Forest flora of Japan. Notes on the forest flora of Japan;

FOREST FLORA OF JAPAN

BT

CHARLES SPRAGUE SARGENT

LANDSCAPE
ARCHITECTURE

Ex

Libris

BEATRIX
JONES

i

REEF POINT GARDENS
LIBRARY

The Gift of Beatrix Farrand

to the General Library
University of California, Berkeley

CSS

o

(-H

£

FOREST FLORA OF JAPAN

NOTES ON THE FOREST FLORA OF JAPAN

BY

CHARLES SPRAGUE SARGENT

DIRECTOR OF THE ARNOLD ARBORETUM OF HARVARD UNIVERSITY

BOSTON AND NEW YORK
HOUGHTON, MIFFLIN AND COMPANY

Dre0g, CambriDge

1894

Copyright, 1894.
BY CHARLES SPRAGUE SARGENT.

All rights reserved.

LANDSCAPE

ARCHITECTURE

The Riverside Press, Cambridge, Muss., V. S. A.
Electrotyped and Printed by H. O. Houghton and Company.

Idd'l
Farrand Gift

••d-

THE following notes, gathered in the autumn of 1892, during a journey through Hondo
and Yezo, were first printed, with the illustrations that accompany them, in " Garden and
Forest." I am indebted to the publishers of that journal for the permission to reprint them

in this form.

C. S. SARGENT.

ARNOLD ARBORETUM, May, 1894-

485

TABLE OF CONTENTS.

Page

INTRODUCTION 1

THE MAGNOLIA FAMILY 8

THE TERNSTRCEMIA, LINDEN, AND KITE FAMILIES 17

THE HOLLY, EVONYMUS, AND BUCKTHORN FAMILIES 23

THE MAPLE FAMILY 28

THE SUMACHS AND THE PEA FAMILY ............. 33

THE ROSE FAMILY 36

THE WITCH-HAZEL AND ABALIA FAMILIES 42

THE CORNELS, HONEYSUCKLES, AND PERSIMMONS, THE STYRAX FAMILY, THE ARBORESCENT MEMBERS OF

THE HEATH FAMILY, THE ASHES, AND THEIR ALLIES 47

THE LAUREL, EUPHORBIA, AND NETTLE FAMILIES 54

THE WALNUTS, BIRCHES, ALDERS, AND HORNBEAMS 60

THE OAKS, CHESTNUTS, WILLOWS, AND POPLARS 67

THE CONIFERS 72

THE CONIFERS, II 79

THE ECONOMIC ASPECTS OF THE JAPANESE FORESTS 85

LIST OF ILLUSTRATIONS.

1. CERCIDIPHYLLUM JAPONICTJM.

2. HEMLOCK FOREST OF LAKE YUMOTO-

3. MAGNOLIA KOBUS.

4. MAGNOLIA SALICIFOLIA.

5. MlCHELIA COMPKESSA.

6. CEKCIDIPHYLLUM JAPONICUM.

7. CERCIDIPHYLLTTM JAPONICUM.

8. TILIA MIQUKLIANA.

9. ACER MIYABKI.

10. AOEH NlKOENSE.

11. GLEDITSIA JAPONICA.

12. PRUNUS MAXIMOWICZII.

13. PYRUS MIYABEI.

14. PYRUS TSCHONOSKII.

15. DlSANTHUS CERCIDIFOLIA.

16. ACANTHOPANAX RICINIFOLIUM.

17. LlNDERA OBTUSILOBA.

18. ULMUS CAMPESTRIS.

19. ZELKOVA KEAKI.

20. ALNUS JAPONICA.

21. CARPINUS CARPIXUS.

22. OSTRYA JAPONICA.

23. QUERCUS DENTATA.

24. AVENUE OF CRYPTOMERIAS AT NIKKO.

25. HEMLOCK FOREST (TSUGA DIVERSIFOLIA), LAKE YUMOTO.

26. LARIX DAHURICA, VAR. JAPONICA.

FOREST FLORA OF JAPAN.

INTRODUCTION.

MANY years ago, in one of the most interesting papers * which has been written on the dis-
tribution of forests, Professor Asa Gray drew some comparisons between the forests of eastern
North America and those of the Japan-Manchurian region of Asia. Here it was shown that,
rich as eastern America is in tree species, Japan, and the regions to the north of it, in spite
of their comparatively small area, are still richer. Professor Gray's Asiatic region included the
four principal Japanese islands, eastern Manchuria, and the adjacent borders of China, while
the contrasted American region embraced the territory east of the Mississippi River, but
excluded the extreme southern point of Florida, inhabited by some sixty tropical trees which
belong to the West Indian rather than to the true North American flora. In the Japan-
Manchurian region he found 168 trees divided among sixty-six genera, and in eastern America
155 trees in sixty-six genera, the enumeration in both cases being confined " to timber-trees,
or such as attain in the most favorable localities to a size which gives them a clear title to the
arboreous rank." In the Japanese enumeration were included, however, a number of trees
which are not indigenous to Japan, but which, as we now know, were long ago brought into
the empire from China and Corea, like most of the plants cultivated by the Japanese. Early
European travelers in Japan, like Thunberg and Siebold, who were unable to penetrate far into
the interior, finding a number of plants common in cultivation, naturally believed them to be
indigenous, and several Chinese plants were first described from individuals cultivated in
Japanese gardens. Later writers2 on the Japanese flora have generally followed the example
of the early travelers, and included these plants in the flora of Japan. Indeed, it is only
very recently that it has been possible to travel freely in all parts of the empire, and to study
satisfactorily the character and distribution of its flora.

The list of Chinese and Corean trees cultivated in Japan, and usually enumerated in Floras
of the empire, includes Magnolia conspicua, Magnolia parvifolia, Magnolia Watsoni, Sterculia
platinifolia, Cedrela Sinensis, Zizyphus vulgaris, Kcelreuteria paniculata, Sapindus Mukorosi,
Acer trifidum, Rhus vernicifera, Sophora Japonica,3 Prunus Mume, Pyrus Sinensis, CrataBgus
cuneata, Eriobotrya Japonica, Liquidambar Formosana (Maximowiczii), Cornus officinalis,

1 Forest Geography and Archaeology, Scientific Papers, ii. through the entire country, especially in the foliaceous for-
204. ests of the north." He had evidently confounded Sophora

2 See Franchet & Savatier, Enum. PI. Jap. — Forbes & with Maackia, a common and widely spread tree, especially
Hemsley, Jour. Linn. Soc., xxiii. and xxvi. in Yezo. Sophora, which is only seen occasionally in gar-

8 Even Rein (The Industries of Japan), usually a most dens, does not appear to be a particularly popular plant with
careful observer, states that Sophora Japonica is " scattered the Japanese.

2 FOREST FLORA OF JAPAN.

Diospyros Kaki, and probably Diospyros Lotus, Chionanthus retusa, Paulownia imperialis,
Catalpa ovata, Linclera strychnifolia, Uhnus parvifolia, Thuja orientalis, Ginkgo biloba, Podo-
carpus Nageia, Podocarpus inacrophylla, and Piuus Koraiensis. If these species,1 twenty-nine
in number, are deducted from Professor Gray's enumeration, there will remain 139 species in
fifty-three genera, or a smaller number of both genera and species than he credited to eastern
America. This, however, does not alter the fact that the Japanese region for its area is unsur-
passed in the number of trees which inhabit its forests.

Indeed, the superiority of the forests of Japan in the number of their arborescent species
over those of every other temperate region, eastern North America included, in proportion to
their area, lias certainly never been fully stated, as perhaps I shall be able to show, having
made two years ago a somewhat extended journey through the northern and central islands,
undertaken for the purpose of studying Japanese trees in their relations to those of North
America. The case, perhaps, can best be stated by following Professor Gray's method, and
making a new census of the inhabitants of the Japan-Manchurian forests and of those of
eastern America, as these two regions extend through nearly the same degrees of latitude, and
possess somewhat similar climates, although Japan has the advantage of a more equally
distributed rainfall and a more equable climate, and offers a far more broken surface than
eastern America, with mountains twice the height of any of the Appalachian peaks.

As the true Atlantic forest extends west to the eastern rim of the mid-continental plateau,
the American region, for purposes of proper comparison, may be extended to the western
limit of the Atlantic tree-growth, although this will add to the American side of the account
a few genera and species of Texas, like Koeberlinia, Ungnadia, Parkinsonia, Prosopis, Acacia.
Chilopsis, and Pithecolobium, which Professor Gray did not include in the enumeration from
which his deductions were made. The south Florida species are again omitted, and those
plants which grow up with a single stem will be considered trees. In eastern North America,
that is in the whole region north of Mexico and east of the treeless plateau of the centre of
the continent, but exclusive of south Florida, 225 species of trees, divided among 134 genera,
are now known. The Japan-Manchurian region includes eastern Manchuria, the Kurile
Islands, Saghalin, and the four great Japanese islands, but for our purpose does not include
the Loochoo group, which, although it forms a part of the Japanese empire politically, is
tropical and subtropical in the character of its vegetation, which, moreover, is still imperfectly
understood. In this narrow eastern border of Asia there are now known 241 trees divided
among ninety-nine genera. The extra Japanese portion of the region contributes but little
to the enumeration. In Saghalin, Fr. Schmidt2 found only three trees which do not inhabit
Yezo, and in Manchuria, according to Maximowicz3 and Schmidt,4 there are only eighteen

1 A number of shrubs, familiar in western gardens, and kerrioides, Cercis Chinensis, Enkianthus Japonicus, Forsy-

usually supposed to be Japanese from the fact that they were thia suspensa, Olea fragrans, Tecoina grandifiora, Daphne

first known to Europeans in Japan, or were first sent from Genkwa, Edgeworthia papyrifera, and Wikstrcemia Japonica.

that country, are also Chinese or Corean, and in Japan are Nandiua domestica, the most universally cultivated orna-

only found in gardens or in the neighborhood of habitations. mental plant in Japan, is probably not a Japanese plant,

Among them are Clematis patens, Magnolia stellata, Mag- although Rein states that it grows wild in Shikoku.

nolia obovata, Berberis Japonica, Citrus Japouica, Prunus - Reisen in Amurland.

tomentosa, Prunus Japonica, Spiraea Thunbergii, Rhodotypos 8 Prim. Fl. Amur. 4 Heisen in Amurland.

INTRODUCTION. 3

trees which do not also occur in Saghalin or in the northern Japanese islands. In the four
islands of Yezo, Hondo, Shikoku, and Kyushu, therefore, we now find 220 trees divided
among ninety-nine genera, or only five less than occur in the immense territory which
extends from Labrador to the Rio Grande and from the shores of the Atlantic to the eastern
base of the Rocky Mountains. Neither Cycas revoluta nor Trachycarpns (Chamserops) excelsa
is included in the Japanese list, as the best observers appear to agree in thinking that these
two familiar plants are not indigenous to Japan proper. I have omitted, moreover, a few
doubtful species from the Japan enumeration, like Fagus Japonica, Maxiniowicz, and Abies
umbellata, Mayr, of which I could learn nothing in Japan, so that it is more probable that
the number of Japanese trees will be increased than that any addition will be made to the
silva of eastern America.

The proportion of trees to the whole flora of Japan is remarkable, being about 1 to 10.14,
the number of indigenous flowering plants and vascicular cryptogams being not very far
from 2,500 species. Still more remarkable is the large proportion of woody plants to the
whole flora. In Japan proper there are certainly not less than 325 species of shrubs, or 550
woody plants in all, or one woody plant in every 4.55 of the whole flora, — a much larger
percentage than occurs in any part of North America.

The aggregation of arborescent species in Japan is, however, the most striking feature in
the silva of that country. This is most noticeable in Yezo, where probably more species of
trees are growing naturally in a small area than in any other one place outside the tropics,
Avith the exception of the lower basin of the Ohio River, where, on a few acres in southern
Indiana, Professor Robert Ridgway has counted no less than seventy-five arborescent species
in thirty-six genera.1 Near Sapporo, the capital of the island, in ascending a hill which rises
only 500 feet above the level of the ocean, I noticed the following trees : Magnolia hypoleuca,
Magnolia Kobus, Cercidiphyllum Japonicum, Tilia cordata, Tilia Miqueliana, Phellodendron
Amurense, Picrasma ailanthoides, Evonymus Europaeus, var. Hamiltonianus, Acer pictum,
Acer Japonicum, Acer palmatum, Rhus semialata, Rhus trichocarpa, Maackia Amurensis,
Prunus Pseudo-Cerasus, Prunus Ssiori, Pyrns aucuparia, Pyrus Toringo, Pyrus Miyabei,
Hydrangea paniculata, Aralia spinosa, var. canescens, Acanthopauax ricinifolium, Acanthopa-
nax sciadophylloides, Cornus macrophylla, Syringa Japonica, Fraxinus Mandshurica, Fraxinus
longicuspis, Clerodendron trichotomum, Ulmus campestris, Ulmus scabra, var. laciniata,
Morus alba, Juglans Sieboldiana, Betula alba, Betula alba, var. Tauschii, Betula alba, var.
verrucosa, Betula Ermani, Betula Maximowicziana, Almis incana, Carpinus cordata, Ostrya
Japonica, Quercus crispula, Quercus gosseserrata, Castanea vulgaris, Populus tremula, Picea
Ajanensis, Abies Sachalinensis, — forty-six species and varieties. Within five miles of this hill
also grow Acer spicatum, var. Kurunduense, Acer Tataricum, var. Ginnala, Styrax Obassia,
Aphananthe aspera, Quercus dentata, Quercus glandulifera, Alnus Japonica, Salix subfragilis,
Salix Caprea, Salix stipularis, Salix acutifolia, Salix viminalis, and Populus suaveolans, — in all
sixty-two species and varieties, or more than a quarter of all the trees of the empire, which are
crowded into an area a few miles square, in the latitude of northern New England, in which,
north of Cape Cod, there are only about the same number of trees.

1 See Proc. U. S. Nat. Mm. 1882, 52. — Garden and Forest, vi. 148.

4 FOREST FLORA OF JAPAN.

A further examination of the trees of the two countries shows that, although the Japan-
Manchurian region possesses more arborescent species than eastern America, the silva of the
latter is much richer in genera, — one hundred and thirty-four to ninety-nine in Japan-
Manchuria. Forty-four genera have arborescent species in the two regions ; forty-five genera
with Japanese representatives have none in the flora of eastern America, and thirty-eight
genera represented in the American flora do not appear in that of Japan. A few genera, five
in eastern America and seven in Japan, are represented by trees in one region and by shrubs
only in the other. Of endemic arborescent genera the silva of eastern America contains
Asimina, Koeberlinia, Cliftonia, Ungnadia, Robinia, Cladrastis, Pinckneya, Oxydendrum, Mohro-
dendron, Sassafras, Planera, Toxylon, Leitneria, Hicoria, and Taxodium, fifteen, while in Japan
there are only five, — Cercidiphyllum,1 Trochodendron, Platycarya, Cryptomeria, and Sciadopitys.

Such a comparison between the silvas of eastern America and Japan is interesting as
showing the great number of arborescent species inhabiting four small islands. The signifi-
cant comparison, however, if it can ever be made, will be between eastern America, as here
limited, and all of eastern Asia from the northern limits of tree-growth to the tropics, and
from the eastern rim of the Thibetan plateau to the eastern coast of Japan. This would
include Corea, practically an unexplored country botanically, especially the northern portions,
and all the mountain ranges of western China, a region which, if it is to be judged from the
collections made there in recent years, is far richer in trees than Japan itself. It is impossible
to discuss with precision or with much satisfaction the distribution of the ligneous plants of
the north temperate zone until more is known of western China and of Corea, where may be
sought the home of many plants now spread through eastern China and Japan, and where
alone outside the tropics the enterprising and industrious collector may now hope to be
rewarded with new forms of ligneous vegetation.

Travelers in Japan have often insisted on the resemblance between that country and eastern
America in the general features of vegetation. But with the exception of Yezo, which is still
mostly uninhabited and in a state of nature, and those portions of the other islands which are
over 5,000 feet above the level of the ocean, it is difficult to form a sufficiently accurate idea
of the general appearance of the original forest-covering of Japan to be able to compare the
aspects of its vegetation with those of any other country, for every foot of the lowlands and
the mountain valleys of the three southern islands has been cultivated for centuries. And
the foothills and low mountains which were once clothed with forests, and could be again,
are now covered with coarse herbage, principally Eulalia, and are destitute of trees, except
such as have sprung up in sheltered ravines, and have succeeded in escaping the fires which
are set every year to burn off the dry grasses. Remoteness, bad roads, and the impossibility
of bringing down their timber into the valleys have saved the mountain forests of Japan,
which may still be seen, especially between 5,000 and 8,000 feet above the level of the sea, in
their natural condition. But these elevated forests are composed of comparatively few species,
and if it were not for the plantations of Conifers, which the Japanese for at least twelve
centuries, it is said, have been making to supply their workers in wood with material, and for

1 Since this was written I have received through M. M. L. de Vilniorm of Paris seeds of a Cereidiphyllum gathered in
the extreme western part of China.

INTR OD UCTION. 5

the trees preserved or planted in the temple grounds in the neighborhood of towns, it would
be impossible to obtain any idea at all of many of the Japanese trees. But, fortunately,
for nearly two thousand years the priests of Buddha have planted and replanted trees about
their temples, which are often surrounded by what now appear to be natural woods, as no
tree is ever cut and no attempt is made to clear up the undergrowth. These groves are
sometimes of considerable extent, and contain noble trees, Japanese and Chinese, which give
some idea of what the inhabitants of the forests of Japan were before the land was cleared
for agriculture.

The floras of Japan and eastern America have, it is true, some curious features in common,
and the presence in the two regions of certain types not found elsewhere show their relation-
ship. But these plants are usually small, and are rare or grow only on the high mountains.
Diphylleia, Buckleya, Epigsea, and Shortia show the common origin of the two floras ; but
these are rare plants in Japan, as they are in America, with the exception of Epigsea, and
probably not one traveler in ten thousand has ever seen them, while the chief elements of the
forest flora of northern Japan, the only part of the empire where, as has already been said,
comparison is possible, — those which all travelers notice, — do not recall America so much,
perhaps, as they do Siberia and Europe.

The broad-leaved Black Oaks, which form the most distinct and conspicuous feature in all
the forests of eastern America, are entirely absent from Japan, and the deciduous-leaved
White Oaks, which, in Japan, constitute a large part of the forest-growth of the north, are of
the European and not of the American type, with the exception of Quercus dentata, which
has no related species in America. The Chestnut Oaks, which are common and conspicuous,
both in the northern and southern parts of eastern America, do not occur in Japan, and the
Evergreen Oaks, which abound in the southern part of that empire, where they are more
common than any other group of trees, are Asiatic and not American in their relationships.

Many of our most familiar American trees are absent from the forests of Japan. The
Tulip-tree, the Pawpaw or Asimina, the Ptelea or Hop-tree, the Loblolly Bay or Gordonia,
the Cyrilla and the Cliftonia, the Plum-trees, which abound here in many forms, the Texas
Buckeye (Ungnadia), the Mesquite, the Locusts, the Cladrastis or Virgilia, the Kentucky
Coffee-tree or Gymnocladus, the Liquidambar, the Tupelos, the Sourwood or Oxydendrum, the
Osage Orange, the Kalmia, the Sassafras, the Persea or Red Bay, the Planera or Water Elm,
the Plane-tree, the Black Walnut, the Hickories, and the deciduous Cypress — all common
and conspicuous in our forests — are not found in Japan. Cratsegus, with a dozen species,
is one of the features of the forest flora of eastern America, while in Japan the genus is repre-
sented by a single species, confined to the northern part of the empire, and nowhere very
common. The Japanese Maples, with the exception of Acer pictum, which is not unlike our
Sugar Maple, have no close resemblance or relationship with the eastern American species ;
the Beech and the Chestnut are European, and not American ; the Birches, with one excep-
tion, are of the Old World type, as are the Lindens, Ashes, Willows, the Celtis, the Alders,
Poplars, and Larches.1

1 Of the arborescent genera of Japan, thirty are represented in Europe, and all of these, with the exception of Buxus, are
also found in eastern America.

6 FOREST FLORA OF JAPAN.

On the other hand, the Japanese will not find in our forests Euptelea and Cercidiphyllum
of the Magnolia family, Trochodendron, Idesia, the arborescent Ternstrcemiacese (Tern-
stroemia, Cleyera, Eurya, and Camellia), Phellodendron and Hovenia, Euscaphis, Maackia and
Albizzia, Distylium, Acanthopanax, Syringa, many arborescent Lauracete (Cinnamomum,
Machilus and Actinodaphne), which, next to the Evergreen Oaks, are the most distin-
guishing features of the forest flora of southern Japan. Nor will they find the beautiful
arborescent Linderas which abound in Japan, while in America the genus is only represented
by two unimportant shrubs, the arborescent Euphorbiaceae, like Buxus, Daphniphyllom,
Aleurites, Mallotus, and Excoecaria, or Zelkova, Aphananthe, Broussonetia, and Debregeasia,
or find anything to remind them of Pterocarya and Platycarya, Cryptomeria, Cephalotaxus,
and Sciadopitys.

The forests of the two regions possess in common Magnolia and J^sculus, which are more
abundant in species and individuals in America than in Japan. The Rhuses or Sumachs are
very similar in the two regions, and so are the Witch Hazel and the arborescent Aralia.
Cornus macrophylla of Japan is only an enlarged Corn us alternifolia of eastern America, and
the so-called Flowering Dogwoods of the two countries are not unlike. The Japanese
Walnut is very like the American Butternut, while, rather curiously, the Japanese Thuya and
the two Chamaecyparis, the Piceas and Abies, resemble species of Pacific North America, a
region whose flora has little affinity with that of eastern Asia. Tumion is common to the
two regions ; in eastern America it is one of the most local of all our trees, while in Japan it
is abundant in the mountainous regions of the central and southern parts of the empire.

Apart from the characters which distinguish related genera and species of Japanese trees
from their American congeners there are many aspects of vegetation which make the two
countries unlike. The number of broad-leaved evergreen trees is much greater in southern
Japan than in the southern United States, there being fifty species of these trees in the
former, and only twenty in eastern America (exclusive always of southern Florida), and the
general aspect of the groves and woods at the sea-level, even in the latitude of Tokyo, is of
broad-leaved evergreens. The number of evergreen shrubs in proportion to the entire flora is
much greater in Japan, too, than it is in America, and plants of this character grow much
farther north in the former than in the latter country. The small number of species of
Pinus in Japan, and their scarcity at the north, is in striking contrast to the number and
distribution of the species of this genus in eastern America, where there are thirteen species to
only five in Japan, including one shrub. In Japan the Hemlock forms continuous and almost
unbroken forests of great extent on the mountain-slopes, which are over 5,000 feet above the
sea, while in eastern America this tree is rarely found except scattered in small groves or as
single individuals through the deciduous-leaved forests. On the other hand, Picea and Abies,
which in America form immense forests almost to the exclusion of other species, grow,
wherever I have seen them in Japan, singly, or, in the case of Abies, in small groves on the
lower border of the Hemlock forests or mingled with deciduous-leaved trees. Picea Ajanensis
is said, however, to form extensive forests in some parts of western Yezo, and Professor
Miyabe informs me that in the extreme northern part of that island there are fine continuous
forests of Abies Sachalinensis. In northern Japan and on the high mountains of the central

PLATE II.

HEMLOCK FOREST OF LAKE YUMOTO.

INTRODUCTION. 7

islands, Birches are more abundant than they are in our northern forests ; and the river banks
at the north, like those of northern Europe and Siberia, are lined with arborescent Willows
and Alders, which are rare in eastern America, where these genera are usually represented
by shrubs. /

The illustration on the opposite page (Plate ii.) gives some idea of the general appearance
of the great coniferous forests which cover the highlands of central Japan. In the fore-
ground, Lake Yuuioto, famous for its thermal springs, nestles, 5,000 feet above the sea, among
the Nikko Mountains. The forests which rise from the shores of the lake are principally
composed of Hemlock (Tsuga diversifolia), among which are Birch (Betula Ermani), Abies
and Picea, Pterocarya, Cercidiphyllum, and the Mountain Ash. In the dense shade by the
shores of the lake grow dwarf forms of the Indian Azalea, Elliottia paniculata, our Canadian
Bunch Berry (Cornus Canadensis), great masses of Rhododendron Metternichii, which in
these forests replaces Rhododendron Catawbiense of the Appalachian Mountains, the dwarf
Ilex rugosa, Clethra canescens, here at the upper limits of its distribution, Panax horrida, and
the dwarf Blueberries which inhabit mountain-slopes in all northern countries, as well as the
ubiquitous Bamboos.

The undergrowth which covers the ground beneath the forests in the two regions is so
unlike that it must at once attract the attention of the most careless observer. In all the
Appalachian region of North America this is composed of a great number of shrubs, chiefly
of various species of Vaccinium and Gaylusaccia, of EpigaBa, wild Roses, Kalmias, dwarf
Pyrus and Lycopodiums ; in Japan the forest-floor is covered, even high on the mountains,
and in the extreme north, with a continuous, almost impenetrable, mass of dwarf Bamboos of
several species, which makes traveling in the woods, except over long-beaten paths and up the
beds of streams, practically impossible. These Bamboos, which vary in height from three to
six feet in different parts of the country, make the forest-floor monotonous and uninteresting,
and prevent the growth of nearly all other nnder-shrubs, except the most vigorous species.
Shrubs, therefore, are mostly driven to the borders of roads and other open places, or to the
banks of streams and lakes, where they can obtain sufficient light to enable them to rise above
the Bamboos ; and it is the abundance of the Bamboo, no doubt, which has developed the
climbing habit of many Japanese plants, which are obliged to ascend the trees in search of
sun and light, for the Japanese forest is filled with climbing shrubs, which flourish with trop-
ical luxuriance.

The wild Grape grows in the damp forests of Yezo with' a vigor and to a size which the
American species do not often attain, even in the semitropical climate of the southern Missis-
sippi valley. Actinidia arguta climbs into the tops of the tallest trees, and nothing is so
un-American or so attracts the attention of the American traveler in Japan as the trunks of
trees clothed to the height of sixty or eighty feet with splendid masses of the climbing
Hydrangeas (H. petiolaris and Schizophragma), or with the lustrous evergreen foliage of the
climbing Evonymus. Wistaria is represented, it is true, in eastern America, but here it is not
common or one of the chief features of vegetation as it is in Japan ; and the Ivy, a southern
plant only in Japan, and nowhere very common, helps to remind the traveler that he is in the
Old and not in the New World.

THE MAGNOLIA FAMILY.

THE general character of the composition of the Japanese forests having been briefly
traced, I shall now say something of the most important Japanese trees ; and, as their botan-
ical characters are already pretty well understood and their economic properties are only of
secondary interest to the general reader, these remarks will relate principally to their quality
from a horticultural point of view. A comparison with allied eastern American species will
perhaps be useful ; it will, at any rate, show that, while Japan is extremely rich in the
number of its tree species, the claim that has been made, that the forests of eastern America
contain the noblest deciduous trees of all temperate regions, can, so far as Japan is con-
cerned, be substantiated, for, with few exceptions, the deciduous trees of eastern America
surpass their Asiatic relatives in size and beauty.

In the Magnolia family Japan possesses five genera, while in the United States there are
only four. In Japan afborescent Magnoliaceae reach the most northern limit attained in any
country by these plants, and one of the most interesting features of the Japanese flora is the
presence in Yezo of two large trees of this tropical and semitropical family as far north, at least,
as the forty-fourth degree, while the representative of a third genus, Schizandra, is found still
farther north on the Manchurian mainland. In eastern America two species of Magnolia reach
nearly as high latitudes as this genus does in Japan, but in the United States Magnolia is
really southern, and has only succeeded in obtaining a precarious foothold at the north, while
in Yezo it is a most important element and a conspicuous feature of the forest vegetation.

Of true Magnolias three species grow naturally in Japan ; two of these belong to the
section of the genus which produces its flowers before the leaves appear, and which has no
representative in the flora of America ; the third, Magnolia hypoleuca, bears some resemblance
to our Magnolia tripetala. This tree is seen at its best in the damp rich forests which cover
the low rolling hills of Yezo, where it sometimes rises to the height of a hundred feet and
forms trunks two feet in diameter ; on the other Japanese islands it is confined to the moun-
tain forests, and apparently does not descend below 2,000 feet above the sea ; and it is only
in Yezo and on the high mountains in the extreme northern part of the main island that
I saw it of large size. In central Japan it rarely appears more than twenty or thirty feet
high, although this can perhaps be accounted for by the fact that all trees in the accessible
parts of the Japanese forests are cut as soon as they are large enough to be used for timber.
Magnolia hypoleuca must be considered a northern species, requiring a cold winter climate
for its best development, and it probably will not thrive in regions where the ground is not
covered with snow during several months of every year.

Magnolia hypoleuca is one of the largest and most beautiful of the deciduous-leaved
Magnolias ; in the early autumn, when the cones of fruit, which exceed those of any of our
species in size and are sometimes eight inches long, and brilliant scarlet in color, stand out on

PLATE III.

C. E. FAXON del.

MAGNOLIA KOBUS, D. C.

THE MAGNOLIA FAMILY. 9

the branches, it is the most striking feature of the forests of Hokkaido, which in variety and
interest are not surpassed by those of any other part of the world. Like Magnolia tripetala,
it is a tree of open habit, with long spreading irregularly contorted branches covered, as well
as the trunk, with pale smooth bark. The leaves, however, are not as much crowded together
at the ends of the flowering branches as they are in the American species, but are placed
rather remotely on the brauchlets ; they are twelve or fourteen inches long and seven or eight
inches broad, and on young vigorous trees are sometimes twice this size. On the upper
surface they are light bright green, and pale steel blue or sometimes almost silvery on the
lower surface, so that when raised by the wind they give the tree a light and cheerful appear-
ance. The flowers are six or seven inches across when expanded, with creamy white petals
and brilliant scarlet filaments ; they appear in May and June, after the leaves are nearly full
grown, and are, very fragrant.

Magnolia hypoleuca is still rare in gardens, although it was sent by Mr. Thomas Hogg to
the United States as early as 1865, and has flowered in the neighborhood of New York and
Boston ; it is probable that it will thrive in any part of the northern United States, although,
like other Hokkaido trees, it may suffer from summer and autumn droughts, which are
unknown in Japan, where the rainfall during August and September is regular and abundant.
As an ornamental tree Magnolia hypoleuca is superior to Magnolia tripetala in the fragrance
of its flowers and in the coloring of its leaves ; it is less desirable than Magnolia macrophylla,
which surpasses its Japanese relative in form and in the size and beauty of its flowers and
leaves, the largest produced on any plant of the Magnolia family, and larger than those of
any other North American tree. As a timber-tree Magnolia hypoleuca is valuable. The
wood, like that of all the Magnolias, is straight-grained, soft, light-colored, and easily seasoned
and worked. It is esteemed and much used in Japan for all sorts of objects that are to be
covered with lacquer, especially sword-sheaths, which are usually made from it ; in Hokkaido
it is employed for the interior finish of houses, and for boxes and cabinets, although harder
woods are generally preferred for such purposes.

In the forests of Hokkaido a second species, Magnolia Kobus,1 occurs. This tree some-
times grows in the neighborhood of Sapporo to the height of seventy or eighty feet, and
develops a tall straight trunk nearly two feet in diameter, covered with rather dark, slightly
furrowed bark. The branches are short and slender, and form a narrow pyramidal head,
which only becomes round-topped when the tree has attained its full size. The branchlets are
more slender than those of most species of Magnolia, and are covered with dark reddish brown
bark. The flowers (Plate iii.) appear near Sapporo in the middle of May, before the leaves,
from acute buds an inch long, half an inch broad, and protected by long thickly matted pale
hairs. They are from four to five inches across when fully expanded, with small acute
caducous sepals and narrow obovate thin creamy white petals. The stamens, with short broad
filaments, are much shorter than the narrow acute cone of pistils. The leaves are obovate,

1 Magnolia Kobus, De Candolle, Syst. i. 456. — Miquel, Magnolia tomentosa, Thunberg, Trans. Linn. Soc. ii. 336

Prol. Fl. Jap. 146. — Maximowicz, Mel. Biol. viii. 507. — (in part).

Franchet & Savatier, Enum. PI. Jap. i. 16. Magnolia glauca, var. a, Thunberg, Fl. Jap. 236.

Kobus, Kaempfer, Icon. Select, t. 42.

10 FOEEST FLORA OF JAPAN.

gradually narrowed below, and abruptly contracted at the apex into short broad points ; they
are pubescent on the lower surface at first, especially on the stout midribs and primary veins,
but at maturity are glabrous, or nearly so, and are bluish green, and rather lighter colored
on the lower than on the upper surface ; they are six or seven inches long, three or four
inches broad, rather conspicuously reticulate-veined, and are borne on stout petioles half an
inch to an inch and a half in length. The fruit is slender, four or five inches long, and is
often contorted or curved from the abortion of some of the seeds ; it is dark brown, the
carpels being conspicuously marked with pale dots.

Magnolia Kobus is exceedingly common in the forests which clothe the hills in the neigh-
borhood of Sapporo, where it grows to a larger size than in any part of Japan which I visited ;
near the shores of Volcano Bay it occurs in low swampy ground and in the neighborhood of
streams, in situations very similar to those selected by Magnolia glauca in the United States.
On the main island Magnolia Kobus is much less common than it is in Hokkaido, and I only
met with it occasionally on the Hakone and Nikko Mountains at considerable elevations above
the sea. This handsome tree was introduced into the United States by Mr. Thomas Hogg,
and was distributed from the Parsons' Nurseries as Magnolia Thurberi, under the belief that
it was an undescribed species. In cultivation it does not flower freely in the young state,
although trees in Pennsylvania, and in the Arnold Arboretum, where it was raised from seed
sent from Sapporo fifteen years ago, have produced a few flowers. In New England Magnolia
Kobus is the hardiest, most vigorous, and most rapid growing of all Magnolias.

I spent the 2d and 3d of October in company with Mr. James Herbert Veitch and Mr.
Tokubuchi, an accomplished Japanese botanist, on Mount Hakkoda, an extinct volcano 6,000
feet high, which rises southeast and a few miles distant from Aomori, the most northern city
of the main island of Japan. Botanically this was one of the most interesting excursions I
made in Japan, and we were able to gather the seeds of a number of plants that we did not
meet with elsewhere. On this mountain, in the very spot, perhaps, where Maries discovered
this fine tree, we found Abies Mariesii covered with its large purple cones ; and on the upper
slopes saw the dwarf Pinus pumila, forming almost impenetrable thickets five or six feet high
and many acres in area, and numerous alpine shrubs like Andromeda nana, Gaultheria pyro-
loides, Epigaja Asiatica, Phyllodoce taxifolia, and Geum dryadoides. On this mountain, too,
we established the most northern recorded station in Asia of the Hemlock (Tsuga diversi-
folia) ; and near the base Ilex crenata, Ilex Sugeroki, a handsome evergreen species with
bright red fruit, the dwarf Ilex integra, var. leucoclada, and Daplmiphyllum humile were
very common ; and here we,were fortunate in finding good fruit and ripe seeds of Magnolia
salicifolia.

On Mount Hakkoda Magnolia salicifolia (see Plate iv.) is a common plant between 2,000
and 3,000 feet above the sea-level. As it appears there it is a slender tree fifteen or twenty
feet high, with stems three or four inches thick, covered with pale smooth bark, and some-
times solitary, or more commonly in clusters of three or four. The branchlets are slender, and
light green at first, like those of Magnolia glauca, later growing darker, and in their third
year becoming dark reddish brown. The leaves are ovate, acute, gradually narrowed or
rarely rounded at the base, contracted into long slender points and sometimes slightly falcate

PLATE IV.

