pets ith we Sear . Peat fesse sastebort f ; { i seatasszatets He 3 é ; $5352534; Hitt ts " : aie Fepepitath muidhes +i strays . fot) ! Nights! * ROCK GARDENS | ‘ t ies { , i Tie Re (Frontispiece. ROCK GARDEN AND STREAM, MOUNT USHER. (From a Photo. by Mrs Delves Broughton.) ROCK GARDENS HOW TO MAKE AND MAINTAIN THEM BY LEWIS B. MEREDITH WITH AN INTRODUCTION BY F. W. MOORE M.A., A.L.S. LONDON WILLIAMS & NORGATE 14 HENRIETTA STREET, COVENT GARDEN IQIO Saale pn ere : : “Cs = oe, ¢ _ I DEDICATE THIS, MY FIRST BOOK, MY MOTHER, wa 7) MY MOST INDULGENT CRITIC wy 1 by 1 ean Dake I i ’ , i ty Regu i * ; ’ ‘ sity » f 4 * ih ae j i > i ry), | wv M4 be Ow it ) y Wt i u . $ a G Dike Hi hb } ait 4% /% y {hy } € } * | Y { hs ‘ iB , nt coy hey ‘\ \ I ') J i) ‘ uv, ot ay Pipi : 1 Y ‘ . 1 4 é rats ; { ‘ ; , \ , \ i » PVE BRE VY iter ii oe oy ae [ode Saieg al! “ % 1 a Wake om " 1 Een ui I hy KS Ay ‘ \ é nq ’ "is aan y i i ; sd / if a5 / y / we \ : A \ ; + wh A 74, om | 7 m1 PREFACE A worp of introduction seems needed to explain the purpose of this book. No branch of horticulture at the present time occupies so much general attention as rock-gardening. Yet not a single book deals with the subject of the cost, which I have attempted to estimate, basing it on my own practical experience. It is superfluous to observe that I have not aspired to any literary graces, but merely to tell simply what I believe to be needed. My aim has been to provide a practical volume which will enable anyone to make and plant a rock garden. When making my own, I know what I had to find out for myself, and in the following pages my endeavour has been to save others the trouble. If I may occasion- ally seem to repeat myself, it is because I wish vii Vill PREFACE to emphasise more clearly the salient points of my subject. A rock garden is one of the most fascinating of possessions. It is an in- exhaustible mine of pleasure; it entails no heavy labour, and is within the most modest means. If others derive half the enjoyment from their rock gardens that I have from mine, they will be abundantly repaid for any trouble they may take. Should I be of use to any unknown reader, I trust he will not hesitate to write to me, as I shall be delighted to reply to any questions he may care to put. My warmest thanks are due to Mr Frederick W. Moore, Curator of the Royal Botanic Gardens, Glasnevin, who not only assisted me throughout by his kind advice, but also for ushering in my book. A preface from an acknowledged authority of his renown dig- nifies my own modest effort and gives it an importance to which, it could not otherwise aspire. JI am also deeply indebted to Mr W. Irving, of the Royal Botanical Gardens at Kew, for so kindly checking the lists of plants at the end of the book. Passed by him, they can be safely commended both to the novice PREFACE Lae and to the more experienced owners of rock gardens. Grateful acknowledgments must also be made to Mr F. W. Moore and Mrs Delves- Broughton, who have so graciously given me photographs, and to the Editor of the Ladies Field for kind permission to reproduce some of them. LEWIS B. MEREDITH. GRAIGUECONNA, Bray, Co. WickLow. INTRODUCTION _ LiTeraTore treating of gardens and of garden- ing has been offered in plenty to garden lovers in recent years. It may fairly be said that much of it had better have been left unwritten, as it really has served no useful purpose, and has only tended to confuse rather than to assist novices, or even those with some experience of plants and their requirements. Such stric- tures cannot apply to this volume. It is a practical work, written by a practical man about a subject which he thoroughly under- stands, and by one who has experienced all the joys and sorrows connected with the cultivation of rock plants and alpines. ‘The rock garden is no longer a feature to be found only in large establishments or in Botanical Gardens. The pleasure which is derived from the successful xl X11 INTRODUCTION cultivation of miniature gems and of dwarf alpine plants is now fully recognised, and rock gardens have become a popular feature in gardening of to-day. Ina properly constructed rock garden many little plants which refuse to grow and live under the ordinary conditions of an herbaceous border, and which resent the ageressive attentions of their more robust neighbours, generally demonstrating their re- sentment by dying, can be cultivated and kept in health and vigour for many years, protected by stones, sheltered from drying winds, and from shade or sunshine, dryness or moisture, accord- ing to their requirements. They can, owing to their raised position, be seen and tended in a manner more conducive to the comfort of the observer and cultivator than would be possible if they were grown on the fiat. To ensure even moderate success in rock gardening two main points are essential: a properly constructed rock garden, and a re- liable guide to the nature and requirements of plants to be grown on it. ‘There existed a demand for a sound practical work, giving explicit and detailed information on these INTRODUCTION xiii points, and this volume meets it. Mr Meredith writes with practical experience. His own rock garden, constructed by himself, is artistic in conception, covers a considerable area, and suits the requirements of a large and varied collection of all classes of alpines. In it many difficult subjects, such as Edrianthus Pumilio, Morisia, Androsace Sarmentose, Carnea, Villosa, Saxifraga retusa, aretiides, cesta, Diapensoides, Daphne cneorum and Blagayana, flourish and grow into good-sized tufts. The details of how this success has been achieved are fully explained, concisely but clearly, and all necessary information is given. F. W. MOORE, M.A., A.LS. Royat Botanic GARDENS, GiasnEvin, DvusBLin. Wert iv CONTENTS RIOR oF CHAPTER I THE SITE Surroundings—What to avoid—Objections to trees —Aspect—Contour—Nature of soil—Chief points to be noted—T ypical site CHAPTER II THE TYPES OF ROCK GARDEN The natural—The artificial—Types of the natural —The old quarry—The amphitheatre—The horseshoe—The valley—Thhe cliff—The rocky bank—The rocky knoll—Points to be noted— Types of the artificial— ‘The sunken — The gravel-pit—Bank and knoll types—The rocky bed—The very large rock garden—Advantages and disadvantages of the natural—Advantages and disadvantages of the artificial CHAPTER III THE NATURAL ROCK GARDEN First preparations—Cleaning ground—Exposing and preparing the rock—Pockets natural and artificial CHAPTER IV THE ARTIFICIAL ROCK GARDEN What can be done with a level field—Laying out the garden—Boldness of design—Preparing banks xV PAGE rs 32 XV1 CONTENTS for rockwork—Treatment of natural valley— Formation of banks—Example of a “valley” garden—Points to be noted in excavations— The hollow garden— Gradient of slopes — Drainage—Aspect—Bank and knoll types and their treatment—Rock bed—Very large rock garden—Difficulties connected with this type of garden—Points to be noted in artificial rock gardens CHAPTER V THE ROCKWORK Kind of rock to use and what to avoid—Principles of particular construction—Types of rockwork to avoid— The placing of individual rocks—Fissures and how to make them—Strata—Principles of general construction — Paths — Steps — The moraine CHAPTER VI THE SOIL Importance of climate—Best type of soil—Supplying deficiencies in soil— Necessity of loosening ground— Drainage ‘ CHAPTER VII THE BOG GARDEN Position—Natural bog aN bog ane —The stream E ‘ CHAPTER VIII THE WILD AND WATER GARDENS The wild garden—Treatment—List of plants suit- able—The water garden—Treatment—List of plants suitable PAGE 38 64 85 95 108 CONTENTS CHAPTER IX THE WALL GARDEN Position and fee aie parle el ae the old wall CHAPTER X PROPAGATION Propagation by seed—By division and cuttings— Methods of raising from seed — Methods of propagation by ied Ai, SOMES ne layers . CHAPTER XI CULTIVATION Climatic difficulties—Contrasting natural and artificial conditions—Tendency to loss of characteristic growth—Methods of overcoming climatic diffi- culties—Importance of proper planting —Time to piant—Importance of top-dressing in spring —Weeding—Garden ae and how tof:deal with them : : : CHAPTER XII PLANTING THE ROCK GARDEN Points for consideration—Difference between alpines and rock plants—Grouping—Examples of same —Planting bulbs—— Ferns— Caution against rampant-growing plants—Importance of pro- portion—Planting shrubs XV1i PAGE 1 126 155 XViil CONTENTS CHAPTER XIIl COST Points affecting the cost—Labour—Clearing scrub— Drainage—Estimated cost of rocks, sand, etc.— Cost of making artificial bog garden—Cost of plants . d : : : , : Pdi Et Alphabetical list of plants suitable for the rock garden A selection of plants for dry and sunny positions, and time of flowering . A selection of plants for full sun or partial sunny positions, and time of flowering A selection of plants for positions in deep shade A selection of plants for partial shady positions. A selection of plants for moist positions in sun or shade . A selection of plants suitable for trailing over rocks . A selection of dwarf shrubs, pines, and conifers A selection of hardy ferns suitable for rock garden in sun or half shade . A selection of hardy ferns for shady position A selection of bulbous plants . Plants for massing INDEX PAGE 201 215 369 37° a7 372 3f3 373 375 376 377 377 379 380 LIST OF ILLUSTRATIONS Rock GARDEN AND STREAM, Mount UsHer Frontispiece TO FACE PAGE WaAttEY Tyre . : ‘ : : : Sauls ee PorTION OF AUTHOR’s GARDEN. 14 Knott Type. ‘ : 4 : d : 20 Bank Type. (In course of construction) . SPOR 10. ts = (Site being cleared) : ‘ see te =“ (Rocks in place) . : ; a2 Rocky BeEp : ‘ : : 3 5: eo HABERLEA RHADOPENSIS. ! ; ‘ Sh NQO Tue Bank. (Roughly formed before rocks are put into place) . : ‘ : : i o 5 AO First Line oF Rocks LAID : : ; EHO, SEconD LINE oF Rose IN POSFRION 0 -) (+2 gO RocKWORK ADVANCING UP THE HILL. : BN (0: NeEarinc ComPLETION : : é : MEd Ist Tue CompieTe Rock work P ; : Boiss. EXAMPLE OF RocKkworRK . i : : a OM AX 6 : : : : oA x1x XX LIST OF ILLUSTRATIONS TO FACE PAGE Boc GARDEN IN AUTHOR’s Rock GARDEN Water-Lity Ponp, Mount UsHER WatTER GARDEN : ; Part oF Watt Garpen, Mount UsHeEr . DIANTHUS ALPINUS . : ROMAZOFFIA SITCHENSIS SAXIFRAGA APICULATA : SAXIFRAGA OPPOSITIFOLIA . ANDROSACE LANUGINOSA, GARDEN. i ? THE AUTHOR’S 98 IIo Igo PART I CHAPTER I THE SITE Surroundings — What to avoid — Objections to trees— Aspect—Contour—Nature of soil—Chief points to be noted—Typical site. ““Wuere shall I lay it out?” This is the first of the many problems to confront one who intends to become the happy possessor of arock garden. In some cases this unfortu- nately presents but few difficulties—unfortu- nately, I say, for then the ground that is avail- able, frequently not of the most suitable, is so limited that there is little, if any, choice in the matter. But others, who have a greater variety of positions to select from, will need to give the question more careful considera- tion. Therefore it would be a mistake to _ lay down any hard and fast rule as to where the site should or should not be, for this 3 4 ROCK GARDENS must obviously greatly depend on what choice of positions there is. A rock garden can be made almost anywhere —almost, be it emphasised ; for under certain conditions, which I shall point out later on, the chances of success would be but small. But no matter what the position is, whether the spot selected be the most ideal, or the most unpromising, it will require considerable thought and no little skill to lay out the garden to the best advantage. The object of this book is to help the reader to approach as near as circumstances will allow to the ideal rock garden: that earthly Paradise, which, alas! the skill of mortal man has not yet achieved, nor ever will, itear. Though there is a great difference of opinion as to what this lovely spot should be like, at the same time there are many points on which all agree. Some indeed are absolutely essential to the well-being of those alpine gems, which one hopes to see grow and thrive as they do in their natural home, though too often one is doomed to disappointment. Yet be not THE SITE 5 discouraged, for there are many and lovely plants which are not in the least difficult to cultivate, and which will amply repay the care devoted to them. But two conditions they do require, light and pure air, which, as with all plants, are essential to success, and indeed I may add a third, sufficient drainage. One has only to remember where most of these treasures make their home in order to realise how important these factors are. Let us for a moment imagine ourselves on some mountain slope where the alpines are in their natural state ; where the Androsaces and Saxifrages carpet the boulder-strewn ground, or beautify the weather-beaten rock with their dainty loveliness ; where by the tiny stream can be seen the lovely Soldanellas and Ranunculus and a thousand other plants to delight the heart of man. Standing there, gazing around, what a picture meets the eye! The craggy height above ; below, the wind-swept pines ; and, far as the eye can reach, range upon range of mountains with their eternal snow-capped peaks glistening in the sunshine. Here is no jarring note ; all is