINIDAD. Hb. Kew.

BIOL. CENT.-AMER., Bot. Vol. 1, Sept. 1879. . b

2 RANUNCULACEA.

4, Clematis dioica, Linn. Ameen. v. p. 398; Sloane, Hist. Jam. i. p. 199, t. 128. fig. 1. South Mexico, region of Orizaba (Bourgeau, 3280), Zimapan (Coulter), 639, Aca- pulco (Lay & Collie).—Braziu, States oF CotomsBia, and the West Inpizs. Hb. Kew. Seemann (Bot. Voy. ‘ Herald,’ p. 267) refers C. caripensis, H. B. K., and C. sericea, H. B. K., hither.

5. Clematis drummondii, Torr. & Gray, Fl. N. Am. i. p. 9. Clematis caudata, Hook. Texas and New Mexico.—Norra Mexico, Chihuahua (Potts), Monterey, Nuevo Leon (Platz), Tamaulipas (Berlandier), Sonora Alta (Coulter, 634 and 637), region of San

Luis Potosi, 6000 to 8000 feet (Parry & Palmer, 2). Hb. Kew. Dr. Seemann refers C. nervata to this species.

6. Clematis filifera, Benth. Pl. Hartw. p. 285. SoutH Mexico, near Leon (Hartweq), Zimapan (Coulter, 642). Hb. Kew. Var. incisa, Hemsl., foliorum segmentis ultimis trifidis vel trisectis. Region of San Luis Potosi, 6000 to 8000 feet (Parry & Palmer, 1). Dr. Seemann unites this with C. reticulata, Walt.

7. Clematis flammulastrum, Griseb. Pl. Wright. Cub. p. 153. SourH Mexico, ruins of Uxmal, Yucatan (Schott, 711)—Cusa. Hb. Mus. Brit.

8. Clematis grahami, Benth. Pl. Hartw. p. 5; Flore des Serres, t. 376. , Sovuta Mexico (Graham, 1830). Hb. Kew. In Hb. Kew. is the following note respecting this species in Mr. Bentham’s hand- writing :—“ A C. caripenst non nisi foliis pubescentibus differt.”

9. Clematis grossa, Benth. Pl. Hartw. p. 33. SoutH Mexico, San Bartolo in the Barranca (Hartweg, 266), Zimapan (Coulter, 699), Orizaba (Bilimek), Mexico (Parkinson). Hb. Kew. This may be a variety of C. sericea.

10. Clematis mociniana, Don, Gen. Syst. i. p. 6. ‘MEXICo.

11. Clematis nervata, Benth. Pl. Hartw. p. 5. Nort Mexico, Zacatecas (Coulter); Sour Mexico, Aguas Calientes (Hartweg, 2) Zimapan (Coulter), Cordillera of Oaxaca (Galeotti). Hb. Kew.

12. Clematis polycephala, Bert. Fl. Guat. p. 24. GuatemaLa, Volcan de Agua (Velasquez).

13. Clematis pauciflora, Nutt. in Torr. & Gray, Fl. N. Am. i. p. 9. CaLirornia and. New Mexico to—Nortu Mexico, Sonora (FI. California).

RANUNCULACE. 3

14. Clematis pubescens, Benth. Pl. Hartw. p. 5. South Mexico, Guanaxuato (Hartweg, 3). Hb. Kew.

15. Clematis reticulata, Walt. Car. p. 156. (C. pitcheri, Torr. & Gray.) Arxansas and Texas to—NortrH Mexico, valley of the Cibolo, Chihuahua (Bigelow).

16, Clematis sericea, H. B. K. Nov. Gen. et Sp. v. p. 37.

Sovran Mexico, valley of Cordova and region of Orizaba (Bourgeau, 1586), Vera Cruz (Galeotti, 4569), Tenancingo, Toluca, 6000 feet (Heller); GuatTEMaLa, San José (S. Hayes, 453); Nicaracua, Chontales (Seemann).—West Inpiss, Perv, and CoLoMBIA. Hb. Kew.

17. Clematis, sp. (aff. C. grahami, Benth.). GuATEMALA, in hedges (Bernoulli, 162, Skinner). Hb. Kew.

18. Clematis, sp. Norta Mzxico, Monterey, Nuevo Leon (Edwards and Eaton). Hb. Kew.

19. Clematis, n. sp.? Sour Mexico, Vera Cruz to Orizaba (F. Miller), Mirador, Vera Cruz (Linden, 965). Hb. Kew. :

20. Clematis, sp. (an C. sericew var.?). Guatema.a, Llano near Duefias, common (Salvin & Godman, 264). Hb. Kew.

Tribe ANEMONE.

2. THALICTRUM.

Thalictrum, Linn. Gen. Plant. n. 697; Benth. et Hook. Gen. Plant. i. p. 4. About fifty species, perennial herbs, natives of the temperate and frigid zones of the

northern hemisphere, a few growing in the mountains of Tropical India, South Africa, and the Andes of South America.

1. Thalictrum densiflorum, H. B. K. Nov. Gen. et Sp. v. p. 39, t. 426. Sourn Mexico, near Moran, 8000 feet (Humboldt & Bonpland).

2. Thalictrum fendleri, Engelm. in Pl. Fendl. p. 5. New Mexico and Catirornra to—NortH Mexico, Sierra del Pajarito, Sonora (Schott). Hb. Kew.

3. Thalictrum hernandezii, Tausch in Presl, Reliq. Henk. ii. p. 69. SoutH Muzxico, on the western side (Henke), Toluca, 8200 feet (Heller).

4. Thalictrum lanatum, Lecoyer, in Bull. Soc. Bot. Belg. XVi. p. 226.

Sourn Mexico, Cordillera of Talea, Oaxaca, 3000 feet ( Galeotti, 4575). 62

4 RANUNCULACEA.

5. Thalictrum longistylum, DC. Syst. i. p.171; Deless. Ic. Sel. i. t. 7% South Mexico, Chiapas (Ghiesbreght), Oaxaca (Ghiesbreght), Orizaba (Bourgeau, 2726),—and southward to Peru. Hb. Kew.

6. Thalictrum mexicanum, DC. Syst. i. p. 187. Mexico (Hernandez).

7. Thalictrum peltatum, DC. Prod. i. p. 11. Mexico (De Bergher, 3).

8. Thalictrum pubigerum, Benth. Pl. Hartw. p. 285. Soutn Mexico, on the banks of rivers, Lagos (Hartweg, 1591). Hb. Kew.

9. Thalictrum rutidocarpum, DC. Syst. i. p. 172. Sourn Mexico (Sallé, Graham), Orizaba (Botteri, 824). —Mountains of Tropica. South America. Hb. Kew.