C. E. FAXON del.

MAGNOLIA SALIC IFOLIA, MAXM.

THE MAGNOLIA FAMILY. 11

toward the apex ; they are thin, light green on the upper and silvery white on the lower
surface, quite glabrous at maturity, five or six inches long, an inch and a half to two inches
broad, and are borne on slender petioles half an inch in length. When bruised they are
more fragrant than those of any species of Magnolia with which I am acquainted, exhaling the
delicate odor of anise-seed. The winter flower-bud is two thirds of an inch long, rather
obtuse, and protected by a thick coat of yellow-white hairs. The flowers of this tree are not
known to botanists, but from the shape and character of the winter-buds they are probably of
good size and produced in early spring before the appearance of the leaves. The fruit is
slender, flesh-color, an inch and a half to two inches long, and half an inch broad.

Magnolia salicifolia 1 grows on Mount Hakkoda in low wet situations, generally near
streams, and is evidently a moisture-loving plant. In November I found a single small plant
of this species near the town of Fnkushima, on the hills which rise above the valley of the
Kisogawa, not far from the base of Mount Ontake, in central Japan. Magnolia salicifolia is
new to cultivation, and we were fortunate in obtaining a good supply of seeds, by means of
which, it is to be hoped, this interesting tree will soon appear in gardens.

Among the Magnolias of Japan there is no evergreen species which resembles the great
evergreen Magnolia of our southern states, or at all equals it in the beauty of flowers and
foliage ; and the nearest approach to an evergreen Magnolia in the empire of the Mikado is
the representative of a closely allied genus, Michelia, differing from Magnolia in the position
of the flowers, which, instead of being terminal on the branches, are, except in the case of
one Indian species, axillary ; and in the number of ovules and seeds, of which there are two
in each carpel of Magnolia and more than two in each carpel of Michelia. A dozen species
are known, inhabitants of southern and southwestern Asia, including the islands of the Indian
Archipelago, and, with the exception of the Chinese Michelia fuscata (or as it is habitually
called in our southern states, Magnolia), which is cultivated for its exceedingly fragrant small
flowers in all warm temperate countries, the genus is not seen in American or European
gardens.

The Japanese species, Michelia compressa2 (see Plate v.), as it appears in the Botanic
Garden of the University of Tokyo, is a tree thirty to forty feet in height, with a trunk
twelve to eighteen inches in diameter, covered with smooth dark bark, and rather slender
branches which form a compact handsome round-topped head. The winter-buds and the
branchlets during their first year are clothed with soft ferruginous or pale hairs ; in their
second season the branchlets are slender, light or dark brown, marked, especially near their
extremities, with large pale lenticels, and conspicuous from the raised nearly circular leaf -scars.
The leaves are oblong or narrowly obovate, gradually contracted into long slender petioles,
and are rounded or short-pointed at the apex, entire, coriaceous, conspicuously reticulate-
veined, and dark green and lustrous on the upper, and pale and dull on the lower surface ;
they are three or four inches long, an inch to an inch and a half broad, with petioles an

1 Magnolia salicifolia, Maximowicz, Mel. Biol. viii. 509. — 2 Magnolia (Michelia) compressa, Maximowicz, Mel.
Franchet & Savatier, Enum. PI. Jap. i. 16. Biol. viii. 506. — Francket & Savatier, Enum. PI. Jap. i. 15.

Biirgeria (?) salicifolia, Siebold & Zuccarini, Fl. Jap.
Fam. Nat. i. 187. — Miquel, Prol. Fl. Jap. 144.

12 FOREST FLORA OF JAPAX.

inch in length, and fall, when a year old. after the appearance of the new shoots. The flowers,
which are very fragrant, are from an inch to an inch and a quarter across when expanded,
with pale yellow narrow ohovate sepals and petals, nearly sessile anthers, and a stipitate
head of pistils, the ovaries, according to Maximowicz, each containing five or six ovules.
The cone of fruit is two inches long, and is raised on a stalk half an inch or more in length :

O

the rusty brown thick-walled carpels, which are marked with large pale circular dots, usually
containing three seeds ; these arc broadly ovate and much flattened l>v mutual pressure.

Michelia compressa, which is the most boreal species of its genus, was discovered near
Nagasaki by Maximowic/., who saw a single tree. Oldham, an English botanist, collected it
near the base of Fuji-san, doubtless from a cultivated tree, and it is said to be found in
several places in the extreme southern part of the empire, although I have never seen it
except in the Botanic Garden of Tokyo, where there are several large trees, from which were
collected the specimens figured in our illustration. It is not improbable that Michelia is at
the northern limit of its range in Japan and that it will be found to be more at home on
the Loochoo Islands or on Formosa when the interesting flora of these islands is carefully

O .

explored. The fact that Michelia compressa flourishes in Tokyo, where our southern ever-
green Magnolia hardly survives, indicates that it may perhaps be grown as far north as
Washington and, possibly. Philadelphia, in southern England and Ireland, and on the west
coast of France, that is. in regions where no other species of this genus can exist in the open
ground, and where broad-leaved evergreen trees are rare and much desired.

The Japanese Illicium is a beautiful and interesting plant. It is the representative of a
genus with two species in our southern states and half a dozen others in India and southern
China, one of which, Illicium verum, supplies the star anise of the pharmacists. Illicium
anisatum. or as it should, perhaps, be called, Illicium religiosum, is a beautiful small evergreen
tree, fifteen or twenty feet high, with brilliant persistent leaves and small fragrant yellow
flowers. It is one of the sacred plants of Japan. Sicbold considered it a native of China or
Corea and an introduction by Buddhist priests into Japan. This may be the correct view,
although Japanese botanists now believe it to be a native of the southern part of the empire,
where Rein found it, as he supposed, growing wild. Sacred to Buddha, it is alwavs planted
in the neighborhood of his temples, and is common in private gardens as far north as the
thirty-fifth degree of latitude. The branches of this tree, especially when it is in flower, are
used to decorate the altars in the temples, or in cemeteries serve to mark the respect of the
living for the dead. From the powdered bark, mixed with resin, are prepared the "smoke
candles" with which incense is made in the temples, and with which the " moxa " is burned
on the human body as a sovereign cure for many of its ills.

The remaining Japanese plants of the Magnolia family, Kadsura and Schizandra, are
woody climbers, Kadsura Japonica is the type of a genus consisting of seven or eight species,
all natives ot southern and western Asia, and its most northern member growing spontane-
ously in the southern islands and at the sea-level in Hondo as far north as the thirty-iifth
degree ol latitude. The flowers are not showy, but it is a plant of extraordinary beauty in
the autumn, when the clusters of scarlet fruit are ripe, their brilliancy being heightened by
contrast with the dark green lustrous persistent leaves. There is a fine specimen of this plant

PLATE V.

C. K. FAXON- del.

MICHEUA COMPRESSA, MAXM.

THE MAGNOLIA FAMILY. 13

in the garden attached to the Agricultural College at Tokyo ; but I have never seen it in any
other, although it might well be grown wherever the climate is sufficiently mild to enable it
to produce fruit.

Schizandra is familiar to American botanists as one species, the type of the genus, Schi-
zandra coccinea, inhabits our southern states ; in Japan two species occur, and one of these,
Schizandra Chinensis, which grows also in Manchuria, carries the Magnolia family farther
north than any of its other members. It is a vigorous plant, with long twining stems, and
small unisexual white flowers, followed by clusters of brilliant red berry-like fruits, which in
September and October enliven the forests of Hokkaido, where this plant is extremely common.
Schizandra Chinensis is now well established in our gardens, flowering freely every year,
although in the neighborhood of Boston it has not ripened its fruit.

The second species, Schizandra nigra, is much less common in Japan than Schizandra
Chinensis, from which it may be distinguished by its broader leaves, larger flowers, and by its
blue-black fruit and pitted seeds. It grows in southern Yezo, where, however, I failed to
find it. Mr. Veitch collected it in September at Fukura, on the west coast of Hondo, and at
the end of October I found a single plant near Fukushima, on the Nakasendo, in central
Japan, from which I had the good fortune to gather a few ripe seeds ; this interesting plant
has not yet been brought under cultivation.

The forests of Japan are distinguished by the presence of a small family related to the
Magnolias, and composed of three genera, Cercidiphyllum and Trochodendron, which are
monotypic and endemic to Japan, and Euptelea, which has also a representative in the Hima-
laya forests. The plants of this genus are principally distinguished from the Magnolias by
their generally dioecious minute flowers, which are either entirely destitute of a perianth or are
furnished with a membranaceous four-lobed calyx. For this family the name of Trochoden-
dracece has been proposed.1

Cercidiphyllum Japonicum is the most important tree of this family, and the largest and
one of the most interesting deciduous trees of Japan, which more than any other of its inhab-
itants gives to the forests of Yezo their peculiar appearance and character (see Plate vi.).
Here it inhabits the slopes of low hills and selects a moist situation in deep rich soil, from
which the denseness of the forest and the impenetrable growth of dwarf Bamboos, which
covers the forest-floor, effectually check evaporation. In such situations the Cercidiphyllum
attains its greatest size, often rising to the height of a hundred feet, and developing clusters
of stems eight or ten feet through. Sometimes it forms a single trunk three or four feet in
diameter and free of branches for fifty feet above the ground ; but more commonly it sends
up a number of stems which are united together for several feet into a stout trunk, and
then gradually diverge. The trunk of a typical Cercidiphyllum of this form appears in the
frontispiece of this work ; it is the reproduction of a photograph made on a hill near Sapporo,
and represents a large but by no means an exceptionally large trunk, which at three feet
above the ground girted twenty-one feet and six inches.

In Cercidiphyllum the leaves on sterile shoots are either alternate or opposite ; in their
axils small acute red buds, covered with four to six thin scarious slightly imbricated scales,

1 Engler & Prautl, Pflanzenfam. iii. pt. ii. 21.

14 FOREST FLORA OF JAPAN.

are formed early in the autumn. The branchlet ends during the winter in a small orbicular
scar between two buds when the leaves are opposite, and at the side of a single bud when
the leaves are alternate. Early in the following spring the buds develop short spur-like
almost obsolete branches, which produce a single leaf and terminal flowers. Later a bud
is formed in the axil of the leaf which, on fruit-bearing trees, appears between the leaf and
the stalk of the fruit-cluster. The branches, therefore, in their second and third years,
appear to be clothed with opposite or alternate leaves, although the leaves are in reality
produced on lateral branches. The leaves are involute and coated on the lower surface in
the bud with pale caducous pubescence, and are furnished with lanceolate acute caducous
stipules slightly connate toward the base. The staminate and pistillate flowers are produced
on separate individuals ; the staminate are subsessile, solitary or fascicled, the pistillate solitary
and pedunculate. The staminate flower is composed of a minute scarious calyx, divided to
the base into four acute apiculate divisions, and of an indefinite number of stamens ; the
filaments are slender, elongated, and inserted on a conical receptacle ; the anthers are oblong-
lanceolate, attached at the base, apiculate by the prolongation of the narrow connective, and
two-celled, the cells opening longitudinally throughout their length. The pistillate flower is
composed of a membranaceous calyx divided into four unequal sepals laciniately cut on the
margins, and of four or sometimes of five or six carpels inserted by their oblique bases on a
prominent pyramidal receptacle ; they are gibbous and acute on the ventral suture, and
straight and rounded on the dorsal suture, and are gradually narrowed into elongated slender
styles stigmatic on their inner face below the middle ; the ovules are inserted in two rows on
the placenta, and are descending and anatropous. The fruit is a cluster of two to six more
or less spreading oblong stipitate follicles tipped with the persistent styles, and splitting
through the ventral suture, which by a twist usually becomes external. The pericarp is thick,
light brown, and lustrous, and separates into two layers ; the outer layer is thin and mem-
branaceous, and the inner layer is hard and woody, and lustrous on the inner surface. The
seeds, which are closely imbricated in two rows, are pendulous, compressed, nearly square,
attached obliquely, and covered with a thin, light brown membranaceous coat, which is
produced into an elongated terminal wing three times as long as the body of the seed and
slightly narrowed at the apex. The embryo is axile in copious fleshy albumen, with plane
cotyledons about as long as the slender superior radicle turned toward the liilnm.

The trunk of Cercidiphyllum Japonicum is covered with thick pale bark, deeply furrowed,
and broken into narrow ridges. Similar bark covers the principal branches ; these are very
stout, and issuing from the stem nearly at right angles, gradually droop, the slender red-
dish branchlets in which they end being often decidedly pendulous. The upper branches
and branchlets are erect, the whole skeleton of the tree showing, even in summer, through the
sparse small nearly circular leaves, which are placed remotely on the branches ; in the autumn
the leaves turn clear bright yellow. In port and in the general appearance of its foliage,
Cercidiphyllum, as it appears in the forests of Yezo, might, at first sight, be mistaken for a
venerable Ginkgo-tree, which in old age has the same habit, with pendulous branches below
and erect branches above ; but the trunk and its covering are very different in the two trees.

Cercidiphyllum Japonicum is distributed from central Yezo southward through nearly the

I

fei

o

a

—

«

THE MAGNOLIA FAMILY. 15

entire length of the Japanese islands. At the north it grows at the sea-level, and is very
common, but on the main island it is confined to high elevations, and is rare. Except in
Yezo, it seldom grows more than twenty or thirty feet high, and I never saw it, in Hondo,
below 5,000 feet elevation, where, as at Yumoto, in the Nikko Mountains, it is scattered
through the lower borders of the Hemlock forest.

Cercidiphyllum Japonicum is a valuable timber-tree, producing soft straight-grained light
yellow wood, which resembles the wood of Liriodendron, although rather lighter and softer,
and probably inferior in quality. It is easily worked, and in Yezo is a favorite material for
the interior finish of cheap houses and for cases, packing-boxes, etc. From its great trunks
the Ainos hollowed their canoes, and it is from this wood that they make the mortars found
in every Aino house and used in pounding grain. In New England, where there are now
trees twenty feet high, Cercidiphyllum is very hardy, and grows rapidly ; in its young state it
is nearly as fastigiate in habit as a Lombardy Poplar, the trunk being covered from the
ground with slender upright branches that shade it from the sun, which seems injurious to
this tree, at least while young. As an ornamental plant, Cercidiphyllum is only valuable for
its peculiar Cercis-like leaves, which, when they unfold in early spring, are bright red, and for
its peculiar habit, as the flowers and fruit are neither conspicuous nor beautiful (see Plate
vii.).

Of the other Japanese trees of this family, Euptelea polyandra is the least desirable as an
ornamental plant, and it will probably never be very much cultivated except as a botanical
curiosity. It is a small tree twenty to thirty feet in height, with a slender straight trunk
covered with smooth pale bark, stout rigid chestnut-brown branchlets, marked with white
spots, and wide-spreading branches, which form an open, rather unsightly head. The leaves
are thin, prominently veined, bright green, sometimes five or six inches long and broad, nearly
circular in outline, and deeply and very irregularly cut on the margins, with long, broad,
apical points; they are borne on long slender petioles, and turn to a dull yellow-brown color
before falling. The minute flowers appear in early spring before the leaves, and are produced
in three or four-flowered clusters from buds formed early in the previous autumn. They have
neither sepals nor petals, and consist of a number of slender stamens surrounding the free
clustered carpels. The fruit, which ripens in November, is not more showy than the flowers ;
it is a small stalked samara half an inch long, and furnished with an oblique marginal
membranaceous wing. The handsomest thing about this tree is the winter-bud, which is
obtuse, half an inch long, and covered with imbricated scales, which are bright chestnut-
brown, and as lustrous as if they had been covered with a coat of varnish. Euptelea poly-
andra is found in the mountainous forests of central Japan, usually on the banks or in the
neighborhood of streams between 2,000 and 3,000 feet above the sea-level, but does not
appear to be anywhere very common.

The third genus of this family, Trochodendron, like Euptelea, produces flowers without
sepals and petals. The only species, Trochodendron aralioides, is a small handsome glabrous
evergreen tree, with alternate broadly rhoniboidal crenulate penniveined leaves, four or five
inches long ; they are borne on elongated stout petioles, and are clustered at the extremities
of the branches. The flowers are produced in short terminal racemes, and consist of numerous

16 FOREST FLOEA OF JAPAN.

anthers raised on slender filaments and surrounding the carpels, which are connate in a
vertical series, and ripen into a small fleshy drupe crowned with the remnants of the persistent
styles. Trochodendron is a tree from fifteen to twenty-five feet in height, and is said to be
very common in some parts of the country, although I never saw it growing wild ; and it is
certainly not an inhabitant of the alpine forests or of Hokkaido, as stated in some works on
the Japanese flora, although, perhaps, it occurs in northern Hondo at the sea-level, as it is
hardy in the gardens of Nikko at an elevation of 2,000 feet above the ocean. Trochodendron
aralioides is often cultivated by the Japanese, and fine specimens of this tree are found scat-
tered through public and private gardens in Tokyo and Yokohama.

PLATE VII.

C. K. FAXON del.

CERCIDIPHYLLUM JAPONICUM, SIKB ET. Zucc.

THE TERNSTRCEMU, LINDEN, AND RUE FAMILIES.

THE chiefly tropical family, Ternstrueuiiaceae, which, in North America, is represented only
by Gordonia and Stuartia, trees and shrubs o£ the southern states, in Japan appears in eight
genera, in which are a number of interesting plants, although none of them become very large
trees. Of these, Camellia Japonica is horticulturally the most important, for its relative,
Camellia theifera, the Tea-plant, is evidently a Chinese or Assam introduction, and not a native
of Japan. In southern Japan the Camellia is a common forest-plant from the sea-level to an
altitude of 2,500 feet, on the east coast growing as far north as latitude thirty-six, and nearly
two degrees farther on the west coast. Here it is a dwarf bush only two or three feet high,
although where the soil and climate favor it, the Camellia becomes a tree thirty or forty feet
tall, with a handsome straight trunk a foot in diameter, covered with smooth pale bark hardly
distinguishable from that of the Beech. In its wild state the flower of the Camellia is red,
and does not fully expand, the corolla retaining the shape of a cup until it falls. In Japan,
certainly less attention has been paid to the improvement of the Camellia than in Europe and
America, although double-flowered varieties are known ; and as an ornamental plant it does
not appear to be particularly popular with the Japanese ; it is sometimes planted, however, in
temple and city gardens, especially in Tokyo, where it is not an uncommon plant, and where
beautiful old specimens are to be seen.

Tsubaki, by which name Camellia Japonica is known in Japan, is more valued for the oil
which is pressed from its seeds than for the beauty of its flowers. This oil, which the other
species of Camellia also produce, is used by the women in dressing their hair, and is an article
of much commercial importance. The wood of Camellia is close-grained, moderately hard,
and light-colored, turning pink with exposure ; it is cut into combs, although less valued for
this purpose than boxwood, and is manufactured into numerous small articles of domestic use.
Sasan-kuwa, Camellia Sasanqua, a small bushy tree of southern Japan and China, is perhaps
more commonly encountered in Japanese gardens than the Tsubaki, and in the first week of
November it was just beginning to open its delicate pink flowers in the gardens of Nikko,
although the night temperature was nearly down to the freezing point.

Ternstrcemia Japonica and Cleyera ochnacea are small bushy trees scattered from India to
southern Japan, where they are considered sacred by votaries of the Shinto religion, and are
therefore planted in the grounds of Shinto temples and in most private gardens. The ever-
green foliage of these two plants is handsome, especially that of Ternstrremia, but the flowers
and fruit possess little beauty, and they owe their chief interest to their association with
Japanese civilization.

Eurya Japonica is another member of the family, of wide range from Ceylon and India to
China, the Fejee Islands, and Japan, where it is exceedingly common in the southern islands
and in Hondo as far north, at least, as the Hakone Mountains. It is usually a shrub only a

18 FOREST FLORA OF JAPAN.

few feet high ; but I saw a specimen in the woods surrounding a temple near Nakatsu-gawa,
on the Nakasendo, which was fully thirty feet in height, with a well-formed trunk nearly a
foot in diameter. Eurya, although not particularly handsome, is interesting from the color of
the leaves, which are yellowish green on the upper surface and decidedly yellow below.

Stuartia is represented in eastern America by two handsome shrubs, one an inhabitant of
the coast region of the south Atlantic states, and the other of the southern Alleghany Moun-
tains ; in Japan there are two and, perhaps, three species. Of these, Stuartia monadelpha,
which inhabits also central China, appears to be a southern plant only ; at any rate, I saw
nothing of it in Japan, nor of the little known Stuartia serrata of Maximowicz. The third
species, Stuartia Pseudo-Camellia, is common on the Hakone and Nikko Mountains between
2,000 and 3,000 feet elevation, where it is a most striking object from the peculiar appearance
of the bark ; this is light red, very smooth, and peels off in small flakes like that of the Crape
Myrtle (Lagerstrcemia) ; to this peculiarity it owes its common name, Saru-suberi, or Monkey-
slider. Stuartia Pseudo-Camellia is often a tree of considerable size ; on the shores of Lake
Chuzenji we measured a specimen whose trunk at three feet from the ground girted six feet,
and which was upward of fifty feet high ; and specimens nearly as large are common on the
road between Nikko and Chuzenji. The flowers of this tree, which resemble a single white
Camellia, are smaller and less beautiful than the flowers of our coast species, Stuartia Vir-
ginica, but they are larger than those of the second American species, Stuartia pentagyna, a
handsome plant, which is not made enough of in our northern gardens, where it is perfectly
hardy and one of the best of the summer-flowering shrubs. Stuartia Pseudo-Camellia was
sent to America nearly thirty years ago by the late Mr. Thomas Hogg, and it appears to have
flowered in the neighborhood of New York several years before it was known in Europe,
where of late it has attracted considerable attention.1 In New England this Japanese species
appears perfectly hardy, and two years ago it flowered in the Arnold Arboretum.

Stachyurus praecox, another Japanese member of this family, is still little known in our
gardens, although it was one of the plants sent by Mr. Hogg to New York soon after the
opening of Japan to foreign commerce. It appears hardy in the neighborhood of New York,
as there is at least one plant established in Prospect Park, on Long Island. In Japan Stachyu-
rus is exceedingly common in the mountain forests and at the sea-level from southern Yezo to
Kyushu, appearing as a tall graceful shrub, with thin semiscandent branches and ovate-
lanceolate acute leaves. In summer or early autumn it forms axillary spikes of flower-buds
two or three inches long, and in very early spring, before the appearance of the leaves, these
buds expand into bell-shaped pale yellow flowers ; these are not more than a third of an inch
long, but they are produced in great profusion, and as they appear so early in the season
Stachyurus will probably prove a popular plant if it is found to flourish in cultivation. The
genus is represented in central China and in the Himalaya Mountains with a second species
described as a small tree.

The genus Actinidia, woody climbers of the Himalayas and eastern Asia, appears in Japan
in three species, of which two, at least, are exceedingly common and conspicuous features
of the mountain vegetation. Of these, the largest and most common, especially at the north,
1 See Rev. Hort. 1879, 430, t. — Gard. Chron. ser. 4, iv. 188, f. 22.— Bot. Mag. cxv. t. 7045.

THE TERNSTE(EMIA, LINDEN, AND RUE FAMILIES. 19

is Actiniclia arguta ; little need be said of this handsome plant, as it is now common and well
established in our gardens, where it grows with great vigor and rapidity, and where it is one
of the best plants of its class. We have heard a good deal of the value of the fruit of this
plant, which is depressed-globular, an inch across, and greenish yellow ; it is eaten in Japan,
but the flavor is insipid, and its merits appear to have been exaggerated. It was offered for
sale in the streets of Hakodate in great quantities, but, of course, green and hard, as the
Japanese use all their fruit before it ripens.

Actinidia polygama, although it inhabits Manchuria and Saghalin, and is common in the
forests of Hokkaido, is more abundant in those which cover the mountains of central Japan ;
it is a slenderer plant than Actinidia arguta, with elliptical acute slightly serrate long-stalked
leaves. The fruit is an inch and a half long, half an inch broad in the middle, and narrowed
at both ends; it is canary-yellow, rather translucent, soft and juicy, with an extremely
disagreeable flavor. Actinidia polygama does not, like Actinidia arguta, climb into the tops
of tall trees ; its weaker stems tumble about and form great tangles, sometimes twenty feet or
more across, and fifteen or twenty feet high. The most remarkable thing about this plant
is that in summer the leaves toward the ends of the branches become pale yellow, either over
their entire surface or only above the middle, not because they are drying up or ripening, but
apparently from an insufficient supply of chlorophyll. The effect that the plants produce at
this time is curious and interesting, and when seen from a distance growing on a mountain-
side or on the banks of a stream, they appear like huge bushes covered with pale yellow
flowers. This fine plant is still little known in cultivation, but if it flourishes in New England
like Actinidia arguta it will form a most valuable addition to our shrubberies.

Actinidia Kolomikta, which is found also in Manchuria and northern China, is much less
common in Japan than the other species. I saw it only on the rocky cliffs of a hill near
Sapporo, where it was growing with Rhododendrons and Menziesia, and where it was a deli-
cate, slender vine, with stems only a few feet in length. Unfortunately, there were no seeds
to be obtained, and I am doubtful if this species has ever been introduced into our gardens,
although the name often appears in nurserymen's catalogues.

In the forests of Japan are found two Lindens. They are both extremely common in
Yezo, but in the other islands are rare, and confined to mountain-slopes of considerable
elevation. The larger of the two, Tilia Miqueliana (see Plate viii.), is a handsome tree, often
growing in central Yezo to the height of one hundred feet, and forming a trunk four or five
feet in diameter. As it is only seen crowded among other trees in the forest, the branches are
short, and the head is oblong and rather narrow. The bark, like that of all the Lindens, is
broken by longitudinal furrows, and is light brown or dark gray. The young branchlets are
unusually stout for a Linden-tree, and in their first season are covered, as are the large ovate-
obtuse winter-buds, with hoary tomentum. The leaves are deltoid or deltoid-obovate, abruptly
contracted at the apex into broad points, obliquely truncate or subcordate at the base, and
coarsely and sharply serrate with incurved callous teeth ; they are four to six inches long,
three or four inches broad, rather light green, and more or less puberulous on the upper
surface, and pale and tomentose on the lower, especially on the prominent midribs and primary
veins, in their axils, and on the stout petioles, which are two or three inches in length. The

20 FOREST FLORA OF JAPAN.

peduncle-bract is rounded at the apex, sessile or short-stalked, three or four inches long, from
one third to two thirds of an inch broad, and, like the slender stems and branches of the
flower-cluster and its bractlets, covered with pale tomentum. The flowers appear in Sapporo
toward the middle of July. Like those of the American Lindens and of two species of eastern
Europe, Tilia petiolaris and Tilia argentea, which Tilia Miqueliana resembles in several partic-
ulars, the flowers are furnished with petal-like scales, to which the stamens, united in clusters,
are attached. The sepals are ovate-acute, tomentose on the two surfaces, especially on the
inner, and shorter than the narrow obovate petals. The style, like the stamens, is longer
than the petals, and is coated at the base with thick pale hairs, which also cover the ovary.
The fruit, which ripens in October, is ovate to oblong, wingless, and nearly half an inch long.
It is from the inner bark of this species that the Aiuos make their ropes.

Tilia Miqueliana 1 is comparatively little known, having at one time been confounded with
Tilia Mandshurica, which does not reach Japan. This noble tree will probably thrive in the
northern states, as plants which have been growing for a few years in the Arnold Arboretum
appear perfectly hardy. In Europe it is cultivated as Tilia Mandshurica (Kew), and as Tilia
heterophylla (Paris), although it does not appear to be much better known there than it is in
the United States.

The second Japanese Linden is a small tree, rarely growing more than fifty or sixty feet
tall in Hokkaido, where, perhaps, it is rather less abundant than Tilia Miqueliana. In books
it appears as Tilia cordata, var. Japonica ; but Tilia cordata is a synonym for Tilia ulmifolia,
a common European and north Asian species, so that unless the Japanese plant is found
specifically distinct, which is not probable, it should be known as Tilia ulmifolia, var. Japonica.
It is a round-headed tree with dark brown bark, slender red-brown branches, glabrous, like
the buds, even when young, and marked with oblong pale lenticels. The leaves are broadly
ovate or nearly orbicular, contracted at the apex into short or long broad points, and usually
cordate, or occasionally oblique, at the base, and sharply serrate with incurved callous teeth ;
they are membranaceous, light green and lustrous on the upper surface, light green, pale, or
nearly white on the lower surface, which is marked by conspicuous tufts of rufous hairs in the
axils of the principal veins, three or four inches long, and two or three inches broad. The
peduncle-bract is from three to three and a half inches long, and half an inch broad, with a
slender stalk sometimes an inch in length. The stem and branches of the flower-cluster are
slender and^ glabrous. The sepals, which are acute, slightly puberulous on the outer surface,
ciliate on the margins, and furnished on the inner surface at the base with large tufts of pale
hairs, are shorter than the narrow acute petals ; the ovary is clothed with white tomentum.
The fruit is oblong, or slightly obovate, and covered with rusty tomentum. The petaloid-
scales, which Maximowicz 2 found developed in some of the flowers of this tree, I have not
seen.

This is the only Linden cultivated by the Japanese, who occasionally plant it in temple
gardens, especially in the interior and mountainous part of the empire. It was introduced in

1 Tilia Miqueliana, Maximowicz, Mel. Biol. x. 585.

Tilia Mandshurica, Miquel, Prol. Fl. Jap, 206 (in part). — Franchet & Savatier, Enum. PL Jap. i. 67 (in part).
8 Mel. Biol. x. 585.

PLATE VIII.

C E. FAXON del.

TILIA MIQUELIANA, MAXM.

THE TERNSTRCEMIA, LINDEN, AND RUE FAMILIES. 21

1886 into the Arnold Arboretum, and has so far proved hardy. It is, however, scarcely
distinct enough from the European plant to make its cultivation as an ornamental tree particu-
larly desirable.

Elseocarpus, a genus of the Linden family distributed with many species through tropical
Asia, Australia, and the Pacific islands, is represented in Japan by two fine trees, found only
in the extreme southern part of the empire. Of these I saw only Elseocarpus photinisefolia, a
noble tree, planted in the gardens of a temple in the seashore town of Atami, where there are
the largest Camphor-trees in Japan and good specimens of a number of other southern trees.

The Rue family has a number of woody plants in Japan. Of these, Skimmia Japonica is
the only one which is much known in our gardens, although Phellodendron Amurense and
Orixa Japonica are now found in most large botanical collections. Evodia rutsecarpa, a shrub
or very small tree, with large, pinnate, strong-smelling leaves and terminal heads of minute
flowers, although not at all handsome, is an interesting plant, as it is from the bark that the
Japanese obtain the yellow pigment which they use in dyeing. For this purpose the bark,
which, with the exception of its thin brown outer coat marked with pale lenticels, is the color
of gamboge, is torn off in long strips, air-dried, and sent to the large cities. Evodia ruta3-
carpa, which also inhabits central China and the Himalayas, is now becoming rare in Japan,
and I saw it only on the coast near Atami ; it is said to be still abundant, however, in Aidzu
and on the peninsula of Yamato. The scarlet aromatic fruit is used by the Japanese in
medicine.

In Xanthoxylum there are four Japanese species. Of these, the most common, and the
most widely distributed at the north and in the mountainous regions of the main island at
elevations of from 2,000 to 4,000 feet above the sea, is Xanthoxylum piperitum. It is a
bushy shrub with many slender steins, or rarely a small tree with a well-developed trunk three
or four inches in diameter ; it is always a handsome plant, with dark or often nearly black
branchlets marked by pale spots and armed with stout straight spines ; with narrow unequally
pinnate leaves of about six pairs of ovate pointed leaflets, very dark green on the upper
surface and pale on the lower ; with small inconspicuous flowers and with heads of handsome
showy fruit four to six inches across ; the pods are rusty brown ; and the seeds, which do not
drop for some time after the pods open, are black and lustrous. The fruits of this plant are
gathered in large quantities by the Japanese before the pods open, and are used as a condiment
and in cooking, as we use pepper. In Hakodate and other northern towns it is commonly
exposed for sale throughout the year.

A nobler plant than Xanthoxylum piperitum, and certainly one of the most beautiful of
the genus, is Xanthoxylum ailanthoides, which I saw only on the Hakone Mountains, where it
is abundant, and near the coast at Atami. It is a round-topped broad-branched tree, some-
times fifty or sixty feet tall, with a trunk twelve to eighteen inches in diameter, covered with
pale bark, upon which the corky excrescences common in many species of this genus are well
developed. The branchlets are stout and pale, and are covered with shdrt stout spines. The
leaves vary from eighteen inches to four feet in length, and are unequally pinnate, with about
ten pairs of lateral leaflets and stout red-brown petioles; the leaflets are dark green, and
conspicuously marked on the upper surface with oil-glands, pale or nearly white on the lower

22 FOREST FLORA OF JAPAN.

surface, ovate-acute, often slightly falcate, long-pointed, rounded or subcordate at the base,
finely serrate, stalked, four to six inches long and two to four inches broad. The flowers,
which are greenish white and small, and inconspicuous like those of all the plants of this
genus, appear on the Hakone Mountains at the end of August or early in September in clus-
ters four or six inches across. The fruit I have not seen. In habit and in foliage this is
one of the most beautiful trees which I saw in Japan ; but as it does not range far north or
ascend to the high mountains, it is not probable that it will prove hardy in our northern
states. The other Japanese species of Xanthoxylum are shrubs of no great beauty or interest.
Simaruba?, a mostly tropical family, to which the familiar Ailanthus of northern China
belongs, appears in Japan only in Picrasma quassioides, a member of a small tropical Asian
genus, which, as an inhabitant of Yezo, seems to have strayed far beyond the limits of its
present home. As Picrasma quassioides appears in the forests near Sapporo, it is a slender
tree twenty to thirty feet in height, with a trunk about a foot in diameter. The branchlets
are stout, dark red-brown, and conspicuously marked by pale lenticels. The leaves are
unequally pinnate, with slender reddish petioles and four or five pairs of lateral leaflets, which
increase in size from the lower pair to the uppermost ; they are membranaceous, very bright
green, ovate-acute, finely serrate, stalked, three to five inches long, and an inch to an inch
and a half broad. The flowers, which are produced in loose, long-branched, few-flowered,
axillary clusters, are yellow-green, and not at all showy ; but the drupe-like fruit is bright
red and handsome in September, when the thickened branches of the corymb are of the same
color. It is, however, for the beauty of the color of its autumn foliage that Picrasma quas-
sioides should be brought into our gardens. The leaves turn early, first orange and then
gradually deep scarlet, and few Japanese plants which I saw are so beautiful in the autumn
as this small tree, which, judging from its northern home in Japan, may be expected to
flourish in our climate. It is a plant of wide distribution, not only in Yezo and Hondo, but
in Corea and in northern and central China ; it occurs on Hongkong and Java, and is
common on the subtropical Himalayas, which in Garwhal it ascends to an elevation of 8,000
feet above the ocean. To the bitterness of the inner bark, which in this particular resembles
that of the Quassia-tree of the same family, it owes its specific name.

THE HOLLY, EVONYMUS, AND BUCKTHORN FAMILIES.