peace and quiet. Nowhere 6 ROCK GARDENS has the hand of man marred the beauty of the landscape. Wesee Nature as she is. With this picture in our minds let us see how best to apply the lesson it has taught us. First we learn that the rock garden should, where possible, be out of sight of all stiff surroundings ; out of sight of the formal garden, with its trim beds and smooth lawn, its close-clipped hedges and rose-clad walls, in some quiet and secluded spot merging from the shrubbery or wild garden. At the same time let it not be so far from the house as to prevent us snatching odd moments amongst our treasures to see what flower has at last bloomed out, or what special plant was raided by the slugs last night. © As variety of outline is one of the keystones of success, endeavour to select a spot that will afford this with the minimum amount of labour. As the spots to choose from are so many and varied, perhaps it would be best, first of all, to find out what should be avoided and what conditions would militate against success. The worst and perhaps really the only hope- THE SITE 7 less site 1s one with large overhanging trees ; for the drip from their branches would all too soon prove fatal to the plants underneath. Although there are some plants that will live and even thrive beneath trees, they are not sufficiently attractive to cultivate exclusively. So to choose such a spot would be but to court disaster; and in addition to the damage caused by the drip from the branches, the roots of the trees would very soon exhaust the soil that had been so carefully prepared for the alpine treasures. It would surprise many to find to what a distance these roots extend. So, above all, let there be no trees overhead, and, if possible, none within at least 15 yards—this is the minimum,—and then only to the north and west. On the other points, south and east, they should be still further away—at least 30 yards,—and even at that distance, as few of them as possible. For if they are closer, the air, which all plants so urgently require, will be shut out, and part at least of the garden will be in constant shade during the winter, with fatal results to many of the plants. Therefore, when feasible, choose a 8 ROCK GARDENS position where there is no chance of trees intercepting the low winter sun, for this is more valuable when the days are short than during the summer months. In cases where, owing to unavoidable circumstances, the trees on the north and west are rather close, provided there is no drip on the garden from their branches; much may be done to overcome the root trouble. A narrow trench sunk rather deeper than the roots of the trees descend, and filled with rough concrete, will form an effectual barrier. Now, having decided that the site must not be overshadowed nor shut in by trees, the next point to be considered is the aspect. This is all-important. The problem is to find out at what point sufficient, but not too much, sunshine can be obtained with due regard to shelter from cold winds. A spot should be selected having an aspect as nearly as possible south-east ; and if there be a wood or a belt of trees some distance off on the north- west, so much the better. This aspect will give the maximum amount of the winter’s sun, and, as south-west is theoretically the hottest THE SITE 9 point, the rock garden will not be so liable to be burnt up during the summer. The next point to be investigated is the contour of the land, and on this will depend to a great extent the type of rock garden to be constructed. Sloping ground, undulating if possible, is undoubtedly the best, for thus more pleasing effects can be obtained, with less labour and expense, than can be expected on level ground, -and the fact that the drainage can be made more effective is by no means the least im- portant consideration. Hollows should at all times be avoided when there are any difh- culties in the way of draining them effectively. Nothing is more fatal to alpines, and indeed to all plant life, than stagnant moisture, which so quickly makes the ground sour. The nature of the soil should be also taken into account. It is a great mistake to imagine that rock plants do not require good soil; the very best fibrous loam, the deeper the better, with a light sandy subsoil, is what they delight and revel in. Therefore, in the selection of a site, avoid as far as possible anything in the 10 ROCK GARDENS nature of a heavy clay soil in which it will be found difficult if not impossible to grow many of the choicer plants, and which will require more drainage and be always hard to work, whether in very dry or wet weather. A stream should be brought through the garden, if it can be obtained by fair means or foul. The merest trickle will suffice, so long as it is constant, but constant it must be, and it will add tenfold to the charm of the garden and afford unlimited possibilities. Having fully described the chief conditions favourable and unfavourable in selecting a site, I briefly recapitulate them as follows :— The garden, when circumstances will permit, should be away from and out of sight of any- thing formal, approached through the wild garden or shrubbery, but still within reasonable distance of the house ; the ground should be undulating, with good light fibrous loam, facing south-east, with, when possible, a stream ; but, most important of all, it must be free from encroaching or overshadowing trees and have plenty of air and sunshine. Bearing these points in mind, a typical THE SITE ey position for this would-be garden of delights may be considered. Perhaps at the foot of a nicely wooded hill facing south-east, with undergrowth around, there may be some open spot, the extent of which is not altogether important ; by removing some shrubs and perhaps a few trees, it can be enlarged, if too small, and, on the other hand, if too big, it will afford scope for judicious planting. This open ground may be in the form of a little valley sloping up the hill, or, again, a small prominence thrust out from the surrounding and gradually rising ground, or undulating with natural hills and hollows. Or, perhaps, the site could be a miniature gorge with wooded heights on either side and a stream flowing through it. ‘These are indeed ideal positions from which to select, provided always that they afford sufficient air and sunshine. | | But with far less promising material—for instance, a sloping field, or even a level one —much can be done. The former is by far the better of the two, for on level ground the question of drainage would probably arise and cause serious difficulties. 12 ROCK GARDENS Sunken ground in the form of a large hollow, such as an old sandpit, can be utilised to some advantage, as can also an old quarry. These, if wooded above, will make charming sites ; but here again attention must be given to the essential necessity of obtaining sufficient fall for the drains. On that point let there be no un- certainty, for, if the drains prove defective, but a poor return will be made for time and money expended. The foregoing are, I think, a few of the most usual sites to select, and their description will, I hope, suggest to the reader how to make the best choice of the ground he has at his disposal. What form the garden will eventually take, and how it can most advantageously be laid out, will in a great measure depend on such details as soil, locality, contour of ground, money to be expended, etc. All these will be dealt with in the following chapters. CHAPTER II THE TYPES OF ROCK GARDEN The natural — The artificial — Types of the natural — _ The old quarry — The amphitheatre — The horse- shoe—The valley—The cliff—The rocky bank— The rocky knoll— Points to be noted — Types of the artificial—The sunken—The gravel-pit—Bank and knoll types — The rocky bed — The very large rock garden — Advantages and disadvantages of the natural — Advantages and disadvantages of the artificial. Tue different types of rock garden may be divided into two sections, which can be called the natural and the artificial. In the Natural garden the rocks are already placed there by Nature, and all that is required is to utilise them to the best advantage. In the Artificial garden—the name speaks for itself—the rocks have to be placed, and in some cases even the banks on which to lay them have to be formed. 13 14 ROCK GARDENS In the first section, the Natural, there are three types—the old quarry, the rocky bank, and the rocky knoll. The second section is more varied, and in it may be included the sunken garden, the old gravel-pit, the bank, the knoll, the rocky bed, and the very large rock garden. The o/d quarry, a favourite and very charming site for a rock garden, and one that has many points to recommend it, is hard to deal with owing to the difficulty of adorning the large masses of perpendicular rocks, which, although capable of giving very striking effects with their cataract of flower and foliage, require years of growth even under the most favour- able conditions. It usually takes one of four forms, which may be called the amphitheatre, the horseshoe, the cliff, and the valley. First let us note the amphitheatre, a very uncommon type. This formation, as the name would imply, is a hollow almost or entirely surrounded by rocky sides, and will seldom be a really suitable site. Most probably it will be found very difficult to drain thoroughly, and even were this done, there would still be that. [Zo face page 14. VALLEY TYPE. VALLEY TYPE. PORTION OF AUTHOR’S GARDEN. (Veronica Lavaudiana and Aubrietia Dr Meeles in foreground. ) THE TYPES OF ROCK GARDEN 15 want of air which causes plants to damp off, especially during the winter. The only occasion on which this form of quarry is admissible is when it is on a very large scale, at least 25 yards across, and even then it will generally prove but a doubtful success. The next and perhaps the commonest form is the semicircle or horseshoe, usually an excava- tion into the side of a hill, giving bold rocky places for planting. This is an excellent form to choose, provided it fulfils certain conditions. It should not open towards the north, for then it would be exposed to the cold winds and the greater part of the rocks would constantly be in shade, and although some shady corners will be found useful, the majority of plants like the sun. In fact, any point but this will do, for so long as the entrance is sheltered from the north it is a matter of no great importance where it may be; but the south- east is preferable, for it gives greater variety of position, both for sun- and shade-loving plants. The valley or defile form can be made very effective. ‘The best way for it to lie is east 16 ROCK GARDENS and west, but this is not so important provided that its course winds sufficiently. The more it winds the better, for it then provides the greatest variety of aspects and sheltered and shady nooks, the advantage of which will be duly appreciated when the time comes for planting. It will also be beneficial if the bottom of the valley slopes gradually from one end to the other, thereby ensuring more efficient drainage. A stream down the centre, when it can be obtained, greatly enhances the charm and considerably enlarges the scope for variety in plants that can be cultivated, because those preferring moisture can be planted on its banks, and those requiring a dry soil will find a home on the heights above. The cif type, as the name implies, is of abrupt formation, rising from comparatively level ground, and is one of the best and most effective forms, provided the aspect is suitable. South-south-east is the best, but any other point will do so long as it does not face towards the north, for with a northern aspect the garden would get but little sun, certainly not the amount all alpines sO imperatively require. THE TYPES OF ROCK GARDEN 17 In this type the surroundings should be care- fully noted, so that they may be in keeping with the proposed rock garden. The ideal approach is through some open and undulating stony ground, with occasional patches of stunted gorse and heather, dotted here and there with oak, holly, or birch trees; then a stretch of grassy sward, with occasional rocks, which become more numerous as the slope gently rises towards the foot of the cliff. Even this is a delightful scene; but picture the face of the rock, now bare, clothed with sheets of Aubrietias, Rock Roses, Dianthus, and in- numerable other equally lovely plants, the Silver Saxifrages springing from some almost invisible chink or cranny, and nestling close to the rock, with their airy blooms waving in the breeze, while from some shady nook the Ramondias peep forth. This is indeed a de- lightful picture, therefore care should be taken to see that the frame is worthy of it. There are now a few points to be noted which apply equally to all the foregoing forms of garden made from an old quarry. The first and most important is to observe Zz 18 ROCK GARDENS the state of the quarried face of the rock. By this is meant, whether it is much broken in outline or presents a sheer, perpendicular face. It should be rough and jagged, so as to give a number of ledges—the more numerous they are and the greater their