10. Thalictrum strigillosum, Hemsley, Diag. Pl. Nov. pars 1, p. 1.

Caule ramisque plus minusve strigillosis, foliolis graciliter petiolulatis subtus hispidis supra sparse puberulis, floribus longe pedicellatis, antheris insigniter aristatis, stylo pubescente longissime exserto, acheeniis fere sessilibus rugis elevatis reticulatis notatis. _

Herba 2-3-pedalis, ramosa, strigilloso-hispida. Radix subcarnosa, fibrosa. Caulis sulcato-striatus,

. strigillosus vel demum fere glaber. Folia (radicalia non vidi) petiolata, subquadri-ternatim divisa, foliolis petiolulatis, cordato-rotundatis vel basi cuneatis, apice 3-5-lobatis, ad 8-12 lineas latis, pracipue subtus secus nervos hispidis. Flores longe pedicellati, pedicellis basi bracteis parvis munitis ; sepala glabra, ovato-oblonga, ad 3 lin. longa; stamina calycem supe- rantia, filamentis capillaceis, antheris aristatis filamentis equilongis. Achenia circiter 7, sessilia vel brevissime stipitata, oblique lunata, 2-3 lin. longa, insigniter costato-reticulata.

SoutH Mexico, rare in ravines among bushes (Schaffner), Tizapan, valley of Mexico (Bourgeau, 276), Zimapan (Coulter, 652), mountains around Mitla, Oaxaca (Andrieus, 546), between San Miguel and La Joya (Schiede). Hb. Kew.

This differs from its allies 7. rutidocarpum, longistylum, &c. in being more or less clothed with hispid hairs, which on the stem are long and on the leaves short, and in the leaflets, anthers, and achenes. Perhaps it will eventually be found necessary to unite several of these forms as one variable species.

11. Thalictrum wrightii, A. Gray, Pl. Wright. ii. p. 7. Norra Mexico, mountain ravine at Santa Cruz, Sonora (Wright). Hb. Kew.

12. Thalictrum, sp. South Mexico, between Chalco and Gonacatepec (Andrieux, 545). Hb. Kew.

13. Thalictrum, sp. Souta Mexico, between San Miguel and La Joya (Schiede). Hb. Kew.

RANUNCULACES, 5

14. Thalictrum, sp. SourH Mexico, Vera Cruz (Galeotti, 4541 and 4549, Linden, 963). Hb. Kew. “ Very near 7. peltatum, DC., but leaflets not peltate.”—Bentham in Hb. Kew.

15. Thalictrum, sp. (foliolis amplis latis). Sour Mzxtco, Oaxaca (Ghiesbreght). Hb. Paris.

3. ANEMONE.

Anemone, Linn. Gen. Plant. n. 694; Benth. et Hook. Gen. Plant. i. p. 4.

About seventy species, chiefly in the temperate andcold regions of the northern hemisphere, a few scattered in the mountainous regions of Central and South America ; three occur in South Africa, and one in Australia.

1. Anemone caroliniana, Walt. Carol. p. 157; Marcy’s Exploration of the Red River, t. 1. Anemone tenella, Pursh. Norra America, from North Carolina through the south-eastern States to—Norti Mexico, Chihuahua (Torrey).

9. Anemone mexicana, H. B. K. Nov. Gen. et Sp. v. p. 41. Nortn America, Illinois, Louisiana, and Texas to—SourH MExico, near Santa Rosa (Humboldt & Bonpland), Oaxaca (Galeotti), Zimapan (Coulter, 654), Orizaba (Botter:, 21), Vera Cruz (Linden, 964). Hb. Kew. |

4. MYOSURUS. Myosurus, Linn. Gen. Plant. n. 394; Benth. et Hook. Gen. Plant. i. p.5.

Two species, both annual herbs. UM. minimus is widely dispersed in the temperate parts of the northern hemisphere, and also occurs in South America and Australia. Probably introduced in some of the localities. J. aristatus grows in Western Extra- tropical America, both North and South, and also in New Zealand, and is likely to occur in Mexico. |

1. ‘Myosurus minimus, Linn. Sp. Pl. 407; Gray, Gen. Ill. t. 8.

Temperate regions of the northern hemisphere, AUSTRALIA and Soura AMERICA;

Nort AMERICA southward to—Norta Mexico, Chihuahua (Bigelow), Sonora (Schott). Hb. Kew.

Tribe RANUNCULEZ.

5. RANUNCULUS. Ranunculus, Linn. Gen. Plant. n. 699; Benth. et Hook. Gen. Plant. i. p. 5.

Annual and perennial herbs ; about 160 species dispersed over nearly all temperate

6 RANUNCULACEA.

and frigid regions, including the mountains within the tropics; but especially abundant in the northern hemisphere.

1. Ranunculus amarillo, Bert. Fl. Guat. p. 24. GuatemaLa (Velasquez).

9. Ranunculus aschenbornianus, Schaver, in Linnea, xx. p. 719. SourH Mexico, mountains about Tula (Aschenborn, 489).

3, Ranunculus cymbalaria, Pursh, Fl. Am. Sept. ii. p. 392; DC. Syst, i. p. 252. Canapa to the ARGENTINE Repustic.—NortH Mexico, region of San Luis Potosi, 6000 to 7000 feet (Parry & Palmer, 3); Sour Mexico, Zimapan (Coulter, 651), marshy pastures near the city of Mexico (Bowrgeau, 2).—Also in NoRTHERN Asia and Evrore. Hb. Kew.

4. Ranunculus delphinifolins, H. B. K. Nov. Gen. et Sp. v. p. 48. Soutn Mexico, wet pastures, Morelia (Hartweg, 267), Real del Monte (Couléer, 648), ‘Zimapan (Coulter, 650), Santa Fé (Bourgeau, 284), Toluca (Heller), Mexico (Graham, 114 and 115). Hb. Kew.