JAPAN and eastern North America are equally rich in species of Holly, there being thirteen
or fourteen in each of the two regions. In Japan, however, Hollies grow to a larger size
than they do in North America, there being eight or nine trees in this genus in the Mikado's
empire, and only four in the United States ; and some of the Japanese Hollies are much
larger and far more beautiful than any of our species. The most beautiful of them ah1 is
certainly the southern Ilex latifolia, an evergreen tree now occasionally seen in the gardens of
southern Europe, where it was first carried more than fifty years ago. Although a native of
southern Japan, Ilex latifolia appears perfectly at home in Tokyo, where it is often seen in
large gardens and temple grounds, and where it occasionally makes a tree fifty to sixty feet in
height, with a straight tall trunk covered with the pale smooth bark which is found on the
stems of most plants of this genus. The leaves are sometimes six inches long and three or
four inches broad, and are very thick, dark green, and exceedingly lustrous. The large scarlet
fruit of this tree, which does not ripen until the late autumn or early winter months, and
which is produced in the greatest profusion in nearly sessile axillary clusters, remains on the
branches until the beginning of the following summer. Ilex latifolia is probably the hand-
somest broad-leaved evergreen tree that grows in the forests of Japan, not only on account of
its brilliant abundant fruit, but also on account of the size and character of its foliage. It
may be expected to prove hardy in Washington, and will certainly flourish in the southern
Atlantic and Gulf states.

Ilex integra is also a beautiful and distinctly desirable ornamental tree, often cultivated in
the temple gardens of Japan, where it frequently reaches a height of thirty or forty feet.
The leaves are narrow, obovate, three or four inches long, and apparently quite entire. The
fruit, which is rather long-stalked, is nearly half an inch in diameter, and very showy during
the winter. A variety of this species (var. leucoclada, Maximowicz), a shrub two to three feet
high, with narrower leaves and smaller fruit, is a northern form, growing as far 'north as
southern Yezo. On Mount Hakkoda, near Aomori, we found this plant in full flower and
with ripe fruit on the 2d of October, and secured a supply of the seeds, so that its hardiness
can be tested in the northern states. It must be remembered, however, that, although this
plant, and several other broad-leaved evergreen shrubs, including two or three species of
Holly, grow in Japan in a higher latitude than Massachusetts, they are protected, as Maximo-
wicz has already pointed out, during the winter by an undisturbed covering of snow, and are
not exposed, therefore, to the changes of climate which endanger the existence of many plants
in eastern America. In Japan, moreover, plants do not suffer from the summer and autumn
droughts, which often sap their vitality in the United States, and are often more directly
responsible for the apparent want of hardiness of many plants than intense winter cold.

A third Japanese evergreen species, Ilex rotunda, is also occasionally cultivated by the

24 FOREST FLORA OF JAPAN.

Japanese, although I saw only two or three specimens of this plant ; these were handsome
trees, thirty to forty feet in height, with well-formed trunks twelve to thirteen inches in
diameter. The leaves are broadly ovate to nearly orbicular, with entire thickened margins,
and are very dark green and lustrous, although not thick nor very coriaceous. The fruit is
smaller than that of the two species already mentioned, and rather oblong in outline.

A very distinct evergreen species, Ilex pedunculosa, is exceedingly common on the Naka-
sendo, the great central mountain road of Japan, in the valley of the Kisogawa. This plant
is sometimes a shrub two or three feet in height, and is sometimes twenty or thirty feet high,
when it is a well-formed tree, with a narrow round-topped head. The leaves are lustrous,
two to three inches long, ovate-acute, entire, and long-petiolate. The stems of the flower-
clusters, from which is derived its specific name and which are longer than the leaves, hold
the large bright-red fruit, which is solitary, or arranged in clusters of three or four, well
outside the leaves, giving to the plants a peculiar and beautiful appearance in the autumn.
Occasionally a tree of this species was seen in the garden of an inn on the Nakasendo ; but it
is evidently little known or cultivated in Japan, and apparently has not been introduced into
western gardens. Ilex pedunculosa will certainly flourish in western and southern Europe,
and I am not without hope that it will survive and possibly thrive in the northern United
States, as in Japan it is found at high elevations in a region of excessive winter cold.

Ilex crenata is the most widely distributed and the most common of the Japanese Hollies
with persistent leaves ; this plant is abundant in Hokkaido, on the foothills of Mount
Hakkoda, and on the sandy barrens near Giffu, on the Tokaido ; and I encountered it in
nearly every part of the empire which I visited. It is usually a low much-branched rigid
shrub, three or four feet high ; but in cultivation it not infrequently rises to the height of
twenty feet, and, assuming the habit of a tree, is not unlike the Box in general appearance.
The leaves, which are light green and very lustrous, vary considerably in size and shape,
although they are rarely more than an inch long, and are usually ovate-acute, with slightly
crenate-toothed margins. The black fruit is produced in great profusion, and in the autumn
adds materially to the beauty of the plant. This is the most popular of all the Hollies with
the Japanese ; and a plant usually cut into a fantastic shape is found in nearly every garden.
Varieties with variegated leaves are common and apparently much esteemed. Ilex crenata
and several of its varieties with variegated foliage were introduced into western gardens many
years ago and are occasionally cultivated, although the value of this plant as an under-shrub
appears to be hardly known or appreciated outside of Japan. Of the broad-leaved Japanese
evergreens, I have the most hope of success with Ilex crenata in this climate ; and if it proves
really hardy it will be a most useful addition to our shrubberies.

Ilex Sugeroki, another evergreen species quite unknown, I believe, in gardens, may be
expected to thrive in Europe, and possibly in the northern United States, as it is an inhabitant
of southern Yezo and northern Hondo, where on Mount Hakkoda we found it in fruit, and
were able to secure a supply of the seeds. It is a spreading bush five or six feet high, with
stout branchlets, light green ovate leaves an inch long, rounded at the apex and coarsely
crenulate-toothed above the middle, and with bright scarlet long-stalked solitary fruit half an
inch in diameter. Ilex Sugeroki is an unusually handsome plant in the autumn, and of consid-
erable horticultural promise.

THE HOLLY, EVONYMUS, AND BUCKTHORN FAMILIES. 25

Of the section of the genus with deciduous leaves (Prinos), represented in eastern North
America by the familiar Black Alder (Ilex verticillata) of our northern swamps and by the
arborescent Ilex Monticola of the Alleghany Mountains, there are several species in Japan.
The largest of these, Ilex macropoda, is a widely distributed, but not a common plant. I saw
it on the cliffs at Mororan on the shores of Volcano Bay, on the hills above Nikkd, and on
the flanks of Mount Koma-ga-take in central Japan, although only a single plant in each of
these widely separated localities. Ilex macropoda is a round-headed tree, twenty to thirty feet
in height, with a trunk sometimes a foot in diameter. It is a well-shaped handsome tree, with
stout branchlets furnished with short lateral spurs and ample, membranaceous, ovate-acute,
long-petioled leaves conspicuously reticulate-veined, which turn bright clear yellow in the
autumn, when they make a beautiful contrast with the bright red long-stalked fruit, which,
although not very large, is exceedingly abundant. Ilex macropoda grows not only far north,
as Professor Miyabe has recently written me of its discovery in the neighborhood of Sapporo,
but in the most exposed situations and at high elevations ; and there is no reason, therefore,
why it should not thrive in our northern states, where it may be expected to add considerably
to the beauty of shrubberies in the autumn and early winter.

A much more common plant than Ilex macropoda is Ilex Sieboldii, although this species
does not reach Hokkaido or ascend to high elevations on the mountains of Hondo. It much
resembles our North American Ilex verticillata and Ilex Isevigata, although much less beautiful
than either of these species, the fruit being smaller and less highly colored. Ilex Sieboldii is
a tall spreading shrub, very common in low grounds and near the borders of streams, with
slender stems often twelve or fifteen feet tall, small ovate-acute sharply serrate conspicuously
veined leaves, and small scarlet fruit clustered on the short lateral spur-like branchlets. In
the autumn the leafless branches of this shrub covered with fruit are sold in immense quan-
tities in the streets of Tokyo for the decoration of dwelling-houses, for which purpose they
are admirably suited, as the berries remain on the branches and retain their color for a long
time. Ilex Sieboldii was introduced many years ago into American gardens by the late
Thomas Hogg ; it is an old inhabitant of the Arnold Arboretum, where it now flowers and
produces its fruit every year. As an ornamental plant, however, it is less desirable than the
related American species, and it will probably only be cultivated in this country and in
Europe as a curiosity, or in botanic gardens.

The other Japanese Hollies with deciduous leaves, Ilex serrata, which is closely related to
and resembles Ilex Sieboldii, and Ilex geniculata, a rare shrub of the high mountains, with
black fruit, I was not fortunate enough to find.

The flora of Japan is rich in Evonymus, there being no less than nine species found within
the limits of the empire. Of these the best known in our gardens is the evergreen Evonymus
Japonicus, now cultivated in all temperate countries, and its climbing variety usually known
as Evonymus radicans. Evonymus Japonicus is a small tree generally distributed at low
elevations, and more common in the south than at the north, although it grows naturally in
the cold climate of southern Yezo, where, however, it does not attain a large size, and where
its presence may be accounted for by the thick covering of snow which protects it in winter.
The scandent variety is a hardier plant found carpeting the ground under the forests of

26 FOREST FLORA OF JAPAN.

Hokkaido, and in the mountain regions of Hondo climbing high on the trunks of trees, which
it encircles with great masses of lustrous foliage borne on stout branches standing out at right
angles sometimes to the length of several feet ; the leaves vary from an inch to four or five
inches in length and correspondingly in width, and show the connection of the climbing plant
with the arborescent type.

There is a second arborescent Evonymus in Japan, a variety of the widely distributed
and variable Evonymus Europseus, to which the name var. Hamiltonianus is given. This
handsome plant, with stout branchlets, large leaves, and showy fruit, was introduced from
Japan several years ago by the late Thomas Hogg, and it is now well established in the
Arnold Arboretum, where it flowers and fruits freely. It is one of the commonest of the
Japanese species in all mountain regions, and grows at least as far north as central Yezo,
where it becomes a tree twenty to thirty feet in height.

Evonymus alatus, a variable plant in the development of the wings on the branches, to
which it owes its specific name, and in the size of the leaves and fruit, in some of its forms, is
also very abundant in the north and on the mountains of central Japan. The wing-branched
variety, which is the only deciduous-leaved Evonymus which I saw in Japanese gardens, where
it is rather a favorite, is now well known in those of the United States and of Europe, where
it is valued for the peculiar pink color the leaves assume in very late autumn. The variety
subtriflora, a more northern plant, with slender terete branchlets and small fruit, is, I believe,
unknown in gardens. It is one of the commonest shrubs in the mountain forests of Japan,
and on the shores of Lake Chuzenji, in the Nikko Mountains, I saw it rising to the height of
fifteen or eighteen feet, with slender diverging stems.

In northern Japan there are three other species of Evonymus, all tall shrubs, with large
leaves and large showy fruit suspended on long slender stalks, which may be expected to
thrive in our climate, and to be decided acquisitions in our shrubberies. Of these Evonymus
Nipponicus and Evonymus oxyphyllus produce globose fruit, and Evonymus macropterus
more or less broadly winged fruit.

Of Celastrus nothing need here be said of the now well-known Celastrus articulatus, which
is one of the commonest plants on the mountains of Japan, except that its leafless branchlets,
covered with fruit, are sold in the autumn in great quantities in all Japanese towns, where
they are used in house decoration, for which purpose they are admirably suited, as the bright-
colored fruit remains on them for many weeks. The second Japanese species, Celastrus
flagellaris, I saw only in the Botanic Garden in Tokyo, where there is a single small plant; it
is a common Manchurian species, but appears to be exceedingly rare in Japan. I judge that
it has no particular horticultural value.

Half a dozen genera of Rhamnacea? are included in the flora of Japan, among them Zizy-
phus, perhaps an introduced plant, often cultivated as a fruit-tree ; Berchemia racemosa, a
twining shrub with long slender branches, very ornamental during the last weeks of summer,
when the half-ripened fruit, which is produced in large terminal clusters, is bright red ; two or
three species of Rhamnus, of no horticultural value, and the curious tree, Hovenia dulcis, an
inhabitant also of China and the Himalaya region, and in Japan often cultivated for the thick-
ened sweetish fruit-stalks, which are edible, although insipid in flavor, and which enjoy among

THE HOLLY, EVONYMOUS, AND BUCKTHORN FAMILIES. 27

the Japanese a certain reputation for curative properties. Hovenia was first introduced into
Europe eighty years ago, and is occasionally seen in the gardens of southern France and
Italy and of our middle states. In general appearance this tree, which is sometimes thirty
or forty feet in height, is not unlike a large-leaved Pear-tree, and as an ornamental plant
possesses little value.

THE MAPLE FAMILY.

IN arborescent plants of the family of Sapinclaceae, Japan is richer than eastern America,
owing to the multiplication of species of Maple in the former country. ^Esculus, on the
contrary, which finds its headquarters in North America, where there are five species, appears
in Japan in only one, — ^Esculus turbinata. This, however, is a noble tree, — one of the
largest and stateliest of all Horse-chestnuts. In the forests of the remote and interior moun-
tain regions of central Hondo, at elevations between 2,000 and 3,000 feet, Horse-chestnuts,
eighty to one hundred feet tall, with trunks three or four feet in diameter, are not uncommon.
These were perhaps the largest deciduous trees which I saw on the main island growing natu-
rally in the forest, that is, which had not been planted by men, and their escape from destruc-
tion was probably due to their inaccessible position and to the fact that the wood of the
Horse-chestnut is not particularly valued by the Japanese. In habit and in the form, venation,
and coloring of the leaves, the Japanese Horse-chestnut resembles the Horse-chestnut of our
gardens, the Grecian J^sculus Hippocastanum, and at first sight it might easily be mistaken
for that tree, but the thyrsus of flowers of the Japanese species, which is ten or twelve inches
long and only two and a half to three inches broad, is more slender ; the flowers are smaller
and pale yellow, with short, nearly equal petals ciliate on the margins ; and the fruit is that
of the Pavias, being smooth and showing no trace of the prickles which distinguish the
true Horse-chestnuts. The Japanese Horse-chestnut reaches southern Yezo, finding its most
northern home near Mororan, on the shores of Volcano Bay, at the level of the ocean ; it is
generally distributed through the mountainous parts of the three southern islands, sometimes
ascending in the south to an elevation of 4,000 or 5,000 feet. There seems to be no reason
why this tree, which has already produced fruit in France, should not flourish in our northern
states, where, as well as in Europe, it is still little known. In northern Japan the fruits are
exposed for sale in the shops, although they are probably used only as playthings for the
children.

To the Maples the forests of Japan owe much of their variety, beauty, and interest. Not
less than twenty species are known in Japan, while in all of North America there are only
nine, with six on the eastern side of the continent. None of the Japanese Maples, however,
grow to the size of real timber-trees, or can be compared in massiveness and grandeur with
some of the American species, which are unrivaled in size and beauty by the Maples of any
other part of the world.

Some of the Japanese Maples are exceedingly common and form a conspicuous feature of
the forest vegetation, and others are rare and confined to comparatively small regions. Sev-
eral of the species I did not see at all, and of others only one or two isolated individuals. The
most common of the Japanese Maples, and the largest, is Acer pictum, a handsome small tree,
not unlike our Sugar Maple in general appearance ; it is one of the most abundant trees in

THE MAPLE FAMILY. 29

the forests of Hokkaido, where it occasionally attains the height of fifty feet, and forms a
trunk eighteen inches in diameter. It is a tree of wide and general distribution in Japan,
Manchuria, China, and northern India, and even in Japan varies remarkably in the size and
pubescence of the five to seven-lobed leaves truncate at the base, and in the size and shape of
the fruit. This tree must be extremely beautiful in May, when the yellow flowers are just
opening, for the large lengthened inner scales of the winter-buds are then bright orange-
color and very showy. The autumn coloring of the leaves I did not see ; it is described as
yellow and red.

Of more interest to the lovers of novelties is Acer Miyabei (see Plate ix.), the latest
addition to the list of Japanese Maples. It is a tree thirty to forty feet in height, with
a trunk twelve to eighteen inches in diameter, covered with pale deeply furrowed bark,
spreading branches which form a round-topped handsome head, and stout branchlets orange-
brown in their first, and ashy gray in their second season. The leaves are five-lobed by
narrow sinuses, with acute entire irregularly two to three-lobed divisions, and are cordate or
almost truncate at the base, five-ribbed, conspicuously reticulate-veined, puberulous on the
ribs and in their axils on the upper surface, and more or less covered with ferrugineous pubes-
cence on the lower surface, especially on the ribs and veins ; they are dark green above, pale
below, and four or five inches long and broad, and are borne on stout petioles enlarged at
the base, two to seven inches in length, and thickly coated while young with pale hairs, which
also cover the unfolding leaves. The flowers, which are yellow, are produced on slender pedi-
cels in few-flowered, short-stalked corymbs. The sepals and petals are narrow, obovate, acute
and ciliate on the margins ; in the male flowers the stamens, composed of filiform filaments
and minute ovate anthers, are inserted between the lobes of a conspicuous disk, and are longer
than the petals ; the pistil is minute and rudimentary ; in the fertile flowers the stamens are
rudimentary and shorter than the ovary, which is coated with long white hairs. The style,
which is described as somewhat shorter than the revolute stigmas, is caducous. The fruit is
two inches long, with broad puberulous nutlets diverging at right angles to the stem, and
thin, slightly falcate, conspicuously veined wings. This fine tree, which is closely related to
the European Acer platanoides, was discovered a few years ago in the province of Hidaka, in
Hokkaido, by Professor Kingo Miyabe, the accomplished professor of botany in the college at
Sapporo and the author of an important work on the flora of the Kurile Islands, in whose
honor it was named in 1888 by Maximowicz.1

On the 18th of September we stopped quite by accident to change cars at the little town of
Iwanigawa, a railroad junction in Yezo, some forty or fifty miles from Sapporo, and, having a
few minutes on our hands, strolled out of the town to a small grove of trees in the hope that
they might prove interesting. In this grove, occupying a piece of low ground on the borders
of a small stream, and chiefly composed of Acer pictum, our Japanese guide recognized at a
glance a number of fine trees of Acer Miyabei covered with fruit, and surrounding the house
of an officer of the Imperial Forest Department, who had been living for years in entire
ignorance of the fact that he was enjoying the shade of one of the rarest trees in Japan.
The find was a lucky one, for Iwanigawa is a long way from the station where this species

1 Mel. Biol. xii. 725.

30 FOEEST FLOE A OF JAPAN.

had been discovered, and mature fruit had not been seen before ; and from these trees I
obtained later from Professor Miyabe a supply of seeds large enough to make this Maple
common in the gardens of this country and of Europe, in which there is every reason to believe
that it will flourish.

In the forests of Yezo eight other species of Maple occur. Among them, growing only in
the extreme north and on the high mountain-slopes, are a variety of our Mountain Maple, Acer
spicatum, so like the New England form of this common tree that it is difficult to distinguish
the two plants, and Acer Tataricum, var. Ginnala, a common Manchurian tree, not rare in
northern Japan, where it grows in low wet ground, near the borders of streams. This little
tree is now well established in American gardens, in which it might be seen more often
to advantage, as its flowers are very fragrant, and the leaves of few trees take on more
splendid autumnal colors. In Yezo, too, Acer capillipes has been found ; this is a species
with small racemose flowers, and thin delicate nearly circular lobed leaves, deeply cut on the
margins. On Mount Hakkoda, in northern Hondo, where Acer capillipes is extremely abun-
dant at elevations of 2,000 to 3,000 feet above the sea, we found it in October, growing as a
stout bush or bushy tree, twelve or fifteen feet in height, with delicate canary-yellow leaves,
and secured a supply of ripe seeds.

In Yezo, Acer Japonicum and Acer palmatum are both common ; these species, next after
Acer pictum, are the most generally distributed Maples in Japan, and the only species which
the Japanese cultivate at all commonly. They are both small trees, rarely, if ever, exceeding
a height of fifty feet, and both, as is well known, vary remarkably in the size, form, and
cutting of their leaves. A few of the varieties of Acer palmatum, particularly the one on
which the leaves are divided into narrow lobes, and the one with pendulous branches, are
favorites in Japan, where few of the numerous and monstrous forms of this tree, with which
we have become familiar of late years in this country, are seen outside of nursery-gardens
with foreign connections. Of these two trees the autumn foliage of Acer Japonicum appears
the more brilliant ; and some individuals of this species which we saw in October, high up
on Mount Hakkoda, were as beautiful in color as a good American Scarlet Maple. These
two trees have not proved very satisfactory in this country, where they have a way of dying
in summer without apparent cause. This is due, perhaps, to the fact that nearly all the plants
brought here have been raised from degenerate nursery-stock, obtained in or near the treaty
ports ; and it will be interesting to watch the behavior here of plants raised from seed gath-
ered in the forests of Yezo. For us these Maples have the advantage of retaining their leaves
later in the autumn than our species, which are bare of foliage before the Japanese trees
assume their brilliant colors ; and this is true of many other Japanese and Chinese plants, like
Ampelopsis tricuspidata and Spiraea Thunbergii, for the autumn in eastern Asia is fully a
month later than it is in this country.

Acer carpinifolium, which is occasionally seen in our gardens, is evidently extremely rare
in Japan. There are a few plants in one of the temple gardens in Nikkd, and I saw a single
wild specimen hanging over the bank of a stream in the mountains above Fukushima, on the
Nakasendo, and was fortunate in obtaining from it a good supply of seeds. In Nikko, Acer
carpinifolium is a handsome round-topped tree, perhaps thirty feet tall. It is well worth

PLATE IX.

C. E. Ftxos del.

ACER MIYABEI, MAXM.

THE MAPLE FAMILY. 31

growing for its beauty as well as for the unusual form of the leaves, which resemble those of
the Hornbeam, for which, at first sight, it might easily be mistaken.

Acer Tschonoskii is common near the margins of Lake Chuzenji in the Nikko Mountains,
and a thousand feet higher is found as a common shrub in the Hemlock forests which cover
the slopes rising from Lake Yumoto. It is a small bushy tree, perhaps twenty feet tall, with
bright red twigs and ample leaves, not unlike those of Acer capillipes in shape and cut-
ting, although in autumn they turn deep scarlet. We could not find a single seed of this
pretty plant, which has probably never been cultivated. Acer rufinerve, hardly distinguish-
able from the Moosewood of our northern forests (Acer Pennsylvanicum), and Acer cratsegi-
folium, both familiar now in our gardens, are rather common, especially the latter, in all the
mountain regions of central Japan, and need no mention here.

Among the rarer and less known species we found Acer diabolicum rather common in the
neighborhood of Nikko, where it is a round-topped tree twenty to thirty feet tall, very like
the European Sycamore Maple in habit and general appearance, with dull yellow-green leaves
four or five inches across, which apparently do not change color before falling, and large
dirty brown fruit covered on the nutlets with fine stinging hairs. This seemed the least
beautiful of the Maples which we encountered in the forests of Japan. Acer distylum I only
saw in the Botanic Garden in Tokyo, and Acer pycnanthum, Acer purpurascens, Acer argu-
tum, Acer parvifolium, and Acer Sieboldianum, the last, probably, only a pubescent-leaved
variety of Acer Japonicum, I looked for in vain.

Of Maples of the section Negundo, with the male and female flowers on separate plants and
pinnate or ternate leaves, there are two species in Japan, — Acer cissifolium and Acer Niko-
ense. The first is said to be common, and widely distributed from southern Yezo through
the mountain ranges of the main island, but I only saw a few small plants in hedge-rows near
Nikko, none of them half the size of specimens which may be seen in some Massachusetts
gardens, where Acer cissifolium is a handsome compact round-headed little tree with slen-
der graceful leaves, of a delicate green in summer, and orange and red in late autumn, and
where it is one of the most distinct and satisfactory of the Japanese trees which have been
tried in our climate.

The second Japanese Negundo (see Plate x.), as it appears in the forests of Japan, is a
distinct and beautiful tree, which, if it thrives in this country, will be a real addition to our
plantations. Acer Nikoense grows to a height of forty or, perhaps, fifty feet, with a trunk
twelve to eighteen inches in diameter covered with smooth dark slightly furrowed bark, and
stout rather slender branches, which form a narrow round-topped head. The branchlets are
thick and rigid, and are coated at first, like the inner scales of the ovate-acute winter-buds, the
young leaf-stalks, the under surface of the young leaflets, the peduncles and pedicels, with
short thick pale or rufous villous tomentum ; at the end of their first season the branchlets
are dark red-brown and are marked with numerous minute lenticular dots. The leaves are
ternate, with stout rigid petioles an inch or an inch and a half in length, and ovate or obovate,
acute, long-pointed, entire, or remotely and irregularly coarsely and crenately serrate leaflets,
the terminal leaflet being long-stalked, symmetrical, and wedge-shaped at the base, the lateral
leaflets rounded on the lower, and oblique on the upper edge at the base, and sessile or nearly

32 FOREST FLORA OF JAPAN.

so. The leaflets are thick and rather rigfid, two and a half to five inches lonsr, an inch and a

O * O'

half to two inches broad, conspicuously reticulate venulose, dark yellow-green on the upper
surface, pale and coated on the lower surface with pubescence, which is rufous on the stout
midribs and broad straight veins ; or sometimes they are bright green on the lower surface,
and glabrous, except on the midribs and veins. In the autumn the leaflets turn brilliant
scarlet on the upper surface, but remain pale on the lower. The flowers are yellow, half an
inch across, and nodding, and are borne in short few, usually three-flowered, subsessile terminal
corymbs on slender graceful pedicels. The sepals and petals are ovate or obovate, rounded at
the apex, and contracted at the base into narrow claws ; in the sterile flower, in which the
ovary is reduced to a minute rudiment, the stamens, which are inserted between the lobes of
the conspicuous disk, are exserted ; the filaments are filiform, and the anthers are large and
oblong ; in the fertile flower the stamens are rudimentary, and not longer than the ovary,
which is coated with thick pale tomentum and crowned with a long stout style with revolute
stigmas. The fruit is three inches long, with remarkably thick and hard-walled puberulous
nutlets, and broad falcate diverging or converging obovate Avings rounded at the apex.

Acer Nikoense is not a common, although a widely distributed, species. I saw a number
of plants in the temple grounds of Nikko and on the road between Nikko and Lake Chuzenji,
a single tree near Agematsu on the Nakasendo, and ten or twelve more on the Yusui-toge
above Yokokawa. According to Maximowicz,1 who distinguished this tree nearly thirty
years ago, it grows as far south as Nagasaki. Acer Nikoense is practically unknown in gar-
dens, although a well-grown specimen exists in the Veitchian collection in London, and a
single small plant was sent from Japan two years ago to the Rixdorf Nurseries in Berlin. A
figure of a leaf taken from this plant has been published in a German periodical.

In September we hunted the Nikko hills in vain for a seed-bearing tree, and had given up
all hope of introducing this species. One day late in October, however, we sat down on the
rocks in the bed of a torrent far up on the side of Mount Koma-ga-take, in central Japan, to
eat our luncheon, when our attention was attracted by some large Maple-seeds which were
new to us floating in a pool at our feet. A search on the bank above discovered a single tree
of Acer Nikoense, from which the wind was scattering showers of seeds. If we had been a
day later, or had selected another resting-place, we should have missed one of the best harvests
we made in Japan, as this single tree yielded at least half a bushel of good seeds.

If Acer Nikoense proves hardy and flourishes in our gardens, it will be particularly
remarked for the brilliancy of its autumn leaves, which are not surpassed in beauty by those
of any other tree which I saw in Japan, and which, unlike those of most trees, are only bright-
colored on one surface.

1 Mel. Biol. vi. 370 ; x. 609 ; Bull. Acad. St. Petersburg, Acer Maximowiczianura, Miquel, Arch. Neer. ii. 473,

t. 76. — Franchet & Savatier, Enum. PL Jap. i. 90.— Pax, 478.
Bot. Jahrb. vii. 205 j Gartenflora, xli. 149.

PLATE X.

C. E. FiXON del.

ACER NIKOEXSE, MAXM.

THE SUMACHS AND THE PEA FAMILY.

IN eastern North America the small family of the Sabiacese has no representative,
although Meliosma, which is mostly a tropical and subtropical Asiatic genus, also occurs in
Mexico and Central America. In Japan there are three species of this genus, of which only
one, Meliosma myriantha, attains the size of a tree. This species grows sometimes to the
height of twenty-five or thirty feet, and produces slender trunks and wide-spreading branches ;
its large thin leaves, which are sometimes eight inches long and three inches broad, of a light
delicate green, are its chief attraction as a garden-plant, for the flowers of Meliosma are
minute, and the terminal panicles in which they are gathered are loose and long-branched.
Only a small portion of the flowers are fertile, so that the fruit, which is a small red berry-
like drupe, is sparse and scattered in the clusters, and not at all showy. This plant is new, I
believe, in cultivation, and its behavior in our climate will be watched with interest. It can
hardly be hoped, however, that Meliosma myriantha will succeed in New England, as in Japan
it does not range far north, and in central Hondo, where, although widely distributed, it is
not common, it does not rise much above 2,500 feet over the sea-level.

From the Rhus family we miss in Japan the Smoke-tree (Cotinus), a familiar European and
western Asiatic type, represented, too, in eastern America by one of the rarest and most local
of all our trees. Of the true Rhuses we have in eastern America a dozen species, including
three small trees, while in Japan there are five indigenous species, and among them three
which can properly be considered trees. The Japanese Lacquer-tree (Rhus vernicifera), which
has played a conspicuous part in the development of the mechanical arts of China and Japan,
and which is certainly the most valuable plant of the genus to man, is not a native of Japan,
where it was carried long ago from China, and although much cultivated, especially in north-
ern Hondo, I saw no indications that it is growing spontaneously or anywhere establishing
itself in the forest.

The Japanese Rhuses are not as ornamental in the autumn as our Sumachs, as none of
them bear fruit covered with the long red hairs which give to the fruit-clusters of the Amer-
ican plants their dense appearance and brilliant color ; but the flowers of the Asiatic Rhus
semialata, a common small tree distributed from the Himalayas to Japan, which are white,
and produced in large terminal panicles, are much more beautiful than the yellow-green flow-
ers of any of our Sumachs, and in August and September, when this tree blossoms in Japan,
it is a striking object in the shrubby coppice-growth which so often covers the low mountain-
slopes. In autumn Rhus semialata is one of the most brilliantly colored plants of the Jap-
anese forest ; and very few Japanese plants succeed so well in our climate. It is from a gall
formed on the leaf of this tree that the dye with which married women in Japan discolor
their teeth, as a sign of domestic bondage, is obtained.

Economically a more important tree, as from it the Japanese obtained their principal

34 FOREST FLORA OF JAPAN.

supply of artificial light before the introduction of American and Russian petroleum, Rhus
succedanea is less interesting in flower, at least, than Rhus semialata ; it is a southern species,
still much cultivated on the southern islands, and in Tokyo seen only in gardens. In habit,
•although it grows to a larger size, it much resembles our Stag-horn Sumach ; the leaflets are
narrower, the flowers are produced in slender few-flowered clusters pendulous in fruit ; and
the drupes covered with a thick coat of the pale waxy exudation, to which this species owes
its name and value, are much larger. Rhus succedanea will, no doubt, flourish in the south-
ern states, and it is not improbable that it will prove hardy as far north as Philadelphia ; it
will certainly never be grown, however, in the United States for the wax it might be made to
yield, and as an ornamental plant, while it is, of course, interesting, it is inferior to the Amer-
ican Sumachs.

Rhus trichocarpa, which, so far as I know, is not in our gardens, should be cultivated for
the extraordinary beauty and brilliancy of the leaves in autumn, when they assume the
brightest scarlet and orange tints. It is a slender tree, sometimes twenty or twenty-five feet
high, and very common in the forests of Yezo and on the mountains of central Hondo. The
leaves are eighteen to twenty inches long, with dark red puberulous midribs and broadly
ovate long-pointed short-stalked membranaceous leaflets, slender panicles of flowers, which
open in July, and pendulous fruit-clusters, with large pale prickly drupes ripening in August
or early in September. Neither the flowers nor the fruit are attractive, and there is nothing
very distinct in the appearance of this tree, except in the autumn, when, however, it is so
beautiful that if it succeeds here I believe it will prove one of the best introductions of recent
years.

Of the poisonous species of Rhus I did not see the pinnate-leaved Rhus sylvestris, which
is said to be a small shrub and a native of the southern part of the empire ; on the Hakone
Mountains, where it is reported to grow, I looked for it in vain ; but our Poison Ivy is one of
the common plants in all the central parts of Hondo and in Yezo, where it grows to its largest
size and climbs into the tops of the tallest trees. The leaves of the Japanese plant are larger
than they usually grow on the American form ; they are thicker, too, and more leathery, and
turn to even more brilliant autumn colors, often to deep shades of crimson, which are rarely
seen on this plant in America. In October no other vine is so handsome in Japan.

Japan is remarkably poor in arborescent Leguminosse, with only three species in three
genera, while here in eastern America there are twenty species in a dozen genera. The best
known Japanese tree of the family is Albizzia Julibrissin, a small Mimosa-like tree which
grows from Persia to Japan, and through cultivation has become naturalized in our southern
Atlantic states and in most other warm temperate countries. Familiar now in this country is
Maackia Amurensis, which, introduced many years ago from the valley of the Amour, is now
sometimes cultivated in northern gardens. This little tree, which, under favorable conditions,
rises occasionally to the height of thirty or forty feet, is common in all the forest regions of
northern Japan, and is not rare on the mountains of central Hondo. The Japanese form
produces larger and more numerous flower-spikes and larger fruit than the mainland tree, as
we see it in this country ; and it is not improbable that it will prove a more desirable garden-
plant. In Yezo, the wood, which is hard, close-grained, and pale brown in color, is manufac-
tured into many small objects of domestic use, and is considered valuable.

PLATE XL

C. E. FAXON del.

GLEDITSIA JAPONICA, MIQ.

THE SUMACHS AND THE PEA FAMILY. 35

The third Japanese leguminous tree is Gleditsia Japonica,1 which, in most essential charac-
ters, resembles our North American Gleditsia triacanthos, but the leaflets are broader and
more lustrous, and the bark, instead of being dark brown, is quite pale. Although it does
not grow to the great size of the American species, the Japanese Gleditsia is, perhaps, a
more beautiful tree. It is (see Plate xi.) a tree sixty to seventy feet in height, with a trunk
occasionally three feet in diameter, and stout branches horribly armed with flattened, often
branched lustrous red-brown spines, two or three inches in length. The branchlets are
remarkably stout as compared with those of our species, and are covered with bright green
bark, marked with orange-colored elevated lenticular spots. The leaves are ten to twelve
inches long, with broad ovate-acute remote leaflets, or they are sometimes bipinnate as on our
species, with smaller leaflets. The male flowers (the female inflorescence I have not seen) are
very similar to those of Gleditsia triacanthos, although they are rather larger and the racemes
are longer and less closely flowered. The pods are compressed and thin-valved, like those of
our northern tree, ten or twelve inches long and an inch and a half broad, but the seeds,
instead of being placed close to the ventral suture of the pod, are sometimes nearer the
middle and surrounded by the pulp, which is more abundant in the Japanese than in the
American species. This pulp is used by the Japanese in washing cloth, and long strings of
the pods are displayed for sale in many towns of northern Japan, where Gleditsia Japonica
grows, not very abundantly, according to my observations, near the banks of streams at the
sea-level. It is common, and reaches its largest size on the banks of the Kisogawa and other
streams of central Japan, at an elevation of some two thousand feet. Here it grows sometimes
close to the water's edge, in rich humid soil, or is as often found at a considerable distance
above the water, growing on dry gravelly slopes. By Rein this tree is said to be often
planted in the neighborhood of villages in Japan, but I saw no specimens, except in the
scientific gardens of Tokyo, which did not have the appearance of growing naturally.