variety the better— forming steps or terraces, by which easy access can be had to all parts. Otherwise it will probably be necessary to plant the alpines and examine them from the steps of a ladder, which, to say the least, is not desirable. But, bad as a too smooth surface may be, an overhanging rock face is even worse. An odd protruding ledge here and there does not so much matter, in fact they may prove of service, but even they should not be of sufficient size to prevent the rain from reaching the plants below. The whole tendency of the rock should be to slope back from the ground at the foot, for it is necessary that rain should have access to every part. ‘Though it is wonderful in what ap- parently dry and arid spots alpines will thrive, it is certain that their roots must find moisture somewhere. Few people realise to what an extent rocks retain moisture even during a dry THE TYPES OF ROCK GARDEN 19 season, so it is important that the slope should be sufficient to catch and store the rain. If it is intended to conduct a stream through the garden, make sure the outlet will be such as to prevent any risk of the garden suffering from flood. The next type is the rocky bank. This, as its name would imply, is a bank of natural rock, and may be the face of a small hill, or form part of a large one. It is, in fact, rather similar to the quarry cliff, already described, except that the rock, being as nature left it, will probably have a more gradual slope, and therefore possess the advantage of being easier to clothe. The advice as regards surroundings applies equally here. The rock garden should be approached through shrubbery or wild garden, and situated in some open, though not too ex- posed, spot facing south-east, with a back- ground of suitable trees, such as oak, holly, or birch, or any of the Conifer family. The rocky knoll is the next and the last of the natural types. It may be a small eminence of rocky ground standing out by itself, or a 20 ROCK GARDENS sort of promontory on the side of a hill, or on sloping ground. This type is practically identical with the last described, except that in this case the whole or the greater part of the hill is dealt with, whereas in the former only one face of it was. The knoll is undoubtedly the better of the two, for with it a much greater variety of aspects can be obtained, and also more scope for developing the different natural undulations of the ground. The aspect is, of course, unimportant, for a hill standing by itself will be open to all points of the compass ; and even if it be part of a larger hill, it will be shaded only on one side, which will not so greatly matter. With this type, also, care should be taken that the surround- ings are suitable. As this concludes the types of the natural rock gardens, a few points in connection with them may be mentioned before proceeding to the artificial. In the first place, large masses of rock have to be dealt with as placed by Nature’s all- powerful hand, and we must therefore adapt ourselves to whatever conditions we may find. [Zo face page 20. OLL TYPE. BANK TYPE. (in course of construction. ) THE TYPES OF ROCK GARDEN 21 Since the chief object is to provide as many spots as possible to plant in, select the type of rock that will give the greatest variety of pockets, remembering at the same time that all pockets, no matter of what size, must be able tobedrained. The necessity of draining having already been emphasised in the first chapter, it is superfluous to dwell further on it here. The quality of the rock should be carefully observed, and if there be any choice in the matter, select a soft and porous stone, such as sandstone, in preference to hard rock. The former has the double advantage of absorbing the moisture essential to plants, and minimising labour when it comes to making drains. Examination of the different types of arti- ficial rock gardens reveals that they are, as one would naturally expect, more difficult and probably more expensive to make, demanding more thought and skill to engineer successfully than do the natural gardens. On the other hand, they possess the great advantage of ex- treme adaptability, for they can be made almost anywhere, from a level field to the side of a mountain. 22 ROCK GARDENS The first and perhaps the most usual is that type which, for the want of a better name, I shall call the sunken garden. One form of this is a sunken path, the centre of a little valley. The other is represented by a large hollow with banks surrounding level ground, rather similar to what was called the horseshoe in the quarry type. Of these two, the former is the more usual and indeed the most attractive ; nor is the reason far to seek, for in this type every aspect can be obtained by making the paths wind sufh- ciently ; this at the same time has the advan- tage of providing, in a limited area, more surface whereon to plant. This form can be made on the level ; but sloping ground is to be pre- ferred, for the effect is better, and, what is more, or at least equally, important, the drainage can thereby be assured. The surroundings, as in all other cases, should be as wild and natural as possible, though much may be accomplished by judicious planting. As the latter, however, takes years to become really effective, it is best, whenever practicable, to have the surroundings provided by natural means. [ Zo face page 22. BANK TYPE. (Site being cleared. ) BANK TYPE, (Rocks in place.) THE TYPES OF ROCK GARDEN 23 The large hollow is not a form which re- commends itself very much, for it is apt to look too artificial and also entails considerable labour to make. But in cases where there is only a very limited extent of ground it is useful, for in this form the maximum amount of surface available for planting can be obtained. It must be on sloping ground, for to dig a large hollow in the level will usually result in the formation of a small lake. The lower side should be open and at least level with the surrounding ground, above it if possible. This entrance should face either east or south; the north is the least desirable aspect, because then so little of the garden would get any sun. The gravel-pit is practically identical with the foregoing, except that the hollow is already made. The lie of the ground here must also be such as to ensure the drainage being efficacious, and the open side should face any point except the north, preferably south or east. The bank and knoll types are exactly the same as the rocky bank and rocky knoll already described, except that in the former the natural rock is already there, while in the 24 ROCK GARDENS latter cases we have only the soil, and the rock will have to be placed. Both these types are attractive, but in a great measure dependent on the natural contour of the ground, which should lend itself to their several requirements. ‘The bank form can be made in a gently sloping field, but the labour entailed would be great. In the case of the knoll, so great would it be that I would not advise the reader to embark on it, not at least on a large scale, for I doubt if the result would sufficiently repay him. Next we will deal with the rocky bed or very small rock garden. This is a most delightful and useful way of growing alpines for those who can only spare a few square yards, or who do not wish to attempt anything larger. The surroundings in a garden of this kind are not, of course, of such importance as in the case of the larger gardens, but all the same it is as well to select as sunny and open a spot as can be obtained, and removed as far as possible from the shade or drip of trees. There is no garden, no matter how small, that cannot have a rocky bed, tiny THE TYPES OF ROCK GARDEN 25 though it may have to be; and it is surpris- ing with what success some of the most diffi- cult alpines can thus be cultivated, creating no small amount of envy amongst the possessors of larger rock gardens. The best and most usual form is a raised bed, the outline of which may be as varied as fancy dictates. It can also be made on the sunken principle, but would then occupy more space and not prove as satisfactory in many ways, whilst it would also entail more labour and greater cost. For all that is aimed at in a garden of the rocky bed type is to have some spot in which to grow the choicest treasures. It is useless to endeavour to imitate larger gardens, because the result would be only disappointing and would look puny and out of keeping. Therefore I most strongly recommend those who cannot, owing to circumstances, have any of the foregoing types, to try the rocky bed, and I feel sure they will never regret the experiment. We will now go to the other extreme and describe the very large rock garden. This garden is diametrically opposed to all other a6 ROCK GARDENS types hitherto dealt with, in which the alpines are grown on the lower banks or slopes, with shrubs and trees above; for in this case the shrubs are planted in the valleys and the alpines on the heights. Without doubt this is copy- ing nature more closely; but to be effective the plan must be on a large scale, and cover at least two or three acres, composed of natur- ally very broken ground, with hills and hollows well defined and as varied as possible. The rockwork must correspondingly be of a bolder formation and its lines more generous. If this is not the case, the result will be but a poor and insignificant imitation of Nature’s handiwork, and consequently be far from pleasing. As, however, bold rockwork entails larger rocks, with cost and labour increased proportionately, I should not advise anybody to whom expense is a matter of importance to embark on a garden of this type. But those who are fortunate enough to be indiffer- ent to cost, and who possess suitable ground, may well be urged to make this style of garden, and I feel sure that they will never have cause for regret, provided it is skilfully laid out. (To face page 26. RGCKY BED. HABERLEA RHADOPENSIS. THE TYPES OF ROCK GARDEN 27 The surroundings are, if possible, even more important than in any of the foregoing types, for here ugly spots cannot be planted out with banks of shrubs, and, as the paths and rockwork are on the heights, there is every opportunity of seeing those walls and hedges which we fain would imagine to be miles away. For these reasons, therefore, the site should be some distance from any of those blots which mar the view, and the approach should be through some wood, or ground of a similar and un- cultivated character. A spot such as the following would be ideal: a large open stretch of very undulating and broken ground at the foot of a wooded hill—with perhaps a stream running through it into a piece of marshy ground at its lower side —which in turn merges into scrub and stunted trees, fully exposed to all points except, per- haps, the north, where it is sheltered by a belt of ancient and majestic pines. This would indeed be a spot worthy of the bestowal of every care and trouble, in order to develop it to the best advantage. The ground for a garden of this sort must be very undulating ; by undulat- 28 ROCK GARDENS ing I do not mean merely uneven ground, but a spot with well-defined hills and hollows—the bolder and more marked they are the better. This concludes a terse survey of the various types of rock gardens, and some consideration of the various advantages offered by each may assist in deciding which of them shall be adopted. Asa rule, though, the choice is but ““ Hobson’s choice,” for the site in most cases decides the type that will have to be adopted. The advantages of all rock gardens belong- ing to the natural section, quarry, rocky bank, or knoll, are that the rock is there ready placed by nature in masses far larger and bolder than could ever artificially be put into position, and that, for this reason, more striking effects can be obtained, both as regards height and boldness of outline. It may also bea very cheap garden to make. But this is a very uncertain quantity, and is difficult to ascertain at first. It may ultimately turn out to be both expensive and troublesome, because so much depends on the formation and description of the rock. One of the disadvantages is that it is often very difficult, and, in fact, at times THE TYPES OF ROCK GARDEN 29 almost impossible to get as much soil into the pockets as the plants that are to inhabit them require. Often, indeed, it is hard to find pockets or ledges of any description where they are wanted, and although rocks are attractive in conjunction with flowers, still they are only the means to an end, and that end the cultivation of the plants. Effective drainage may at times be found difficult and expensive to carry out successfully, but as this depends enormously on the texture of the rock and other circumstances, it would be a mistake to lay down any hard and fast rules regarding it. © _ Itisa great fallacy to imagine that alpines, because they are dwarf-growing, do not re- quire a deep rooting medium, for there is no class of plants whose roots penetrate so deeply in comparison to their height. In addition, as all the rockwork has to be artificial, every description of pocket and ledge can be made, as varied and numerous as