5. Ranunculus dichotomus, Moe. et Sessé, in DC. Syst. Veg. i. p. 288; Calques des Dess. Fl. Mex. 1.

Norta Mexico, region of San Luis Potosi, 6000 to 8000 feet (Parry & Palmer, 5); Sourn Mexico, in wet grassy places near Jalapa (Schiede), between San Miguel del Soldado and La Joya (Schiede), Oaxaca (Galeotti, 4568), Real del Monte (Coulter, 647), Tacubaya, valley of Mexico(Bowrgeau,1), Mexico(Harris).—CotomBia; Perv. Hb. Kew. —

6. Ranunculus donianus, Pritzel, in Walp. Rep. ii. 740. Ranunculus humilis, D.Don, in G.Don’s Gen. Syst. i. p. 34, nec Collie in Hook. Bot. Beechey’s Voy.

Mexico (Hb. Lambert).

7. Ranunculus geoides, H. B. K. Nov. Gen. et Sp. v. p. AT, t. 429. Sourn Mexico, peak of Orizaba, 9000 to 12,000 feet (Linden, 960); near Moran and

Real del Monte (Humboldt & Bonpland). Hb. Kew.

8. Ranunculus hookeri, Schl. in Linnea, ix. p. 610.

Nort Mexico, region of San Luis Potosi, 6000 to 8000 feet (Parry & Palmer, 6) ; South Mexico, Oaxaca ((raleotti, 4567), Orizaba, abundant (Bourgeau, 2414), Vera Cruz to Orizaba (Miller), Zimapan and Real del Monte (Coulter, 645 and 48) in grassy places near Jalapa (Schiede). Hb. Kew.

9. Ranunculus hydrocharis, Spenner, Fl. Frib. iv. p. 1007 (BR. longirostris, Godr.,= R. aquatilis, Linn., var.). Soura Mexico, Rio de Actopan (Schiede), canal of Santa Anita, near Mexico, and road from Pueblo Viejo to Real del Monte. Hb. Paris. This appears to be the only form of the Batrachian Ranuneuli collected in Mexico.

RANUNCULACEA. 7

10. Ranunculus hydrocharoides, A. Gray, Pl. Thurb. p. 306. Norta Mexico, wet meadows, Mabibi, Sonora (Thurber, 441). Hb. Kew.

11. Ranunculus llaveanus, Schl. in Linnea, x. p. 233. SourH Mexico, in grassy places about Jalapa (Schiede). This should perhaps be referred to B. delphinifolius, H. B. K.

12. Ranunculus “ microcarpus, Presl?” Sourn Mexico, near Yotla (Andrieur, 547). Hb. Kew.

13. Ranunculus multicaulis, D. Don in G. Don’s Gen. Syst. i. p. 34. Mexico. Hb. Lambert.

14. Ranunculus oxynotus, A. Gray, in Proc. Am. Acad. x. p. 68. CALIFORNIAA—NortH Mexico, Sonora Pass.

15. Ranunculus ornithorrhynchus, Hook. Fl. Bor.-Am. p. 21, t. 9. SoutH Mexico, Toluca, 8200 to 9000 feet (Heller). Hb. Kew. Probably the same as &. dichotomus, DC. The typical plant is a native of Oregon.

16. Ranunculus pennsylvanicus, Linn. fil. Suppl. p. 272. Ranunculus hispidus, Michx. CanapDa southward to—Nortu Mexico. Hb. Kew.

17. Ranunculus petiolaris, H. B. K. Nov. Gen. et Sp. v. p. 45, t. 428. Sour Mexico, near Santa Rosa, 8400 feet (Humboldt & Bonpland).

18. Ranunculus peruvianus, Pers. Ench. ii. 103; Hook. Ic. viii. t. 745. Soutn Mexico, Peak of Orizaba, 12,000 to 12,500 feet (Linden, 961).—CoLomBra ; Perv. Hb. Kew.

19. Ranunculus sibbaldioides, H. B. K. Nov. Gen. et Sp. v. p. 48. Sour Mexico, margin of the crater-lake of Toluca, 12,000 feet (Galeotti, 4960) ; meadows about Toluca (Heller).—Mountains of Tropica, South America. Hb. Kew.

20. Ranunculus stolonifer, Hemsley, Diag. Pl. Nov. pars alt. p. 17.

Pusillus glaber stolonifer, foliis subintegris radicalibus longe petiolatis, floribus parvis, toro conico, petalis circiter 6 oblongis longe unguiculatis, unguis apice nectario amplo producto, acheniis levibus subrotundis compressis.

Herba perennis, stolonifera, tota glaberrima, caulibus erectis, 2-6-pollicaribus vel vix evolutis. Folia integra vel interdum calloso-crenata ; radicalia longe petiolata reniformia rotundata ~ elliptica usque lanceolato-oblonga, lamina 3-12 lin. longa, petiolo 1-2-pollicari basi scarioso- dilatato; caulina sessilia, lmearia. Flores parvi, flavi, longiuscule pedunculati ; torus conicus, glaber, sepala oblongo-elliptica, 1-14 lin. longa; petala 5-6, oblongo-elliptica, subtrivenia, circiter 1} lin. longa, longe unguiculata, unguis apice nectario conspicuo producto; stamina

8 RANUNCULACEA.

numerosa, petalis Iongiora, filamentis dilatatis. Achenia circiter 20, levia, subrotunda, valde compressa leviter carinata.

North Mexico, region of San Luis Potosi, 6000 to 8000 feet (Parry & Palmer, A). Hb. Kew.

This apparently undescribed species of Ranunculus is allied to R. ophioglossifolius, pusillus, &c., but differs in habit, torus, achenes, &c., and in the petals having a relatively longer claw.

21. Ranunculus trachyspermus, Engelm. and Gray, Pl. Lindh. i. p. 3. Texas.—Nortiu Mexico, Sonora (Schott). Hb. Kew.

92. Ranunculus tridentatus, H. B. K. Nov. Gen. et Sp. v. p. 42. Sout MExico, ditches, Comancipilla (Hartweg), Mexico (Graham, Berlandier, 452). Hb. Kew. Var. B. Near Carpio at Lake San Cristobal (Humboldt & Bonpland). CoLomBIA; PERU.

23. Ranunculus, sp. SourH Mexico, valley of Mexico, Desierto Viejo (Bourgeau, 759). Hb. Kew.

24. Ranunculus, sp. (? R. repentis, var.). GuaTEMALA, Volcan de Fuego (Salvin). Hb. Kew.

Tribe HELLEBOREA.

6. AQUILEGIA. Aqguilegia, Linn. Gen. Plant. n. 684; Benth. et Hook. Gen. Plant. i. p. 8. Perrennial herbs. A large number of forms have been described as species; but Bentham and Hooker (J. ¢.) think they should be reduced five or six species. Baker, in

Gard. Chron. 1878, defines twenty-seven species. They are distributed all round the temperate zone of the northern hemisphere.