As an ornamental tree, Gleditsia Japonica, as it appeared on the mountains of Japan, is a
more beautiful tree than any of the species common in cultivation, and it may be expected to
become a valuable addition to the list of exotic trees suitable for the decoration of the parks
and avenues of the United States and Europe.

1 Gleditsia Japonica, Miquel, Prol. Fl. Jap. 242. — Franchet & Savatier, Enum. PI. Jap. i. 114 ; ii. 325. — Maximowicz,
Mel. Biol. xu. 452.

THE ROSE FAMILY.

TREES of the Rose family in the flora of Japan are not numerous as compared with that
of eastern America, and among them there is not one of first-rate value as a timber-tree.
Horticulturally they are more important, and Japanese gardens owe much of their interest to
species of Prunus. Although the most popular garden-tree in Japan, Primus Mume is proba-
bly not Japanese at all, but a native of Corea, where Mr. Veitch found it planted as a shade-
tree along the borders of the highroads. This is the tree which all foreign writers upon
Japan speak of as the Plum, although it is really an Apricot. In cultivation Prunus Mume
produces white, rose-colored, or red, and often double flowers, which appear before the leaves
in February and March, and are revered as harbingers of spring. The Mume is planted in
nearly every Japanese garden of any pretensions, and is one of the most universally used pot-
plants. Care and labor are often expended in producing dwarfed, contorted, or pendulous-
branched specimens, which sometimes command what seem exorbitant prices.

A more interesting tree than Prunus Mume is the Japanese Cherry, Prunus Pseudo-Cerasus,
the largest tree of the Rose family in ^he empire, and, next to the Apricot, more cultivated
for flowers by the Japanese than any other tree. In the forests of Yezo, Prunus Pseudo-
Cerasus occasionally rises to the height of eighty feet, and forms a trunk three feet in diam-
eter. In the character of the bark, in habit and general appearance, it much resembles the
European Cherry, the wild type of the familiar Cherry-tree of our gardens and orchards, and
as it appears in the forest it might well be mistaken for that species. The Japanese Cherry is
common in Yezo and in all the mountain regions of Hondo up to 5,000 or 6,000 feet above the
sea-level, and often forms a considerable portion of the forest-growth, although, in Hondo, all
large trees appear to have been cut. In the early autumn it is conspicuous in the landscape
and very beautiful, as the leaves turn deep scarlet and light up the forest before the Maples
assume their brightest colors. For centuries the Japanese have planted these Cherry-trees
in all gardens and temple grounds, and often by the borders of highways, as at Mukojima,
near Tokyo, where there is an avenue of them more than a mile in length along the banks of
the Sumi-da-gawa, and at Koganei, where, a century and a half ago, 10,000 Cherries were
planted in an avenue several miles long. The flowering of the Cherry-tree is an excuse for a
holiday, and thousands of men, women, and children pass the day under these long avenues
in more or less hilarious contemplation of the sheets of bloom. The flowers of the wild tree
are single, white, and of the size of those of the garden Cherry, but, not unnaturally, many
varieties have been produced during the centuries it has been a garden-plant. Bright red
and pink single-flowered varieties are common in Japan, as well as many double-flowered
forms. Of these several have been introduced into this country and Europe, and are now
well known in our gardens, where, however, they do not flower as freely as they are repre-
sented to flower in their native land. Prunus Pseudo-Cerasus is a plant of cold climates, and

PLATK XII.

C. E. FAXON deL

PRUNUS MAXIMOWICZII, RUPR.

THE HOSE FAMILY. 37

great summer heat evidently does not suit it, as in Tokyo the trees never grow to a great size,
and by midsummer are leafless ; so that, except during the short blooming season, the exces-
sive use of this tree is a real injury to the appearance of the gardens and promenades of the
capital and of other southern cities.

Prunus Pseudo-Cerasus is of some value as a timber-tree, producing hard, close-grained red
wood, which is hardly to be distinguished from that of the European Cherry. It is used in
considerable quantities for all sorts of wooden dishes and other small articles of domestic use.
Rather curiously, perhaps, no attention has been paid to improving the size and quality of the
fruit, which is not larger than a small pea, with a thin layer of flesh.

The pendulous-branched Cherry-tree, with precocious pink flowers, now common in our
gardens, where it is known as Prunus pendula, is often cultivated by the Japanese, who,
however, do not appear to feel the same regard for this graceful tree that the Apricot and the
Cherry inspire. I never saw it growing wild, and cannot refer the cultivated plants to a wild
type. Specimens fifty or sixty feet high, with wide-spreading fountain-like heads, are not
uncommon in old temple gardens in many of the cities of Hondo. This beautiful tree thrives
perfectly in our climate, and in early spring, when its branches are covered with its pale pink
pendulous flowers, no tree is more beautiful.

I did not see the Cherries in bloom, and most of them had dropped their fruit before I
reached Japan ; several species described by botanists I did not see at all, and of several
others I obtained a very superficial idea ; and there is evidently still much to be learned of
the proper limitation of described east Asian species and varieties, and of their geographical
distribution. Among the little known species, Prunus Maximowiczii (see Plate xii., made
from material for which I am indebted to Professor Miyabe), seems to deserve the attention
of horticulturists.

As I saw it in Yezo, Prunus Maximowiczii l is a tree twenty-five to thirty feet in height,
with a slender trunk and branches covered with smooth pale or light red bark. The young
branchlets and petioles, the under surface of the unfolding leaves, and the branches of
the inflorescence are coated with rusty pubescence which only partly disappears during the
season. The leaves are elliptical or elliptical-obovate, contracted at the apex into long slender
points, wedge-shaped or rounded at the base, long-petiolate, coarsely and doubly serrate, thin,
light green on the upper, and paler or rufous on the lower surface. The stipules are folia-
ceous, lanceolate-acute, coarsely serrate, an inch long, or rather shorter than the petioles, and
deciduous. The flowers, which, in the neighborhood of Sapporo, appear in May, are produced
on long slender pedicels in axillary racemes three or four inches long, and conspicuous from
their large foliaceous bracts, which are coarsely serrate with gland-tipped teeth ; they are half
an inch across when expanded, with leafy serrate hairy calyx-lobes and obovate or orbicular
white petals. The fruit ripens in July, and is oblong and rather less than a quarter of an
inch long. In Japan, Prunus Maximowiczii does not appear to be a common tree. I saw a
few specimens on the hills near Sapporo and a single tree on the main island, where it is said

1 Prunus Maximowiczii, Ruprecht, Bull. Phys. Math. Acad. Franchet & Savatier, Enum. PI. Jap. i. 118. — Forbes &
St. Petersbourg, xv. 131. — Maximowicz, Fl. Amur. 89 ; Mel. Hemsley, Jour. Linn. Soc. xxiii. 219.
Biol. xi. 700. — F. Schmidt, Reisen in Amurland, 125. —

38, FOREST FLORA OF JAPAN.

by Maximowicz to grow in several of the mountain provinces ; it also inhabits Saghalin, Corea,
and eastern Manchuria, where it was discovered.

The common Prunus Padus of Europe and northern Asia reaches northern and central
Yezo, where it is not rare in low ground in the neighborhood of streams, and where it grows
to a considerable size. A much more common tree in Yezo and in the elevated forests of
Hondo is Prunus Ssiori, another Bird Cherry, always easily distinguished by its pale, nearly
white bark. It is a handsome glabrous tree with oblong membranaceous leaves and long
graceful racemes of small flowers, and is well worth introducing into our plantations as an
ornamental plant. It grows also in Saghalin, where it was discovered by Schmidt, in Man-
churia, and in western China. The wood of Prunus Ssiori is very hard and close-grained,
and is used by the Ainos for numerous domestic purposes.

Prunus Grayana, the third Japanese Bird Cherry, is common in all the mountain forests
of Hondo, and extends across the straits of Tsugaru into southern Yezo. It is a small tree,
twenty to thirty feet high, with a slender trunk, ample membranaceous long-pointed setaceo-
serrate leaves, biglandular at the base, but without glands on the petioles, a peculiarity which
best distinguishes this species from Prunus Padus, although the hair-like teeth of the leaves
are characteristic and apparently constant.

Of true Plums there are in the flora of eastern America no less than nine or ten indige-
nous species, of which six are considered trees ; in some parts of the country these plants are
exceedingly common, and in early spring enliven forest glades, or the seacoast of the north
with their profuse and fragrant flowers ; but Japan apparently possesses no indigenous Plum-
tree, and although Plums are sometimes cultivated in the neighborhood of Japanese houses,
they are by no means common, and the fruit which is offered for sale in the markets is not
abundant or of good quality.

In recent years a good deal has been heard in this country of Japanese Plums, which are
now successfully cultivated in the southern states. Some of these varieties have possibly been
made in Japanese gardens, but the original stock from which they have all been derived is
probably some south China or Indian species of doubtful identity — perhaps, as has been
suggested, the Prunus triflora of Roxburgh, an obscure plant, which is possibly a form of
Prunus domestica. But the parentage of the so-called Japanese Plums will not be satisfacto-
rily settled until competent botanists have explored western and southwestern China, where
must be solved many of the problems which relate to the origin and geographical distribution
of many cultivated plants.

The Japanese form of the Old World Mountain Ash, Pyrus aucuparia, is a common tree in
Yezo and on all the high mountain ranges of Hondo, where it is sometimes twenty or thirty
feet in height, and always conspicuous in autumn from the brilliant orange and scarlet colors
of the foliage. This peculiarity will give to this tree a horticultural value, although, except
in its more glabrous buds, it does not vary in any marked way from the Mountain Ash of
Europe and northern Asia.

Pyrus sambucifolia, which is much like the American plant, although described as a small
shrub, is said to be abundant in northern and eastern Yezo and on the Kurile Islands. I only
saw it in the Botanic Garden at Sapporo, where Professor Miyabe has established a remarkable

PLATE XIII.

C. E. FAXOS del.

PYRUS MIYABEI, SARG.

THE ROSE FAMILY. 39

collection of Hokkaido plants. The third Japanese Mountain Ash, Pyrus gracilis, is, I
believe, unknown in cultivation ; it is a particularly well-marked species, with woolly buds,
leaves only four or five inches long, oval or oblong leaflets rounded or acute at the apex, and
pale on the lower surface, orbicular incisely serrate stipules an inch or more across, minute
flowers in small few-flowered clusters, and oblong fruit barely an eighth of an inch long.
Pyrus gracilis inhabits mountain forests in Kyushu and in central Hondo, where, however, I
did not succeed in finding it.

Aria is represented in Japan by two handsome trees, the first, Pyrus lanata l (I follow
Hooker in the Flora of British India in referring the Japanese plant to the Pyrus lanata of
Don, which grows also on the Himalayas from Cashmere to Kumaon), is not rare in central
Japan, where it is principally found at about 5,000 feet elevation above the sea-level on the
lower edge of the great Hemlock forest. Here it is a tree thirty or forty feet in height, with
a trunk six to eight inches in diameter, slender light red branchlets marked with white dots,
and oblong obtuse winter-buds covered with pale chestnut-colored imbricated scales. The
leaves are three or four inches long, two or three inches wide, broadly oblong to ovate-lance-
olate, acute at the apex, slightly lobulate and serrate, dark green and mostly glabrous on the
upper surface, and silver-white and more or less thickly coated with tomentum on the lower
surface. The flowers I have not seen, but the fruit is subglobose to oblong, one third of an
inch long, bright scarlet, and marked with pale lenticels. The second Japanese species of
Aria (see Plate xiii.) is a tree fifty or sixty feet in height, with a trunk covered with pale
smooth bark, and occasionally a foot and a half in diameter, slender branches, which form a
narrow oblong head, and red branchlets marked by oblong lenticular dots. The leaves are
ovate, acute, often long-pointed at the apex, rounded or sometimes wedge-shaped at the base,
serrate with incurved teeth, or often coarsely and doubly serrate above the middle, thin, or
subcoriaceous at maturity, dark green on the upper surface, pale on the lower, two or three
inches long and one or two inches broad, with thick prominent midribs, straight parallel
veins, and slender petioles one or two inches in length. The flowers, which appear near
Sapporo early in June, are borne in loose spreading long-branched few-flowered corymbs, and
are half an inch in diameter. The calyx-lobes are ovate, acute, densely coated on the inner
surface with thick white tomentum, and much shorter than the oblong white petals, rounded
at the apex and contracted at the base into short claws more or less covered with tufts of long
white hairs. The stamens are exserted, with filiform filaments enlarged at the base, and
rather longer than the two spreading styles. The fruit ripens in September, and is oblong or
subglobose, the size of a pea, light red, and conspicuously marked by the scar left by the
deciduous calyx. Unfortunately the name Pyrus alnifolia, which has been given to this tree,
is not applicable, it having been previously applied by Sprengel, in 1825, to an entirely
different plant, Amelanchier alnifolia, and as a new name must be found for it, I am glad of
the opportunity of associating with this fine tree that of Professor Kingo Miyabe, whose

1 Pyrus lanata, Don, Prodr. Fl. Nepal. 237. — Hooker f. Sorbus Aria, var. Kamaonensis, Maximowicz, Mel. Biol.

Fl. Brit. Ind. ii. 375. ix. 173.

Sorbus lanata, Wenzig, Linncea, xxxviii. 61.

40 FOREST FLORA OF JAPAN.

knowledge of the flora of Hokkaido is unrivaled. Pyrus Miyabei1 is one of the common
trees of the forests of central Yezo, and, according to Maximowicz, it inhabits the province of
Nambu, in Hondo, and southern Manchuria. So far as I know, it has .not been introduced
into our gardens, where it may be expected to flourish.

Of true apple-trees, there is apparently only a single indigenous species in Japan, the
Pyrus Toringo of Siebold. This is the tree which is often cultivated in American and Euro-
pean gardens as Pyrus Malus floribunda, Pyrus microcarpa, Pyrus Parkmani, Pyrus Halleana,
Pyrus Sieboldii, and Pyrus Kingo. It is a common and widely distributed plant in Japan,
growing from the sea-level in Yezo to elevations of several thousand feet in central Hondo,
usually in moist ground in the neighborhood of streams. Sometimes it is a low bush, but
more often a tree fifteen to thirty feet in height, with a short stout trunk and spreading
branches. The leaves are exceedingly variable, and on the same plant are often oblong,
rounded or acute at the apex, or broadly ovate or more or less deeply three-lobed. The
fruit, which, like that of the Siberian Pyrus baccata, loses the calyx before it is fully ripe,
resembles a pea in size and shape, and in color varies from bright scarlet to yellow. In early
spring Pyrus Toringo is one of the most beautifid of the trees found in our gardens, where it
is perfectly hardy, and where it covers itself every year with fragrant pink or red single or
semi-double flowers.

Pyrus Sinensis, the common cultivated Pear-tree of Japan, although now growing sponta-
neously in some mountain regions, is probably a native of northern China and Manchuria ;
and the only indigenous Pear-tree is Pyrus Tschonoskii. This interesting and handsome tree
was first described by Maximowicz,2 whose collector, Tschonoski, brought to him from the
slopes of Fuji-san a single fruit and a portion of a leaf, now preserved in the herbarium of
the Imperial Botanic Garden at St. Petersburg. Nothing more was seen of it until Mr. J. H.
Veitch and I encountered in the woods near Nikko a single tree of a Pyrus, which by
subsequent comparison with Tschonoski's specimen, proved to have been this tree. It is
evidently rare, for I only saw it in two other localities, — in the grounds of a temple near
Nekatsu-gawa, where there was a single specimen, and in the woods at the head of the Ysui-
toge, near Karuizawa, at the base of the volcano Asama-yawa, in central Hondo, where there
were two or three trees. Pyrus Tschonoskii (see Plate xiv.), which is a Pear-tree, rather than
an Apple as described by Maximowicz, is, as we saw it, a tree thirty to forty feet in height,
with a trunk about a foot in diameter, covered with smooth pale bark, and a narrow round-
topped head. The branchlets are stout and terete, and are marked with small oblong or
circular orange-colored lenticels ; during their first summer they are red-brown, rather lus-
trous, covered with loose pale tomentum, and encircled at the base by the conspicuous ring-
like scars left by the falling of the inner scales of the winter-buds ; later they grow darker,
and sometimes nearly black. The winter-buds are ovate, obtuse, and rather less than a
quarter of an inch long, and are covered with loosely imbricated chestnut-brown lustrous
scales, tomentose above the middle, and ciliate on the margins. The leaves are ovate, acumi-

1 Pyrus Miyabei. Maximowicz, Mel. Biol. ix. 173. — Wenzig, Linncea, xxxiii.

Cratsegus alnifolia, Siebold & Zuccarini, Abbild. Acad. 61.
Milnch. iv. 130. — Regel, Act. Hort. Pelrop. i. 125. Aronia alnifolia, Decaisne, Nouv. Arch. Mus. x. 100.

Sorbus alnifolia, Miquel, Ann. Mus. Lugd. Bat. i. 249. — 2 Mel. Biol. xii. 165.

PLATE XIV.

C. E. FAXON del.

PYRUS TSCHONOSKIT, MAXM.

THE ROSE FAMILY. 41

nate, unequally rounded or wedge-shaped at the base, and coarsely and unequally serrate with
rigid glandular teeth, which are largest and most unequal above the middle of the leaf ; they
are thick and firm, dark green, lustrous and pilose on the upper surface, coated on the lower
surface and on the petioles with thick pale, close tomentum, four to five inches long and two
to three inches broad, with stout midribs impressed on the upper side, and five or six pairs
of conspicuous oblique veins running to the principal teeth and connected by reticulate cross
veinlets ; they are borne on slender terete petioles an inch and a half in length. The flowers
are unknown. The fruit, which is usually solitary, or is sometimes in clusters of two or
three, is obovate, pointed at the base, and crowned with the thickened and partly immersed
calyx-lobes, which are triangular, obtuse, and covered with a thick coat of dense white
tomentum ; it is an inch long, two thirds of an inch broad, of a dull yellow color, and rosy-
red on one side, with a thick skin covered with pale lenticels, and austere coarse granular
flesh. The seed is a quarter of an inch long, obliquely obovate, acute at the base, and cov-
ered with a light red-brown shining coat. The fruit is borne on a stout rigid stem an inch to
an inch and a half long and coated with pale loose tomentum, especially toward the much
thickened apex.

Cratsegus, which, in eastern America, abounds with many species which are conspicuous
features of vegetation in different parts of the country, is only represented in Japan by
Cratsegus chlorosarca, one of the black-fruited group related to Cratsegus Douglasii of our
Pacific states, which it much resembles. It is not rare in the neighborhood of Sapporo,
where it grows near streams in low wet soil, and apparently does not range south of Yezo.
The flowers are not large, and as a garden-plant this species has little to recommend it.

The Saxifrage family, which is conspicuous in Japan with a large number of shrubs,
including some which have become important features in our gardens, has only a single
Japanese arborescent representative ; this is the now well-known Hydrangea paniculata, which
is one of the most common northern and mountain plants, and which occasionally in favorable
situations, especially on the hills of central Yezo, becomes a tree twenty-five to thirty feet in
height, with a short well-formed trunk a few inches in diameter and branches stout enough
for a man to climb into. From the branches the Ainos make their pipes.

THE WITCH-HAZEL AND ARALIA FAMILIES.

IN the Witch-hazel family, Distylium racemosum, an evergreen tree of the southern islands
and of southern China, with peculiar and exceedingly hard dark-colored valuable wood, will
require in this country the mild climate of the extreme southern states and of California.
The Japanese Hamamelis, however, is already an inhabitant of our gardens, where, unlike the
American species which flowers in the autumn, it produces its orange or wine-colored flowers
in March. Hamamelis Japonica is one of the common forest-shrubs or small trees in its native
country, where specimens occasionally occur thirty or forty feet in height, with stout straight
trunks and broad shapely heads. In the autumn the leaves turn bright clear yellow ; but on
one form which we found on Mount Hakkoda, near Aomori, with small thick often rounded
leaves (Hamamelis arborescens of Hort., Veitch), they were conspicuous from their deep rich
vinous red color. This may, perhaps, prove to be a second Japanese species.

We were fortunate in securing a good supply of ripe seeds of the little known Disanthus
cercidifolia of Maximowicz, a curious and interesting member of the Witch-hazel family (see
Plate xv.). Disanthus, of which only one species is known, is a shrub with slender spreading
branches, eight or ten feet high, stout terete red-brown branchlets conspicuously marked with
pale lenticels,and obtuse buds covered with chestnut-brown imbricated scales. The leaves are
suborbicular, rounded and minutely mucronate at the apex, or rarely orbicular-ovate and
sharp-pointed, cordate or rarely truncate at the base, entire, palmately five or seven nerved,
dark blue-green on the upper surface, pale on the lower, thick and firm, or ultimately sub-
coriaceous, and three or four inches long and broad, with reticulated veinlets and stout
petioles one or two inches long and thickened at the base. In the autumn they turn deep
vinous red or red and orange. The flowers appear in October, when the fruit developed from
the flowers of the previous year ripens ; they are dark purple, sessile, base to base, in two-
flowered heads on slender-ridged peduncles produced from scaly buds, and are each surrounded
by three thick ovate obtuse woolly closely imbricated bracts which form the apparent connec-
tive between the two flowers. The calyx is five-parted ; the divisions which are imbricated
in aestivation are ovate, obtuse, latitudinally unequal, reflexed, and much shorter than the
five lanceolate acute petals imbricated in estivation, spreading into a star-shaped corolla, and
slightly incurved at the apex. The stamens are as long as the lobes of the calyx and are
inserted on its base opposite the petals ; the filaments are short and broad, and as long as the
anthers, which are nearly as broad as long, attached on the back, two-celled, and extrorse, the
cells opening longitudinally. The ovary is superior, ovate, compressed, two-celled, gradually
contracted into two short spreading styles stigmatic at the apex ; the ovules are numerous in
each cell, suspended from its apex, and anatropous. The fruit is a woody ovoid two-celled
capsule, which opens loculicidally, with a thin cartilaginous inner coat separable from the
thick hard outer covering. The seed, of which there are a number in each cell, is ovate,

PLATE XV.

C. E. FAXON del.

DISANTHUS CERCIDIFOLIA, MAXM.

THE WITCH-HAZEL AND ARALIA FAMILIES. 43

acute, compressed, angled by mutual pressure, with a thick hard dark chestnut-brown
lustrous coat, an oblong pale lateral hilum,and thin albumen surrounding the terete embryo,
with its long erect radicle and thick ovate cotyledons.

Disanthus cercidifolia J is not rare in the valley of the Kisogawa on the Nakasendo, in
central Hondo, where it is occasionally found, covering steep hillsides with thickets sometimes
a quarter of an acre in extent. In habit and in the autumn color of its leaves Disanthus is
one of the most beautiful shrubs which I saw in Japan, and if it flourishes in our gardens it
should prove one of the best plants of its class recently introduced into cultivation.

The Aralia family has no representative in eastern North America outside of the genus
Aralia and only one woody plant, Aralia spinosa, a small tree of the middle and southern
states. In Japan the family appears in no less than eight genera. The Ivy of Europe
reaches Japan, where it is rather common in the south, although we did not meet with it
north of the Hakone Mountains and the region about Fuji-san. Helwingia, a genus with
two species of shrubs, remarkable in this family for the position of the flowers which are
produced on the upper surface of the midribs of the leaves, is Japanese and Himalayan. In
Japan Helwingia ranges to southern Yezo, where, in the peninsula south of Volcano Bay, in
common with a number of other plants, it finds its most northern home.

In the flora of Japan, Fatsia is represented by the handsome evergreen plant, Fatsia
(Aralia) Japonica, now well known in our conservatories, an inhabitant of the extreme
southern part of the empire, although often cultivated in the gardens of T5kyo, both in the
open ground and in pots ; and by Fatsia horrida, a low shrub with stout well-armed stems,
large palmately lobed leaves and bright red fruit, which is also common on the mountains
of the northwest coast region of North America, from Oregon to Sitka. In Japan we found
it growing under the dense shade of the Hemlock forests on steep rocky slopes above Lake
Yumoto, in the Nikko Mountains, at an elevation of 5,000 feet above the sea-level, and in
Yezo. The third member of the genus, Fatsia papyrifera, from the thick pith of whose
branches the Chinese rice-paper is made, and an inhabitant of central and southern China
and of Formosa, is frequently seen in Tokyo gardens, as it is in those of the United States
and Europe. In Yezo is found a representative of the Manchurian and Chinese genus Eleu-
therococcus, a shrub still to be introduced into our gardens, and Panax repens, a delicate
herb with trailing stems and bright red fruit, which manages to live on mountain-slopes
under the dense shade of Bamboos, while Dendropanax, a tropical genus of trees and shrubs
of the New World, as well as of the Old, reaches southern Japan with a single shrubby
species, Dendropanax Japonicum.

Aralia is more multiplied in species in eastern America, where six are known, than it is in
Japan, whose flora contains only two, although a third, the Ginseng (Aralia quinquefolia), a
native of Manchuria, northern China, and the United States, has been cultivated for centuries
in Japan for the roots, which the Chinese esteem for medicine and buy in large quantities,
sometimes paying fabulous prices for them, especially for the wild Manchurian roots, which
are considered more valuable than those obtained from North America or from plants culti-
vated in Japan, or in Corea, where Ginseng cultivation is one of the most important branches

1 Maximowicz, Mel. Biol. vi. 21 (1866).

44 FOREST FLORA OF JAPAN.

of agriculture. Curiously enough, this North American and Chinese species was first made
known to the outside world by Kaempfer's description of the plants cultivated in Japan.

Of the indigenous Aralias of Japan, Aralia cordata is an herb with large pinnate leaves
and long compound racemose panicles of white flowers, which are followed by showy black
fruits. In habit and general appearance it resembles our North American Spikenard, Aralia
racemosa; but it is a larger and handsomer plant, and well worth a place in the wild garden.
In Japan Aralia cordata is often cultivated in the neighborhood of houses for the young-
shoots, which, as well as the roots, are cooked and eaten. The second Japanese Aralia only
differs from our American Aralia spinosa in its rather broader and more coarsely serrate
leaflets, and in the character and amount of pubescence which covers their lower surface.
Aralia spinosa, var. elata, is a common tree in Yezo and in all the low mountain region
of northern and central Hondo. It usually selects rather moist soil, and sometimes, under
favorable conditions, rises to the height of thirty or forty feet and forms a straight well-
developed trunk. In Hondo large plants are rare, probably owing to the fact that the forests
on the low and accessible mountain-slopes are frequently cut off, but the shrubby covering of
such hills is almost always brightened in September by the great compound clusters of the
white flowers of the Aralia which rise above it. The Japanese form does not appear to be
much known in gardens, although young plants have lately been raised in the Arnold Arbore-
tum from seed sent from Japan a few years ago by Dr. Mayr ; and it is the Manchurian
variety, known as Aralia Chinensis, or as Dimorphanthus Manchuricus, that is usually seen in
our gardens, from which the American form, the type of the species, appears to have pretty
nearly disappeared, although the name is common enough in nurserymen's catalogues.

But of all the Araliaceae of Japan, Acanthopanax is the most interesting to the student of
trees. It is a small genus of about eight species of trees and shrubs, all members of tropical
Asia, and of China and Japan, where half a dozen of them have been found. The most
important of the Japanese species are Acanthopanax ricinifolium and Acanthopanax sciado-
phylloides. Of the other species, Acanthopanax innovans is a small tree, of which I saw
young plants only on the Nakasendo, without flowers or fruit, and which is still to be intro-
duced. Acanthopanax aculeatum, a shrub or small tree, with lustrous three or five parted
leaves, is much planted in Japan in hedges, and is hardy in southern Yezo, where, however,
it has been introduced. Acanthopanax trichodon, of Franchet & Savatier, a doubtful species,
which, from the description, must closely resemble Acanthopanax aculeatum, we did not see ;
but Acanthopanax sessililiflorum of Manchuria and northern China, and an old inhabitant of
the Arnold Arboretum, we found evidently indigenous near Lake Yumoto, in the Nikko Moun-
tains, on the Nakasendo and in Yezo.

Acanthopanax sciadophylloides is still unknown in our gardens, and we were fortunate in
securing an abundant supply of seeds. It is a handsome shapely tree sometimes forty feet
in height, with a trunk a foot in diameter, covered with pale smooth bark, short small
branches which form a narrow oblong round-topped head, and slender glabrous unarmed
branchlets. The leaves are alternate, and are borne on slender petioles with enlarged clasp-
ing bases and four to seven inches in length, and are composed of five, or rarely of three,
leaflets ; these are oval or obovate, long-pointed, wedge-shaped at the base, coarsely serrate

PLATE XVI.

ACANTHOPANAX RICINIFOLIUM, SEEM.

THE WITCH-HAZEL AND ARALIA FAMILIES. 45

with incurved teeth tipped with long slender mucros, membranaceous, dark green on the
upper surface, and pale or sometimes almost white on the lower surface, quite glabrous at
maturity, five or six inches long and two or three inches broad, with stout petiolules some-
times an inch in length, broad pale midribs, and about seven pairs of straight primary veins
connected by conspicuous transverse reticulate veinlets. The flowers appear in early summer,
on slender pedicels in few-flowered umbels arranged in terminal panicles five or six inches
across, with slender branches, the lower radiating at right angles to the stem, the upper erect.
According to Franchet & Savatier, the flowers are five-parted, with acute calyx-teeth, oblong-
obtuse greenish white petals, and two united styles. The fruit, which is the size of a pea, is
dark blue-black, somewhat flattened or angled, crowded with the remnants of the style, and
contains two cartilaginous flattened one-seeded stones. This handsome species inhabits the
mountain forests of Nikko, where it is not common. Later we found it in great abundance
on Mount Hakkoda, in northern Hondo, and in central Yezo, where it is common in the
deciduous forests which clothe the hillsides. Here it apparently attains its largest size, and
grows with another species of this genus, Acanthopanax ricinifolium, the largest Aralia of
Japan. I have followed the Japanese botanists in referring this tree to the Panax ricinifolia
of Siebold & Zuccarini, although the plant cultivated in our gardens and in Europe as Acan-
thopanax ricinifolium or Aralia Maximowiczii is distinct from the Yezo tree in the more deeply
lobed leaves with much broader sinuses between the lobes. A single individual similar to
the plant of our gardens I saw growing in the forest near Fukushima, in central Japan, but,
unfortunately, it was without flowers or fruit. And as I was unable to find any leaves on
the Yezo trees with the broad sinuses of this plant or any intermediate forms, it will not be
surprising if the forests of Japan are found to contain two species of simple-leaved arbores-
cent Acanthopanax, in which case it will be necessary to examine Siebold's specimens to deter-
mine which species he called Panax ricinifolia.

In the forests of Yezo, where it is exceedingly common, Acanthopanax ricinifolium (see
Plate xvi.), as it will be called for the present at least, is a tree sometimes eighty feet in
height, with a tall straight trunk four or five feet in diameter, covered with very thick dark
deeply furrowed bark, and immense limbs which stand out from the trunk at right angles,
like those of an old pasture Oak, and thick reddish-brown mostly erect branchlets armed with
stout straight orange-colored prickles with much enlarged bases. The leaves are nearly orbi-
cular, although rather broader than long, truncate at the base, and divided to a third of their
width or less by acute sinuses into five nearly triangular or ovate acute long-pointed lobes
finely serrate with recurved callous-tipped teeth ; they are five to seven ribbed, seven to ten
inches across, dark green and very lustrous on the upper surface, and light green on the
lower surface, which is covered, especially in the axils of the ribs, with rufous pubescence.
The small white flowers are produced on long slender pedicels in many-flowered umbels
arranged in terminal compound flat-topped panicles with long radiating branches, which are
sometimes two feet in diameter ; they appear in August and September, and are very conspicu-
ous as they rise above the dark green foliage, giving to this fine tree an appearance entirely
unlike that of any other inhabitant of northern forests.

Acanthopanax ricinifolium is common in Saghah'n and Yezo, and I saw it occasionally

46 FOREST FLORA OF JAPAN.

on the mountains of central Hondo, where, however, it does not grow to the great size it
attains in the forests of Yezo ; here it is associated with Lindens, Magnolias, White Oaks,
Birches, Maples, Cercidiphyllum, Walnuts, Carpinus, and Ostrya. The Avood is rather hard,
straight-grained, and light brown, with a fine satiny surface. In Yezo it is highly valued,
and is used in considerable quantities in the interior finish of houses, and for furniture,
cases, etc.

The illustration (Plate xvi.) is made from a photograph taken two years ago on the
wooded hill near Sapporo, and represents a large although by no means an exceptionally
large or remarkable specimen. At the right of the Acanthopanax two young Magnolias
have sent up their trunks in search of light, and on the left appear a number of stems of the
noble Japanese Grape-vine (Vitis Coigneti*), which have climbed into its upper branches.

THE CORNELS, HONEYSUCKLES, AND PERSIMMONS, THE STYRAX

FAMILY, THE ARBORESCENT MEMBERS OF THE HEATH

FAMILY, THE ASHES, AND THEIR ALLIES.

CORNUS, which is exceedingly common in North America, where sixteen or seventeen
species are distinguished, is less abundant in Japan than in the other great natural botan-
ical divisions of the northern hemisphere. In the northern regions of eastern America
different species of Cornus often form a considerable part of the shrubby undergrowth which
borders the margins of the forest or lines the banks of streams, lakes, and swamps. In Japan
these shrubby species, or their prototypes, do not exist. High up among the Nikko Moun-
tains, on rocks under the dense shade of Hemlocks, we saw a few dwarf sprawling plants of
the Siberian and north China Cornus alba, but did not encounter in any other part of the
empire a shrubby Cornel. High up on these mountains, too, the ground is carpeted with the
little Bunch-berry, the Cornus Canadensis of our own northern woods, which is also common
in some parts of Yezo, and on the Kurile Islands, where a second herbaceous Cornel, with
large white floral scales, Cornus Suecica, is found. This is a common plant, too, in all the
boreal regions of North America from Newfoundland and Labrador to Alaska, and in north-
ern Europe and continental Asia. Of arborescent Cornels the flora of Japan possesses only
two species, Cornus Kousa and Cornus macrophylla, and neither of these is endemic to the
empire.

Cornus Kousa represents in Japan the Cornus florida of eastern America and the Cornus
Nuttallii of the Pacific states. From these trees it differs, however, in one particular ; in our
American Flowering Dogwoods, the fruits, which are gathered into close heads, are indi-
vidually distinct, while in the Japan tree and in an Indian species they are united into a
fleshy strawberry-shaped mass, technically called a syncarp. Owing to this peculiarity of the
fruit, botanists at one time considered these Asiatic trees generically distinct from the Amer-
ican Flowering Dogwoods, and placed them in the genus Benthamia, which has since been
united with Cornus. In Japan, Cornus Kousa is apparently not common ; certainly it is not
such a feature of the vegetation in any part of the empire which we visited as Cornus florida
is in our middle and southern states. Indeed, we only saw it in one place among the Hakone
Mountains, and on the road between Nikko and Lake Chuzenji, where it was a bushy, flat-
topped tree, not more than eighteen or twenty feet high, with wide-spreading branches. The
leaves are smaller and narrower than those of our eastern American Flowering Dogwood ; the
involucral scales are acute and creamy white, and the heads of flowers are borne on longer
and much more slender peduncles. Cornus Kousa also inhabits central China ; it was intro-
duced into our gardens several years ago, and it now flowers every year in the neighborhood
of New York, where it was first cultivated in the Parsons' Nursery at Flushing. As an orna-

48 FOEEST FLORA OF JAPAN.

mental plant' it is certainly inferior in every way to our native Flowering Dogwood, and in
this country at least it will probably never be much grown except as a botanical curiosity.