the needs of the plants demand, the benefit of which will be found when the time comes for putting the alpines into their home. 30 ROCK GARDENS The chief disadvantage is purely a matter of cost, though this is often a rather important point. The labour of digging out the garden and, in all probability, of having to cart the rocks some distance means expense. Those people are fortunate indeed who have the rocks close at hand, and only those who have had to draw their supplies of stones from a distance can fully realise how great a saving it Is. The merits of the sloping bank and knoll types are, that if the ground has the necessary conformation, the only labour entailed is to clear the surface of scrub and weeds, and to place the rocks in position ; therefore, unless any unforeseen difficulty should arise, they are probably the least costly types to make, but at the same time they are not so effective as are some of the other forms. As regards the old sand- or gravel-pit, pretty well the same remarks apply to them as the foregoing, except that in these cases soil may have to be procured in addition to rocks, and perhaps, the outline being bolder, better effects may be obtained. The small rocky bed (we can scarcely call THE TYPES OF ROCK GARDEN 31 it a rock garden) has this advantage over all the other types, that it can be made in a very limited space ; but it does not give scope for the effective massing of plants; its purpose is merely to satisfy those who wish to grow alpines, but do not desire to embark on any large undertaking. We now come to the last, namely, the large rock garden with its shrub-filled valleys. This form, provided that the space is large enough and its undulations sufficiently well and boldly defined, can be made most effective ; but it is, without doubt, the most difficult to lay out really well, and the cost, owing to its size, must be considerable, for in order to get the proper effect it must be made on very generous lines. Big, bold masses of rockwork are essential to success, and these should so combine with the undulations of the ground as to give it that natural appearance which should be the characteristic feature of all rock gardens. CHAPTER III THE NATURAL ROCK GARDEN First preparations—Cleaning ground—Exposing and pre- paring the rock—Pockets natural and artificial. Tus chapter will be devoted to the develop- ment of the natural rock garden, the various types of which have been dealt with in the previous chapter. As may be expected, very similar treatment will be required for all types, in order to convert them from rocks, either bare or covered with rank vegetation, into gardens in which the choicest flowers can be grown. The follow- ing directions will therefore equally apply to all natural rock gardens, irrespective of their various forms. When the scene of operations has been decided on, the first thing to be done is, if necessary, to expose the surface of the rocks 32 THE NATURAL ROCK GARDEN _ 33 and clean out every nook and cranny, remov- ing all weeds or other vegetation growing there. This will at times be found no light task, for it is surprising how roots, especially those of weeds, penetrate into almost invisible fissures. It is difficult to oust such a plant as a dandelion when it gets its long and fleshy tap-root down several inches into some crack in solid rock, for it will not be sufficient merely to cut off its crown—its severed crown will promptly reappear worse than before. All overhanging trees should be cut down which are likely to cause a drip on to the plants below, or which would in any way shut out light and air from the garden. It is, however, advisable, in the initial stages at least, to leave any small stunted trees or shrubs springing from fissures in the face of the rock, or clinging to its summit, especially if these include such kinds as mountain ash, oak, holly, or any of the Conifer family. ‘They are characteristic of and in keeping with mountain plants, and can easily be removed later, if it be deemed advisable, 34 ROCK GARDENS When the work of clearing and exposing the rock has been done, the next operation is to prepare receptacles for the soil in which to grow the plants. These may be obtained by opening up and developing natural fissures, by making new ones, or by forming pockets among the rocks. The only means of accomplishing this is with the crowbar or cold chisel, assisted at times with some explosive, such as dynamite or gunpowder; but this latter method is dangerous in every sense. Besides requiring very careful manipulation, it is very apt to do more harm than good by removing rocks that would have been better left, or by impercep- tibly loosening others, the first intimation of whose unstable condition is finding them one morning at the foot of the cliff burying some of our choicest plants in their dééris. It is well to remember, when splitting rocks, whether it be for the purpose of making or enlarging some crevice, or opening up some pocket, to first find out how the “grain” of the rock runs, and split the rocks with, not across it. For rock, like wood, has a regular THE NATURAL ROCK GARDEN 3; and defined dip or grain, though perhaps not as well marked as the latter. As it is obviously impossible that any definite rules can be laid down as to how such pocket or fissure is to be formed, the reader must exercise his own judgment in the matter, and utilise the grain of the rock to the best advantage. In no case should a pocket or a crevice, whether per- pendicular or horizontal, be made in an overhanging rock face. Of even greater importance is the ‘necessity of ensuring that the bottom of each pocket can be drained thoroughly, even should this entail cutting through solid rock. If this, however, be found impossible owing to position, or other circumstances, in order to prevent water lodging, fill up the pocket with rough concrete to the point where the draining becomes effective. When there is a difficulty in getting a sufficient number of natural pockets in the rock a good deal may be accomplished with the assistance of cement, yet it should never appear in the completed work, but always PS ROCK GARDENS be most carefully hidden by either plants or soil. . Examples of this kind of work are the building of a small terrace of stones along the front of some ledge, incapable of holding any or a sufficient depth of soil, or the fixing of wedge-shaped pieces of stone with their thin ends up in some large vertical fissure to keep in the soil, which otherwise it would be found difficult, if not impossible, to prevent being washed out by heavy rain. Bare rocky ledges which slope outwards in such a way that the rain is thrown off, instead of being directed back towards the soil behind, can have this fault remedied by raising the front with a layer of concrete. Steps may also have to be cut to give access to the plants. These should never be formal, but as uneven as possible, and merely form stepping-stones from one ledge to another. Where a stream comes over the face of the tock, its volume and course should be so con- trolled that even in time of flood it can do no damage to the plants growing beside it. The two chief things to remember in the THE NATURAL ROCK GARDEN 37 preparation of natural rockwork are first to ensure thorough and efficient drainage of every pocket, and next to provide space sufficient to give that depth of soil all rock plants so urgently require. Great care should also be taken, when putting soil into the pockets and crevices, to work it thoroughly into every corner, so that there is no possibility of leaving any air space, which so often proves fatal to plants. With regard to the soil required, full particulars will be found in Chapter VI. CHAPTER IV THE ARTIFICIAL ROCK GARDEN What can be done with a level field—Laying out the garden -—Boldness of design—Preparing banks for rockwork — Treatment of natural valley—Formation of banks— Example of a “valley” garden—Points to be noted in excavations— The hollow garden—Gradient of slopes — Drainage—Aspect—Bank and knoll types and their treatment—Rocky bed—Very large rock garden— Difficulties connected with this type of garden—Points to be noted in artificial rock gardens. Even more care and study are required to lay out the artificial rock garden to the best advantage than is the case with the natural rock garden; for whereas in the latter the position chosen will to a great extent, if not altogether, determine the type it has to be, this is not so with the artificial garden, or at least not to the same extent. ‘Take, for example, a narrow field, sloping gradually from one end to the other, which all will agree does not 38 THE ARTIFICIAL ROCK GARDEN 39 appear the most promising material on which to work. Yet even this can be developed into three different types of garden. By levelling a piece of ground across the face of the slope, and raising the upper part with the soil removed, we have the founda- tion for the bank type; again, by sinking a path up the slope of the field a little valley is obtained ; and, lastly, by excavating into the slope the foundation for a garden is made, which will in a great measure correspond to the horseshoe type. It is therefore quite evident that it is necessary to devote con- siderable thought and exercise no little skill in laying out the ground, if it is to be done in the best possible, and at the same time most economical manner, and full advantage be taken of all its natural formations. The fact that the same piece of ground can be laid out in several very different ways, all equally good, or nearly so, does not simplify the task. It should also be realised that one cannot expect to get quite such bold effects in the artificial garden as are found in the natural, and 40 ROCK GARDENS the reason of this is not far to seek, for in order to obtain them large masses of rocks, weighing at least several tons, would have to be put into position, a Herculean task involving no little cost, and a bigger undertaking than the majority of people would care to attempt. But let it not be for a moment supposed that these large boulders are an absolute necessity, for quite as attractive gardens can be made by using com- paratively small rocks, so long as these rocks are skilfully placed. On the placing of them the ultimate result will in a great measure depend. As it would be best perhaps to take the several types by themselves, the little va//ey shall be dealt with first. Having selected a site of ground sloping, if possible, for reasons already stated, the first thing to be done is roughly to lay out its general lines. Place a few stakes in the ground along the pro- posed lines and study the effect from various points. There are several things to be noted when doing this. Let the valley lie as nearly east and west as possible, and especially let it lead from one definite point to another, and not look as if it had come there by chance and with no [ To face page 40. THE BANK. (Roughly formed before rocks are put into place.) FIRST LINE OF ROCKS LAID, THE ARTIFICIAL ROCK GARDEN 41 object. Study the contour of the ground, and if, owing to the surroundings, it appears as though a path should come from one special place and Pdead to another, mark these two points and lay out the garden between them. The garden should neither begin nor end too abruptly ; let it commence with a barely perceptible hollow, the height of the banks on either side of the path gradually increasing ; while in a similar manner they should decrease in height at the further end, merging by degrees into the rising ground beyond. Nothing could look more unnatural than to come suddenly from comparatively level and probably rockless ground upon some bold rockwork which is there for apparently no reason, and which ceases equally abruptly and inconsequently. The same rule applies to all artificial rock gardens ; avoid anything sudden in the way of rockwork, but lead up to it gradually, with odd rocks scattered here and there, increasing in number as you approach your garden proper. If this be done it will only be copying nature, which should always be your guide. If the space available for working upon be 42 ROCK GARDENS rather limited, and it is desired to obtain the utmost possible surface on which to plant, resist the temptation to make two small valleys in. preference to one large one. This is most im- portant and one of the most essential points to be noted, and I cannot impress it too strongly on the reader ; for in laying out a rock garden, one of the chief objects is to obtain the boldest effects the site will permit of. This cannot be achieved if more is attempted than the space conveniently allows. If space is available and it is desired to have two parallel valleys, they should be at least 20 or 30 yards apart, in order to allow for planting on the top. The path should so wind about that it is impossible to see both ends simultaneously from any spot. The garden will be far more attractive and its size apparently very much increased if we see only a small portion of it at a time, and if round every corner some fresh attraction can be provided. | Having laid out the general lie of the main path, the next procedure is to