1. Aquilegia chrysantha, A. Gray, in Gard. Chron. 1873, p. 1335.

Aquilegia leptoceras, var. flava, A. Gray, Pl. Wright. ii. p. 9. Aquilegia leptoceras, var. chrysantha, Hook. f. Bot. Mag. t. 6073.

Rocky Mountains to—NortH Mexico, banks of rivers, Sonora (Thurber & Smith), Mabibi (Parry). Ub. Kew. |

2. Aquilegia skinneri, Hook. Bot. Mag. t. 3919. .A. mexicana in the description. GuaTEeMALA (Skinner). There is no specimen bearing either of these names in Hb. Kew. 3. Aquilegia, sp. SoutH Mexico, Mount Tanga, Oaxaca, 8000 feet (Bourgeau, 4574). Hb. Paris.

DELPHINIUM. 9

7. DELPHINIUM. Delphinium, Linn. Gen. Plant. n. 681; Benth. et Hook. Gen. Plant. 1. p. 9.

Perennial and annual herbs, natives of the temperate regions of the northern hemi- sphere. About 40. species are known.

1. Delphinium bicornutum, Hemsley, Diag. Pl. Nov. pars alt. p. 17.

Foliis palmatim 5-partitis, segmentis latiusculis 3—5-lobatis, floribus circiter pollicaribus, calcari calycino basi distincte bicornuto fere recto sepalis longiore, calcaribus petalinis calcari calycino eequilongis basi tantum tubulosis, petalis lateralibus basi appendiculatis, carpellis 3 immaturis pubescentibus.

Herba 2-3-pedalis, caulibus subsimplicibus crassiusculis, glabris, levibus. Folia (radicalia mihi ignota) caulina inferiora longe petiolata, palmatim 5-partita ; segmentis latiusculis, 1-14-poll. longis, 3-5-lobatis, subtus sparse puberulis, petiolo usque 6-pollicari. Pedunculi pauciflori, bracteis linearibus. Flores pollicares, puberuli, pedicellati, pedicellis puberulis bibracteolatis, bracteolis subulatis; sepala oblonga, acuta vel obtusa, calcari fere recto basi distincte bicornuto deorsum spectante; petalorum superiorum laminis angustis obliquis, apice rotundatis breviter bifidis, calcaribus calcari calycino equilongis basi tantum tubulosis; petala lateralia ungui- culata, alte bifida valde barbata, ungue basi appendice squamiformi ornato ; stamina leviter puberula. Carpella 3, immatura, puberula.

Soutn Mexico, Oaxaca (Ghiesbreght). Hb. Kew.

Characterized by the distinctly bifid, relatively long calyx-spur, the rather long basal appendage of the lateral petals, the short tubular portion of the spurs of the upper petals, &c. &

2. Delphinium latisepalum, Hemsley, Diag. Pl. Nov. pars alt. p. 17.

Foliis palmatim 5-partitis, segmentis latiusculis semel vel bis lobatis, bracteis linearibus, floribus subvillosis circiter pollicaribus, calcari calycino basi integro leviter curvato sursum spectante sepalis eequali vel breviore, calcaribus petalinis calcari calycino equilongis basi tantum tubu- losis, petalis lateralibus exappendiculatis, carpellis 3 immaturis pubescentibus.

Herba sesquipedalis, plus minusve pubescens vel villosa, caulibus subsimplicibus. Folia palmatim 5-partita, 14-3 poll. diametro; radicalia segmentis latiusculis, 3-5-lobatis, petiolo 2-3-pollicari ; caulina segmentis linearibus integris. Pedunculi pauciflori, pedicellis gracilibus. Flores villosi, circiter pollicares ; sepala late oblongo-elliptica, rotundata, calcari basi integro leviter curvato sursum spectante sepalis quali vel breviore; petalorum superiorum laminis angustis vix obliquis, apice brevissime bifidis, calcaribus calcari calycino eequilongis basi tantum tubulosis ; petala lateralia anguste unguiculata, basi exappendiculata, alte bifida, utrinque valde barbata ; stamina glabra. Carpella 3, immatura, cano-pubescentia.

Sourn Mexico, Cordillera of Oaxaca, 8000 to 8500 feet (Galeotti, 4547). Hb. Kew. This differs from the other Mexican species described here in its relatively short up- turned calyx-spur, broad sepals, &c.

3. Delphinium leptophyllum, Hemsley, Diag. Pl. Nov. pars alt. p. 18. Foliis alte palmatim 5-partitis, segmentis angustis alte trisectis, lobis ultimis linearibus, floribus

BIOL. CENT.-AMER., Bot. Vol. 1, Sept. 1879. Cc

10 RANUNCULACEA.

sesquipollicaribus, calcari basi integro leviter curvato deorsum spectante sepalis longiore, calcaribus petalinis calcari calycino equilongis ultra medium tubulosis, petalis lateralibus exap- pendiculatis, carpellis 3 puberulis.

Herba biennis seu perennis, 2-8 ped. alta, caule tereti glabro levi vix ramoso. Folia glabra, alte palmatim 5-partita, segmentis alte trisectis, laciniis ultimis linearibus obtusis, lamina 14-3 poll. diametro, petiolo foliorum radicalium gracili 5-6-pollicari, superiorum gradatim breviore. Pedunculi elongati, 2—4-flori vel interdum probabiliter pluriflori. Flores distantes, puberuli, longe graciliterque pedicellati circiter sesquipollicares, pedicellis puberulis 2—85-pollicaribus, bracteis bracteolisque lineari-subulatis ; sepala glabrescentia, oblongo-elliptica, apice rotundata, ‘gemipollicaria, calcari basi integro leviter curvato deorsum spectante sepalis longiore ; petalorum superiorum laminis oblique oblongis, apice breviter bifidis fere glabris quam sepala brevioribus, calcaribus calcari calycino equilongis, ultra medium tubulosis; petala lateralia unguiculata, basi exappendiculata, intus leviter barbata, alte bifida, lobis acutis ; stamina glabra. Carpella 3, immatura, puberula. | a

Norra Mexico, region of San Luis Potosi, 6000 to 8000 feet (Parry & Palmer, 7). Hb. Kew.

This species is characterized by the long narrow lobes of its palmately-divided leaves, by the spurs of the petals equalling the calyx-spur and tubular from the base to above the middle, and by the lateral petals being deeply bifid, not furnished with an appendage at the base, and only slightly bearded.