The second arborescent Japanese Cornel, Cornus macrophylla, often known by its syno-
nym, Cornus brachypoda, is also an inhabitant of the Himalayan forests, where it is common
between 4,000 and 8,000 feet above the sea-level, and of China and Corea. It is one of the
most beautiful of the Cornels, and in size and habit the stateliest and most imposing member
of the genus. In Japan, trees fifty or sixty feet in height, with stout well-developed trunks
more than a foot in diameter, are not uncommon, and when such specimens rise above the
thick undergrowth of shrubs which in the mountain regions of central Japan often cover the
steep slopes which descend to the streams, they are splendid objects, with their long branches
standing at right angles with the stems, and forming distinct flat tiers of foliage, for the
leaves, like those of our American Cornus alternifolia, are crowded at the ends of short lateral
branchlets which grow nearly upright on the older branches, so that in looking down on one
of these trees only the upper surface of the leaves is seen. These are five to eight inches
long and three or four inches wide, dark green on the upper surface, but very pale, and some-
times nearly white, on the lower surface. The flowers and fruit resemble those of Cornus
alternifolia, although they are produced in wider and more open-branched clusters ; and, like
those of this American species, they are borne on the ends of the lateral branchlets, and,
rising above the foliage, stud the upper side of the broad whorls of green.

Cornus macrophylla is exceedingly common in all the mountain regions of Hondo, where
it sometimes ascends to 4,000 feet above the sea, and in Yezo, where it is scattered through
forests of deciduous trees, usually selecting situations where its roots can obtain an abundant
supply of moisture. This fine tree was introduced into the United States many years ago
through the Parsons' Nursery, but I believe has never flourished here. In the Arnold Arbo-
retum, where numerous attempts to cultivate it have been made, it has never lived more than
a few years at a time. Raised from seed produced in the severe climate of Yezo, Cornus
macrophylla may, however, succeed in New England, where, if it grows as it does in Japan, it
will prove a good tree to associate with our native plants.

Cornus officinalis, as it was first described from plants found in Japanese gardens, has
usually been considered a native of that country. But, although it has been cultivated in
Japan for many centuries on account of its supposed medical virtues, it is probably Corean.
It may be considered, perhaps, a mere variety of the European and Asiatic Cornelian Cherry,
Cornus Mas, from which the Corean tree is best distinguished by the tufts of rusty brown
hairs which occupy the axils of the veins on the lower surface of the leaves. In the Botanic
Garden in Tokyo, which includes the site of a physic-garden established in the early days of
the Tokugawa dynasty, there is a group of trees of Cornus officinalis, which appear to have
attained a great age ; they are bushy plants, perhaps thirty feet tall, with bent and twisted
half-decayed trunks, and contorted branches, which form broad thick round heads, and in
October were loaded with the bright Cherry-like fruit.

The Honeysuckle family is represented in Japan by seven genera and a large number of
species, especially of Viburnum ; but none of them can be considered trees, although Diervilla
Japonica is occasionally almost arborescent in size and habit. The Japanese Viburnums are

THE HEATH FAMILY. 49

now all pretty well known in our gardens, with the exception of Viburnum furcatum, a
common northern and mountain plant, so similar to our American Hobble-bush, Viburnum
lantanoides, that some authors have considered the two plants identical, and Viburnum
Wriglitii, a distinct, black-fruited species of northern Japan, where the American botanist,
Charles Wright, detected it when the Wilkes Expedition explored the shores of Volcano
Bay. Viburnum furcatum is distributed through the mountain regions of the empire, and is
one of the commonest species. Sometimes it grows to the height of fifteen feet ; and it is
always conspicuous from its great thick reticulate-veined, nearly circular leaves, which, in the
autumn, turn to marvelous shades of scarlet, or to deep wine-color. If this fine plant takes
kindly to cultivation it will prove a real acquisition to our gardens.

Ericaceae abound in Japan, where we miss, however, such familiar American types as
Kalmia, Oxydendrum, and Gaylussacia. Vaccinium is multiplied in species ; but, with the
exception of the red-fruited Vaccinium Japonicum and the black-fruited Vaccinium ciliatum,
they are not very abundant, and are mostly confined to alpine summits, where the species are
found which, in the extreme north, encircle the globe ; and Blueberries nowhere cover the
forest-floor with the dense undergrowth which is common in our northern woods. The broad-
leaved evergreen true Rhododendrons are not very common in Japan, where there are only
two species, and, being mostly confined to high elevations, they nowhere make the conspicuous
feature in the landscape which Rhododendron maximum produces in the valleys of the south-
ern Alleghany Mountains, or Rhododendron Catawbiense makes around the summit of Roan
Mountain, in North Carolina and Tennessee. Most of the Japanese Azaleas produce purplish
or brick-colored flowers; and in spite of all that travelers have said of the splendor of
Japanese hillsides at the time when the Azaleas are in bloom, it is doubtful if they compare in
beauty with some Alleghany mountain-slopes when these are lighted up with the flame-colored
flowers of Azalea calendulacea, or with the summit of Roan Mountain during the last days of
June, when one of the greatest flower shows of the world is spread there for the admiration
of travelers.

None of the Japanese Azaleas, excepting, perhaps, Rhododendron Sinense, the Azalea
mollis of gardens, produce such beautiful flowers as those of American species like Rhodo-
dendron (Azalea) viscosum, Rhododendron nudicaule, or Rhododendron arborescens. None
of the Japanese Rhododendrons can be considered trees, although one or two of the deciduous-
leaved species grow to the height of twenty or, possibly, thirty feet.

Andromeda Japonica, now common in our gardens, is properly a tree, for in the temple
park of Nara, where it grows in profusion, there are specimens at least thirty feet in height,
with stout well-formed trunks six or eight feet in length. Enkianthus campanulatus, the rep-
resentative of a small genus of southern and eastern Asiatic trees, may be expected to become
an ornament in our gardens of much interest and beauty ; and as it grows as far north as the
shores of Volcano Bay in Yezo, and up to over 5,000 feet in central Hondo, it may flourish
in the climate of New England. Enkianthus campanulata is a slender bushy tree, sometimes
thirty feet in height, with a smooth light red trunk, occasionally a foot in diameter, and
thick smooth round branchlets. The leaves are mostly oval, sharply serrate, firm, dark green
above and pale yellow-green below, about three inches long, and one inch wide ; they are

50 FOREST FLORA OF JAPAN.

deciduous, and in the autumn, before falling, turn clear light yellow. The flowers are
campanulate, pure white, and are borne on slender stalks in many-flowered drooping racemose
panicles. By Japanese botanists it is spoken of as one of the most beautiful flowering trees
in Japan, and we considered ourselves fortunate in securing a supply of ripe seed, for this
species is still very rare in cultivation. There is but one other Japanese plant of the Heath
family which can pass as a tree ; this is the handsome Clethra canescens, or, as it is more
generally known in Japan, at least, Clethra barbinervis, a more recent name. It is a beautiful
small tree, occasionally twenty-five or thirty feet in height, with a slender trunk, a narrow
oblong head, long-stalked obovate pointed leaves, four to six inches in length, and very dark
green on the upper surface, and pale on the lower surface with hoary pubescence, which also
covers the branches of the inflorescence and the outer surface of the calyx of the flowers.
These are white, and are produced in slender upright terminal panicled racemes six to twelve
inches long, and open in succession during several weeks in August and September. In
southern Yezo Clethra canescens grows nearly down to the sea-level, and on the mountains of
the southern islands ; in central Hondo, where it is a common forest-plant, growing usually
near the borders of streams and lakes, it reaches an elevation of over 5,000 feet, so that there
is reason to believe that this fine species will thrive in our climate if plants are raised from
seed produced at high elevations, although up to the present time those which have been sent
to the Arnold Arboretum have never been very satisfactory. Clethra canescens grows, not
only in Japan, but in China, Java, the Philippines, and Celebes.

Although we have learned to look upon Japan as the home of the Persimmon, which is
intimately associated with the expression of modern Japanese art, it is doubtful if either of
the species of Diospyros commonly encountered in that country is really indigenous in the
empire, where they were both probably introduced, with many other cultivated plants, from
China. The more common and important of the two species is, of course, the Kaki, Diospy-
ros Kaki, which is planted everywhere in the neighborhood of houses, which in the interior
of the main island are often embowered in small groves of this handsome tree. In shape it
resembles a well-grown Apple-tree, with a straight trunk, spreading branches which droop
toward the extremities and form a compact round head. Trees thirty or forty feet high are
often seen ; and in the autumn, when they are covered with fruit, and the leaves have turned
to the color of old Spanish red leather, they are exceedingly handsome. Perhaps there is no
tree, except the Orange, which, as a fruit-tree, is as beautiful as the Kaki. In central and
northern Japan the variety which produces large orange-colored ovate thick-skinned fruit
is the only one planted, and the cultivation of the red-fruited varieties with which we have
become acquainted in this country is confined to the south. A hundred varieties of Kaki, at
least, are now recognized and named by Japanese gardeners, but few of them are important
commercially in any part of the country which we visited, and, except in Kyoto, where red
kakis appeared, the only form I saw exposed for sale was the orange-colored variety, which,
fresh and dried, is consumed in immense quantities by the Japanese, who eat it, as they do
all their fruits, before it is ripe, and while it has the texture and consistency of a paving-
stone.

Diospyros Kaki, or an allied species, is hardy in Peking, with a climate similar to that of

THE STYRAX FAMILY. 51

New England, and fully as trying to plant life ; it fruits in southern Yezo, and decorates
every garden in the elevated provinces of central Japan, where the winter climate is intensely
cold. There appears, therefore, to be no reason why it should not flourish in New England
if plants of a northern race can be obtained; and, so far as climate is concerned, the tree,
which, in the central mountain districts of Hondo, covers itself with fruit year after year, will
certainly succeed in all our Alleghany region from Pennsylvania southward. In this country
we have considered the Kaki a tender plant, unable to survive outside the region where the
Orange flourishes. This is true of the southern varieties which have been brought to this
country, and which may have originated in southeastern Asia, in a milder climate than that
of southern Japan, for the Kaki is a plant of wide distribution, either natural or through
cultivation. But the northern Kaki, the tree of Peking and the gardens of central Japan,
has probably not yet been tried in this country. If it succeeds in the northern and middle
states, it will give us a handsome new fruit of good flavor, easily and cheaply raised, of
first-rate shipping quality when fresh, and valuable when dried, and an ornamental tree of
extraordinary interest and beauty.

Diospyros Lotus, which is probably a north China species, and which is naturalized or
indigenous in northwestern India, and naturalized in the countries bordering the Mediterra-
nean, is occasionally cultivated in northern Japan, where, however, as it does not appear to be
more hardy than the Kaki, it does not seem to be much esteemed. The fruit is small and of
an inferior quality. Diospyros Lotus may be expected to endure the climate of our northern
states.

In Japan, Styracea? is represented by Symplocos with half a dozen species, all shrubs
rather than trees, by Pterostyrax, which replaces our Mohrodendron, from which the Japanese
genus only differs in its terminal paniculate inflorescence, five-parted flowers, and small fruits ;
and by Styrax with two species. Neither of the two species of Pterostyrax equals in size our
Mohrodendron Carolinum, which, under favorable conditions, becomes a tree eighty to a hun-
dred feet high on the southern Alleghany Mountains, and neither of them approaches our two
arborescent Mohrodendrons in the beauty of their flowers, which, although produced in ample
clusters, are individually small. Pterostyrax corymbosum, which I believe to be almost exclu-
sively a southern species, I saw only in the Botanic Garden in Tokyo, where there is a bushy
plant eighteen or twenty feet in height. Pterostyrax hispidum, which is now beginning to
be known in our gardens, where it is hardy from Boston to Philadelphia, is a bushy tree or
shrub, which we saw wild in Japan only on the banks of a stream among the mountains above
Fukushima, on the Nakasendo, where we found a single plant twenty or twenty-five feet in
height.

As an ornamental plant, the most valuable of this family, as represented in Japan, is cer-
tainly Styrax Obassia, a tree which grows as far north as Sapporo, in Yezo, and which may
therefore be expected to be as hardy as Cercidiphyllum, Syringa Japonica, Magnolia Kobus,
or any of the Yezo trees, with which it grows and which flourish here in New England.
Styrax Obassia, as it appears in Yezo and on the mountains of central Hondo, where it is
common between 3,500 and 4,000 feet above the sea, is a tree twenty to thirty feet in height,
with a straight slender stem, long and graceful branches well clothed with nearly circular

52 FOREST FLORA OF JAPAN.

leaves dark green on the upper surface, pale on the lower surface, and often more than six
inches across. The white bell-shaped beautiful flowers, which are nearly an inch in length,
are borne in long drooping racemes, and are produced in the greatest profusion. The second
species, Styrax Japonica, is a common plant in the mountain forests of Hondo and in southern
Yezo, and is a shrub or occasionally a small tree twenty to thirty feet high. It is now well
known in American and European gardens.

From the Olive family we miss, in Japan, Forestiera, an exclusively American genus, and
Chionanthus, which is eastern American and Chinese. Fraxinus and Osmanthus are common
to the floras of Japan and eastern America, and in Japan, Ligustrum and Syringa, both Old
World genera, are represented. In eastern America, Fraxinus appears in nine species ; in
Japan there are probably not more than two indigenous species, and only Fraxinus longi-
cuspis is endemic. This is a tree of the Ornus section, which is rather common in the elevated
Hemlock forests of Hondo, and ranges northward into Yezo. It is a slender tree, twenty to
thirty feet in height, with thin rigid ashy gray branchlets, black buds, and leaves with five
stalked ovate acute and finely or coarsely serrate leaflets, which in the autumn are conspicu-
ous from the deep purple color to which they change.

Fraxinus Manchurica, which is also common in Manchuria, Saghalin, and Corea, is a
noble tree in Yezo, where it is exceedingly abundant in low ground near the borders of swamps
and streams, and where it often rises to the height of a hundred feet, and forms tall strain-lit

o / o

stems three or four feet in diameter ; its stout orange-colored branchlets, large black buds,
ample leaves, with lanceolate acute coarsely serrate leaflets, and great clusters of broad-winged
fruits, well distinguish this species, which is certainly one of the noblest of all the Ashes, and
one of the most valuable timber-trees of eastern Asia. For many years Fraxinus Manchurica
has inhabited the Arnold Arboretum, where it is hardy, and where it promises to grow to a
good size. Another Ash-tree commonly cultivated along the borders of Rice-fields near
Tokyo is referred by the Japanese botanists to the Fraxinus pubinervis of Blume. This has
every appearance of being an introduced tree in Japan ; but I was unable to obtain fruit or
any satisfactory information with regard to it.

Syringa Japonica is rather common in the deciduous forests on the hills of central Yezo,
and I saw it occasionally on the high mountains of Hondo. In its native country, the Jap-
anese Lilac, when fully grown, is an unshapely straggling tree, twenty-five to thirty feet in
height, with a trunk rarely twelve or eighteen inches in diameter, and does not display the
beauty of foliage and the compact handsome habit which we associate with this plant in our
New England gardens, where it is far more beautiful than in its native forests.

Osmanthus Aquifolium, or, as it is more commonly called in gardens, Osmanthus ilicif olius,
is usually supposed to be a Japanese tree. I saw it in city gardens, and in greater perfection
in the mountain region of central Hondo, where it is often planted near dwellings and by the
roadside, and where it sometimes grows to the height of thirty feet, and makes a trunk a foot
or more in diameter, and a broad compact round head, loaded in October with fragrant flow-
ers. But where I saw it, it had evidently been planted, and if it is a Japanese species, which
is doubtful, it is only indigenous in the extreme south.

Ligustrum is poorly represented in Japan, and of the three species found within the borders

THE ASH FAMILY. 53

.

of the empire only the evergreen Ligustrum Japonicum of the south becomes a tree. Of the
other species, Ligustrum medium is much more common than Ligustrum Ibota ; at the north
it is found in moist low forests, but farther south ascends to high elevations^ where, in central
Japan, Ligustrum Ibota is also found.

In the remaining Gamopetalous orders of trees Japan possesses only Ehretia acuminata,
which inhabits the Loochoo Islands, and possibly reaches the southern shores of Kuyshu, a
small tree, which I saw only in the Botanic Garden of Tokyo, and the beautiful Clerodendron
trichotomum, which in late summer enlivens the banks of streams with its great masses of
tropical foliage and brilliant flowers, and in Yezo often attains to the size and habit of a small
tree.

THE LAUREL, EUPHORBIA, AND NETTLE FAMILIES.

THE American traveler landing for the first time in Yokohama is surprised at the abun-
dance of arborescent Lauracese, which here, with evergreen Oaks and Celtis anstralis, make
the principal features of the woods which cover the coast-bluffs and surround the temples.
The most abundant of the Lauraceae in this part of Japan appears to be the Camphor-tree,
Cinnamomum Camphora, one of the most beautiful of all evergreens, and probably indigenous
in southern Japan. In that part of the country which we visited, however, it had every
appearance of having been planted. Even at Atami, on the coast some distance below Yoko-
hama, a popular winter resort famed for the mildness of the climate and for the geyser, which
attract many visitors, the Camphor-tree is probably not indigenous. Near the town is the
grove of Kinomiya, where may be seen what is popularly supposed to be the largest Camphor-
tree in Japan. It is really a double tree, as the original stem has split open, leaving irregular
faces, which have become covered with bark. Between the two parts there is sufficient space
for a small temple. The larger of the two divisions at five feet from the surface of the
ground, and well above the greatly swollen base, girths thirty-three feet eight inches, and the
smaller twenty-seven feet six and a half inches. This remarkable tree, Avhich has every
appearance of great age, is still vigorous and in good health. Atami is celebrated for the
skill of its workers in wood, and for the production of many small articles made from the
wood of the Camphor-tree. The hills along the coast are covered with groves of Orange-
trees, and in the temple gardens were many southern trees which we did not see in perfection
in other parts of the empire.

On the coast of this part of Japan, Cinnamomum pedunculattim becomes a tree thirty or
forty feet in height, and on the neighboring Hakone Mountains ascends to elevations of a
couple of thousand feet. It is a handsome tree, with ample ovate acute lustrous leaves, pale
or nearly white on the lower surface, and long-stalked flowers and fruit. In the same region
two other arborescent Lauraceae grow naturally — Litsea glauca and Machilus Thunbergii.
They are both evergreens, and are handsome trees, especially the Litsea, which bears oval
leaves pointed at both ends, silvery white on the lower surface, and often six inches long,
and near the ends of the branches abundant clusters of black fruit. These two trees are the
most northern in their range of the Lauracese of Japan with persistent foliage, and they may
be expected to thrive in this country where the evergreen Magnolia and the Live Oak flourish.

The flora of eastern Asia is rich in Linderas, no less than twenty species having already
been found in the Chinese empire and in Corea, while in North America there are only two —
Lindera Benzoin, the common Spice-bush of northern swamps, and the southern Lindera
melissaefolia. Japan possesses half a dozen indigenous Linderas, although none of them are
endemic, and in the mountain regions of Hondo several species are common, and make nota-
ble features in the shrubby growth which covers hillsides and borders streams and lakes.

PLATE XVII.

C. E. FAXON del.

LINDERA TRILOBA, BMIME.

THE LAUREL, EUPHORBIA, AND NETTLE FAMILIES. 55

The most beautiful, perhaps, of the Japanese species is Lindera umbellata, a southern
plant, found also in central China, which I saw only in the Botanic Garden at Tokyo, where
it forms a stout bush eight or ten feet high. The leaves, which appear in the spring with the
flowers, are lanceolate-acute, very gradually narrowed at the base, rounded at the apex, entire,
and often six or eight inches in length ; they are lustrous on the upper surface, and pale and
covered on the midribs and veins on the lower surface with rufous pubescence. The fruit,
which is a quarter of an inch in diameter and brilliant scarlet, is produced in great quantities
in dense axillary clusters on the branches of the previous year, and ripens in August and Sep-
tember. As I remember it, this seems one of the most beautiful plants which I saw in Japan.
It may be expected to thrive in the southern states and in southern Europe, but it will prob-
ably not be able to support the cold of the north.

As a garden-plant for this region, Lindera obtusiloba is, perhaps, the most promising ; and
we were fortunate in securing a sufficiently large quantity of seeds, gathered at high eleva-
tions in central Hondo, to give it a good trial. Lindera obtusiloba (see Plate xvii.) often
becomes a bushy tree, twenty to twenty-five feet in height, with a short stout trunk, terete
brown branchlets, and conspicuous winter-buds covered with imbricated chestnut-brown scales.
The leaves appear with the flowers and are broadly ovate, palmately three-nerved, mostly
three-lobed at the apex, three or four inches long and broad, thick and firm, lustrous above,
and pale and often puberulous on the veins below. In the autumn, before falling, they turn
to a beautiful clear yellow color, and make a handsome contrast with the shining black fruits,
which are borne on hairy stalks in few-fruited axillary clusters, produced on short spur-like
lateral branchlets of the previous year. This handsome plant, although it grows to its largest
size in central Hondo at four or five thousand feet above the sea, does not, so far as we
observed, range north of the Nikko Mountains, and, therefore, does not reach Yezo, where
only Lindera sericea is found. This is a small slender shrubby species, with precocious
flowers, oval entire pointed leaves, silky-canescent at first, and at maturity dark green on the
upper and pale on the lower surface, and small black fruit.

The other species of Lindera, which may possibly prove hardy in our northern gardens,
are Lindera triloba and Lindera prsecox. The first is a common plant in Hondo, where it
does not, however, ascend to the heights reached by Lindera obtusiloba, which is a more
northern and a hardier plant. Lindera triloba often grows twenty feet tall, and produces
trunks six inches in diameter, from which spring numerous slender divergent branches well
clothed with leaves. These appear with the flowers, and are elliptical or oblong, wedge-
shaped at the base, and divided at the apex into three acute lobes separated by deep broad
sinuses rounded in the bottom ; they are three-nerved, membranaceous, light green above,
pale, and covered below on the ribs with rufous pubescence, three or four inches long, and
two or three inches broad, and are borne on slender petioles. The fruit is half an inch
in diameter, and is produced in few-fruited umbels on short stout club-shaped stalks.

Lindera prsecox, like our American species, flowers before the leaves appear ; it is a bushy
tree fifteen to twenty feet in height, with stout divergent light brown branches, and is con-
spicuous in midsummer from the large size of the flower-buds, which are then fully grown,
and which probably open during the winter or in earliest spring. The leaves are ovate, long-

56 FOREST FLORA OF JAPAN.

pointed, rather thin, dark green above, pale and often pubescent below, and two or three
inches in length, with long slender stalks. The fruit, which is nearly an inch in diameter, is
reddish brown and marked with many small white dots ; the flesh is thin, papery, and very
brittle. Lindera prsecox is common on the Hakone Mountains ; we found it near Agematsu,
on the Nakasendo, and Mr. Veitch collected it on Mount Chokai-zan, on the northwest coast
of Hondo. If it inhabits the Nikko Mountains we missed it there, and also on Hakkoda,
near Aomori, where Lindera sericea was the only species seen. The other Japanese species,
Lindera glauca, is a southern black-fruited plant with precocious flowers and the habit and
general appearance of Lindera sericea, from which it differs in its larger leaves and more rigid
branches.

Of Elaeagnus, the sole representative of its family in Japan, we only saw growing natu-
rally Elseagnus umbellata, a variable plant in the size and shape of its leaves and fruit, and
one of the commonest shrubs in Japan from the level of the sea up to elevations of 5,000
feet. In the mountainous regions and at the north it is often planted near houses for the sake
of its small acid fruit ; and in cultivation it not infrequently rises to the size and dignity of a
small tree. ElaBagnus umbellata is now well established in our gardens, where it flowers and
fruits as freely as it does in Japan.

The now well-known Ekeagnus longipes was often seen in gardens, especially among the
mountains and in Yezo, but we did not notice it growing wild. In old age it sometimes
attains the height of twenty or twenty-five feet, and forms a stout straight trunk a foot in
diameter. Such a plant, evidently of great age, may be seen in the Botanic Garden at
T5kyo. The beautiful Elseagnus pungens, with its long wand-like stems, now a familial-
object in several varieties in the gardens of southern Europe, was seen in the temple grounds
at Nara and by the roadside near Kyoto, where it appeared to have escaped from cultivation
rather than to be an indigenous plant. Of the other reputed Japanese species we could
hear nothing

The flora of Japan, although it is comparatively rich in Euphorbiacese, does not contain
any important trees belonging to this family. One species of Daphuiphyllum, a Malayan
genus with beautiful lustrous evergreen foliage and handsome fruit, and now known in the
gardens of temperate Europe in the shrubby Daphniphyllum glaucesens, attains the size of a
small tree ; this is Daphniphyllum macropodum, which we saw not far from Gif u, growing, as
it seemed, naturally. It is interesting to note that one species of this tropical genus, Daph-
niphyllum humile, grows far north in Yezo, where, as well as on the mountains of northern
Hondo, it is a common under-shrub in the forest of deciduous trees. We obtained a supply
of seeds of this handsome plant, although it is hardly to be expected that it will be able to
survive our northern winters, as it will miss here the continuous covering of snow under which
it is buried in Yezo during many months of the year.

Of the smah1 genus of Aleurites of eastern tropical Asia and the Pacific islands, one
species reaches southern Japan, Aleurites cordata, which we saw only in the Botanic Garden
at Tokyo. This little tree has large long-stalked three-lobed leaves, inconspicuous flowers in
terminal panicles, and large black drupe-like fruit ; it may be expected to grow in the southern

u

O!

THE LAUREL, EUPHORBIA, AND NETTLE FAMILIES. 57

states ; but it will be valued for its botanical interest, and not for its beauty. And tbis is
true of tbe otber Japanese tree of the Euphorbia family, Exccecaria Japonica, which may
possibly prove hardy here in New England, as we found it growing on high elevations on the
Nakasendo near Agematsu in central Japan, as well as on the high Otome-toge in the
Hakone Mountains ; and Mr. Veitch gathered specimens on Mount Chokai-zan, on the north-
west coast of Hondo. It is a small tree with thick firm dark green leaves which vary from
oval or obovate to obovate-lanceolate, and are sometimes six or seven inches long, and three-
lobed fruit three quarters of an inch in diameter.

Of the Nettle or Elm family, Japan possesses some important trees, although in Elms
themselves the flora of Japan is poor as compared with that of eastern North America, where
there are five well-distinguished species, while in Japan there are only two ; these are both
continental, reaching in Japan their most eastern home. In Hondo Elm-trees are not com-
mon, and in that island are nowhere such features of vegetation as they are in our New
England and middle states and in Europe, and it is only in mountain forests between 3,000
and 5,000 feet above the sea-level that occasional small plants of Ulmus campestris, with
branchlets often conspicuously winged, appear. In Yezo, however, this tree is much more
abundant, growing on the river-plains nearly at the sea-level and in the forests which cover
the low hills, not infrequently becoming a prominent feature of the landscape. In Sapporo,
where many fine old specimens were left in the streets by the American engineers who laid
out the town, individuals seventy or eighty feet tall, with trunks three or four feet in diam-
eter, may be seen. The broad heads of graceful pendent branches reminded us of New
England, for this Japanese form of the Old World Elm has much of the habit of the American
White Elm. The portrait of one of these trees (see Plate xviii.), although not a large one,
growing a mile or two from Sapporo, gives a fair idea of the habit of this tree in Yezo.

The second Elm of Japan grows in all the mountain-woods near Sapporo. The Russian
botanists have considered it a variety of Ulmus scabra, to which the name laciniata has been
given, and which is principally distinguished by the peculiar shape of the leaves, which are
coarsely serrate, often six or seven inches long, three or four inches broad, and three-lobed at
the wide apex. It is a small tree, barely more than thirty feet tall, as we saw it, and very
fragrant, like our American Slippery Elm, which in habit it much resembles. It is from the
touofh inner bark of this tree that the Ainos weave the coarse brown cloth from which their

O

clothes are made. The process is a simple one ; the bark is stripped from the trees in early
spring, and is then soaked in water until the bast, or inner bark, separates from the outer in
long strips, which are twisted by the women into threads, and are then ready for use. This
interesting tree is not in cultivation, I believe, and we reached Yezo too late to obtain its
seeds. It is desirable, however, that it should be brought into our gardens, not only on
account of the curious appearance of the leaves, but that its development may be watched ;
for when it can be compared in a living condition with the European and Siberian forms of
Ulmus scabra, it may prove sufficiently distinct to be regarded as a species.

The Keaki, Zelkova Keaki, a member of this family, is, perhaps, the largest deciduous-
leaved tree of Japan; it is its most valuable timber-tree. The Keaki may be described as a
Beech, with the foliage of an Elm. The bark is smooth and pale, like the bark of a Beech-

58 FOREST FLOEA OF JAPAN.

tree, and the dense compact round head of slender branchlets resembles the crown of that
tree, while the leaves, which are large, ovate-acute, coarsely serrate, and roughened on the
upper surface, are like those of the Elm.

Zelkova is a genus with three arborescent species ; one inhabits Crete, another the Cauca-
sus, and the third Japan. The flowers are very similar to those of the Elm, and are unisexual
or rarely polygamous, and are produced in early spring on branchlets of the previous year,
the males clustered in the axils of the lower leaves, and the females solitary in those of the
upper leaves. The fruit is a small drupe, more or less irregularly oblique in shape and two-
beaked with the remnants of the eccentric style; it has a membraneous or slightly fleshy
outer covering and a thin hard endocarp or stone, containing a single compressed concave
horizontal seed, without albumen, the thick embryo filling its cavity.

Zelkova Keaki sometimes grows to the height of a hundred feet, and produces a trunk
eight to ten feet in diameter. Such specimens are often found in the gardens surrounding
temples in the large cities, and by village roadsides in the interior provinces. If any wild
Keakis are left in the forests of Japan they must be rare, and I am not sure that we saw this
tree growing naturally, although it is everywhere one of the most commonly planted deciduous
trees. Large specimens, which we saw on the Nakasendo, near Agematsu, in one of the
mountain provinces of central Hondo, and a very remote region, may have been growing
naturally ; but even this is doubtful, for the Nakasendo has been a traveled highway for at
least twelve hundred years, and a thousand years ago was probably more frequented than it is
now. Of the range and habitat of this tree I have, therefore, no idea whatever.

The wood is more esteemed by the Japanese than that of any of their other trees. It is
noted for its toughness, elasticity, and durability, both in the ground and when exposed to
the air ; it is considered the best building material in Japan, although it has become so scarce
and expensive that keaki is not now used for this purpose, except in temples, where the large
round light brown highly polished columns which support the roof are always made of this
wood. It is still much used in cabinet-making and turnery, and in the manufacture of many
small articles, which always command high prices.

Zelkova Keaki is probably the only Japanese tree which is worth introducing into this
country on a large scale as a timber-tree ; that it will thrive here at least as far north as
southern New England, the plants in Dr. Hall's garden in Warren, Rhode Island, indicate.
There are several of these raised from seed sent home by Dr. Hall in 1862 ; they have
received no special care, the soil in which they were planted is not exceptionally good, and
their growth has been no doubt checked by overcrowding. They are now, however, at least
fifty feet high, and have produced trunks a foot in diameter ; they flowered and fruited last
year, and our illustration (see Plate xix.) is made from specimens taken from these trees,
with the exception of the large single leaf, which has been drawn from a specimen gathered
in Japan.

The Zelkova, of all Japanese trees, should be better known in eastern America, where it
may, perhaps, become an imported timber-tree, and produce wood as strong as our best oak,
which it surpasses in compactness, durability, and lightness, for keaki, in comparison with its
strength, is remarkably light.

PLATE XIX.

C. F. FAXON del.

ZELKOVA KEAKI, SIEB.

THE LAUREL, EUPHORBIA, AND NETTLE FAMILIES. 59

There is little to be said of the other Japanese trees of the Elm family. Celtis Sinensis,
with its thick coriaceous leaves and dull red berries, is one of the first trees to greet the
traveler landing in Yokohama, where it is common in the groves which cover the shore-bluffs,
growing with the Camphor-tree and the evergreen Oaks. It is a southern species of wide
range in southeastern Asia, which we cannot hope to grow in this country, except in the
southern states.

Aphananthe aspera, a Celtis-like tree with ample bright green leaves and black fruit,
ranges as far north as central Yezo, and may be expected to give interest and variety to
dendrological collections in the United States and Europe, although to the mere lover of trees
with peculiar foliage or with showy flowers and fruit it will not appear sufficiently distinct
from our native Nettle-tree.

Of the Broussonetias or Paper Mulberries, of which two or three species are included in
the flora of Japan, I saw specimens only in the Botanic Garden at Tokyo. They are all trees
of the south, or more probably introductions from China. The White Mulberry, Morus alba,
however, is certainly a Japanese species, as it grows as a small tree in the remote and primeval
forests of Yezo, although the numerous forms cultivated by the Japanese as food for the silk-
worm are usually of Chinese origin. Of the Fig-trees which appear in the flora of Japan, I
saw nothing at all, with the exception of one or two cultivated shrubby plants. They all
belong to the extreme south, and inhabit regions we did not visit.

THE WALNUTS, BIRCHES, ALDERS, AND HORNBEAMS.

IN nut-bearing trees the forests of Japan are poor in comparison with those of eastern
North, America. The Hickory, if it ever existed in the ante-glacial forests of Asia, has
entirely disappeared from them, and the Walnut family now appears in Japan in three
genera — Juglans, Pterocarya, and Platycarya ; the last two belong exclusively to the Old
World. In Japan, Juglans is represented by Juglans Sieboldiana, a common forest-tree in
Yezo and in the mountain regions of the other islands. As a timber-tree it is much less
important than either of the two eastern American Walnuts, as specimens more than fifty
feet high are uncommon ; it is a wide-branched tree, resembling our Butternut in habit and
in the color of its pale furrowed bark, as it does in the pubescent covering of the youno-
branches, the lower surface of the leaves, and the fruit. The nuts are arranged in long racemes,
and resemble those of the Asiatic, or, as it is familiarly called in commerce, the English Wal-
nut (Juglans regia), rather than our American walnuts, which are deeply sculptured into
narrow ridges, while the surface of the Japanese nut is smooth, or sometimes more or less
pitted ; it is pointed at the apex with thickened wing-like sutures, and is often an inch and a
half long and about an inch broad, although it varies considerably both in size and shape ; in
flavor the kernel resembles that of the English walnut. The walnut is evidently an important
article of food in Japan, as the nuts are exposed for sale in great quantities in the markets of
all the northern towns. Juglans Sieboldiana is perfectly hardy here in New England, where
it ripens its fruit ; it is not worth growing, however, as an ornamental tree, as the Black
Walnut surpasses it in size and beauty. It will produce fruit, however, in regions of greater
winter cold than the English Walnut can support, and as a fruit-tree it may find a place in
northern orchards, although the abundance and cheapness of English walnuts seem to forbid
its cultivation as a source of profit.

I am unable to throw any light upon the curious Juglans cordiformis of Maximowicz, dis-
tinguished by its flattened, long-pointed, and more or less heart-shaped nuts. The tree which
produces these peculiar nuts is not recognized by the Japanese botanists, who consider them
extreme varieties of their common walnut. I looked in vain for nuts of this form in the
markets of Hakodate, where they were first seen by the Kussian naval officer Albrecht ; after-
ward, however, I found them offered for sale by the Nurserymen's Association of Yokohama,
and was told that they were collected on the sides of Fuji-san. A plant raised from one of
these heart-shaped nuts has been growing for a number of years in the Arnold Arboretum,
and has produced fruit. In habit and in foliage it is not distinguishable from plants of the
same age of Juglans Sieboldiana. Juglans regia, although included in most works on the
flora of Japan, is not a native of the empire ; it is occasionally cultivated in the neighborhood
of temples and as a fruit-tree, but we saw no evidence of its being anywhere indigenous, and
it is probable that it was introduced from northern China, where one form of this tree appar-
ently grows naturally.