mark out lines on either side about 20 feet from it, following its course throughout, which we will call lines THE ARTIFICIAL ROCK GARDEN 43 of section. Then from the ground, between the lines of section, the sod must be removed, skinning it off as thinly as practicable and stacking in a heap in some spot convenient, but well out of the way. The next operation is to begin to dig out the valley. But should the work be done after the method practised in railway and other similar cuttings, the top, and most valuable, soil would become buried Bi Bas. rane Bs ASTIN eo mR Ground Level - D cota hae =x A 2 - beneath the almost useless subsoil. Therefore some other method must be adopted. A practical way of doing this is to divide the ground between the lines of section into plots about 10 yards long ; each of these will then measure about 30 feet long by 40 feet wide. From the first plot remove the top spit to say a depth of 6 inches, putting it in heaps (B, B,, fig. 1) on either side, and about 10 yards away. Next, dig out the soil down to the subsoil, also putting it in heaps (B, B,) on either side, and about 5 yards from ground 44 ROCK GARDENS marked out. Thus a cutting will be obtained (A A) about 40 feet wide, and probably from 12 to 18 inches deep, with all soil of any value in heaps (B, B,) on either side. But the banks must be quite 6 feet high, and at least another 3 feet will have to be sunk in order to achieve this. In the centre of the excavation dig a trench (C) about 3 feet deep and g feet wide, the bottom of which will be the path, and throw the soil back on either side over the ground (D) between the heaps (B, B,) and lines of section. _ Now cut out the portion E E, and throw the soil over what has just been removed from C. This will probably raise the bank to from 4 to 5 feet above the bottom of the path, with its sides sloping towards it. At this stage it is well to work out the general formation of the banks in the plot being dealt with ; by this I mean that where it is ultimately intended to have a long and gentle slope, keep the subsoil further back and roughly make the desired gradient, or if a bold and almost per- pendicular bank is required, work accordingly. For further particulars with regard to the formation see page 47. THE ARTIFICIAL ROCK GARDEN 45 Having done this, the second plot is ap- proached. Remove top spit as before, but the rest of the soil, down to the subsoil, instead of being put into a heap, should be thrown over the two sloping banks of plot 1. Then proceed to dig out the subsoil as before, again making the general formation of the banks as the work proceeds. Plot 3 is dealt with in ‘a manner similar to plot 2, and so on till the last plot, when the soil removed from plot 1 and left in a heap (B,) must be put over it. The result of all this will be a cutting with sloping banks on either side about 5 to 6 feet high, with the good soil over the subsoil, and with heaps of the top and best soil available for use later on when the time comes to place the rocks in position. This method undoubtedly entails consider- able labour. The only alternative is to treat the whole in a way similar to that adopted for plot 1, and after having roughly formed the banks to throw back the soil removed. This, of course, has the advantage of saving the labour of carting the soil from the first to the last plot, and also it is easier to work out the general 46 ROCK GARDENS formation of the banks when all excavations are completed ; but, on the other hand, there is more labour in throwing the good soil into heaps and then putting it back again than in placing it directly over that plot which has just been dug out. It is, however, a matter of individual taste which method to choose, and though perhaps the latter alternative does entail a little more work, it is the better of the two, on account of its being easier to lay out the general scheme when all digging is completed. I have described these operations rather fully, as I wish them to serve as a basis for the reader to work on; but he will, of course, have to modify them according to the circumstances of each individual case. Where, however, there is a natural valley it may only be necessary to make it a little deeper. This can easily be done by sinking the centre path and throwing the soil up on either side, having first removed the sods, or whatever was growing on it, for it is essential to have clean ground to start with. The chief point to look to when excavating is that the subsoil shall be in its proper place beneath the upper soil, and THE ARTIFICIAL ROCK GARDEN 47 also that the latter shall be of the greatest possible depth. It has been already pointed out that the course of the valley should wind about consider- ably. The slopes forming the banks must also vary as much as possible, in order to avoid pre- senting a too artificial appearance.- In some places let the bank recede in a gradual slope, whilst in others a more abrupt formation will lend a pleasing contrast. Here and there some flat patches may skirt the path and be almost on the same level with it; further on they may be raised considerably above the path level. These are a few suggestions whereby to break a stiff and formal outline and give the variety which should be aimed at. As I said before, it is far more satisfactory, and will probably in the end save considerable trouble, if the rough formation of the banks is made when cutting out the valley and before the rockwork proper is commenced. Nothing is more annoying than to find, when making a hollow in a certain spot, and when it is nearly completed, that the subsoil appears, for this entails removing the good soil for some distance round the spot and 48 ROCK GARDENS sinking the hollow deeper, with every chance of leaving it in a form that will retain water—the last thing to be desired. For these reasons let the rough scheme be worked out in the initial stages. It may perhaps assist the reader and be the simplest way of showing what is meant by these variations in the slope of the banks, if I describe one artificial valley that I know. This valley is made in a piece of ground about 85 yards long by 35 yards broad, with a fall of about 1 foot in 16. Owing to the cir- cuitous course the valley takes, its actual length is about 105 yards, with a path varying from 7 feet to 12 feet wide, while the banks on either side vary from 5 feet to g feet in height. Particulars of the formation of the right bank may now be given, beginning from the lower end of the valley. For the first 15 yards or so the slopes are very gradual, but increasing in gradient until some bold rockwork at A (fig. 2) is reached. This is almost perpendicular, and is from 8 to g feet high. Round this the path turns sharply to the right, and then follows an almost semi- THE ARTIFICIAL ROCK GARDEN 49 circular course to the left for about 15 yards da ores’ @ & @ @ 4 a rg a ‘ 4 Th ¢, te’ . i s § wy a YY saoi \! sex hy * wa ae C3 a) a & | ee (Wi * S(esepe Gh & eRe ed & im SS 4 e8> )E7 jz a & ye & ! Lag fate. Bo & < a aes Bip CELLO Ge " Sees e yy e Be om DEB QSaS : Contour Lines ae b-) Dwarf Shrubs 4 t3$t. SS Tall Shrubs Sto 8Sty a 4 Trees, Oak, Birch, Fine: Fic. 2. After passing the corner at A, the bank gradu- ally takes a more gentle slope receding from 4 50 ROCK GARDENS the line of the path, thereby forming an almost level piece of ground at B between the base of the bank and the path. This continues to about the centre of the semicircle, when at C the bank again advances towards the edge of the path, at the same time increasing its gradient until, by the time the end of the semi- circle is reached at D, the slope again reaches the line of the path. Here it bears rather sharply to the right and continues to wind on gently for the next 17 yards, when it takes a decided bend to the left at E for about 20 yards, then once more curves gently to the right and gradually dies away towards the rising ground beyond. Beyond the rather steep part at D, the bank by degrees assumes the form of two terraces, the lower one about 3 feet above the level of the path, and of a very gentle gradient, while the upper one is much steeper. This formation continues until about the point E, when the terraces once more merge into one bank, with a gentle gradient, which does not vary very much for some distance, when it again becomes steeper as it winds round the curve, after [Zo face page 50. SECOND LINE OF ROCKS IN POSITION. ROCKWORK ADVANCING UP THE HILL. ie vy = ae wh me a ihe tate 8 : dite et ; j * -~ ara e - x 7 ai s £ ; fuses , ca | - - we es i OAS > rin 7 AY ip —oo Hy 4 see . ed 3 ~ a | oe f, . ak s THE ARTIFICIAL ROCK GARDEN 51 which it gradually dies away into the rising ground beyond. If the reader has been able to follow this description, he will see the necessity of working out the general scheme of the banks as he proceeds with his excavations. There are a few things to be observed which may help him in deciding where the level or where the steeper places should be. It is generally found to be more effective if the sharper gradients are made on the con- vex curves, while on the concave are the level spots and more gentle slopes. The advantage of this is fairly obvious, for in the former case the steep rockwork will stand out boldly and naturally form a corner, while in the latter the banks form a hollow which lends itself to the formation of a level spot at their base. If these ideas are carried out, there will be no danger of having the same formations opposite each other on either side of the path, for where the curve is concave on one side it follows that it must be convex on the other, and vice versa. It will give a pleasing variety if in some py ROCK GARDENS places the valley has the appearance of forming a gorge, while in others it may pass between two hills ; but even then the steep parts need not be exactly opposite, for the high ground can be cut diagonally. It is only the general outline of the banks that is being dealt with at present. In the chapter dealing with the placing of rocks in position, it will be shown in more detail how they should be worked up to give that appearance of natural rockwork which is to be aimed at. There is yet another point to be noted before leaving this type of garden, though, indeed, it applies generally to ali gardens which have to be artificially dug out. It is this, when excavations have to be made, always be careful to get the top soil well away from where actual work is being done. You will thus ensure there being plenty of room, if, as may easily happen, it is found desirable to slope the top of the bank further back than had originally been intended. If the good soil be in the way, extra labour will be entailed clearing the required space. THE ARTIFICIAL ROCK GARDEN 53 The next type of garden is what may be called the hollow. If it is to be made on the face of a hill or rising ground, lay out the site in a manner similar to that adopted in the case of the valley ; the outline should be as varied as possible, and be careful to avoid the appearance of its being marked out with a compass. Remove sods or whatever may be growing on it, then take off top soil to a depth of about 6 inches, putting it in heaps at a convenient distance from the ground marked out. After this has been done, dig out the remaining soil to the subsoil, also putting it aside. Then continue the excava- tions, throwing the subsoil removed on to the ground above until the required depth has been attained. The banks must not be made too steep; they should have a general slope of not more than 1 foot in 4 feet, because the steeper the formation the more numerous and the larger are the rocks required. In places, of course, the slope may be fairly perpendicular ; but even there a gradient of 2 feet in 5 feet is quite enough, and effects as good, if not better, can be obtained than if 54 ROCK GARDENS a miniature cliff be constructed. After having dug out the garden in the rough, the good soil can be spread over the bank that has been made, keeping the first 6 inches for use when making the rockwork. In this case also it is advisable to work out the formation of the banks in a general scheme before commencing to build in the rocks. Make it as varied as possible—a steep’ place here, there a little hollow looking like a tiny gorge which might at one time have been a watercourse. In another spot let the slope of the bank be very gentle, merging into a level stretch at its foot. Variety should always be aimed at, and the sequence of formations should not be repeated. When deciding on the depth to which the hollow is to be sunk, always look to the drainage to see that there will be a good fall towards the entrance, in order that the water from the surrounding rocks may be carried off. This type of garden should always face south, or as nearly towards that point as possible. If there is a natural hollow to work on, then the labour of making the garden is consider- THE ARTIFICIAL ROCK GARDEN 55 ably reduced. All that is required to be done is to first clear the ground and then to develop it as far as is deemed necessary, treating it in much the same way as in the more arti- ficial type. For a natural hollow merely means