4, Delphinium pedatisectum, Hemsley, Diag. Pl. Nov. pars alt. p. 18.

Foliis caulinis pedatisectis, segmentis 7-3 integerrimis, floribus 1-1} poll. longis, calcari calycino basi integro leviter curvato deorsum spectante sepalis longiore, calcaribus petalinis calcari calycino fere tertia parte brevioribus vix ad medium tubulosis, petalis lateralibus basi minute

& appendiculatis, carpellis 3 immaturis pubescentibus.

Herba erecta, ramosa, ramis teretibus glaberrimis levibus. Folia radicalia ignota, caulina inferiora pedatisecta 7-loba, superiora trisecta, lobis linearibus 1-2 poll. longis, subtus pubescentibus. Pedunculi elongati, circiter 4-flori. Flores distantes, puberuli, longe graciliterque pedicellati, circiter pollicares, pedicellis puberulis, bracteis bracteolisque lineari-subulatis ; sepala glabres- centia oblique oblonga, calcari basi integro leviter curvato deorsum spectante ; petalorum superiorum laminis angustis alte bifidis, lobis acutis, calcaribus calcari calycino multo brevi-

_ oribus non vel vix ad medium tubulosis extus intusque puberulis ; petala lateralia unguiculata, intus barbata, limbo bipartito, ungue basi appendice squamiformi ornato; stamina glabra. Carpella 3, immatura, tomentosa, stylis elongatis.

Mexico (Parkinson), without locality. Hb. Kew.

Distinguished by the pedatisect leaves, by the spurs of the deeply bifid upper petals being shorter than the calyx-spur, and by the bipartite lateral petals having a scale- like appendage at the base.

5. Delphinium wislizeni, Engelm. Bot. Wisliz. Exp. p.22. Nortu Mexico, Cosiquiriachi (Wislizenus).

DILLENIACEA. 11

Order II. DILLENIACEZ. Dilleniacee, Benth. et Hook. Gen. Plant. i. p. 10.

This order consists of about eighteen genera and 180 species of trees, shrubs, and a few herbaceous plants. They inhabit the tropics of both worlds, but are most numerous in Extratropical Australia. A few are found in Eastern Temperate Asia and North America.

Tribe DELIMEZA. 1. DAVILLA. Davilla, Vahl, DC. Syst. i. p. 404 (Hieronia, Vell. Fl. Flum. v. t. 116); Benth. et Hook. Gen. Plant. i, p. 12. About fourteen species of climbing or twining shrubs, natives of tropical America. 1. Davilla kunthii, A. St.-Hil. Plant. Us. Bras. sub t. 22, p. 6. Daviila lucida, Presl, Reliq. Henk. u. r(aHA Seem. Bot. Voy. ‘ Herald, t. 13.

Davilla ovata, Presl. Panama, Rio Grande railway-station (S. Hayes), Chagres (Fendler, 26), Veraguas

(Seemann).—CoLomBIA to Brazit. Hb. Kew. 2. Davilla rugosa, Poir. Enc. Sup. ii. p. 457, et in St.-Hil. Plant. Us. Bras. t. 22. Davilla brasiliana, DC. Deless. Ic. Sel. i. t. 71. Nicaraaua, Chontales (Seemann, 2); Panama, Chagres (Fendler, 26), Remedios (See- mann).—West Inpiss, and CoLtompra to Brazit. Hb. Kew. 8. Davilla sagrezana, Rich. Fl. Cab. (Spanish ed.), i. p. 8. . Davilla multifiora, Seem., not of St.-Hil.

Tetracera multiflora, DC. ; Nicaragua, Chontales (Tate, 283); Panama, San Lorenzo (Seemann, 1621).—Cusa,

Peru, and Brazit. Hb. Kew. 4, Davilla, sp. (1). GuATEMALA (Skinner). Hb. Kew. 5. Davilla, sp. (2). GuaTEMALA (Friedrichsthal). Hb. Kew. | 6. Davilla, sp. (3). Panama, Island of Taboga (Sinclair). Hb. Kew.

OO 2. CURATELLA. Curatella, Linn. Gen. Plant. n. 679; Benth. et Hook. Gen. Plant. i. p. 12. A genus of two’species of small trees or shrubs peculiar to tropical America.

1. Curatella americana, Linn. Sp. Pl. p. 248; Aublet, Pl. Guian. t. 232.

Sours Mexico, Acapulco (Lay & Collie); Panama, ‘in meadows and stony places in : c2

12 — DILLENIACEZ.

Veraguas and Panama (Seemann), Island of Taboga (Hinds).—Widely dispersed in Tropical Sours America, and also occurring in TRinipAD. Hb. Kew.

3, DOLIOCARPUS.

Doliocarpus, Roland, DC. Syst. Veg. i. p. 405 (Othlis, Schott) ; Benth. et Hook. Gen. Plant. .i. p- 12.

Soramia and Calinea of Aublet and Ricaurtea of Triana are referred hither. - About eighteen shrubby species, restricted to Tropical America.

1. Doliocarpus pubens, Mart. in Flora, xxi., Beibl. P. 49,

Doliocarpus semidentatus, Garcke. _ Panama, Chagres (Hendler, 50 and 309), village of La Mesa, Veraguas (Seemann).— Brazit. Hb. Kew. 2. Doliocarpus, sp. (?.D. pudentis var.). | | PanaMA, in woods, Mamei railway-station (8. Hayes, 497).—TrinipaD; GUIANA. Hb. Kew. 3. Doliocarpus, sp. Panama, Bujio railway-station (S. Hayes, 143). Hb. Kew.

4, TETRACERA.

Tetracera, Linn. Gen. Plant. n. 683 (Euryandra, Forst.; Wahlbomia, Thunb.) ; Benth. et Hook. Gen. Plant. 1. p. 12. . | .

A genus of about twenty-five species of climbing shrubs or trees, inhabiting the tropical regions of both the Old and New worlds. One species occurs in New — Caledonia. Presl’s are all doubtful species.

1. Tetracera alata, Presl, Reliq. Henk. p. 71. Mexico (Henke).

2. Tetracera erecta, Sessé et Mogino, DC. Syst. Veg. i. p. 404; Calques des Dess. Fl. Mex. 2 ((=T. sessilifiora, Tr. et P1.). | Mexico (Mogino & Sessé).

3. Tetracera oblongata, DC. Prodr. i. p. 67% Deless. Ic. Sel. i. t. 67. Panama, near the town (Seemann, 337); Braz. Hb. Kew.