THE WALNUT FAMILY. 61

Pterocarya, the curious genus with leaves like those of a Hickory, and long slender spikes
of small hard nut-like fruits surrounded by foliaceous bracts, appears in Japan with one
species ; a second inhabits China, and a third, the type of the genus, the Caucasus. The
Japanese Pterocarya rhoifolia is a large and important timber-tree. We first met with it on
the lower margin of the Hemlock forest about LakeYumoto, in theNikko Mountains, where it
grows to no great size ; and it was not until we ascended Mount Hakkoda, in the extreme
northern part of Hondo, that we saw this fine tree to advantage. On the slopes of this moun-
tain it is exceedingly common at elevations of from 2,500 to 4,000 feet above the sea, and,
next to the Beech, is the largest deciduous tree of the region, often rising to a height of
eighty feet, and producing trunks two and a half feet in diameter. It is a broad-topped
tree, with stout branches, which spread nearly at right angles to the stem, and form a dense
leafy crown. In winter the Japanese Pterocarya may be readily recognized by its orange-
colored branchlets, thickly beset with small light-colored lenticels, and by the stout acute
buds, three quarters of an inch long, covered with apiculate black puberulous scales conspic-
uously marked with clusters of pale hairs. The leaves are unequally pinnate, eight or ten
inches long, and four to six inches broad, with stout hairy petioles, and six or seven pairs
of lateral leaflets, which are acute, unequally rounded at the base, long-pointed, finely serrate,
yellowish green, and covered on the lower surface of the midribs with pale or rusty brown
pubescence. In the first days of October, when the fruit was fully ripe and just ready to
drop, the leaves were beginning to turn yellow ; a month later, in the forest above Lake
Yumoto, the trees were bare of foliage. A specimen of the wood of Pterocarya rhoifolia,
for which I am indebted to the officers of the Forestry Department at Aomori, is white, soft,
yery light, and straight-grained, with bands of open ducts marking the layers of annual
growth ; it might be mistaken at the first glance for a piece of our American white pine.

The other Japanese member of the Walnut family, Platycarya strobilacea, we saw only in
the Tokyo Botanic Garden, where there is a tree fifteen or twenty feet high, which two years
ago was covered with the curious cone-like heads of fruit which distinguish this genus. In
the mountain regions of Kyushu it is said to become a large and stately tree. Platycarya is
occasionally cultivated in the botanic gardens of southern Europe, but I am not aware that it
is growing in the United States.

Myrica Gale, in a distinct pubescent form, is as common in low marshy ground in Yezo
as it is in the same latitude in North America, and a second species of Myrica, akin to our
Bayberry, inhabits the sandy coast, although it does not range far north of the thirty-fifth
parallel. This is the handsome evergreen Myrica rubra, a small shapely tree, now well known
in California gardens, and occasionally cultivated in the southern Atlantic states.

In Japan, as in all other temperate northern lands, the Cupuliferse abound, and the decid-
uous forests of the northern islands are principally composed of Oaks, Beeches, Hornbeams,
Alders, and Birches. The mountain forests of Hondo and those of Yezo contain many Birch-
trees, which are also important elements of the forest in all northern and northeastern Asia.
The Old World White Birch, Betula alba, in at least three of its forms, is common in central
Yezo, and we saw also a number of trees of the typical form on the plains between Chuzenji
and Yumoto, in the Nikko Mountains. The most distinct of the Japanese forms of Betula

62 FOREST FLORA OF JAPAN.

alba is that which botanists call var. Tauschii, and which is distributed from southern Siberia
through the Amour country, to Yezo, where it is a slender tree, sometimes eighty feet in
height ; it is distinguished by its larger and rather thicker leaves, which are of a deeper and
more lustrous green on the upper surface than those of the other forms of the White Birch
with which it is associated. It is certainly worth a place in our plantations. The variety
verrucosa, well distinguished by the warts which beset the young branches, appears to be
confined in Japan to Yezo, where, so far as we were able to observe, it is an exceedingly rare
plant.

In the forests of Yezo, too, we saw, for the first time, Betula Maximowicziana. This is
certainly one of the handsomest trees in Japan, and one of the most distinct and beautiful of
the Birches ; and its introduction into our plantations was alone well worth the journey to
Japan. In Yezo, Betula Maximowicziana is a shapely tree, eighty or ninety feet in height,
with a trunk two or three feet in diameter, covered with pale smooth orange-colored bark.
Toward the base of old individuals the bark becomes thick and ashy gray, separating into
long narrow scales. The branchlets are stout, covered with dark red-brown bark and marked
by many pale lenticels. The leaves, however, are the most distinct feature of this tree ; in size
they are not equaled by those of any other Birch-tree, and as they flutter on their long
slender stalks they offer a spectacle which can be compared with that which is afforded by
our Silver-leaved Linden waving its branches before some Hemlock-covered hill of the southern
Alleghany Mountains. The leaves of Betula Maximowicziana are broadly ovate, cordate at
the base, coarsely and doubly serrate, very thin and membranaceous, dark green and lustrous
on the upper surface, pale yellow-green on the lower surface, four to six inches long and four
or four and a half inches broad. The flowers and fruit I have not seen. The male catkins
in September are an inch and a half long, very slender, with bracts rounded and apiculate at
the apex. From the seeds, for which I am indebted to the Forestry officials of Hokkaido, a
large number of seedlings of this fine tree have been raised in the Arboretum. Specimens
collected in the Nikko Mountains by Mayr indicate that it is an inhabitant of Hondo, where,
however, we did not see it. From Yezo it ranges northward through Saghalin into Man-
churia. The tough thin bark is used by the Ainos for many domestic purposes.

The most common Birch of the high mountain forests of Hondo is Betula Ermani, a hand-
some species now well known in European and American collections, into which it has been
introduced through the agency of the St. Petersburg Botanic Garden. In Hondo, where it is
found scattered through the coniferous forests, it is common at elevations of from four to six
thousand feet above the sea, and is conspicuous from the white bark of the trunk and the
bright orange-colored bark of the principal branches. From the different forms of the White
Birch this species can be readily distinguished in the herbarium by the long spatulate middle
lobe of the bract of the female flower ; in the forest the color of the bark of the branches
well distinguishes it.

On the shores of Lake Yumoto we found a single individual of a black-barked Birch-tree,
much like our American Betula lenta, with the same Cherry-like flavor in the bark of the
branchlets. From Betula lenta it differed in its larger, more obtuse and paler winter-buds, in
the more prominent midribs and veins of the leaves covered on their lower surface with silky

PLATE XX.

C. E. FAXON del.

ALNUS JAPONICA, SIEB. F.T Zucc.

THE ALDER FAMILY. 63

pubescence, and in the shorter cones of fruit, the lateral lobes of the bracts being narrow and
acute, instead of broad and rounded, as in the American species. With considerable hesita-
tion I have referred this tree to the Betula serra of Siebold & Zuccarini. The seedling
plants which have been raised in the Arboretum will, perhaps, throw some light upon its true
position. Betula ulrnifolia, Betula Bhojpattra, and Betula corylifolia, included in the flora
of Japan, we did not see.

In Japan Alders are more numerous in species, and grow to a much larger size than in
eastern America. Aluiis incana, which is only a shrub here, in Japan becomes in some of its
forms a stately tree fifty or sixty feet in height, forming trunks often two feet in diameter.
Trees of this size of the varieties glauca and hirsuta, the latter well characterized by the pale
pubescence which covers the lower surface of the leaves, are common in Yezo, where they are
found on low slopes in moist rich ground, but not often close to the banks of streams, which
are usually occupied by Alnus Japonica. This is the largest and most beautiful of the
Japanese Alders. It is a pyramidal tree, often sixty to eighty feet tall, well furnished to the
ground with branches clothed with large dark green lustrous leaves. This species has been
confounded with the rare North American Alnus maritima, from which it differs in habit and
in the size and color of the leaves. The fruit of the two species is very similar, but the
Japanese tree flowers in the spring, ripening its fruit in the autumn of the same year, while
the American tree flowers in the autumn, and does not perfect its fruit until a year later.
The figure of Alnus Japonica (see Plate xx.) has been made from a drawing of a wild speci-
men gathered in Yezo.

Alnus Japonica is sometimes found in our collections, and is generally cultivated under
the name of Alnus firma. It is perfectly hardy in New England, where it grows rapidly, and
promises to become a large and handsome tree. The true Alnus firma, which is largely planted
along the margins of the Rice-fields near Tokyo to afford support for the poles on which the
freshly cut rice is hung to dry, was not seen growing under what appeared natural conditions ;
but the beautiful mountain-tree, distinguished by the thick conspicuously veined leaves, which
has been considered a variety of Alnus firma (var. multinervis), we often saw on the moun-
tains of Hondo, where it grows on dry rocky soil, and reaches elevations of some 5,000 feet
above the sea-level. It is a graceful tree, sometimes twenty or thirty feet high, with slender
spreading branches and thin flexible branchlets covered with ample thick dark green acute
leaves with from sixteen to twenty-four pairs of pale conspicuous straight veins. When better
known, this handsome tree will probably prove to be specifically distinct, and a garden-plant
of value.

What has been considered a form of Alnus viridis (var. Sibirica) is a very distinct-looking
plant in Japan, with broadly ovate cordate leaves fully twice as large as those produced by
Alnus viridis in America or Europe. At high elevations on Mount Hakkoda we found it
groAving as a bushy tree from twenty to twenty-five feet tall, and forming a short stout trunk.
A review of all the known forms of Alnus viridis will probably necessitate the separation of
this Japanese plant from that species.

Of the Oak family it is in Carpinus only that the forests of eastern Asia are superior to
those of America, where we have a single species of Hornbeam, a small tree confined to the

64 FOREST FLORA OF JAPAN.

eastern side of the continent. Europe possesses, also, a single species which extends to the
Orient, where a second species is found. The forests which cover the Himalayas contain two
species ; at least two or three others are found in the Chinese empire ; and to the flora
of Japan six species are credited. One of the Japanese species, however, Carpinus erosa of
Blume, is a doubtful plant; another, the Carpinus Tschonoskii of Maximowicz, from the
Hakone Mountains and the region of Fuji-san, I have never seen ; and a third, Carpinus
Yedoensis, a small tree cultivated in gardens in the neighborhood of Tokyo, is, perhaps, like
many of the plants cultivated by the Japanese, a native of China. Three species are certainly
indigenous to the Japanese soil.

Carpinus laxiflora resembles, in the character of the bark, in the size and shape of the
leaves, and in the structure of the flowers, the European and American Hornbeams. It is a
graceful tree, occasionally fifty feet in height, with a trunk eighteen to twenty inches in
diameter, covered with smooth pale, sometimes almost white, bark, and slender branches.
The leaves are ovate or ovate-elliptical, rounded or subcordate at the base, contracted at the
apex into long slender points, and doubly serrate ; they are dark green above, pale yellow-
green below, three to four inches long, an inch to an inch and a half broad, and prominently
many-veined, and in the autumn turn yellow or red and yellow. The fruit is produced in lax
hairy catkins four or five inches long, with spreading oblique prominently veined bracts,
which are obscurely lobed, more or less infolded at the base round the fruit, and nearly an
inch long. This fine tree is common in all the mountain forests of Hondo, where it is most
abundant at elevations between two and three thousand feet above the sea ; in Yezo it reaches
the southern shores of Volcano Bay, where, near the town of Mori, it is common in Oak
forests, and grows to its largest size.

The other Japanese species of Carpinus differ from Carpinus laxiflora and from the Ameri-
can and European species in their furrowed scaly bark, in the stalked bract of the male flower,
in the closely imbricated bracts of the fruiting catkins, which look like the fruit of the Hop-
vine, and in the form of these bracts, which are furnished at the base with a lobe which covers
the fruit, and is more or less inclosed by the infolding of the opposite side of the bract. On
account of these differences these two trees are sometimes referred to the genus Distigocarpus,
founded by Siebold & Zuccarini to receive their Distigocarpus Carpinus.

The figure (see Plate xxi.) shows flowering and fruiting branches of this tree, a staminate
flower, and a bract of the fruiting catkin. Botanists now pretty generally agree that the
characters upon which Distigocarpus was founded are not of sufficient importance to justify
its separation from Carpinus ; and Distigocarpus Carpinus, if the oldest specific name is used,
becomes Carpinus Carpinus. By Blume, who first united Distigocarpus with Carpinus, it was
called Carpinus Japonica, the name under which it has appeared in all recent works on the
Japanese flora. It is a tree forty to fifty feet in height, with a trunk often twelve to eighteen
inches in diameter, and wide-spreading branches which form a broad handsome head. The
branches are slender, terete, and coated at first with long pale hairs, and later are covered
with dark red-brown bark often marked with oblong pale lenticels. The winter-buds are half
an inch long, acute, and covered with many imbricated thin light brown papery scales ; with
the exception of those of the outer rows they are accrescent on the growing shoots, and at

PLATE XXI,

C. E. FAXON del.

CARPINUS CARPINUS, SARG.

THE HORNBEAM FAMILY. 65

maturity are nearly an inch long and hairy on the margins. The leaves are ovate, long-
pointed, slightly and usually obliquely cordate at the base, coarsely and doubly serrate, thick
and firm, dark green on the upper surface, paler on the lower surface, three or four inches
long and about an inch and a half wide, with stout midribs and many straight prominent
veins slightly hairy below and deeply impressed above. The stipules are linear, acute, scari-
ous, an inch long, and covered with pale hairs. The male inflorescence is an inch long, with
stalked lanceolate-acute bracts half an inch in length, and more or less ciliate on the margins.
The female inflorescence is two thirds of an inch long, and is raised on a slender stem coated,
like the bracts which subtend the ovaries, with thick white tomentum ; the outer bracts are acute,
scarious, a quarter of an inch long, and early deciduous ; the inner bracts are oblique, coarsely
serrate toward the apex, conspicuously many-ribbed, and furnished at the base with a minute
ovate serrate lobe which covers the ovary. Before the fruit ripens these inner bracts enlarge
until they are two thirds of an inch long and one third of an inch broad, and are closely
imbricated into a cone-like catkin which resembles in shape, color, and texture that of our
American Hop Hornbeam ; it is, however, often two or two and a half inches long. The
nutlet is slightly flattened, with about ten straight prominent ridges extending from one end
to the other.

Carpinus Carpinus is common on the Hakone and Nikko mountains between two and three
thousand feet elevation above the sea ; it apparently does not range very far north in Hondo
or reach the island of Yezo. This interesting and beautiful tree, which is remarkable among
Hornbeams in the character of the bark and in the female inflorescence, appears to be
perfectly hardy in New England. For a number of years it has inhabited the Arnold
Arboretum, and during the last two seasons has produced flowers and fruit here. In its
young state it makes a handsome, compact, pyramidal, bushy, and very distinct-looking tree.

But the most beautiful of the Hornbeams of Japan, as it appears in the forests of Yezo, is
Carpinus cordata, which often attains the height of forty feet, with a stout trunk sometimes
eighteen inches in diameter, covered with dark deeply furrowed scaly 'bark. The stout
branchlets are orange-color, or light brown when they are three or four years old, and are
covered with large oblong pale lenticels. This species is remarkable in the size of its winter-
buds, which are fully grown by midsummer, and are sometimes nearly an inch in length, and
acute, and covered with light chestnut-brown papery scales. The leaves are thin, broadly
ovate, pointed, deeply cordate, doubly serrate, six or seven inches long, and three or four
inches broad ; they are light green on both surfaces, although rather lighter colored on the
lower, with conspicuous yellow midribs and veins slightly hairy below and impressed above.
The catkins of fruit are often five or six inches long and an inch and a half wide, with
broadly ovate, remotely serrate bracts ; their basal lobe is proportionately much larger than
that of the last species, and is sometimes a third of the length of the bract, to which it is
often united along nearly its entire length, while in Carpinus Carpinus the lobe is only
attached at the base.

This is the only species of central Yezo, where it is one of the common forest-trees,
growing with Oaks, Magnolias, Ashes, Walnuts, Acanthopanax, Birches, etc. ; it also grows
in Hondo at high elevations, although it is here much less common than farther north. This

66 FOREST FLORA OF JAPAN.

fine tree is apparently still unknown in American and European gardens ; it is one of the
largest of the Hornbeams, and certainly one of the most distinct and beautiful of them all.
As it grows in its native forests with a number of trees which flourish here in New England,
it may be expected to grace our plantations with its stately habit, large leaves, and long
clusters of fruit.

Ostrya, the Hop Hornbeam, appears in North America with two species, one, a small
forest-tree of the eastern states, the second known only in the Grand Caiion of the Colo-
rado in Arizona ; a third species inhabits southern Europe, Asia Minor, and the Caucasus ;
the genus has no representative in the Himalaya forest region ; and, so far as I know, has
not been found within the borders of the Chinese empire or in Manchuria. It appears
again in northern Japan, however, where the Hop Hornbeam is one of the rarest of
Yezo trees. Maximowicz, who found it in the southern part of that island, considered the
Japanese Ostrya a variety of our American species and called it Ostrya Virginica, var.
Japonica.1 The American and the Japanese trees are very similar in botanical characters ;
indeed, it is difficult to find characters to separate satisfactorily the species of this genus,
which might all be considered geographical varieties of one. The Japanese and American
trees, however, look very differently in the forest, and there are differences in the bark
which are not easy to express in words. The leaves of the Japanese tree are thinner and
the heads of fruit are smaller than those on the American species (see Plate xxii.). Unfor-
tunately, I have not had an opportunity to examine the flowers of the Japanese tree ; it is
not probable, however, that they would afford a character by which the species could be dis-
tinguished. In the forests of Yezo I felt no doubt of its specific distinctness ; the meagre
and unsatisfactory material of the herbarium rather shakes than confirms this opinion. But,
all things considered, it is, perhaps, best to consider the Japanese tree as specifically distinct.
Not until it has been grown here during many years side by side with the American species
will it be possible to reach any opinion on this subject worthy of much consideration. If it
proves to be distinct it should bear the name of Ostrya Japonica. In the neighborhood of
Sapporo the Japanese Ostrya is rare ; here in low moist woods, growing with Oaks, Acantho-
panax, and Aralia, it sometimes attains a height of eighty feet, and forms a tall straight trunk
eighteen inches in diameter. We saw only one such tree, in the grounds attached to the
headquarters of the Forest Department of Hokkaido, and only two or three other individuals;
these were much smaller, perhaps not more than twenty feet high, and were scattered over
the Sapporo hills. We saw nothing of this tree in southern Yezo or in northern Hondo,
where Tschonoski, Maximowicz's servant and collector, found it on the high mountains of the

province of Nambu.

i Mil. EM. xi. 317.

PLATE XXII.

C. E. FAXON del.

OSTRYA JAPONJCA, SARG.

THE OAKS, CHESTNUTS, WILLOWS, AND POPLARS.

ALTHOUGH poorer in species and less important in the number, size, and value of individ-
uals than in eastern America, Quercus- furnishes one of the principal elements of the forests
of Japan. The types are all of the Old World, and there is nothing in Japan which corre-
sponds with our Red, Black, or Scarlet Oaks, or with the Black Jack, the Willow Oak, the
Shingle Oak, the Turkey Oak, the Spanish Oak, the Water Oak, or the Pin Oak, the Blue
Jack, or with our Chestnut Oaks. In the north and on the high mountains of Hondo there
are four White Oaks, and in the south a number of species with evergreen foliage of sections
of the genus, which are not represented in the United States. In the south, too, there
are a couple of deciduous-leaved species with biennial fructification of the Turkey Oak
(Quercus Cerris) sort.

The best known of the Japanese Oaks to European and American planters is Quercus
dentata (the Quercus Daimio of gardens). This tree is remarkable for the great size of its
leaves, which are often a foot long and eight inches broad, obovate in outline, and deeply
serrately lobed, and for the long loose narrow chestnut-brown scales of the large cup which
nearly incloses the small-pointed acorn. In central Hondo this tree is found only on the high
mountains, and it is not at all common ; but in the extreme northern part of the island it
appears in great numbers on dry gravelly slopes, at no great elevation above the sea. Here,
apparently, however, it does not reach the size it attains farther north, and the finest trees we
saw were on the gravelly plain south of Volcano Bay, and in the neighborhood of Sapporo.
The illustration (see Plate xxiii.) represents a group of these trees growing just outside of
Sapporo, and shows their habit at maturity. Although Quercus dentata grows to the height
of at least eighty feet, and forms a thick trunk more than three feet in diameter, it is not
an imposing or handsome tree in its maturity, and is only beautiful in youth. Old trees
lack symmetry and the appearance of strength, and are sprawling in habit, without being
picturesque. The bark is rather dark for a White Oak, and not unlike that of our Rock
Chestnut Oak (Quercus Prinus) ; it is valued for tanning leather, but the wood is considered
worthless. Quercus dentata appears to be the only deciduous-leaved Oak cultivated by the
Japanese, and small trees are common in the gardens of Toky5 and other southern cities,
where, however, they seem to languish. A variety (pinnatifida), with deeply divided leaves, is
cultivated in the Botanic Garden at Tokyo, and has, I believe, been introduced into Europe.

In central Yezo two noble White Oaks, Quercus crispula and Quercus grosseserrata, form
a considerable part of the forest-growth. The Dutch botanist Miquel considered them forms
of one species ; but Professor Miyabe, who has had the best opportunity for studying these
trees under the most favorable conditions, believes them to be distinct in their fruit, although
similar in foliage. In Quercus crispula he finds " the cup deeper, embracing about half the
cylindrical nut, falling off with it when ripe'; while in the latter, Quercus grosseserrata, the

68 FOREST FLORA OF JAPAN.

cup is hemispherical, inclosing about a third of the oblong ovoid nut, which falls off free
when ripe." His view, too, that Quercus grosseserrata cannot be specifically distinguished
from the Saghalin and Manchurian Quercus Mongolica, will probably be found to be correct.
Quercus crispula appears to range farther south than Quercus grosseserrata, which extends
north to the Kurile Islands, and was not recognized by us in Hondo ; on the Nikko Moun-
tains, on the road to Lake Chuzenji, we saw, however, fine forests of Quercus crispula. In
central Yezo, where the two species grow side by side on the hills, Quercus crispula appears
the more common tree on low ground, near the banks of streams. Both have elliptical or
obovate-oblong coarsely and irregularly lobed leaves, resembling in color and texture those of
the common Oak of Europe. Their bark is pale or sometimes dark, and scaly ; and both
species under favorable conditions rise to a height of eighty to a hundred feet, and produce
stems three to four feet in diameter. Both are timber-trees of the first class, and both, should
they thrive in this country, may be expected to add beauty and interest to our parks and
plantations. The smaller, shorter acorn of Quercus crispula appears to offer the only char-
acter for distinguishing the two trees; in their port, bark, and foliage they were indistin-
guishable to my eyes.

The fourth Japanese White Oak, Quercus glandulifera, ranges in Yezo nearly as far north
as Sapporo, although it is only south of Volcano Bay that it is really abundant. This, the
common Oak of the high mountains of central Japan at elevations over 3,000 feet, is probably
the most widely distributed species of the empire ; it is a pretty tree, rarely more than thirty
or forty feet high, although on the hills above Fukushima, on the Nakasendo, we saw
specimens nearly twice that height. The leaves are narrowly obovate or lanceolate-acute,
glandular-serrate, pale or nearly white on the lower surface, and from one to four inches in
length. The acorns are small, acute, and inclosed at the base only by the shallow thin-
walled cups covered with minute appressed scales. Like many American Oaks, this species
varies remarkably in the size of individuals, and in some parts of the country traversed by the
Nakasendo we found plants only a foot high covered with acorns. This Oak was sent to the
Arnold Arboretum many years ago from Segrez by Monsieur Lavallee. It is perfectly hardy
here, and has flowered for years, although it remains a bush, making no attempt to grow into
a tree.

Of the other deciduous-leaved Oaks, Quercus serrata, one of the most widely distributed
of the Asiatic species, ranging, as it does, from Japan to the Indian Himalaya, is common in
dry soil near the coast below Yokohama and on the foothills of the mountains of central
Hondo. It is a small tree, twenty to forty feet high, with a slender black-barked trunk and
beautiful dark green lustrous oblong acute leaves, their coarse teeth ending in long slender
mucros, and with small acorns inclosed in cups covered with long, loose, twisted, and reflexed
scales coated with soft pale tomentum. In Japan this tree appears to spring up in waste
lands in great numbers ; it is only valued for the charcoal which is made from it.

Quercus variabilis, a nobler tree of the same general character, we saw only in the grounds
of a temple near Nakatsu-gawa, on the Nakasendo, where there were specimens fully eighty
feet high, with tall straight trunks three or four feet in diameter, covered with thick pale
corky bark, which is sometimes used by the Japanese for the same purposes that we use the

PLATE XXIII.

^/7w^M-^*^-/;

: / /, ,. fwitofeT^MftMU ,F^E£rA,l

QUERCUS DENTATA, THUNB.

THE OAK FAMILY. 69

bark of the Cork Oak. The leaves are oblong-oval, pointed, less coarsely toothed than those
of Quercus serrata, dark green and lustrous above, and pale, or nearly white, below. From
Quercus serrata, too, it differs in the smaller cups and in their shorter thicker scales. A
number of plants have been raised in the Arboretum from the acorns which we picked up
under these trees, and if they are not hardy here in New England they will certainly thrive in
the middle states.

It is impossible to know whether many of the evergreen Oaks which we saw in Japan were
growing naturally or had been planted. In the gardens and temple grounds of Tokyo,
Yokohama, Kyoto, and other southern cities evergreen Oaks are the commonest trees ; but we
did not see them growing in the forest except near temples. The species most frequently
seen in Tokyo and Yokohama are Quercus cuspidata and Quercus glauca ; they are both large
and beautiful trees, said to be particularly conspicuous in early spring from the bright red
color of their young shoots and new leaves, which at that season make a charming contrast
with the dark and lustrous green of the older foliage. They should be introduced into our
southern states, where, probably, all the Japanese evergreen Oaks will flourish. The wood
produced by Quercus cuspidata and Quercus glauca does not appear to be valued in Japan,
but the acorns of the latter are of considerable commercial importance, and are cooked and
eaten by the Japanese.

Quercus acuta, which is also much planted in Tokyo, we saw growing to the height of
more than eighty feet, with Quercus variabilis, in the temple grounds at Nakatsu-gawa, and
also near the temple of Higane, near Atami, on the coast. It is a noble tree, with ovate,
acute, long-pointed, dark green, thick, and lustrous leaves. Quercus acuta has been intro-
duced into English gardens, with a number of other evergreen Japanese Oaks, through the
efforts of the Veitches, who obtained it some years ago from their collector, Maries. But the
finest Oak-tree, and perhaps the finest tree which we saw in Japan, was a specimen of Quercus
gilva in the temple grounds at Nara, where there are a number of remarkable specimens of
this beautiful species, which is distinguished by its lanceolate-acute leaves, which are glandular-
serrate only above the middle, bright green on the upper, and thickly coated on the lower
surface, like the young branches, with pale or slightly ferrugineous tomentum. The largest
of these Nara trees was probably a hundred feet high, with a trunk covered with pale scaly
bark, which, breast-high from the ground, girthed just over twenty-one feet ; it rose without
a branch, and with little diminution of diameter, for something like fifty feet, and then
separated into a number of stout horizontal branches, which had not grown to a great
length, and formed a narrow cylindrical round-topped head.

Of the other Japanese Oaks, Quercus Thalassica, Quercus Vibrayiana, and Quercus glabra,
we only saw occasional plants in gardens. The Quercus lacera of Blume we did not see
at all.

The Chestnut-tree is widely distributed through the mountain forests of Japan, and seems
to have received some attention as a fruit-tree from the Japanese, who recognize a number of
large-fruited varieties. Very large chestnuts appear in profusion in the markets of Aomori,
and are said to be produced in the immediate neighborhood of that northern town. But the
largest chestnuts of Japan, which equal in size the best marrons of southern Europe, are

70 FOREST FLORA OF JAPAN.

found in the markets of Kobe and Osaka. It is these Kobe marrons which are now sent to
San Francisco in considerable quantities. Rein, whose book on the industries of Japan con-
tains the fullest and most exact account of Japanese rural economy which has yet been written,
believed that the chestnut was less used in Japan as an article of human food than in Europe,
but I have never seen chestnuts offered in such quantities in the markets of any American or
European city as in those of Tokyo and other Japanese towns. The Chestnut-trees which we
saw had the appearance of growing spontaneously ; and we saw nothing like an orchard of
these trees, which, so far as we were able to observe, are not planted near dwellings or temples
or for shade. In Japan the Chestnut-tree grows at least as far north as central Yezo, and is
scattered through the mountain forests of Hondo, where it is most abundant at elevations of
about 2,500 feet above the ocean, growing on steep slopes in small open groves or mixed with
trees of other kinds. We saw no evidences that the Chestnut-tree grows in Japan to the
noble dimensions it sometimes reaches in Europe and on the slopes of the southern Alleghany
Mountains, and specimens over thirty feet high, with trunks more than a foot in diameter,
were rare in that part of the country which we visited. The Japanese Chestnut appears
to be more precocious than the American tree, and saplings only ten or twelve feet high are
often covered with fruit. The large-fruited northern form from the neighborhood of Aomori
should be brought to this country, as it may be expected to support a greater degree of
cold than the French or Kobe Marrons, and, therefore, to be available for cultivation much
farther north here. By its introduction it is possible that marron-growing may become a
profitable industry in states with climates as severe as those of Wisconsin, Michigan, and
Massachusetts.

The Beech in Japan is one of the noblest trees of the forest, as it is in eastern North
America and in Europe. Its range is similar to that of the Horse-chestnut, in the north
appearing on the shores of Volcano Bay in Yezo only a few feet above the level of the
ocean, and extending southward along the mountains of the other islands. It is, perhaps, the
commonest deciduous tree of the mountains of Hondo, where, between 3,000 and 4,000 feet,
or toward the upper limits of the deciduous forest, it sometimes covers wide areas, nearly to the
exclusion of other trees, or sometimes grows mixed with Oaks, Chestnuts, and occasional Firs
and Spruces. Trees eighty or ninety feet tall, with trunks more than three feet in diameter,
are not uncommon. The fact that Beech-wood is little used by the Japanese, and the com-
paratively inaccessible situations where it is mostly found, account, no doubt, for the abundance
of this tree in Japan and the existence of so many large individuals. This, the Asiatic form
of the European Fagus sylvatica, is hardly to be distinguished from the European tree, which
it resembles in every essential character. The variety Sieboldii (the Fagus Sieboldii of
Endlicher) I looked for in vain, and I hazard the opinion that it will turn out to be a tree of
the herbarium and not of the forest.

I can throw no light upon the Japanese Willows which abound at the north in numerous
continental mostly shrubby forms ; they require more careful investigation than it was possi-
ble to give them during our hurried autumnal visit, when the flowers and fruit had disap-
peared. On the streets of Europeanized Tokyo, Willows are now chiefly planted as shade-
trees ; they are the Weeping Willow (Salix Babylonica), an inhabitant of China,and a favorite

THE WILLOW FAMILY. 71

with the Japanese, and Salix eriocarpa of Franchet & Savatier, a species which we saw
growing by river-banks on the Nakasendo, and which looks too much like Salix alba to be
distinct from that species which might be expected to reach Japan. But the handsomest
Willow we saw in Japan, and certainly one of the most beautiful of all Willows, is Salix
subfragilis, which appears to be confined to Japan, where it was discovered in the neighbor-
hood of Hakodate by Charles Wright. We were first struck by the beauty of this tree
between Nikko and Lake Chuzenji, where there are a few specimens on the banks of the
mountain torrent which the road follows in ascending the mountains. It was at Sapporo,
however, that this Willow appeared in its greatest beauty ; here on the banks of streams Salix
subfragilis forms trees at least fifty feet in height, with short stout trunks three or four feet
in diameter, covered with thick deeply furrowed bark, and stout branches which spread nearly
at right angles, like those of an old pasture Oak. The leaves are oblong, acute, rounded at
the base, and coarsely crenulate-serrate ; they are borne on stems an inch and a half in length,
and are six or seven inches long, two or two and a half inches broad, dark green and lustrous
on the upper surface, and silvery white on the lower ; the stipules are foliaceous, obliquely
rounded, and rather more than half an inch across. This Willow appears to be one of the
most desirable trees to introduce into our collections, and the only Japanese Willow we saw
of real value, from a horticultural point of view.

Populus is poorly represented in Japan ; the two species which are found in the empire are
both of Old World types, and there is nothing which corresponds to the Cottonwoods, which
line the river-banks in all the central and western regions of this continent. The Aspen of
Europe appears in one of its forms in Japan (Populus tremula, var. villosa), looking, however,
so distinct from the continental Aspen that it is hard to believe that it is not specifically
distinct ; it is the Populus Sieboldii of Miquel, the oldest name. This tree is not rare in
southern Yezo, where it grows to the height of twenty or thirty feet, springing up in consid-
erable numbers on dry, gravelly soil. We saw it in the greatest perfection on the plains
south of Mori, on Volcano Bay, and less commonly on the mountains near Aomori in Hondo.
Of the second species, the Populus suaveolens of Fisher, we encountered a few individuals in
southern Yezo, where it is probably near the southern limit of its range, it being a tree of
Saghalin and the Amour country. It is evidently only a form of the Balsam Poplar, which is
found in all northern regions, where, especially in some parts of British America, it constitutes
by far the largest part of the forest-growth. In Japan the Balsam Poplar grows to an
immense size, and some individuals which we saw were certainly eighty and perhaps a
hundred feet tall, with long trunks five or six feet in diameter, rising like sentinels above
the low mostly second-growth forests of southern Hokkaido.

THE CONIFERS.

IN cone-bearing plants Japan is somewhat richer than eastern America. All of our gen-
era, with the exception of Taxodium, the Bald Cypress of the southern states, are represented
in the empire, where two endemic genera occur, Cryptomeria and Sciadopytis, and where
Cephalotaxus and, perhaps, Cupressus have representatives. In cone-bearing species, too,
owing to the greater multiplication of forms of Abies, Japan is richer than eastern America,
where we have only two indigenous Firs. The genus Pinus, which furnishes a very consider-
able part of the forest-growth on the Atlantic seacoast, and which is represented here by
thirteen species, has only five in Japan ; and of these two are small trees of high altitudes
and one is an alpine shrub. Japan is richer than eastern America in Spruces, of which we
have only two, in Chamsecyparis, and in Juniperus. The two floras each contain a single
Thuya, a Taxus, a Tumion, two Hemlocks, and a Larch. In Japan, Conifers are more
planted for shade and ornament than they are in America, or, perhaps, in any other country,
although, except above 5,000 feet in Hondo, where there are continuous forests of Hemlock,
they form a small part of the composition of indigenous forest-growth ; and forests of Pines,
Spruces, or Firs, such striking features in many parts of this country, do not occur, except,
perhaps, in northern Yezo, which we did not visit, and where there are said to be great forests
of Abies Sachalinensis.