that Nature has already done a certain amount of the work, and man will have to take it up where she left off. Now to deal with the 4ank and knoll types. These, as may be expected, do not require the same amount of work in the earlier stages as those types already described, for, instead of the general formation having to be made, the outline is already there and only requires judicious development. If they are not sufficiently high, this can be rectified by sinking the ground at the base and raising that above with the soil dug out, having previously cleared the ground of whatever was growing on it. Very probably there will be occasional and solitary rocks just showing above the sur- face. Dig round these to ascertain their size and how they can eventually be worked into the general scheme of the rockwork. The banks in gardens of this type are usually 56 ROCK GARDENS higher than in the case of the valley gardens, and it will be expedient to make paths over them ; it will therefore be advisable that their general lie should be planned out, and the formation of the banks made in the rough. The rocky bed requires comparatively little work in the preparatory stage. It usually is made on ground that is level or fairly so; therefore make sure that it can be properly drained. When the site has been marked out, making its outline as varied as possible, dig out the soil to a depth of about 24 feet, and put in 6 to g inches of broken stones, coarse first and smaller above, which will facilitate drainage. At the same time take care that there is a proper outlet for the water to get away. ‘The soil can, if necessary, be replaced, adding more then to bring it to the required height ; for the top should be quite 3 feet above the surrounding ground. In dealing with the /arge garden it is difficult to give any very definite instructions for its treatment, as so much depends on the natural formation of the ground. One of the more important points is to lay out the garden THE ARTIFICIAL ROCK GARDEN 57 in correct proportions: a most difficult thing to achieve. Try to avoid having a number of small and abrupt hills with wide valleys between, looking like a collection of ant-hills dotted over the ground. If the valleys are wide, let the hills be large, not necessarily high, but covering a good extent of ground. Three or four, or even two, are quite enough, rising gradually in long and gentle gradients. For example, we have two fairly large hills, each covering about half an acre or more, with a curving valley about 30 feet wide between them. As we advance up this valley another and rather larger hill appears in front, round _ the foot of which the valley winds towards the left, while on the right a miniature pass crosses the range. These three hills, with the valley and pass, are quite enough to cover a couple of acres of ground ; any more would look out of proportion and give a fussy effect, which should always be avoided. This is merely a description of the main scheme, for the hills will need to be broken up and varied in outline. As the rockwork is on the higher ground, 58 ROCK GARDENS the paths leading through it will probably have to be on almost the same level as the plants, or at least more nearly so than in the types of gardens hitherto described. There-— fore, in order to show the plants to advantage, and to be in keeping with the surroundings, the rocks should be considerably larger and the whole scheme worked out on bolder and broader lines. Where, for instance, in the valley type, the level spots cover a couple of square yards or so, three or four times that extent would be scarcely sufficient here. Again, when in the former case the bank rises 6 feet in a distance of as many yards, here it may take 15 or 20 yards, and the eventual height from the bottom of the valleys to the tops of the hills will be anything from 10 feet to 30 feet. The inclines should always be more gradual than in any of the gardens thus far noticed. In fact, to put the whole matter concisely, in the former types the idea was to copy some small portion of an alpine scene, while in this it is to produce, though on a very reduced scale, the scene in its entirety. [Zo face page 58. NEARING COMPLETION. THE COMPLETE ROCKWORK, (Note general dip of stones from right to left. ) THE ARTIFICIAL ROCK GARDEN 59 How difficult it is, only those who have tried can fully realise ; but if it is successful, a masterpiece of engineering and artistic skill will have been achieved. ‘There are few who, on account of expense, could attempt a garden of this description unless they had ground that lends itself naturally to it, when it would perhaps only be necessary to make the valley or valleys deeper and slightly raise and vary the heights and contours of the hills. In this case one need clear only the ground that is intended to be rockwork, namely, the heights, unless indeed the valleys are overgrown with trees or scrub, when it should be thoroughly cleaned. As the surface to be planted is so large, and as it is always necessary to reach all parts easily and without fear of walking on the plants, a number of paths will be required. Not paths in the generally accepted term— nice gravelled walks with stone edges—but almost imperceptible ways winding in and out amongst the boulders, starting from the main valley up a gentle slope for perhaps a few yards, then round some large rock and 60 ROCK GARDENS on, ever gradually rising, through some tiny gorge, then perhaps a rock-strewn valley, bending its way round some lesser peaks of the main hill, till at last the top is reached. Then let this summit, while bold and conspicuous, be no mere incongruous mass of piled-up rocks. The object of a rock garden is to grow plants, and the rocks are merely an adjunct, though an important one, but still only the means to an end. So unlimited is the scope for variety in a garden of this sort, that space will not admit of the enumeration of all its possibilities, so much must be left to the reader’s discretion. Almost every type of garden thus far described can be worked into it more or less ; the object being to get the greatest variety of aspect and formation. ‘To him who is in want of ideas there is no better teacher to go to than Nature. There is, however, no need to travel all the way to Switzerland. Let him but study the rocks and hills at home, and on his return try to reproduce the formation or scene on a smaller scale in his garden. More ideas can THE ARTIFICIAL ROCK GARDEN 61 be obtained in a day’s walk ina wild and rocky country than in a week’s reading, and a few studies of natural rock made with a camera will prove of untold assistance. The chief difficulty is found when one comes to try to reproduce and adapt these pictures on the reduced scale: such at least has been my experience. Those who, undeterred by expense, or the magnitude of the undertaking, wish to make such a garden, but have only sloping ground out of which to form it, will have to dig out the valleys and raise the hills. To such I should recommend working on lines somewhat similar to those laid down for the valley, only, of course, on a very much larger scale, the valleys wider in proportion to their depth and the hills with a more gradual incline. To get the soil in its proper place and to avoid burying what is wanted on the top and to distribute it evenly will be even more difficult than before. For while alpines require at least 12 inches of good soil as a general rule, the shrubs in the valleys require far more, and great ingenuity will have to be exercised to provide all this. 62 ROCK GARDENS Keep the very best fibrous loam for the alpines, because shrubs, though requiring more depth, are not so particular as to quality. Even if the soil is mixed with a little subsoil, “‘ weathering ” will soon improve this and do no harm. In all artificial rock gardens aim at variety of outline, but avoid too sudden transitions from one formation to another. Let the change be gradual, and always try to make it look as if there was a reason for it, for such is invariably the case in nature. Try to keep the lines of the garden in proper propor- tion. It is not an easy thing to do, for in ten yards you must reproduce what in nature covered ten times that distance ; but as pro- portion is the keynote to success, it is worth taking considerable trouble to achieve it. Do not attempt more than the size of the ground will allow. One good wide valley looks far better than two narrow ones. ‘There should always be the feeling of openness and expanse. Also remember that all-important point, drainage. Whenever making a hollow or forming a level plateau, see that there is no chance of water lodging: this is the stage THE ARTIFICIAL ROCK GARDEN 63 at which to do it, and not after the rock- work is finished. It is hardly necessary to say that the real effect cannot be judged until after your rocks have been placed in position ; but please remember that the hills and banks are not made to grow rocks, but that the rocks are placed there to keep up the soil in which the plants are to grow. Without the rocks the soil on any steep gradient would soon be washed down, although there is an angle, called >) “‘the angle of repose,” at which soils of every description will remain stationary. But this angle is generally of a far more gentle gradient than is suitable for a rock garden, therefore it will be necessary to adopt the other alternative and build the slope up in the form of terraces, each of which will, in all probability, have a gradient considerably less acute than the “‘ angle of repose,” and there will, therefore, be less chance of the soil being washed down by heavy rains. CHAPTER V THE ROCKWORK Kind of rock to use and what to avoid — Principles of particular construction—Types of rockwork to avoid —The placing of individual rocks—Fissures and how to make them—Strata—Principles of general con- struction—Paths—Steps—The moraine. Havine made the garden in the rough, the next step will be to construct the rockwork, and the first thing one naturally wants to ascertain is the best description of rock for this purpose. On this point there is a diversity of opinions, but the majority consider that sand- stone, which has special properties of retaining moisture, is the best all round, though others prefer mountain limestone. But almost any kind will do provided that it has a sufficient variety of forms, the greater the variety the better for our purpose. The stone from the nearest quarry will, generally, be found most desirable, if for 64 [ To face page 64. EXAMPLE OF ROCKWORK. EXAMPLE OF ROCKWORK. THE ROCKWORK 65 economical reasons only. _Quarried stone will give more pleasing results, and is easier to build, the faces of the stones working in better together than is the case with those rounded by the action of the weather or water. Some people indeed object to the freshly hewn rock as looking too new, but the advantage of having rocks that will join closely together more than makes up for this, and a couple of years will give the desired weather-beaten appearance. It is impossible to build rocks with rounded surfaces properly together, for, being invariably convex, it follows that on either side of the point where they touch there must be a gap, which has to be filled up somehow or other with small stones. ‘This is seldom satisfactory, for frost and rain have a marvellous way of working soil through any opening, no matter how small. The rock for building should consist of solid blocks, with as much variety of surface as possible, and must have a good base to rest upon. ‘The area of the base should be about two-thirds of the surface exposed, and 5 66 ROCK GARDENS quite one-third of the rocks should be buried in the ground ; some people say even as much as half, but this is rather more than is necessary. As I have already stated, almost any class of rock will do; but avoid anything in the shape of thin slabs or round boulder stones—the former because they have an ugly and mono- tonous face and are difficult to make steady in the ground, the latter on account of the difficulties already mentioned in working them together satisfactorily ; though a few indeed may be half buried on the summit of a mound to give the idea of the point of a rock just appearing over the surface of the ground. Above all things, avoid anything in the nature of artificial rock, clinkers, burrs, bricks, or the like. The bottom of a drain is the only place for such as these. Equally to be shunned are old tree stumps or wood in any form, for it so quickly rots away and promotes the growth of fungus, which is pretty certain to kill any plant growing near it. Nothing but good rough stones should be used. The size will in a great measure depend on the facilities for dealing with them, and also on the nature THE ROCKWORK 67 and size of the garden; the larger the rocks, the bolder the effects obtained. All sizes can be used, weighing from a few pounds to a ton or more. A good building material is about 15 or 20 rocks to a ton. Not only do I recommend quarried stone, but go so far as to say that it should all come, not only from the same quarry, but even from the game part of it; for you will then get both a uniform