4. Tetracera portobellensis, Beurling, in Vetenskaps Acad. Handl. 1854, p. 113. PANAMA, in woods on the road to the city of Panama (Biliberg).

5. Tetracera rhamnifolia, Presl, Reliq, Henk. ii. p. 72. Panama (Henke).

DILLENIACEA, | | 18

6. Tetracera salicifolia, Presl, Reliq. Henk. ii. p. 71. Mexico (Henke).

7. Tetracera sessiliflora, Tr. et Pl. in Ann. Sc. Nat. 1862, p. 21. Tetracera volubilis, H. B. K. non Linn. Delima mexicana, Mos. et Sessé, DC. Syst. Veg. i. p. 407 ; Calques des Dess. Fl. Mex. 3. Soura Muxico, Acapulco (Lay & Collie), San Blas (Hinds); Guatemata (Hriedrichs- thal) ; Panama (Weddell), Chagres (Fendler, 27).—Cotomsia. Hb. Kew.

8. Tetracera volubilis, Linn. Sp. Pl. p. 617; Lam. Encycl. t. 485; Tr. et Pl. fe. Panama (Duchassaing). [ Recchia, Mos. et Sessé, Fl. Mex. ined., DC. Syst. Veg. i. p. 44, is referred to Rzgic- stachys by Planchon, a genus placed in Simarubee by Bentham and Hooker].

Order ITI. MAGNOLIACE.Z. Magnoliacee, Benth. et Hook. Gen. Plant. i. p. 16.

_ Trees or shrubs, about seventeen species, belonging to nine genera. Natives of Tropical and Eastern Temperate Asia, North America, a few occurring in Tropical and Temperate South America, New Zealand, and Australia. None hitherto discovered in Africa.

Tribe MAGNOLIEA.

1. MAGNOLIA. Magnolia, Linn.. Gen. Plant. n. 690; Benth. et Hook. Gen. Plant. i. p. 18.

About fourteen species known, whereof six are Asiatic (China, Japan, and the Himalaya Mountains), and the remainder North-American and Mexican.

OL. Magnolia dealbata, Zucc. in Abhandl. bayer. Akad. ii. p. 369, t. 3 et 4. Sourn Mexico, Cumbre de Obispo (Schiede).

2. Magnolia schiedeana, Schl. in Bot. Zeit. 1864, p. 144, Sout Mexico, between San Salvador and Jalapa (Schiede).

2, TALAUMA. | Talauma, Juss. Gen. Plant. p. 281; Hook. et Benth. Gen. Plant. i. p. 18. Trees, about fifteen species, three or four of which inhabit Tropical America, and the

remainder Tropical Asia, especially the eastern part, northward to Japan.

1. Talauma mexicana, Don, Gen. Syst. i. p. 85. IF oN Magnolia mexicana, DC., Calques des Dess. Fl. Mex. 6. Mexico (Mogino & Sessé).

4 | MAGNOLIACER.

2. Talauma macrocarpa, Zucc. in Abhandl. bayer. Akad. ii. p. 369 ad 478, t. 1 et 2.

- Sout Mexico, Cumbre de Obispo (Schiede), on the Pacific slope of the mountains, a at an altitude of about 1500 feet (ex Zuccarint).

This may be the same as Magnolia dealbata.

Tribe WINTERER.

3. DRIMYS. Drimys, Forst. Char. Gen. p. 42; Benth. et Hook. Gen. Plant. i. p. 18. _ Evergreen shrubs and trees, six species, whereof one or two are American, ranging from

Mexico to Chili, one is a native of Borneo, one of New Zealand, and two of Australia.

1. Drimys granatensis, H. B. K., var. sylvatica, St.-Hil. Pl. Us. Bras. t. 27. South Mexico, Sierra Colorada near San Andres, and Cuesta Grande de Jalacingo (Schiede & Deppe).—Couomsia to Braz. Hb. Kew.

2. Drimys mexicana, Mocino et Sessé, DC. Syst. i. p. 444; Calques des Dess. Fl. Mex. 5. SoutH Mexico, Laguna de Tanetze (Hartweg). Hb. Kew. Probably not specifically different from D. granatensis, which has a very wide distri- bution in South America.

Order IV. ANONACEE. Anonacee, Benth. et Hook. Gen. Plant. i. p. 20.

Trees or shrubs, often climbing, comprising about fifty genera and 450 species, nearly all within the tropics of Asia, Africa, and America, a few occurring in North America, and Subtropical South America, Africa, and Australia.

Tribe UVARIE.

1. SAPRANTHUS.

Sapranthus, Seem. Journ. Bot. iv. p. 369; Benth. et Hook. Gen. Plant. i. p. 956 (where it is referred to the genus Porceka).

The following is the only species. 1. Sapranthus nicaraguensis, Seem. Journ. Bot. iv. p. 369, t. 54.

GuaTeMALA (Friedrichsthal) ; Nicaracua, between Leon and Granada (Seemann). Hb. Kew.

ANONACEZ. 15

2. UVARIA. Uvaria, Linn. Gen. Plant. n. 692; Benth. et Hook. Gen. Plant. i. p. 23. About thirty-five species, shrubby climbers. Bentham and Hooker (U. ¢.) limit the genus to the tropics of the Old World; and the affinities of the following plant are uncertain, as its flowers are unknown. |

1. Uvaria (Porcelia ?) hahniana, Baill. Adans. viii. p. 347. Sour Mexico, mountain woods, Coachilote (Hahn, 239). Hb. Paris.

8. ASIMINA. Asimina, Adans. ex Dunal, Anon. p. 81; Benth. et Hook. Gen. Plant. i. p. 24.

Shrubs or small trees: four North-American species, and two or three Mexican and Central-American.

1. Asimina campechiana, H. B. K. Nov. Gen. et Sp. v. p. 61. Soutn Mzxico, about Campeche (Humboldt & Bonpland).

2. Asimina, sp. ? Soura Mexico, Chinitan, between Tehuantepec and the river Goazacualcos (Andrieux, 543). Hb. Kew.

4, GUATTERIA. Guatteria, Ruiz et Pav. Prodr. p. 85, t.17; Benth. et Hook. Gen. Plant. 1. p. 23.

About fifty species, trees or shrubs, confined to the warmer parts of America. The Asiatic species of various authors are referred to Polyalthia by Bentham and Hooker.