The Japanese Arbor-vitse, Thuya Japonica, which is sometimes found in our plantations
under the name of Thuyopsis Standishii, and which is more like the species of the northwest
coast (Thuya gigantea) than our eastern Arbor-vita? or Yellow Cedar, appears to be a rare tree
in Japan, and we saw only a few solitary individuals on the shores of Lake Chuzenji and of
Lake Yumoto in the Nikko Mountains. Here it was a formal pyramidal tree twenty or thirty
feet high, with pale green foliage and bright red bark.

Thuyopsis dolobrata, which is, perhaps, best considered a Cupressus rather than a Thuya,
is a tree of high altitudes. In the Nikko Mountains above Lake Yumoto it is common between
5,000 and 6,000 feet above the sea-level, growing as an under-shrub under the shade of dense
Hemlock forests, and here, in favorable positions, sometimes finally rising to the height of
forty or fifty feet, with a slender trunk covered with bright red bark, long pendulous grace-
ful lower branches, and a narrow pyramidal top. This handsome tree finds its northern home
on the mountains which surround the Bay of Aomori, in northern Hondo. It is a species
which evidently requires shade, at least while young, and even the older plants, where we saw
them, were always surrounded and overtopped by taller trees. The elevation at which this
tree grows indicates that it should prove hardy here if properly protected from the sun, espe-
cially during the winter, for in Japan the young plants are not only shaded by the coniferous
forest above them, but are buried during several months under a continuous covering of
snow. Under proper conditions this tree will be found to be one of the best plants to form

THE CONIFERS. 73

undergrowth in coniferous forests. The wood is considered valuable, and, owing to its dura-
bility, is used in boat and bridge building. We saw planted trees in the coniferous forests
on the mountain-slopes near Nakatsu-gawa, in the valley of the Kisogawa, but no other
indication that it is valued as a timber or ornamental tree by the Japanese, who, according
to Dupont,1 have produced a number of varieties, of which the one with variegated foliage
only has reached our gardens.

Of all the Japanese Conifers the most valuable is the Hi-no-ki, Chama3cyparis (Retinospora)
obtusa. In the forests planted on the lower slopes of mountains in the interior of Hondo, and
in some of the temple groves, notably in those of Nikko, this fine tree attains a height of a
hundred feet, with a straight trunk without branches for fifty or sixty feet, and three feet
through at the ground. At elevations between 2,000 and 3,000 feet above the sea, usually
on northern slopes and in granitic soil, which it seems to prefer, the Hi-no-ki is largely
planted as a timber-tree ; indeed, only the Cryptomeria, which seems to be less particular about
soil and exposure, is more planted for timber in Japan. The tree is sacred among the disci-
ples of the Shinto faith, and is, therefore, cultivated in the neighborhood of all Shinto temples,
which are built exclusively from Hi-no-ki wood. The palaces of the Mikado in Kyoto were
always made of it, and the roof was covered with long strips of the bark. It is considered the
best wood to lacquer ; at festivals food and drink are offered to the gods on an unlacquered
table of this wood, and the victim of harakari received the dagger upon a table of the
same material. It is used for the frames of Buddhist temples and for the interior of the
most carefully finished and expensive houses. The wood is white or straw-color, or sometimes
pink, and in grade, texture, and perfume resembles that of the Alaska Cedar, Chamsecyparis
Nootkatensis. Like the wood of that tree, it has a beautiful lustrous surface, and is straight-
grained, light, strong, and tough, and remarkably free from knots and resin. In America we
have no wood of its class which equals it in value, with the exception of that furnished by the
two species of Chamsecyparis of the Pacific coast, and the Hi-no-ki might be introduced with
advantage as a timber-tree into those parts of the eastern states where it could find condi-
tions which would insure its growth. It has proved perfectly hardy in this country as far
north as Massachusetts, but the sea-level or the dry summers here do not suit any of the
Retinosporas, which give no promise of long life or great usefulness anywhere on the Atlantic
seaboard. They should be tried, however, on the slopes of the southern Alleghanies where
they could find conditions not very unlike those in which they flourish in their native land.

The second species of this genus, Chanisecyparis pisifera, the Sawara, is a less valuable tree
than the Hi-no-ki, although the two species are always found growing together in plantations
and in temple gardens ; indeed, they can only be distinguished after some practice, unless the
cones are examined, although after a few days among them the more ragged crown, with its
looser and more upright branches, of the Sawara stands out clearly to the eye in contrast with
the Hi-no-ki with its rounder top and more pendulous branchlets. The wood is of a reddish
color, of a rougher grain, and less valuable than that of the Hi-no-ki, although the two trees
are planted in about equal numbers. As it grows here in our gardens, Chamaecyparis pisifera
is a less ornamental plant than Chamaecyparis obtusa ; it grows, however, more vigorously, and

1 Les Essences Forestieres du Japan.

74 FOEEST FLORA OF JAPAN.

promises to live longer and attain a greater size. All the other Retinosporas of our gardens
are juvenile or monstrous forms of these two trees. Some of the dwarf forms are much culti-
vated in Japan especially as pot-plants, but they are not as popular there as I had heen led
to expect, and are most often seen in the nursery-gardens of the treaty ports, where they are
collected to please the fancy of foreign purchasers.

The most generally planted timber-tree of Japan is the Sugi, Cryptomeria Japonica, and its
wood is more universally used throughout the empire than that of any other Conifer. It is
one of the common trees of temple gardens and roadside plantations, and, when seen at its
best, as in the temple groves of Nikko or Nara, where it rises to the height of a hundred or a
hundred and twenty-five feet, with a tall shaft-like stem tapering abruptly from a broad base,
covered with bright cinnamon-red bark and crowned with a regular conical dark green head,
it is a beautiful and stately tree which has no rival except in the Sequoias of California.
Great planted forests of the Cryptomeria appear all over Hondo on broken foothills and
mountain-slopes up to elevations of nearly 3,000 feet above the sea, low valleys and good
soil being usually selected for such plantations, as the trees need protection from high winds.
The plantations decrease in size and luxuriance in northern Hondo, and the cultivation of the
Sugi does not appear to be attempted north of Hakodate, where there is a grove of small
trees on the slope of the hill above the town. The wood is coarse-grained, with thick layers
of annual growth, dark reddish heartwood, and thick pale sapwood ; it is easily worked,
strong and durable, and is employed in all sorts of construction. The bark, which is careful'ly
stripped from the trees when they are cut down, is an important article of commerce, and is
used to cover the roofs of houses. A large round bunch of branchlets covered with their
leaves hung over the door of a shop is the familiar sign of the dealer in sake.

Japan owes much of the beauty of its groves and gardens to the Cryptomeria. Nowhere
is there a more solemn and impressive group of trees than that which surrounds the temples
and tombs at Nikko, and the long avenue of this tree, under which the descendants of
leyasu traveled from the capital of the Shoguns to do honor to the burial-place of the
founder of the Tokugawa dynasty, has not its equal in stately grandeur. This avenue, if the
story told of its origin is true, can teach a useful lesson, and carries hope to the heart of the
planter of trees, who will see in it a monument more lasting than those which men sometimes
erect in stone or bronze in the effort to perpetuate the memory of their greatness. When the
body of leyasu was laid in its last resting-place on the Nikko hills, his successor in the Shogun-
ate called upon the Daimyos of the empire to send each a stone or a bronze lantern to deco-
rate the grounds about the mortuary temples. All complied with the order but one man,
who, too poor to send a lantern, offered instead to plant trees beside the road, that visitors to
the tomb might be protected from the heat of the sun. The offer was fortunately accepted,
and so well was the work done that the poor man's offering surpasses in value a thousand-fold
those of all his less fortunate contemporaries.

Something of the stateliness of this avenue appears in our illustration (Plate xxiv.)
although, without the aid of colors, it is impossible to give an idea of the beauty of the Cryp-
tomeria. The planted avenue extends practically all the way from Tokyo to Nikko, but it
is only when the road reaches the foothills that it passes between rows of Cryptomerias, the

a

5

THE CONIFERS. 75

lower part being planted, as is the case with the other great highways of Japan, with Pine-
trees ; nor is this avenue continuous, as has often been stated, for whenever a village occurs
or one of the roadside tea-houses, which are scattered all along the road, there is a break in
the rows of trees, and it is only in some particular spots that a long view of continuous trees is
obtained. The railroad, which follows parallel and close to the avenue for a considerable dis-
tance and then crosses it just before the Nikko station is reached, is a serious injury to it.
The trees, as will be seen in the illustration, are planted on high banks made by throwing up
the surface-soil from the roadway ; they are usually planted in double rows, and often so close
together that sometimes two or three trees have become united by a process of natural graft-
ing. Young trees are constantly put in to fill gaps, and every care apparently is taken
to preserve and protect the avenue. How many of the trees originally planted when the
road was first laid out in the beginning of the seventeenth century are left it is impossible
to say, but I suspect that most of those now standing are of much later date. One of the
trees close to the upper end of the road which had been injured by fire was cut down during
our visit to Nikko. The stump, breast-high above the ground, measured four feet across inside
the bark, and showed only one hundred and five layers of annual growth. Few of the trees
in the avenue were much larger than this, although in the neighborhood of the temples there
are a few which girt over twenty feet ; these were probably planted when the grounds were
first laid out.

The two, Chamsecyparis and the Cryptomeria, the most valuable timber-trees in Japan, are
now almost unknown in a wild state. They may, perhaps, be found growing naturally on
some of the southern mountains which we did not visit ; but wherever we went, we saw only
trees that had been planted by man, although some of the plantations had evidently lived
through several centuries.

Cephalotaxus drupacea is the only Japanese member and the type of a genus of half a
dozen species distributed from Japan, through China to northern India. It is widely and quite
generally scattered through the mountain regions of the empire, extending north to central
Yezo, where it appears on the low hills as an under-shrub only two or three feet high, while
on the Hakone Mountains, in Hondo, it occasionally grows into a bushy tree twenty or twenty-
five feet in height. Cephalotaxus drupacea is now common in our gardens, although it is not
very hardy or satisfactory here in New England, where it often suffers in winter, missing,
no doubt, the thick and continuous covering of snow which protects it in Yezo. Like its
relative, the Gingko, the same individual does not produce male and female flowers, and
the fruit, like that of the Gingko, is an almond-like nut inclosed in a fleshy covering. A res-
inous oil, used in lamps, is pressed from the seeds, and the wood is occasionally employed in
cabinet-making.

The Gingko, although we are in the habit of associating it with Japan, is in reality not a
native of that country, into which it was brought with their religion by the Buddhist priests.
It is still unknown in a wild state, and it is possible that this genus, Avhich was widely distrib-
uted with many species through the temperate and colder parts of the northern hemisphere in
tertiary times, has become exterminated from its native forests, and has only been preserved
through the agency of the priests of Buddha, who seem to hold it in particular respect. The

76 FOREST FLORA OF JAPAN.

hardiness of this beautiful tree, which thrives under the most trying conditions and in the
severest climates, indicates that it originated in some northern interior region ; and if it is
ever seen growing without cultivation it will probably be in some remote district of Mongolia.
There are noble great broad-branched specimens in the neighborhood of the temples in
Tokyo fully a hundred feet high, with tall massive trunks six or seven feet in diameter. The
Gingko is, perhaps, the most beautiful, as it is certainly the most interesting tree which is
to be seen in Japan ; and in the autumn, especially when the sunlight flutters through the
bright yellow leaves, these great trees, with their broad heads of graceful semipendulous
branches, are magnificent objects. The fleshy covering of the fruit has a rancid and most
disagreeable flavor, but the kernel of the almond-like stone is delicate and is esteemed a lux-
ury in both China and Japan, where it is found in the markets in considerable quantity.
The wood, which is light yellow in color, is soft and brittle, and as the trees grow to a very
great age and are planted only for ornament in Japan and rarely cut down, it has no economic
importance there.

Torreya, or, if the custom which now prevails among American botanists is followed,
Tumion, Rafinesque's name, which also appears in eastern and western America and in China,
occurs in Japan in its largest and most beautiful representative, Tumion nuciferum, one of the
handsomest of all coniferous trees. Although nowhere very common, the Kaya, as this tree
is called in Japan, was seen in all the mountainous regions of central Hondo which we visited.
It often grows as an under-shrub in the forest, or as a small tree twenty or thirty feet tall,
but occasionally rises to the dignity of a tree of the first class, as on the banks of the
Kisogawa, near Agematsu, where we saw specimens fully eighty feet high, with great trunks
four or five feet in diameter. Such trees, with their bright red bark and compact heads of
dark green, almost black lustrous foliage, possess extraordinary beauty. No other Yew-like
tree which I have seen equals it in massiveness and depth of color, and the Kaya should be
cultivated wherever the climate permits it to display its beauty. The elevation above the sea
at which it flourishes in Japan indicates that it will be hardy in the middle states, although
we cannot expect to see it grow to any size in New England. An oil used in cooking, kaya-
no-abura, is an article of considerable commerce in Japan, and the kernels of the nuts, which
possess an agreeable, slightly resinous flavor, are sold in great quantities in the markets in
the autumn, and are a favorite article of food. The wood is strong, straight-grained, light
yellow, and valued in building and cabinet-making.

Taxus, which has two species in eastern America, one in the north and another, almost
the rarest of American trees, in the south, which is represented in Avestern North America,
and is widely distributed through Europe and continental Asia, appears in Japan with a noble
tree, Taxus cuspidata, which, to judge by our observations, is confined to the island of Yezo,
where it is not uncommon on the low hills of the interior. Here it often attains the height of
forty or fifty feet, and forms a trunk two feet in diameter, covered with bright red bark.
The Yew is often employed by the Japanese to ornament their gardens, and the wood, which
is exceedingly hard, tough, and of a bright red color, is used by the Ainos for their bows, and
is valued in cabinet-making and for the interior decorations of expensive houses. This beau-
tiful tree, as is now well known, flourishes in this country, where it has proved itself perfectly
hardy, promising to be really valuable as an ornamental plant.

THE CONIFERS. 77

It is not thought now that Podocarpus, a genus of the tropical and subtropical regions of
both hemispheres and of Tasmania, is indigenous in Japan, although two species are often
cultivated there. The more common is Podocarpus macrophylla, a small tree with lanceolate
acute leaves, and a common hedge-plant in Tokyo gardens, in which it is often cut into
fantastic shapes. It is a much less beautiful, although a hardier, tree than Podocarpus
Nageia, with its thick broad glossy leaves and beautiful purple trunks, the second species
seen in Japan. It is one of the favorite subjects, especially in a variety in which the leaves
are marked by broad white stripes, for dwarfing and pot-culture. The real beauty of this
tree is only seen, however, when it has become large and old and the trunk is covered with its
peculiar smooth purple bark. A grove of these trees on the hill behind the Shinto temples
at Nara is one of the most interesting spots in Japan, and in solemn dignity and beauty is
only surpassed by the grove of Cryptomerias which surrounds the mausoleums of leyasu and
lemitsu at Nikko.

Like Crvptomeria, Sciadopitys is monotypic and endemic to Japan. It is one of the most
curious and interesting of trees, with scale-like leaves in whose axils are produced the phylloid
shoots, which are generally mistaken for the leaves, and which are arranged near the ends of
the branches like the ribs of an umbrella, — a peculiarity to which this tree owes its familiar
English name, the Umbrella Pine. Like the Gingko, the Sciadopitys was for a long time
known only from a few individuals cultivated in temple gardens and from the grove on the
hill in Kyushu, where the ancient monastery town of Koya stands, to which the Sciadopitys
owes its Japanese name, Koya-maki. There is said to be a remarkable grove of these trees
here, which was once supposed to be the original home of the species ; but Rein and other
writers now agree in thinking that they were originally planted by the monks. Dupont found
what he considered indigenous trees on Chimono and in the province of Mino. In this prov-
ince, on the Nakasendo, below Nakatsu-gawa, we saw young plants of the Koya-maki in all the
roadside gardens, a pretty sure indication in this remote region that the tree was growing in
the woods not very far off, and here for the next two or three days we saw it sending up its
narrow pyramidal heads above the Pines and other trees of the forest, growing, as we thought,
quite naturally, and leading us to believe that we had found the true home of this tree,
although in a country like Japan, which has been densely populated for centuries, and in
which tree-planting has been a recognized industry for more than .a thousand years, it is not
easy to determine whether a forest has been planted by man or not. But whether these trees
had been planted or whether they were the offspring of trees brought from some other region,
or the indigenous inhabitants of the forest, the Sciadopitys grows on the mountains of Mino
in countless thousands, often rising with a tall straight trunk to the height of nearly a hun-
dred feet, and remarkable in its narrow compact pyramidal head of dark and lustrous foliage.
The wood, which is nearly white, strong, and straight-grained, is a regular article of commerce
in this part of Japan, and from Nakatsu-gawa is floated in rafts down the Kisogawa to Osaka,
where it is said to be chiefly consumed. Except in the neighborhood of Nakatsu-gawa, the
Sciadopitys is not very much cultivated as a garden-plant in Japan ; and it is not often found
in old gardens, except in the immediate neighborhood of temples, where picturesque old speci-
mens may occasionally be seen occupying a place of honor within the fence which incloses

78 FOREST FLOE A OF JAPAN.

the principal buildings, and carefully protected by low stone railings. There is a remarkable
specimen with pendulous branches standing before one of the mortuary temples in the Shiba
Park in Tokyo.

In Japan, Junipers are much less common than they are in eastern America, and although
five or six species are included in the Floras of the empire, the genus does not make an impor-
tant element of the landscape, and one misses the dark spires which Juniperus Virginiuna
sends up so frequently in many parts of eastern America. Juniperus Chinensis appears to
grow to a larger size than the other Japanese species, although we saw it in only one region
growing, as it appeared, without cultivation. This was on the high volcanic ridge which
dominates the Chikuma, one of the streams which flow from Asama-yama, in central Japan.
Over this elevated and inhospitable region occasional Junipers are scattered, the largest
attaining a height of thirty or forty feet, their wind-swept heads and straggling branches,
covered with gray-green foliage, adding to the dreariness of the scene. Before the Buddhist
temple of Zenkogi, in Nagano, the principal city in this part of Japan, two venerable Junipers
show to what size plants of this species can grow, and how picturesque they can become.
These trees are seventy or eighty feet high ; their hollow trunks, which are rather more than
six feet in diameter, support narrow heads of twisted and contorted branches clothed with
scanty foliage, and indicate that centuries may well have passed since the roots of these
marvelous trees first penetrated this sacred soil. On the rocky cliffs and grassy slopes of the
coast, fully exposed to the spray of the ocean, a prostrate variety of Juniperus Chinensis forms,
with its long creeping stems, dense mats, often of considerable size. It is said to be a feature
of the littoral vegetation of Japan ; but we saw only a few plants in Yezo, between Mororan
and the Aino village of Horobetsu, where they receive the unbroken sweep of the Pacific.
On the sandy dunes of the Bay of Hakodate, opposite that city, another littoral Juniper was
found by the American botanist Charles Wright, and later by Maximowicz. This is the
Juniperus conferta of Parlatore (Juniperus littoralis of Maximowicz), a species distinguished
by its stout crowded leaves and large globose fruit. We saw it on Hakodate Bay at the end
of September, and Mr. Veitch collected it earlier near Han jo, on the west coast of Hondo,
but we were too early to obtain ripe fruit of this species or of the prostrate form of Juniperus
Chinensis. The only other Juniper we saw in Japan, Juniperus rigida, is a small tree some-
times twenty feet high, but more often a low spreading bush. It is common in the barrens
near Gifu, and appears to be generally distributed at low elevations in central Japan, although
it grows only on dry sterile gravelly soil. This is the Juniper which is most commonly
cultivated by the Japanese, and which is not infrequently an inhabitant of temple gardens.
Its long slender rigid leaves and small fruit, tipped with a minute niucro, serve to distinguish
it from Juniperus conferta.

THE CONIFERS, II.

Or the true Pines of Japan two species are valuable timber-trees ; these are Finns densi-
flora and Pinus Thunbergii ; both bear an important part in the decoration of Japanese
gardens, and one at least has had its influence in all expressions of the artistic temperament
of the people. All the Pine-woods of Japan, except those found on the upper slopes of some
of the high mountains of central Hondo, have evidently been planted. Such planted woods
are often seen covering sandy plains near the coast, and the principal highways of the empire
are shaded by avenues of these Pines, except where Cryptomerias replace them when moun-
tains are crossed. Of the two species the Black Pine, Pinus Thunbergii, appears to be the
most commonly cultivated, and to grow to the larger size. Of its distribution and appearance
when growing naturally I was not able to get an idea, as all the plants I saw had evidently
been planted by man. It is of this species that the plantations of the coast are mostly formed,
although the two species are generally found mixed together in all plantations ; and it is this
species which is usually selected by the Japanese gardener when he wants to make the
branches of a Pine-tree cover an arbor or hang suspended over the sides of a moated wall,
and which is found in every garden and is most revered by the Japanese. Pinus Thun-
bergii is one of the most picturesque of Pines, with a broad head of stout contorted some-
what pendulous branches, often growing to the height of eighty feet, and producing trunks
three feet through. Its dark deeply furrowed bark, darker colored and thicker leaves and
white buds serve to distinguish it from the Red Pine, Pinus densiflora, which is a tree of
high altitudes, and which, although planted in large plantations and by the sides of high-
ways, does not appear to be such a favorite in gardens as the Black Pine. The Red Pine
we saw growing wild high up on Mount Koma-ga-take, in central Hondo, and on the Nikko
Mountains, where, at about 3,000 feet above the sea-level, it is not rare. It is a more
slender tree than the Black Pine, with thinner lighter green leaves. The bark on the upper
part of the trunk and on the main branches is light red, separating in thin scales, so that
a forest of these trees presents a bright and cheerful appearance. Several varieties of the
two species recognized by Japanese gardeners are described by Mayr, who also found what he
thought was a hybrid between them.1 The wood of the two species is very similar, and,
apparently, is not distinguished in Japanese lumber-yards. It is coarse-grained, resinous, and
moderately strong, and is used in great quantities in all sorts of coarse construction, and as
fuel, the rapid growth of the trees on soil too poor to produce more valuable crops to
advantage rendering it exceedingly cheap. These two Pines have long inhabited our gardens,
where they are hardy and grow with great rapidity, some of the oldest plants of the Red Pine
here already beginning to show the picturesque habit which in their native country constitutes
the charm of these trees.

1 Die Abietineen des Japanischen Reiches, 83, t. 7, f. 2, 3, 4.

80 FOREST FLORA OF JAPAN.

The other Pines of Japan belong to the group in which the species produce their leaves in
clusters of fives. The largest and the most widely distributed is Pinus parviflora, a beautiful
small tree of high mountain forests, through which, at elevations over 5,000 feet above
the sea-level, it is found scattered, either singly or in small groves, sometimes growino- to
a height of sixty or seventy feet, although it is usually much smaller. In those parts of
Japan which we visited, it was most common and grew to the largest size on the slopes of
Mount Hakkoda, in northern Hondo, where its dark pyramidal heads of slender spreading
branches, rising above the forests of Oaks and Beeches, break the sky-line, just as its relative,
our eastern White Pine, raises its nobler head high above the Oak forests of New England.
The wood of Pinus parviflora is soft, straight-grained, light-colored, and of considerable value,
but so difficult to obtain that it is little known or used by the Japanese. This beautiful Pine
flourishes in our gardens, where it appears to be perfectly at home, and where it grows rapidly
and every year covers itself with cones.

In southern Yezo, a second species of the same group, Pinus pentaphylla, has been distin-
guished by Mayr. This is an exceedingly rare tree, found in a few isolated situations, and
distinguished from Pinus parviflora by its longer cones and stouter leaves. We saw only a
cultivated tree at the hot springs of Kakumi, near the shore of Volcano Bay, being prevented
by bad weather from reaching a small grove of these trees growing on the mountains in the
neighborhood. This Pine has not been introduced into our gardens, where it may be expected
to flourish.

The fifth Japanese Pine is interesting from the fact that it is the only Japanese Conifer
which grows naturally in North America. It is the Pinus pumila of Kegel, a species so similar
to the Stone Pine of Europe that it has been considered by many authors a variety of that
tree. We saw it only on the summit of Mount Hakkoda, where it forms, at 6,000 feet above
the sea-level, impenetrable thickets a few feet in height and hundreds of acres in extent ;
it occurs on the summits of some of the high mountains of Yezo, ranges north through
Saghalin and eastern Manchuria to Kamtschatka, and by the Kurile Islands reaches those of
the Alaska coast.

Of Spruces, there appear to be four species in Japan, where, except, perhaps, in some
parts of Yezo, they are exceedingly rare. The first, Picea polita, we saw only in two or three
individuals in the Nikko Mountains, on the hills below Lake Chuzenji. The trees were small,
much torn and stunted by the wind, and of such a miserable appearance that it was difficult to
realize that the young trees in perfect health and beauty which decorate our gardens belonged
to the same species. For the second Spruce of the mountain forests of central Hondo, to
which it appears to be confined, Mayr proposes the name of Picea bicolor, this specific name
having, he finds, been first used by Maximowicz for this tree. This is the beautiful Spruce
with blue-green leaves, silvery white on the under surface, which is usually cultivated under
the name of Picea Alcockiana, and which is easily distinguished in the spring by the bright
red color of the young shoots.

It is not my purpose to discuss here the synonymy of the Japanese Firs and Spruces, upon
which such a mass of names have been heaped in almost hopeless confusion, that only a
critical examination of all the specimens which have been studied by European botanists can

1

<J

8
s

w

C3

2

o

S

s

w

a

THE CONIFERS. 81

make it possible to reach any useful conclusions on the subject, and I shall only speak of the
trees as I saw them growing in the forests of Japan. Picea bicolor, of which we saw but
three or four specimens, is evidently a rare and local tree, found only at high elevations,
scattered through the Oak and Beech forests, and, like Picea polita, presenting in its home a
wretched and forlorn appearance. The leaves are nearly equally four-sided, and the cones
are four to six inches long, with narrow pointed more or less laciuiate scales. These two
species, so far as I was able to observe, are the only Spruces which grow on the island of
Hondo, the other species finding in Yezo their most southern home. These are Picea Ajanen-
sis, a tree with smaller cones than the last, and short broad flat leaves, dark green above
and pale on the lower surface. This is the common Spruce of Yezo, occurring on the hills
near Sapporo, which is the only place where I saw it, in isolated individuals scattered through
the forests of deciduous trees. According to Mayr, this tree forms in the western part of
the island considerable forests on low swampy ground, not much raised above the level of the
ocean. This appears to be the common Spruce of Saghalin and of the Manchurian coast.

The fourth species, Picea Glenhi, discovered by F. Schmidt in Saghalin, has been found
in a few situations in southern Yezo. This tree, which is still to be introduced into our gar-
dens, we did not see growing. In many characters it resembles the Siberian Picea obovata,
and in the herbarium it is not easy to find characters by which it can be satisfactorily sepa-
rated from that species. Like the species of the White Spruce group of North America, which
appear to pass one into another by gradual transitions, the Spruces of northeastern Asia are
difficult to distinguish with the material found in herbaria, and it will only be possible to study
them satisfactorily when all the various forms have been planted side by side in some arbore-
tum and allowed to grow to maturity.

Of the Hemlocks found in Japan, one is northern and the other southern ; both are com-
mon at high elevations, and one at least forms extensive forests. The great forest, which
covers the Nikko Mountains at an altitude of more than 5,000 feet above the ocean, is
composed almost entirely of the northern Hemlock, Tsuga diversifolia, which is distin-
guished by its bright red bark, small leaves, and cones. This Hemlock forest, which is the
only forest in Hondo which seems to have been left practically undisturbed by man, is the
most beautiful which we saw in Japan. The trees grow to a great size, and while they stand
close together are less crowded than the trees in an American Hemlock forest, under which
no other plants can grow, and light enough reaches the forest-floor to permit the growth of
Ferns, Mosses, and many flowering under-shrubs which clothe the rocky slopes up which this
forest stretches. One of the most beautiful spots which we saw in Japan is the walk cut
through this forest which follows along the shores of Lake Yumoto, and I am fortunate in
being able to reproduce a photograph of it (Plate xxv.), for which I am indebted to Pro-
fessor Mayr, of Munich, who, during an official residence of several years in Japan, explored
the forests of all parts of the empire more thoroughly than any other foreigner. We found
Tsuga diversifolia in scattered groups on the rocky cliffs of Mount Hakkoda, in the extreme
north of Hondo, the most northern station which has been recorded for this tree, which is still
to be introduced into our gardens ; but, south of Nikko, it was replaced by the second species,
Tsuga Tsuga, which we saw in great beauty on Koma-ga-take, where, however, it does not

82 FOREST FLORA OF JAPAN.

form a continuous forest, but is scattered in groves of considerable extent among deciduous
trees and Pinus densiflora. It is this southern species, Tsuga Tsuga, which is cultivated in
our gardens, where it appears to be as hardy as our native species, which it surpasses in its
more graceful habit, and in its broader and darker colored leaves.

Of the Firs of Japan we saw only four species, and one of these, Abies firma, only as a
cultivated plant. This is the largest and the most beautiful of the Japanese Firs, often grow-
ing in cultivation to the height of one hundred and twenty feet, and producing clean tall
stems four or six feet in diameter. Writers on Japanese forests speak of Abies firma as com-
mon south of latitude 40°, north, in the upper belt of deciduous trees, but we never saw it
except in parks and temple gardens, or in the immediate neighborhood of houses. It is this
species which is chiefly called Momi by the Japanese, although the name is applied generally
to all Firs, and which, in Hondo, supplies the fir wood of commerce. This is soft, straight-
grained, and easily worked, and hardly distinguishable from the wood of the European Fir ; it
is used for building purposes and cheap packing-cases, but is not greatly valued. This species
has usually proved a disappointment as an ornamental tree in this country and in Europe, but
it is certainly, as it grows in Japan, one of the most beautiful of all Firs, distinguished by the
nobility of its port and by its bright green and very lustrous long rigid leaves, which are
sometimes sharply pointed, and sometimes divided at the apex. It probably needs a warmer
and moister climate than that of the northern United States in which to develop all its beau-
ties ; farther south it should, however, make a fine tree.

The Fir of which we saw the most in Japan is the Abies homolepis of Siebold and Zucca-
rini. This is the plant which is now often cultivated in our gardens under the name of Abies
brachyphylla, a more recent name. It is the common Fir of central Japan, and abounds on
the Nikko Mountains between 4,000 and 5,000 feet elevation above the sea, although it does
not form continuous forests, but is scattered singly, or in small groups, through the Birch
and Oak woods which cover the ground just below the Hemlock belt. It is a massive
although not a very tall tree, apparently never growing to a greater height than eighty or
ninety feet ; in old age it is easily distinguished from all other Firs by its broad round head,
the branches near the tops of the trees growing longer than those lower down on the stems.
This peculiarity is seen even on young plants in our gardens, on which the lower branches,
which soon stop growing, are shaded by the longer ones produced above them. The pale
bark, the long crowded leaves, dark green above and silvery white below, and the large purple
cones make this a handsome tree. In cultivation here it is very hardy, and grows with remark-
able rapidity. The inaccessibility of the places where Abies homolepis grows in Japan pre-
cludes the general use of the wood, which we found employed in the little alpine village of
Yumoto for building material.

The chief object of our visit to Mount Hakkoda, in northern Hondo, was to find Abies
Mariesii, which the botanical collector, whose name this tree bears, discovered there several
years before. It is common on this mountain at about 5,000 feet above the sea-level, scat-
tered among deciduous trees, and, so far as we observed, it is the only Fir of northern Hondo.
As we saw it, Abies Mariesii forms a compact pyramid about forty or fifty feet high, with
crowded branches covered with short dark foliage, pale below, and many large dark purple

THE LAECH FAMILY. 83

cones. It is a handsome but in no wise a striking or remarkable tree, which in all probabil-
ity will flourish in severe climates. It is only known on the high mountains of northern
Hondo and in one place on the shores of southern Yezo, where it was discovered during the
summer of 1892 by Mr. Tokubuchi.

On the hills of central Yezo, Abies Sachalinensis is not rare, and in the northern part of
the island and on Saghalin this fine tree is said to form extensive forests. It is a tall pyram-
idal tree with pale bark, long slender dark green leaves, and white buds which make it pos-
sible to distinguish it readily from the other Japanese Firs. A curious form has been noticed
by Professor Miyabe, growing near Sapporo, with red bark, dark red wood, and red cone
bracts ; it grows with the common form and is probably merely a seminal variety, although
Professor Miyabe considers it specifically distinct and proposes to call it Abies Akatodo. We
were fortunate in securing a supply of seeds of the white and of the red barked varieties, and
the seedlings, perhaps, will show whether they should be considered distinct. Abies Sachali-
nensis produces wood of fair quality, which is used in Sapporo in building and for packing-
cases ; it has only a local consumption. The young plants of Abies Sachalinensis in our
gardens are perfectly hardy and grow more rapidly than those of any other species of Fir-
tree.

Abies Veitchii, discovered many years ago on the slopes of Mount Fuji-san, we looked for
everywhere, and although it is said to grow among the Nikko Mountains, we saw nothing of
it there or elsewhere. From Abies homolepis, which this species most resembles, it may be
distinguished by its shorter and more crowded leaves, its more slender pubescent shoots and
smaller cones. This tree is an old inhabitant of our gardens, having been sent many years
ago by Mr. Thomas Hogg to the Flushing Nurseries, where it was cultivated under the unpub-
lished name of Abies Japonica long before it was known in Europe. Of Dr. Mayr's Abies
umbellata, a species probably too near Abies homolepis, we saw nothing.

The forests of Hondo contain at least one Larch, Larix leptolepis. It is a fine tree, sev-
enty or eighty feet tall, with pale green foliage and massive trunks covered with reddish
bark, and in habit not unlike the European species. The Japanese Larch is not rare at
elevations of from 5,000 to 6,000 feet in the central part of the island, although we saw it
nowhere growing in continuous forests, but always scattered in small groves, mixed with other
deciduous trees. Larix leptolepis was introduced into American gardens many years ago ; it
grows in this country with great rapidity, and the oldest trees here have for many years
produced abundant crops of seeds. The wood, like that of other Larch-trees, is hard, heavy,
and strong. The trees, however, are so difficult to reach that it is little used in Japan, except
for the timbers of mountain mines.

Maximowicz describes a variety of this species, var. Murrayana (Larix Japonica, Murray),
which grows as a low shrub near the timber-line of Fuji-san ; and my companion, Mr. Cod-
man, made a special trip late in the autumn for the purpose of securing specimens and seeds
of this plant. In this he was successful, but his specimens gathered from plants but a
few feet high, growing at an elevation of 8,500 feet above the sea-level, differ from those
of the common arborescent form only in the smaller size of the cones and in the shorter
leaves.

84 FOREST FLORA OF JAPAN.

A variety of Larix Dahurica, a species widely distributed through Siberia, northern China,
Manchuria, Kamtschatka, and Saghalin, reaches the extreme northern part of Yezo and the
Kurile Islands. This form has been called var. Japonica by Maximowicz, and by Mayr Larix
Kurilensis. We were not fortunate enough to see this tree in its native forests, but some idea
of its appearance as it grows in the island of Iturup can be obtained from our illustration
(Plate xxvi.), which is produced from a photograph for which I am indebted to Dr. Mayr, who
visited the Kurile Islands during his residence in Japan.

The other arborescent plants of Japan, Cycas revoluta, a favorite garden-plant, especially in
the south, where it often grows to a great size, and Trachycarpus (Chamserops) excelsa, natu-
ralized in some parts of the south, appear to have been introduced from the Loochoo or other
southern islands, or from Formosa or southern China.

PLATK XXVI.

LARIX DAHURICA, VAK. JAPONICA, MAXM.