colour and strata, which will to a great extent simplify the building and also give the ultimate rockwork a more natural appearance. It should never look like a collection of geo- logical specimens. The fundamental principle governing the construction of rockwork is to make the visible rock appear as if it is merely a part, and only a small part, of what is hidden by the soil. To carry out this idea, there must be uniformity in the arrangement of the strata. No rock should look isolated, but part and parcel of those adjoining. If this is carried out, a very different effect will be obtained from that found in those so-called rock gardens, now fortunately becoming less common, which 68 ROCK GARDENS consist of a mound of earth and a number of stones. Long, flat, spiky ones are generally chosen, and stuck up on end like so many solitary gravestones, without the least connec- tion one with the other; in another, though perhaps less objectionable, form, the stones are laid flat ; or, in yet another type, a number of large and ugly boulders are strewn over the ground, apparently for no special purpose, and ‘certainly with no sense of cohesion. In none of these types are the stones of the least use, either for keeping the soil in position, or for showing off the alpines planted there. The general result is that the owners of these gardens, despairing of ever making them “a thing of beauty ” or “‘a joy for ever,” cover them as best they can with Ferns, Sedums, Nasturtiums, or anything that can be found to quickly hide their ugli- ness. Mr Reginald Farrar, in his book My Rock Garden, so aptly designates these forms as the “‘almond pudding,” the “ dog’s grave,” and the “ devil’s lapful.” But bad as the above types are, they are yet a great advance on that form, remains of which are still occasionally seen—nhideous arches, THE ROCKWORK 69 grottos, or bridges, without a spot to hold more than a handful of soil. The constructors of such atrocities must have thought, if indeed they were guilty of thinking at all, that alpines lived on air and stones alone. For nothing else was provided. It is needless to say no self-respecting plant would grow under such conditions. The only remaining hope was that some friendly ivy would soon cover their nakedness. These, then, are some of the most usual forms to be avoided ; but there are others more or less on the same lines, all hideous and unnatural. Why they were ever made will always remain a mystery. It may be presumed that the object in view was to grow alpines in a place somewhat akin to that from which they originally came. But how was this attempted ? A home was prepared for them as diametrically the opposite in every way as human ingenuity could contrive. The result was a foregone conclusion. It is only within comparatively recent years that the more natural and reason- able method has been adopted, with the result that success in place of failure has crowned 70 ROCK GARDENS the efforts of those copying Nature more faithfully. Having now pointed out how the rocks should not be placed, there is no need to dwell further on the subject, and I will proceed to show how they may be used to the best advantage. There are a few rules that should be observed in the construction of rockwork. In the first place, every rock, big or little, should be made quite firm in the ground; as this is often rather loose owing to the bank having so recently been made, the bed for each rock should be well trampled in order to get as solid a foundation as possible and minimise its chances of moving. The soil also should be rammed in well around it; for although the rocks will get firmer in time as they settle, it is better not to rely too much on this. Each rock should be sunk for at least one- third of its height below the surface ; it should be able to bear the weight of a man without moving, for it will often be necessary to stand on the rocks when working in the garden, and if they should move, considerable, if not fatal, THE ROCKWORK 71 damage might be done to the plants growing over and around them. Another advantage also gained by sinking the rocks is that most alpines like to get their roots amongst stones. To all rock plants moisture is most essential, but at the same time it must be accompanied by sharp drainage to prevent any water lodging at the roots. For this reason the top surface of all rocks should slope down towards the bank behind, in order that the moisture may be directed towards the roots of the plants growing there. The faces of the rocks should not as a rule overhang so as to prevent the rain reaching the ground at their foot, for this would prove fatal to many kinds of plants. There are indeed some, especially those species having woolly leaves, that require some shelter over- head, for damp lying on their leaves soon kills them ; but even these require that their roots be kept moist, which can be achieved by sink- ing flat stones in such a manner as to catch the rain from the rock above and direct it to the roots of the plant. In the construction of rockwork, more 72 ROCK GARDENS especially when negotiating the bolder masses, it may often be found necessary to deal with Spaces, or, to use a more correct term, fissures, between two rocks. These are of three kinds, oblique, vertical, and horizontal. The oblique fissure looks externally very FIG. 3. much the same as the horizontal, the differ- ence being that in the case of the former it slopes down towards the back, while the latter keeps to the same plane. In both of these it is very important to make sure that the under rock projects beyond the upper, otherwise rain falling on the sloping face of the rock above will miss the fissure THE ROCKWORK 73 altogether, and whatever is planted in it will in all probability soon die; for if no moisture reaches the roots from the front, they will have to extend 12 inches or more to the back of the rock before they obtain any. This is more than most are capable of doing, certainly in the juvenile stage, in which they must be planted. It is necessary to keep the rocks slightly apart in order to give room for a good layer of soil between them, and this should be done, while building the rockwork, by inserting some pieces of hard stone. ‘The layer of soil should be filled in before the upper rock is put in position, so that no air space may be left, which it may be found very difficult to avoid later on. Vertical fissures should, when feasible, be made wider at the top than the bottom, for then, as the soil settles, its tendency will be to compress itself against the sides of the rocks, instead of leaving them, as would occur were the shape reversed. Where the fissure is sufficiently large, wedge-shaped pieces of stone should be inserted, and care should be taken 74 ROCK GARDENS to place them in such a way that their thin ends are uppermost. The object of placing the stones in this manner is to separate up the fissure in such a way that each of the divisions may form an integral part of the whole and become com- Vertical Fissure. Vertical Fissure. Wrong Method of Building. Right Method of Building. Fic. 4. pressed as the soil sinks under the action of the rain and its own weight. This indeed is the theory on which a vertical fissure is built. It is, however, in practice far from easy to carry out with any degree of success, for it will generally be found that, no matter how carefully it is made, nor how tightly the stones are rammed into posi- THE ROCKWORK 75 tion, a few winters’ frosts and rains will work them loose and a quantity of the soil be washed out, to the great detriment of whatever is growing in it. This applies chiefly to the larger fissures, for in the smaller ones the soil is less likely to be dislodged. A number of the rarer alpines, however, so dislike moisture lying about their crown that they can only be grown successfully in this type of fissure ; the chief difficulty is to get the plant well established ; but once it is established, the roots will help to keep the soil in position and the leaves protect it from being washed away by the action of the rain. The oblique and horizontal fissures are, on the other hand, quite easy to make, and prove very satisfactory. Of the two the oblique is to be preferred, as it ensures utilising all available moisture. | | On studying the face of a quarry, it will be seen that there is a certain uniform grain or strata running through it; this may vary greatly in form, and is very much more marked in some kinds of rock than in others. In limestone it is very apparent, and in a less 76 ROCK GARDENS degree in granite or sandstone. Now this grain should, in a great measure, rule the construction of the rockwork. When in its natural position the strata runs in a certain uniformity ; therefore when building the rock- work each stone should be so placed that its grain coincides with that of those next it. Right Method. FIG. 5. Examine carefully the formation of each rock, and, if its strata is vertical, the rock adjoining must be the same; if oblique, it should be built into one having a similar formation. In this case it is necessary to note that its dip should continue in the same plane, as otherwise the lines of grain would meet in a V shape (fig. 5). If, as is sometimes the THE ROCKWORK 77 case, a certain amount of “live” rock exists and it is necessary to add to it, the rocks added must observe the same “dip” or forma- tHenvas the << live” rock. If these rules be carefully followed, little if any difficulty should be experienced in building the rocks into one another. Each will, how- ever, have to be chosen with some care, in order that it may correspond both with its neighbour and also with the general formation that has been laid out ; that is to say, if it is required to form a corner, the stone chosen should have one of its faces to correspond with and be in the same line as the last previously placed stone, whilst another face will give the direction in which to continue the rockwork. Having now dwelt at sufficient length on the way in which the individual rocks should be treated, the next thing is to show how to construct the rockwork in general. As was pointed out in a previous chapter, the primary object of the rocks is to keep the soil in position and prevent it being washed down the sides of the garden by heavy rains. The most usual and effective and also the 78 ROCK GARDENS simplest method of accomplishing this is, with the help of the rocks, to make the banks into a series of terraces; at the same time the rock- work should follow the same line or dip of the strata throughout. These terraces should vary in every con- ceivable manner: height, width, outline, and gradient. This last is important, for if it is too steep the moisture will run off too quickly and not percolate the soil, and if it is too flat the ground is apt to get sodden. A fair gradient, such as 6 inches in 2 feet, is about right. It will also be necessary during the construction to make a number of pockets or compartments in order to give a variety of aspects, the advantage of which will be fully appreciated when the time for planting comes. In order to avoid getting the terraces in regular planes, let them, in places, merge into each other ; perhaps where they join a fresh one may be started on yet another level. In the long and gentle gradients the terraces will naturally be wider and not so deep, very different from the formation in the bolder and steeper parts, where each rock almost touches THE ROCKWORK 29 that below it. Keep the largest rocks for the boldest and most perpendicular formation. Where the banks slope back with a gentle gradient there should be no hard and defined line of rockwork along the top to give the impression of a ridge, but rather they should recede gradually, with the rocks becoming fewer and more scattered, until eventually the forma- tion is lost in the shrubs beyond. On the other hand, in the bold and precipitous forma- tion, the top of the rockwork may be abrupt and well defined, but even in this case the effect is more pleasing if the ground beyond rise gradually towards the shrubs forming the background. Build up the main rockwork first, keeping in view the chief object, which is to maintain the soil in position. Afterwards, if there appear to be any large extent of ground with- out rocks, it will be easy to put in a few here and there to break the monotony, or form a position for some plant requiring a special aspect or treatment. Do not fall into the rather popular error of imagining that a rock garden must be covered 80 ROCK GARDENS with rocks: it is a mistake from every point of view. The cost will be greater, for more rocks must be used, and the effect be less pleasing. The rocks should monopolise the ground in only a few places, as is the case in nature, and, as a general rule, in the wider terraces an odd one appearing here and there is quite sufficient. In the case of the large garden, where the shrubs are in the valleys instead of being on the heights, the rocks should increase in number as you ascend. They should also become larger and the construction of the rockwork bolder as the summit is approached, for this is how it is in nature. Once more, let me advise the would-be con- structor of rockwork to go to that master- builder, Nature, and study some natural rock formation ; it will prove of great assistance in carrying out the directions I have just given. As a parting word of advice, have some definite scheme and stick to it. Do not let the rockwork meander about in an aimless sort of fashion, but endeavour all through to carry out some well-defined geological forma- THE ROCKWORK » 81 tion ; if this is done, the extra trouble it entails will be more than repaid. Paths are an important and necessary factor to be considered. ‘The main paths, such as those that go through the centre of the valley or across the bank and knoll types, may be from 7 feet to 12 feet wide, and should be well made. A good surface may be obtained by : laying a foundation of about 4 inches of stones, broken fairly small, and covering it with about 2 inches of the finest riddlings from a ‘stone-crusher.