1. Guatteria amplifolia, Tr. et Pl. in Ann. Sc. Nat. xvii. (1862), p.32. Panama, Lion-Hill railway-station (S. Hayes, 346), Chagres (Fendler, 3). Hb. Kew.

9. Guatteria bibracteata, Hemsley, Diag. Pl. Nov. pars i. p. 1. Anona? bibracteata, Hook. Ic. Pl. iv. t. 828 (char. emend.).

Ramis gracilibus, junioribus rufo-pubescentibus, foliis breviter petiolatis oblongo-lanceolatis obtuse acuminatis vel gradatim attenuatis petiolisque sparse pubescentibus, floribus longe pedunculatis, pedunculis infra medium bibracteatis, bracteis foliaceis.

Arbor . . . Yamis gracilibus glabrescentibus. Folia breviter petiolata, oblongo-lanceolata, 3-4 poll. longa, integerrima, obtuse acuminata, basi attenuata, subtus secus costam mediam et marginem leviter pubescentia demum glabra, supra preter costam canaliculatam glabra, nitida, petiolis 1-2 lineas longis. Pedunculi graciles, oppositifolii, 1-2-flori, infra medium 2-bracteati, 2-8 poll. longi; bracteis amplis foliaceis inequalibus, inferiore majore late cordato-ovata polli- cari. Flores virides, parvi, pubescentes ; sepala ovata, 1 lin. longa; petala oblonga, obtusa, sub- coriacea, ad 8 lin. longa. Carpella ovata, semipollicaria, brevissime stipitata. |

Sourn Mexico, Consoquitla, Vera Cruz (Linden, 982); Mexico, without special locality (Harris). Hb. Kew.

16 ANONACE.

This species is near the Brazilian G. macropus and the Colombian G. longipes, but readily distinguished from both by its extra-axillary flowers and the shape of its bracts. |

3. Guatteria diospyroides, Baill. Adans. viii. p. 269. Sout Mexico, Chinantla (Liebmann). Hb. Kew.

4, Guatteria galeottiana, Baill. Adans. viii. p. 268. SoutH Mexico, Comaltepec, Tuitalungo, and Lobani (Liebmann), Campeche (Linden). Hb. Paris.

5. Guatteria jurgensenii, Hemsley, Diag. Pl. Nov. pars i. p. 1.

Ramis teretibus, foliis amplis lanceolatis acuminatis breviter petiolatis subtus sparse pilosulis, floribus solitariis axillaribus pedunculatis, pedunculis ad medium articulatis infra articulationem multibracteatis, bracteis minutis (?) cito deciduis.

Arbor? Ramiteretes, glabrescentes. Folia subcoriacea, lanceolato-oblonga, 6-9-poll., longe et obtuse acuminata, basi cuneata, petiolata, integerrima, supra glabra nitida reticulato-venosa, subtus sparse pilosula costa elevata venis reticulatis prominentibus ; petiolis crassis, 3-4 lin. longis. Pedunculi ferrugineo-pubescentes, 9-12 ]ineas longi, infra medium articulati, supra articulationem ebracteati, sursum gradatim incrassati, infra articulationem graciliores pluribracteati, bracteis deciduis. Flores sericeo-pubescentes, circiter 1 poll. diametro ; sepalis 8, subrotundatis, acutis, 2-3 lin. longis; petalis 6, fere eequalibus, coriaceis, obovato-ellipticis, acutis, extus intusque seri- ceis; stamina numerosissima, styli equilonga, connectivo ultra loculos antherarum late truncato dilatato; ovaria numerosissima, sericea, stylis brevissimis. Fructus nobis ignotus. |

SoutH Mexico, Sierra San Pedro Nolasco, Talea, &c. (Jurgensen, 718). Hb. Kew. This species appears to be closely allied to G. galeottiana, Baillon, from which, how- ever, it differs in various characters, judging from his description.

6. Guatteria macrantha, Presl, Relig. Henk. ii. p. 78. Mexico (Henke).

» 7. Guatteria schomburgkiana, Mart. Fl. Bras. Anon. p. 38, in adnot. Panama, San Lorenzo (Seemann); Guiana and Braziu. Hb. Kew.

4

8. Guatteria, sp. Nicaraeua, Chontales (Tate, 2). Hb. Kew.

9. Guatteria, sp. | Nicaragua, Chontales (Tate, 4). Hb. Kew. 10. Guatteria, sp. Costa Rica (Endres, 176). Hb. Kew. 11. Guatteria, sp. Costa Rica, 8500 feet (Endres, 13). Hb. Kew.

ANONACEZ. | : 17

Tribe UNONE.

5, ANAXAGOREA.

Anaxagorea, St.-Hil. in Bull. Soc. Philom. 1825, ex Blume, Fl. Jav. Anon. p. 64, t. 32; Benth. et Hook. Gen. Plant. i. p. 25. Trees. Seven species, three inhabiting Tropical Asia and four Tropical America.

1. Anaxagorea crassipetala, Hemsley, Diag. Pl. Nov. pars i. p. 2.

Ramis teretibus petiolisque rufo-puberulis, foliis vix coriaceis lanceolatis acuminatis, floribus gemi- natim pedunculatis, petalis crasse coriaceis 3 exterioribus ovatis obtuse acuminatis, 3 interioribus lanceolatis acutis brevioribus, carpellis graciliter stipitatis.

Arbor? Rami teretes, rufo-puberuli. Folia petiolata vix coriacea, glabra, lanceolata, 6-9 poll. longa, acute acuminata, costa venisque lateralibus distantibus infra prominentibus, petiolis rufo- puberulis teretibus ad 4 lin. longis. Pedunculi axillares, brevissimi, biflori; pedicellis rufo- puberulis, 4-6 lin. longis, juxta flores bibracteolatis, bracteolis latis brevibus persistentibus. Flores parvi, rufo-puberuli ; sepala 3, triangularia, 2-3 lin. longa; petala 6, crasse coriacea vel fere carnosa, 3 exteriora ovata, obtuse acuminata, 5-6 lin. longa, 3 interiora lanceolata, acuta, ad 3 lin. longa. Carpella 5-10, rufo-puberula, demum glabra, longe graciliterque stipitata, semi- nibus nitidis rufo-fuscis.

Nicaragua, Chontales (Tate, 8, Seemann, 3). Hb. Kew. A very distinct species, if the petals of the other American species have been cor- rectly described. In other characters it approaches very closely to A. acuminata, which has yellowish branches, thicker leaves, shorter and stouter petioles, and the lateral veins distinctly run together, and the stipes of the carpels is stouter. ‘

6. UNONA.

Unona, Linn. f. Suppl. p. 270.