THE ECONOMIC ASPECTS OF THE JAPANESE FORESTS.

IN few other countries are the forests of greater importance to the prosperity of the nation
than they are in Japan. The formation of the islands, with their high central mountain
ranges and short precipitous swift flowing rivers, make floods particularly prevalent and
dangerous, and the necessity of preserving the forest-covering of the upper mountain-slopes
proportionately great ; and no other race, with the single exception, perhaps, of the people of
the United States, is such a consumer of forest products as the Japanese, all their houses and
most of their articles of domestic use being entirely made of wood. The traveler, therefore,
watches with some interest, as bearing upon the future of Japan, the condition and prospects
of her forests.

According to the most reliable statistics available, those compiled by Rein, and based in part,
at least, upon the report of the Japanese Forestry Exhibit at Edinburgh in 1884, thirty-seven
per cent, of the three southern islands — that is, of the whole empire, with the exception of the
practically unsettled island of Yezo — is desert or unproductive land. Twenty-three per cent,
is occupied by the mountain forests, eighteen per cent, by the cultivated forests, while rather
less than twenty per cent, is devoted to agriculture, the remainder being taken up by build-
ings, roads, etc. The cultivated forests, in which it is presumed that the areas surrounding
the temples are included, which, although covered with splendid groves of trees, are unpro-
ductive except so far as they are made to furnish material for repairing or rebuilding the
temples, are well stocked with coniferous trees, — Retinosporas, Cryptomerias, and Pines, —
and furnish all the building material used in the empire. It is said that the Japanese have
been making these plantations for twelve hundred years ; and if this is true, they began plant-
ing trees for timber before any other people with whose agriculture we are acquainted.
Scientific methods might, perhaps, make these plantations, which are mostly the property of
individuals, rather more productive, but any great increase of forest supplies can only follow
the better management of the mountain forests or the replanting of desert lands. The moun-
tain forests, which are the property of the state, may be roughly divided into two belts, the
upper composed principally of Hemlocks, and extending from about 5,000 feet above the
sea-level to the timber-line, the lower stretching between 2,000 and 5,000 feet elevation, and
composed of Beeches, Oaks, Maples, Birches, Pines, and a few Firs and Spruces. The upper
belt, owing to its inaccessibility and the bad condition of the mountain roads, is practically
untouched, except where mining operations have created a local demand for timber. Scien-
tific management and good roads would make this upper coniferous forest yield quantities of
valuable material. The deciduous belt below it, which ought to be the most productive part
of the Japanese forest, is, wherever we entered it, in a deplorable condition, and although
Japan has supplied herself with a Forest Department, we saw no evidence that it is seriously
occupying itself with the care of this part of the public domain. There appear to be no rules

86 FOREST FLORA OF JAPAN.

about cutting the trees in this belt, which is invaded by bands of wood-choppers, who cut
without system any trees that appear large enough to answer their purpose ; and the only
mature trees it contains are those growing in inaccessible positions, or of sorts which are not
considered valuable. No attention is paid to reproducing the valuable species, with the result,
of course, that such species, being the most cut, have become the least common. Reproduction
is chiefly by coppice-growth which is cut at irregular intervals ; and the Japanese deciduous
forests display all the bad effects of an indiscriminate and long-continued system of jardinage.
The application of the block system would in time, of course, increase the output of these
forests and supply large quantities of valuable timber where only fuel or small sticks are now
produced.

Still better results might be expected from covering some parts of the desert land with
forests. These so-called deserts consist of sandy seashore planes and dunes, often capable of
producing a moderate growth of Pine ; the alpine summits of mountains, their lava-covered
slopes, bare mountain-ridges from which the forests have been artificially removed, and the
hara. This is the rolling foothill region about the base of the high mountains or below the
mountain forest-belt, and must form a very large part of the thirty-seven per cent, of desert ;
it is covered with a mat of coarse bunch-grass (Eulalia) and with many other perennial plants.
Here the Japanese cut the fodder for their animals and cure their hay for winter use. Every
spring the whole hara is burnt over to destroy the dried vegetation of the previous year and
start a new growth of grass. That fires have made these foothills treeless by the destruction
of all seedling trees as fast as they appear seems to be shown in the fact that where ravines
or other depressions occur among the hills, which the fire cannot easily reach, they are covered
with a vigorous growth of trees of many species. It is not easy to find in existing conditions
of Japanese life any cause for the original destruction of the foothill forests, but once
destroyed it is easy to see why they have not been able to grow again. Much of the hara
region is suited in soil and elevation to produce Retinosporas and Cryptonierias, the most val-
uable of the Japanese timber-trees, and its conversion from unproductive prairie — for not
one per cent, of the hara is used for hay — into forest would add enormously to the wood
product of the empire.

Japan is well situated geographically to supply a vast number of people living in foreign
countries with timber ; and its soil and climate are preeminently suited to produce forests. It
could easily send, if it had it to spare, coniferous timber to China, where the demand for
building material is practically inexhaustible, to the Straits Settlements and Australia ; and
oak-staves for wine casks to California, which is now supplied from the fast vanishing forests
of the Mississippi valley.

From the changed conditions which have followed the hasty and often ill-considered intro-
duction of European methods into Japan, grave economic questions are rising. The cessation
of civil wars which followed the abolition of the Shogunate and the deposition of the Daimyos,
and the introduction of western medical practices, have caused a great increase of population in
the last twenty years, and the question of food supply is becoming a vital one to Japan. The
limit to the production of rice, the one great staple, has been practically reached, and all
efforts to induce the superfluous population of the southern islands to colonize Yezo have

THE ECONOMIC ASPECTS OF THE JAPANESE FORESTS. 87

utterly failed, in spite of the great sums of money spent by the Imperial government during
the last twenty-five years to encourage its settlement. A few thousand coolies leave home
annually to work in other countries, but this movement is comparatively small, and many of
these emigrants return to their homes at the end of a few years. Starvation threatens Japan
unless it can import food from other countries, and this it will only be able to do by increas-
ing its exports. There is still room to increase the product of tea if the demand in this
country for low grades of Japanese tea justifies it, but the ground fit to grow Mulberry-trees
advantageously is practically all taken up, and the silk product cannot therefore be very mate-
rially increased. Curios, of course, can be made in unlimited quantities, but the demand for
them in the United States and Europe is more likely to decrease than to increase ; and wood is
really the only product upon which Japan can depend to greatly increase the volume of her
exports. The care of her existing forests and the planting of her waste lands would give
employment to thousands of coolies, and in time would add important sums to the national
exchequer.

The forests of Yezo are still intact, except where here and there a struggling settlement has
broken into the forest-blanket which covers this noble island. Here are great supplies of oak
and ash of the best quality, of cercidiphyllum, walnut, fir, acanthopanax, cherry and birch
— a storehouse of forest wealth, which, if properly managed, could be drawn upon for all
time, and which, if the timber is not needed in Japan, may become, when the trans-Asiatic
railroad is finished, an important factor in the development of southern Siberia and some of
the treeless countries of central Asia.

INDEX.

Names of admitted Genera and Species and other Proper Names in romau type ; of Synonyms in italics.

Abies Alatodo, 83.

Abies brachyphylla, 82.

Abies, distribution of, in Japan, 6.

Abies flrma, 82.

Abies firma in American gardens, 82.

Abies firma, wood of, 82.

Abies homolepis, 82.

Abies bomolepis, distribution of, 82.

Abies Japonica, 83.

Abies Mariesii, 10, 82.

Abies Saclialinensis, 83.

Abies Saclialinensis, red wooded vari-
ety of, 83.

Abies umbellata, 83.

Abies Veitchii, 83.

Acantbopanax, 44.

Acantbopanax aculeatum, 44.

Acauthopanax innovans, 44.

Acauthopaiiax ricinifolium, 44, 45.

Acanthopauax ricinifolium, wood of,
46.

Acanthopanax sciadophylloides, 44.

Acautbopanax sessiliflorum, 44.

Acanthopanax tricbodon, 44.

Acer argutuui, 31.

Acer capillipes, 30.

Acer carpinifoliuui, 30.

Acer cissifolium in American gardens,
31.

Acer cratsegifolium, 31.

Acer diabolicum, 31.

Acer distylum, 31.

Acer Japonicura, 30.

Acer Japonicum, foliage of, in autumn,
30.

Acer Ufaximowiczianum, 32.

Acer Miyabei, 29.

Acer Nikoense, 31.

Acer Nikoense in European gardens,
32.

Acer palmatum, 30.

Acer paluiatum, varieties of, in Jap-
anese gardens, 30.

Acer parvifoliuin, 31.

Acer pictum, 28.

Acer purpurascens, 31.

Acer pycnanthum, 31.

Acer rufinerve, 31.

Acer Sieboldianum, 31.

Acer spicatum, 30.

Acer Tataricum, var. Ginuala, 30.

Acer Tataricum, var. Ginnala, in
American gardens, 30.

Acer trifidum, 1.

Acer Tschonoskii, 31.

Actiuidia arguta, 7, 19.

Actinidia arguta, fruit of, 19.

Actinidia in Japan, 12.

Actinidia Kolomikta, 19.

Actinidia polygama, 19.

Actinidia polygama, peculiarity of tbe
leaves of, 19.

vEsculus in North America, 28.

^Eseulus turbinata, 28.

Albizzia Julibrissin, 34.

Alder, the black, 25.

Alders, arborescent, abundance of, in
Japan, 7.

Alders in Japan, 63.

Aleurites cordata, 56.

Alnus firma, 63

Alnus jirma, 63.

Alnus firma, var. multinervis, 63.

Alnus incana, 63.

Alnus incana, var. glauca, 63.

Alnus incana, var. hirsuta, 63.

Alnus Japonica, 63.

Aluus maritima, 63.

Alnus viridis, var. Sibirica, 63.

American forests, composition of, com-
pared with Japanese, 1.

Andromeda Japonica, 49.

Andromeda nana, 10.

Aphananthe aspera, 59.

Aralia Chinensis, 44.

Aralia cordata, 44.

Aralia family in Japan, 43.

Aralia family in eastern America, 43.

Aralia family in North America, 43.

Aralia Maximoieiczii, 45.

Aralia quinquefolia, 43.

Aralia racumosa, 44.

Aralia spinosa, var. elata, 44.

Arboixjcent genera, common to the
forests of eastern North America
and Japan, 6.

Arbor vitre, the Japanese, 72.

Aria in Japan, 39.

Aronia alnifolia, 40.

Ash, the Mountain, 78.

Aspen, the, 71.

Azalea calendulacea, 49.

Azalea mollin, 49.

Azaleas in Japan, 49.

Balsam Poplar, the, in Japan, 71.

Bamboos, prevalence of, in Japanese

forests, 7.

Beech, the, in Japan, 70.
Beech-wood in Japan, 70.
Benthamia, 47.
Berberis Japonica, 2.
Bercbemia racemosa, 26.
Betula alba, 61.
Betula alba, var. Tauschii, 62.
Betula alba, var. verrucosa, 62.
Betula Bhojpattra, 63.
Betula corylifolia, 63.
Betula Ermani, 7, 62.
Betula Maximowicziana, 62.
Betula serra, 63.
Betula ulmifolia, 63.
Birches, abundance of, in Japan, 7.
Birch-trees, black-barked, 62.
Birch-trees in Japan, 61.
Black Alder, 25.
Black Pine, 79.
Bronssonetias in Japan, 59.
Bunch-berry, the, 47.
Biirgeria ? salicifolia, 11.

Camellia Japonica, 17.

Camellia Japonica, oil of, 17.

Camellia Japouica, wood of, 17.

Camellia Sasanqua, 17.

Camellia theifera, 17.

Camphor-tree, the, 54.

Camphor-tree at Atami, 54.

Carpinus Carpimis, 64.

Carpinns Carpiuus in American gar-
dens, 65.

Carpinus cordata, 65.

Carpinus erosa, 64.

Carpinus Japonica, 64.

Carpinus laxiflora, 64.

Carpinus Tschonoskii, 64.

Carpinus Yedoensis, 64.

Carpinus, distribution of, 64.

Catalpa ovata, 2.

Cedrela Sineusis, 1.

Celtis Sinensis, 59.

Cepbalotaxus drnpacea, 75.

Cercidiphyllum Japonicum, 13.

Cercidiphyllum from western China,
4.

Cercidiphyllum, its introduction into
American gardens, 15.

Cercidiphyllum, wood of, 15.

Cercis Chinensis, 2.

90

INDEX.

Celastrus articulatus, 26.

Celastrus articulatus, decorative uses
of, in Japan, 26.

Celastrus flagellaris, 26.

Chaniaecyparis obtusa, 73.

Chamsecyparis obtusa, plantations of,
in Japan, 73.

Chamsecyparis obtusa, uses of the wood
and bark of, in Japan, 73.

Chamsecyparis obtusa in the eastern
United States, 73.

Chamaecyparis pisifera, 73.

Chamsecyparis pisifera in the eastern
United States, 73.

Chamsecyparis pisifera, wood of, 73.

Chainserops excelsa, 3, 84.

Cherry, the Japanese, 36.

Cherry-tree cultivated in Japan, 36.

Cherry-tree, Pendulous-branched, 37.

Cherry-tree, varieties cultivated in Ja-
pan, 36.

Chestnut-tree, the, in Japan, 69.

Chestnuts as food in Japan, 70.

Chinese rice-paper, 43.

Chinese shrubs cultivated in Japan, 2.

Chinese trees cultivated in Japan, 1.

Chionanthus, 52.

Chionanthus retusa, 2.

Cinnamonium Camphora, 54.

Cinnamoraum pedunculatum, 54.

Citrus Japonica, 2.

Clematis patens, 2.

Clerodeudron trichotomum, 53.

Clclhra barbinervis, 50.

Clethra canescens, 50.

Cleyera ochnacea, 17.

Climbing plants in Japanese forests, 7.

Climbing Hydrangeas in Japan, 7.

Composition of Japanese and eastern
American forests compared, 1.

Conifers in Japan, 72.

Conifers planted in Japan, 72.

Corean shrubs cultivated in Japan, 2.

Corean trees cultivated in Japan, 1.

Cornus alba, 47.

Cornus brachypoda, 48.

Cornus Canadensis, 7, 47.

Cornus Kousa, 47.

Cornus Kousa in American gardens, 47.

Cornus macrophylla, 47, 48.

Cornus officinalis, 1, 48.

Cornus officinalis cultivated in Japan,
48.

Cornus Suecica, 47.

Cornus in North America, 47.

Cratcegus alnifolia, 40.

CratsBgus chlorosaca, 41.

Crataegus cuneata, 1.

Crataegus in eastern North America, 41.

Cryptomeria Japonica, 74.

Cryptonieria Japonica, plantations of,
in Japan, 74.

Cryptonieria Japonica, groves and ave-
nues of, at Nikko, 74.

Cryptomerias at Nikko, age of, 75.

Cryptomeria Japonica, uses of the wood
and bark of, in Japan, 74.

Cupuliferse in Japan, 61.
Cycas revoluta, 3, 84.

Daphne Genkwa, 2.
Daphniphyllum glaucesens, 56.
Daphniphylluiu humile, 10, 56.
Daphniphyllum macropodum, 56.
Dendropanax Japouicum, 43.
Desert lands, areas of, in Japan, 85.
Desert lands, nature of, in Japan, 86.
Diervilla Japonica, 48.
Dimorphanthus Manchuricus, 44.
Diospyros Kaki, 2, 50.
Diospyros Lotus, 2, 51.
Disanthus cercidifolia, 42.
Disligocarpus, 64.
Distigocarpux Carpinus, 64.
Distylium racemosuin, 42.

Economic aspects of Japanese forests,
85.

Edgeworthia papyrifera, 2.

Ehretia acumiiiata, 53.

Eheagnus lougipes, 56.

Elseagnns pungens, 56.

Ekeagnns umbellata, 56.

Elseagnus umbellata in American gar-
dens, 56.

Elaeocarpus in Japan, 21.

Elaeocarpus photinisefolia, 21.

Eleutherococcus in Japan, 43.

Elliottia paniculata, 7.

Elm-bark cloth, made in Yezo, 57.

Elm-family in Japan, 57.

English walnuts, 60.

Enkianthus campanulatns, 49.

Enkianthus Japonicus, 2.

Epigaea Asiatica, 10.

Ericacese in Japan, 49.

Eriobotrya Japonica, 1.

Euphorbiaceae in Japan, 56.

Euptelea polyandra, 15.

Eurya Japonica, 17.

Evergreen shrubs, prevalence of, in Ja-
pan, 6.

Evodia rutiecarpa, 21.

Evodia rutieearpa, economic uses of, 21.

Evonymus alatns, 26.

Evonymus alatus in American and Jap-
anese gardens, 26.

Evouymus alatus, var. subtriflora, 26.

Evonymus Europaeus, var. Hamiltonia-
nus, 26.

Evonymus Europaeus, var. Hamiltoni-
anus in American gardens, 26.

Evonymus in Japan, 25.

Evonymus Japonicus, 25.

Evouymus macropterus, 26.

Evonymus Nipponicus, 26.

Evonymus oxyphyllus, 26.

Evonymus radicams, 25.

Evonymus, the climbing, 7.

Exccecaria Japonica, 57.

Fagus Sieboldii, 70.
Fagus sylvatica, 70.
Fagus sylvatica, var. Sieboldii, 70.

Fatsia horrida, 43.

Fatsia Japonica, 43.

Fatsia papyrifera, 43.

Features of vegetation in Japan and
eastern North America compared, 4.

Fig-trees in Japan, 59.

Firs in Japan, 82.

Forestiera, 52.

Forest-region, the Atlantic, extent of, 2.

Forest-region, the Japanese-Manchuri-
aii, 2.

Forests, Japanese, character and divi-
sions of, 85.

Forests, Japanese, composition of, com-
pared with eastern American, 1.

Forests, Japanese, economic aspects of,
85.

Forests, Japanese, methods of reproduc-
tion, 86.

Forests of Yezo, 87.

Forest, the Atlantic, census of, 2.

Forest, the Hemlock, of Lake Yumoto, 7.

Forest, the Japanese-Mauchurian, cen-
sus of, 2.

Forest undergrowth, character of, in
Japan, 7.

Forest undergrowth, character of, in
North America, 7.

Forsythia suspensa, 2.

Fraxinus in eastern America, 52.

Fraxinus longicuspis, 52.

Fraxinus Manchurica, 52.

Fraxinus Manchurica in American gar-
dens, 52.

Fraxinus pubinervis, 52.

Gaultheria pyroloides, 10.

Genera, arborescent, common to the
forests of eastern North America and
Japan, 6.

Genera, endemic in eastern North
America, list of, 4.

Genera, endemic in Japan, list of, 4.

Genera, number of arborescent, in east-
ern North America, 4.

Genera, number of arborescent, in Ja-
pan, 4.

Geum dryadoides, 10.

Giukgo biloba, 2.

Gingko, fruit of, 76.

Gingko, the, 75.

Gingko, the, in temple grounds of To-
kyo, 76.

Ginseng, cultivation of, in Japan and
Corea, 43.

Ginseng, the, 43.

Gleditsia Japonica, 35.

Gleditsia Japonica, use of the fruit in
Japan, 35.

Grape, the wild, in the forests of
Yezo, 7.

Hakkoda, Mount, vegetation of, 10.
Hamamelis arborescens, 42.
Hamamelis Japonica, 42.
Hamamelis, the Japanese, 42.
Kara, the, 86.

INDEX.

91

Helwingia in Japan, 43.

Hemlock forest of Hondo, 81.

Hemlock forest of Lake Yuruoto, plants
in,' 7.

Hemlock, forests of, in Japan, 6.

Hemlocks, distribution of in Japan, 81.

Hi-no-ki, the, 73.

Hollies, the, in Japan and North Amer-
ica, 23.

Honeysuckle family, the, in Japan, 48.

Hop Hornbeam, distribution of, 65.

Horse-chestnut, the Japanese, 28.

Horse-chestnut, use of the fruit in Ja-
pan, 28.

Hovenia dulcis. 20.

Hovenia dulcis, cultivated as a fruit-
tree in Japan, 26.

Hovenia dulcis, cultivated in America
and Europe, 27.

Hydrangea pauiculata, 41.

Hydrangea petiolaris, 7.

Hydrangeas, climbing, in Japan, 7.

Ilex crenata, 10, 24.

Ilex crenata in Japanese gardens, 24.

Ilex geniculata, 25.

Ilex integra, 23.

Ilex integra, var. leucoclada, 10, 23.

Ilex latifolia, 23.

Ilex macropoda, 25.

Ilex Monticola, 25.

Ilex pedunculosa, 24.

Ilex pedunculosa in Japanese gardens,
24.

Ilex rotunda, 20.

Ilex serrata, 25.

Ilex Sieboldii, 25.

Ilex Sieboldii in American gardens,
25.

Ilex Sieboldii, used for decoration in
Japan, 25.

Ilex Sugeroki, 10, 24.

Ilex verticillata, 25.

Illicium anisatum, 12.

Illicium religiosum, 12.

Illicium, the Japanese, in religious fes-
tivals, 12.

Illicium verum, 12.

Ivy, the, in Japan, 7, 43.

Ivy, the, Poison, 34.

Japan as a producer of timber, 85.

Japanese Arbor-vitse, 72.

Japanese Cherry, 36.

Japanese and North American forests,
composition of, compared, 1.

Japanese Lacquer-tree, 33.

Japanese Plums cultivated in the
United States, 38.

Juglans cordiformis, 60.

Juglans regia, cultivated in Japan, 60.

Juglans Sieboldiana, 60.

Juglans Sieboldianas in American gar-
dens, 60.

Junipers in Japan, 78.

Juniperus Chinensis, 78.

Juniperus Chinensis at Nagano, 78.

Juniperus Chinensis, prostrate form of,

78.

Juniperus conferta, 78.
Juniperus I'Moralis, 78.
Juniperus rigida, 78.

Kadsura Japouica, 12.

Kaki, the, 50.

Kaki, varieties of, 50.

Kaya-no-abura, 76.

Kaya, the, 76.

Keaki, the, 57.

Kobus, 9.

Kcelreuteria paniculata, 1.

Koya-maki, 77.

Lacquer-tree, Japanese, 33.

Larix Dahurica, distribution of, 84.

Larix Dahurica, var. Japonica, 84.

Larix Japonica, 83.

Larix Kurilensis, 84.

Larix leptolepsis, 83.

Larix leptolepsis in American gardens,

83.

Larix leptolepsis, var. Murrayana, 83.
Larix leptolepsis, wood of, 83.
Lauracese in Japan, 54.
Leguminosa? in Japan, 34.
Leitneria, 4.
Ligustruiu Ibota, 53.
Ligustrum in Japan, 52.
Ligustrum Japonicum, 53.
Ligustrum medium, 53.
Lindera Benzoin, 54.
Lindera glauca, 56.
Lindera in eastern Asia, 54.
Lindera, melisssefolia, 54.
Lindera obtiisiloba, 55.
Lindera pracox, 55.
Lindera sericea, 55.
Lindera strychnifolia, 2.
Lindera triloba, 55.
Lindera umbellata, 55.
Liquidambar Formosana, 1.
Liquidambar Maximowiczii, 1.
Litsea glauca, 14.

Maachia Amurensis, 34.

Maachia Amureusis, wood of, 34.

Machilus Thuiibergii, 54.

Magnolia compressa, 11.

Magnolia conspicua, 1.

Magnolia family, the, 8.

Magnolia family, the, in Japan and
Manchuria, 8.

Magnolia family, the, in North Amer-
ica, 8.

Magnolia fuscata, 11.

Magnolia glauca, var. a, 9.

Magnolia hypoleuca, 8.

Magnolia hypoleuca as a timber-tree, 9.

Magnolia hypolenca, introduction of,
into American gardens, 9.

Magnolia Kobus, 9.

Magnolia Kobus, introduction of, into
American gardens, 10.

Magnolia obovata, 2.

Magnolia parvifolia, 1.
Magnolia salicifolia, 10.
Magnolia stellata, 2.
Magnolia Thurberi, 10.
Magnolia lomentosa, 9.
Magnolia Watsoni, 1.
Maples in Japan, 28.
Maples, Japanese, in American gar-
dens, 30.

Marrons of Kobe, 70.
Meliosma, 33.
Meliosma myriantha, 33.
Michelia compressa, 11.
Michelia, distinctive characters of, 11.
Michelia fuscata, 11.
Mohrodendron Carolinum, 51.
Momi, the, 82.
Monkey-slide, the, 18.
Morns alba, 59.
Mountain Ash, 38.
Mulberry, the White, 59.
Myrica Gale, 61.
Myrica rubra, 61.

Nandina domestica, 2.

Nara, grove of Podocarpus Nageia,

at, 77.

Negundo in Japan, 31.
Nikko, groves and avenues of Crypto-

1 1 M TIII Japouica at, 74.

Oaks, evergreen, in Japan, 67.

Olea fragrans, 2.

Olive family, the, in Japan, 52.

Orixa Japonica, 21.

Osmanthus Aqm'folium, 52.

Osmanthus Aquifolium in Japanese

gardens, 52.

Osmanthus ilicifolium, 52.
Ostrya Japonica, 66.
Ostrya, the distribution of, 66.
Ostrya Virginica, var. Japonica, 66.

Pauax horrida, 7
Panax repens, 43.
Panax ricinifolia, 45.
Paper Mulberries in Japan 59.
Paulownia imperialis, 2.
Pear-tree of Japan, 40.
Pendulous branched Cherry-tree, 37
Persimmon, the, in Japan, 20.
Phellodendron Amurense, 21.
Phyllodoce taxifolia, 10.
Picea Ajanensis, 81.
Picea Ajanensis, distribution of, 81.
Picea Alcockiana, 80.
Picea, bicolor, 80, 81.
Picea, distribution of in Japau, 6.
Picea Glenhi, 81.
Picea polita, 80.
Picrasma quassioides, 22.
Pines, Japanese, in American gardens, 79.
Pine, the Black, 79.
Pine, -the Red, 79.
Pine, the Umbrella, 77.
Pine-wood, nature and uses of, iu Ja-
pan, 79.

92

INDEX.

Pine-woods, planted in Japan, 70.

Finns desittora, 79.

Pinus, distribution of, in Japan, 72.

Pinus Koraieusis, 2.

Piiius parviflora, 80.

Finns parviflora in American gardens,

80.

Finns pentaphylla, 80.
Finns pnmila, 80.

Pinus puiuila on Mount Hakkoda, 10.
Pinus, scarcity of species in Japan, 6.
Finns Thunbergii, 79.
Plantations connected with Buddhist

temples, 5.

Plants, climbing, in Japanese forests, 7.
Platycarya strobilacea, 41.
Plums, Japanese, cultivated in the

United States, 38.
Plum-trees in eastern North America,

38.

Plum-trees in Japan, 38.
Podocarpus, distribution of, 77.
Podocarpus macrophylla, 2, 77.
Podocarpus Nageia, 2, 77.
Podocarpus Nageia, variegated form

in Japanese gardens, 77.
Poison Ivy, the, 34.
Poplar, the Balsam, in Japan, 71.
Populns in Japan, 71.
Populw Sieboldii, 61.
Popiilus stiaveolens, 71.
Populus treinula, var. villosa, 71.
Primus Grayana, 38.
Prunus Japouica, 2.
Primus Maximowiczii, 37.
Primus Mume, 1, 36.
Prunus Mume in Corea, 36.
Prunus Padus, 38.
Prunus Pseudo-Cerasus, 36.
Prunus Pseudo-Cerasus, wood of, 37.
Prunus Ssiori, 38.
Primus Ssiori, uses of the wood of,

38.

Prunus tomentosa, 2.
Prunus triflora, 38.
Fterocarya, distribution of, 61.
Pterocavya rhoifolia, 61.
Pterocarya rhoifolia, wood of, 61.
Pterostyrax corymbosum, 51.
Pterostyrax hispidum, 51.
Pyrus alnifolia, 57.
Pyrus aucuparia, 38.
Pyrus gracilis, 39.
Pyrus Halleana, 40.
Pyrus lanata, 39.
Pyrus Alalus Jloribunda, 40.
Pyrus microcar/ia, 40.
Pyrus Miyabei, 40.
Pyrus Parkmani, 40.
Pyrus Rintjo, 40.
Pyrus sambucifolia, 38.
Pyrus Sieboldii, 40.
Pyrus Sinensis, 1, 40.
Pyrus Toringo, 40.
Pyrus Toringo in American gardens,

40.
Pyrus Tschonoskii, 40.

Quercus acuta, 69.

Qiiercus crispnla, 67.

Quercus cuspidata, 69.

Quercus Daimio, 67.

Quercus dentata, 67.

Quercus dentata, var. pinnatifida,
67.

Quercus gilva, 69.

Quercus glabra, 69.

Quercus glandulifcra, 68.

Quercus glandulifera in American gar-
dens, 68.

Quercus glauca, 69.

Quercus grosseserrata, 67.

Quercus in Japan, 67.

Quercus lacera, 69.

Quercus serrata, 68.

Quercus Thalassica, 69.

Quercus variabilis, 68.

Quercus Vibrayiana, 69.

Red Pine, the, 79.

Retinospora obtusa, 73.

Retinosporas, monstrous forms of, in

Japan, 74.
Rhamuaceje, 26.
Rhododendron arborescens, 49.
Rhododendron Catawbiense, 7, 49.
Rhododendrons, evergreen in Japan.

49.

Rhododendron maximum, 49.
Rhododendron Metteruichii, 7.
Rhododendron nudicaule, 49.
Rhododendron Sinense, 49.
Rhododendron viscosum, 49.
Rhodotj'pos kerreoides, 2.
Rhus family, the, in eastern North

America, 33.

Rims family, in Japan, 33.
Rhus semialata, 33.
Rhus succedanea, 34.
Rhns sylvestris, 34.
Rhus trichocarpa, 24.
Rhus vernicifera, 1, 33.
Rice-paper, Chinese, 43.
Rose family, the, in Japan, 36.
Rue family, the, 21.

Sabiacese, 33.
Salix Babylonicu, 70.
Salix eriocarpa, 71.
Salix subfragilis, 71.
Sapindacese, 28.
Sapindus Mukorosi, 1.
Saru-suberi, 18.
Sasan-kuwa, 17.
Sawara, the, 73.
Saxifrage family in Japan, 41.
Schizandra, 8.
Schizamlra Chinensis. 13.
Schizandra nigra, 13.
Schizophragma, 7.
Sciadopytis cultivated in Japan, 77.
Sciadopytis, the, 77.
Sciadopytis, the wood of, 77.
Shrubs, Chinese, cultivated in Ja-
pan, 2.

Shrubs, Corean, cultivated in Japan, 2.

Shrubs, evergreen, in northern Japan,
23.

Shrubs, evergreen, prevalence of, in
Japan, 6.

Shrubs, number of, in Japan. 3.

Simarubie, 22.

Skimmia Japonica, 21.

Sophora Japouica, 1.

Sorbus alnifolia, 40.

Sorbus Aria, var. Kamaonensis, 39.

Sorbus lanata, 39.

Spice-bush, the. 54.

Spikenard, the North American, 44.

Spirsea Thunbergii, 2.

Spruces in Japan, 80.

Stachyurus prsecox, 18.

Stachyurus prsecox in American gar-
dens, 18.

Sterculia platinifolia, 1.

Stuartia in America, 18.

Stuartia monadelpha, 18.

Stuartia pentagyna, 18.

Stuartia Fseudo-Cainellia, 18.

Stuartia Pseudo-Camellia in American
gardens, 18.

Stuartia serrata, 18.

Stuartia Virginica, 18.

Styracere in Japan, 51.

Styrax Japouica, 52.

Styrax Obassia, 51.

Sugi, the, 74.

Syringa Japonica, 52.

Syringa Japonica in New England gar-
dens, 52.

I
Taxus euspidata, 76.

Taxus cuspidata in American gardens.
76.

Taxns cuspidata, wood of, 76.

Taxus, distribution of, 76.

Tea-plant, the, 17.

Teconia grandiilora, 2.

Temples, Buddhist, plantations con-
nected with, 5.

Ternstro3iniaceie, 17.

Ternstreemia Japonica, 17.

Thuja Japonica, 72.

Thuja orientalis, 2.

Tbayopsu dolobrata, 72.

Thuyopsis dolobrata, varieties of, 72.

Thuyopsis Standishii, 72.

Tilia cordata, var. Japonica, 20.

Tilia heterophylla, 20.

Tilia Mandshurica, 20.

Tilia Miqueliana, 19.

Tilia ulmifolia, var. Japonica, 20.

Tila ulmifolia, var. Japonica, in Amer-
can gardens, 20.

Torreya, 76.

Toxyiou, 4.

Trachycarpus excelsa, 3, 84.

Treeless foothill region, 86.

Trees, aggregation of species, in south-
ern Indiana, 3.

Trees, aggregation of species in Ja-
pan, 3.

INDtiX.

93

Trees, American, list of, absent from

the forests of Japan, 5.
Trees, Chinese, cultivated in Japan, 1.
Trees, Corean, cultivated in Japan, 1.
Trees, Japanese, list of, absent from

the forests of eastern North Amer-
ica, 6.
Trees, list of, on hills near Sapporo,

3.

Trochodendracese, 13.
Trochodendron aralioides, 15.
Tsubaki, 17.

Tsuga diversifolia, 7, 81.
Tsuga diversifolia, northern station of,

10.

Tsuga Tsuga, 81.
Tsuga Tsuga in American gardens,

82.

Tumion, distribution of, 76.
Tumion nuciferum, 76.
Tumion nuciferum, fruit of, 76.
Tumion nuciferum, wood of, 76.
Types of vegetation common to the

floras of eastern North America and

Japan, 5.

Ulmus campestris, 57.

Ulmus parvifolia, 2.

Ulmus scabra, laciniata, 17.

Umbrella Pine, the, 77.

Undergrowth of the forests of eastern
North America and Japan contrast-
ed, 7.

Vacciniiim ciliatum, 49.

Vaccinium in Japan, 49.

Vacciniiim Japonicum, 49.

Vegetation, features of, in eastern
North America and Japan com-
pared, 4.

Vegetation of Mount Hakkoda, 10.

Vegetation, types of, common to the
floras of eastern North America and
Japan, 5.

Viburnum furcatum, 48.

Viburnum Wrightii, 49.

Vitis Coignetiffi, 46.

Walnuts as food in Japan, 60.

Walnuts, English, 60.

Walnut family, the, in Japan, 60.

White Birch, the, 61.

Wikstrcemia Japonica, 2.

Willows, arborescent, abundance of, in
Japan, 7.

Willows in Japan, 70.

Willow, the Weeping, planted in To-
kyo, 70.

Wistaria in Japan, 7.,

Witch Hazel family in Japan, 41.

Xanthoxylum ailanthoides, 21.
Xanthoxylum in Japan, 21.
Xanthoxylum piperitum, 21.
Xanthoxylum piperitum, uses of the
fruit in Japan, 21.

Yezo, forests of, 87.

Zelkova, distribution of, 58.
Zelkova Keaki, 57.
Zelkova Keaki in Rhode Island, 58.
Zelkova Keaki, wood of, 38.
Zizyphns vulgaris, 1.
Zizyphus vnlgaris, cultivated as a fruit-
tree in Japan, 2,

14 DAY USE

RETURN TO DESK FROM WHICH BORROWED

LA.NDSGAP: • -j. IE LIBRARY

This book is due on the last date stamped below, or

on the date to which renewed.
Renewed books are subject to immediate recall.

DEC 1 9 1958

AUG I 7 1959

LD 21-100m-6,'56
(B9311slO)476

General Library

University of California

Berkeley

I