This genus, according to Bentham and Hooker, Gen. Plant. i. p. 24, is confined to the tropics of the Old World. The following plant is not quoted under any other genus. It is probably a Sapranthus or an Asimina.

1. Unona violacea, Dunal, Monogr. Anon. p. 105, t. 25. South Mexico (Mogino & Sessé). ;

Tribe MITREPHOREAL. | This tribe is chiefly Asiatic.

7. CYMBOPETALUM. ~ Cymbopetalum, Benth. in Journ. Linn. Soc. v. p. 69; Benth. et Hook. Gen. Plant. i. p. 27.

Small trees. Two or three species, the third one a native of Brazil and Peru. - BIOL. CENT.-AMER., Bot. Vol. 1, Sept. 1879. d

18 ANONACEZ.

1. Cymbopetalum penduliflorum, Baill. Adans. viii. p. 268. Unona penduliflora, Dunal, Monogr. Anon. t. 28. Sourn Muxico, Pital (Liebmann). Hb. Kew.

2. Cymbopetalum, sp. SoutH Mexico, Chiapas ( Ghiesbregit 721). Hb. Kew.

Tribe XYLOPIE. 8. ROLLINIA. Rollinia, St.-Hil. Fl. Bras. Mer. i. p. 28, t. 5; Benth. et Hook. Gen. Plant. 1. p. 27.

Trees or shrubs. 20 species, natives of the warmer parts of America.

1. Rollinia mucosa, Baill. Adans. viii. p. 268. Rollinia sieberi, A. DC. Anona mucosa, Jacq.

South Mexico, Mecapulco and Mirador (Liebmann); without locality (Ocampo). —Also in Trinipap and St. Vincent. Hb. Paris. :

2, Rollinia, sp. _ GuatemaLa (Friedrichsthal). Tb. Kew.

8. Rollinia, sp. ? | | | ’ South Mexico, valley of Cordova (Bourgeau, 2448). Hb. Kew.

9, ANONA. Anona, Linn. Gen. Plant. no. 693; Benth. et Hook. Gen. Plant. 1. p. 27. About fifty species, trees and shrubs, widely spread in Tropical America, two or three

having a wide range in Tropical Africa and Asia, though originally introduced from America.

1. Anona cherimolia, Mill. Dict. ed. 9, no 5; Bot. Mag. t. 2011. Anona tripetala, Ait. South Muxico, region of Orizaba (Bourgeau, 2451, Bottert), Jalapa (Linden) ; Panama, Chiriqui (Seemann). Hb. Kew. | Widely spread in Tropical America, frequently as an escape from cultivation. Naturalized in some of the West-Indian Islands, according to Grisebach.

9, Anona cinerea, Dunal, Monogr. Anon. p. 72, t. 8. Sour Mexico, Vera Cruz (Schiede & Deppe). Perhaps the same as A. sguamosa.

3. Anona? depressa, Baill. Adans. viii. p. 267. SoutH Mexico, Tozamapa (Liebmann).

ANONACEZ. 19

4. Anona echinata, Dunal, Monogr. Anon. p. 68, t. 4. Widely dispersed in Tropica, Sours AMERICA and reaching—Panama, Bujio railway- station (S. Hayes, 142). Hb. Kew.

5. Anona excelsa, H. B. K. Nov. Gen. et Sp. v. p. 59. Sout Mexico, near La Venta del Exido (Humboldt & Bonpland).

6. Anona globiflora, Schl. in Linnea, x. p. 235.

Sourn Mexico, Papantla (Ziebmann), near the Hacienda de la Laguna (Schiede). Hb. Kew.

7. Anona involucrata, Baill. Adans. viii. p. 265. Sovuta Mexico, Tlatatla (Liebmann). Hb. Kew.

8. Anona liebmanniana, Baill. Adans. viii. p. 266. Sovrta Mexico, Comaltepec (Liebmann). 9, Anona muricata, Linn. Sp. Pl. p. 756; Jacq. Obs. i. t. 5. Anona bonplandiana, H. B. K. Native of Tropical America, and commonly cultivated. Perhaps not indigenous in CENTRAL AMERICA and Mexico. Propanche de la Concepcion (Liebmann). Hb. Kew. Grisebach thinks this may be indigenous only in some of the West-Indian Islands.

10. Anona palustris, Linn. Sp. Pl. p. 757. Also commonly cultivated. SoutH Mexico, Vera Cruz (Schiede).

11. Anona purpurea, DC. Prodr. i. p. 84; Dunal, Monogr. Anon. t. 2. Sour Mexico (Mogino & Sessé); Panama (Duchassaing). Hb. Kew.

12. Anona reticulata, Linn. Sp. Pl. p. 757. Panama, Island of Taboga (Seemann). Hb. Kew. Commonly cultivated for its fruit, and widely dispersed in Tropical America.

13, Anona squamosa, Linn. Sp. Pl. p. 757; Jacq. Obs. i. t. 6. fig. I. Cultivated. Hb. Kew.

14. Anona, sp. Panama, Chagres (Fendler). Hb. Kew.

15. Anona, sp. (aff. A. sylvatica, A. St.-Hil.). . SoutH Mexico, region of Orizaba (Bourgeau, 2482). Hb. Kew.

16. Anona, sp. SoutH Mexico (Hahn). Hb. Kew.

17. Anona, sp. (2.4. sericee var. foliis pedalibus). Panama, Obispo Falls (S. Hayes, 127). Hb. Kew.

a2

20 . ANONACEA.—MENISPERMACEZ.

10. XYLOPIA. |

Xylopia, Linn. Gen. Plant. no. 1027 (Waria, Aubl. Guian. p. 604, t. 243) ; Benth. et Hook. Gen, Plant. i. p. 28.

Trees and shrubs. About thirty species, whereof five are Indian, seven or eight African, and the remainder Tropical-American.

1. Xylopia frutescens, Aubl. Guian. i. p. 602, t. 292. ~ Panama (8. Hayes, 670), Veraguas and the island of Coiba (Seemann, 312).— Guiana; Brazit. Hb. Kew.

2. Xylopia grandiflora, St.-Hil. Fl. Bras. Merid. i. p. 40, t. 8. Panama, Savannas from Panama to Veraguas (Seemann).—A|so in J aMatca, TRINIDAD, Cusa, CotomBia, Peru, and Braziz. Hb. Kew. Planchon and Triana refer this to X. longifolia.