¢ ‘They should be rolled well together to make them bind properly. The path should have a decided fall from one end to the other, in order to ensure efh-— cient drainage of the banks on either side. This is a point that should be carefully noted. | ae It is also very advisable to lay flat stones throughout the rockwork at convenient dis- tances apart, in order to facilitate getting at the plants and to avoid treading on them. The main paths will require something in the way of a rocky edging, if only to prevent the soil from being washed over them from 6 82 ROCK GARDENS the banks ; but anything the least formal must, of course, be most carefully avoided. In order to prevent formality the line of rocks should be varied as much as possible. In some places, especially where the construc- tion of the rockwork is abrupt, they may be as much as acouple of feet high, while in other places they need not be more than a couple of inches above the level of the path. Steps are often a necessary adjunct, and, if well made, will prove a very attractive feature. Here again there must be no formality, nothing built up with cement. Large flat stones, the more varied in size and shape the better, may be just laid on the ground, after it has been made level and solid, to form the steps. The commoner Sedums, Saxifrages, Campanulas, and other dwarf-growing species should be planted between each of the steps, so that in a short time but little of them is visible. In conclusion, when building the rockwork, make quite certain that the soil is well packed in behind and between each rock, and that there is not the least chance of an air space being left. _Myriads of plants have been lost through THE ROCKWORK 83 neglecting this apparently unimportant detail. It must be done in the building stage, for later on it may be found very difficult, if not impossible, to do it properly. Some of the higher-growing alpines would, if planted in anything like good soil, quickly succumb to over-feeding. Many of these plants are found growing in the moraines deposited by the action of the glaciers, the soil of which consists chiefly of grit and small stones, with just a trace of vegetable loam in it. There some of the rarest alpines will be found growing in a state of health never arrived at ‘in this country. For them a special place will have to be prepared, to supply as nearly as possible the conditions to which they have been accustomed. Choose some bank the drainage of which is ensured, facing south, about 3° to 4 feet above the level of the path, and having a gentle gradient of not more than 1 foot in 16 feet. Remove the soil to a depth of about 18 inches, and put. a layer, about 4 inches thick, of coarse stone in the bottom ; then fill up with small broken stones, into which grit and a little, very little, 84. ROCK GARDENS light fibrous loam have been mixed, and a moraine is made. The alternative method is to build up instead of sinking. This can be done with large blocks of stone, filling up the intervening space with the required compost. ‘Though somewhat more expensive to make, it will probably give better results. The one essen- tial need of the moraine is drainage of the most perfect description ; for this reason let it be fairly high and exposed, when it will also be less liable to suffer from ground damp. Many people might imagine that in such a spot as this plants would suffer greatly from drought during a dry summer, but this is not the case, for so well does the stone retain moisture that even in a dry season the soil a few inches below the surface remains damp. CHAPTER VI THE SOIL Importance of climate—Best type of soil—Supplying de- ficiencies in soil_Necessity of loosening ground— Drainage. Cumate has more effect on rock plants than either soil, aspect, elevation, or any other natural factor. This may not sound encourag- ing, for, whatever else we may do, we cannot control the climate, so all that remains is to adapt ourselves to, and make the best of, existing conditions. Perhaps it is just as well that it is so, for the result would indeed be curious were we able to secure the amount of sunshine or rain that each desired. Uncertain as the climate of the British Isles now is, it would then be a hundred times more so. Next to the climate, the soil, as might be expected, is of the chief importance. But 85 86 ROCK GARDENS with regard to it we can assist or adapt the natural soil to meet our needs. Leaving for the present those plants which require a special receipt for their nutriment, as a general rule the majority of alpines like a cool, deep, light, and gritty soil, rich in humus, such as a good light fibrous loam. Anything in the nature of a clay-soil is to be avoided, for in winter it is liable to get sodden, owing to its being very difficult to drain effec- tively, and in summer to bake and crack. It would be difficult to get anything better than the top 6 inches or so of an old pasture, if the soil is of a loamy description. Unless the soil is naturally very light and gravelly, sand will have to be mixed with it: good sharp river sand is the best. Small broken stones, such as would pass through a 3-inch riddle, are beneficial. Alpines like, above all things, to get their roots round and between stones ; it keeps them cool and damp, as stones retain moisture much longer than the soil and are not so easily affected by change of temperature. If the garden is of large extent and the THE SOIL $7 soil is not of a gravelly nature, spread sand and broken stones over it to the depth of about 2 inches. This, when well dug in, about 18 inches deep, ought to give a good material to work on. Some plants will re- quire an even lighter mixture, but it can be prepared for them when they are being planted. It is a fallacy, and I fear rather a popular one, to imagine that alpines do not require good soil. This may have arisen from the fact that they are often found, and in a thriv- ing condition, in places where apparently other herbage could not exist. To a great extent this is due to the deep-rooting powers of the alpines, which enable them to obtain moisture and nourishment in places where other species of a more shallow-rooted nature would soon die. | But although these alpines require and appreciate a good light soil, anything rich and heavy is very detrimental, for, if they do not damp off in wet weather, they will probably run to leaf so much that the flowers will be but sparse, and, in addition to this, the foliage 88 ROCK GARDENS will lose that hard and compact form which it should have, and which is the joy and pride of those who are successful in the cultivation of rock plants. If the soil be deficient in humus it may be supplied in the form of leaf-mould, but it should be old and well decayed. Many and various are the opinions about lime-hating and lime-loving plants, but on one point all are agreed, namely, that some plants, such as the peat-lovers, will not live if there is any lime in the soil ; this unfortunately cannot be eradicated if there naturally, but it is easy to supply artificially if required ; it is clear, therefore, that a soil free from lime is to be preferred. Lime can be supplied either in the form of old lime-rubbish or well-slaked lime ; but in order to be sure that the latter is quite “‘dead,”’ it should be dug into the ground some weeks before planting. It is recommended by some authorities, such as Mr Robinson, that separate portions of the garden should be prepared for those plants exclusively requiring lime, or peat, or so forth, in preference supplying the THE SOIL 89 necessary ingredients to the individual plant. This method, while having much to recom- mend it, has, I think, the disadvantage of rather restricting one as to the position of certain plants, irrespective of the suitability of aspect or how they will work into the general scheme of planting. On the whole it seems a better plan to supply each plant, or group of plants, with whatever they need at the time they are being put in. This, however, is entirely a matter of individual taste, so long as each plant gets what it specially requires. But it will not do to just put the soil each plant likes round its roots; there must be sufficient for them to extend through. It has been proved by experience that some plants, although found in nature growing in soils of a special character, will thrive quite satisfactorily in ordinary soil; as an example, most of the Ericas, Rhododendrons, or Azaleas, which are essentially peat-loving plants, will thrive quite satisfactorily in good fibrous loam, provided it has no trace of lime. In fact, the only exception I know of to the above rule is that those plants disliking lime will not, as go | ROCK GARDENS already mentioned, live if there is any present in the soil. There are some plants indeed, and amongst them some of the most difficult alpines to grow, that require the very poorest of soil, composed chiefly of grit and small stones, and to these anything in the nature of high feeding quickly proves fatal. For these the moraine, as described in the previous chapter, is eminently adapted, ensur- ing as it does the quick and efficient drainage so essential to maintaining them in health. Alpines, even those that are found most difficult to cultivate in this country, grow freely in their native habitat, and it is now becom- ing generally recognised that this condition of things is due, not so much to the soil as to the climate, the altitude, rainfall, temperature, length of growing season, duration of snow, humidity of the air, etc. Although the winter in high altitudes is so much longer and far more severe than is ever experienced in this country, the alpines lie dormant during it, nursing their strength and vigour under a thick blanket of snow, THE SOIL gl which keeps them dry and warm and protects them from sudden changes of temperature or extremes of cold. When the spring comes again, melting the snow, they are found ready to shoot up and adorn those arid slopes with the glory of their bloom. How very different are the conditions in this country, where the winters are made up of sudden changes, one day dry and warm, the next bitterly cold, and always the ever-present damp, which is more fatal than anything else ! Owing to the mild days the plants never become really dormant, and are, therefore, far more susceptible to cold than they would be if they had ceased growing. As, however, the climate cannot be altered, all that can be done is to protect those plants which are most affected by the damp, such as the Androsaces and other of the woolly-leaved species. ‘This protection can be obtained by placing sheets of glass over the plants during the winter months, say from November to April, which, though unattractive in appear- ance, will often save some treasure from extinction. g2 ROCK GARDENS Many plants are not only affected by damp in winter, but also may suffer from drought in the summer, and generally it will be those kinds which are found fully exposed to the sun in their native homes. This may perhaps, to a certain extent, be accounted for by the differ- ence of altitude, for there is a theory that the sun, owing to the air being so much more rarefied, does not have the same parching effect as at a lower level. The only remedy for this is watering, which in a large garden may prove rather an arduous task, but is preferable to losing the plants. It is also advisable in dry weather to keep the surface of the ground loose. It is rather a popular error to imagine that stirring the soil conduces to evaporation; the result is quite the reverse, for by loosening the surface the capillary action of the soil is encouraged and moisture is drawn up from below. A very useful tool for this purpose is the > ‘“‘ Baby Bucco,” a miniature form of that very excellent cultivator, “The Bucco,” so widely advertised in all gardening papers. If one comes to analyse it, it will be seen THE SOIL 93 how impossible it is to supply the climatic needs of a collection of plants gathered from all parts of the globe, such as is found in the modern rock garden, where such extremes as Mesembryanthemums and Androsaces grow side by side, the former almost a _ semi- tropical plant, while the latter comes from the snowline of the Alps. So it can scarcely be a matter of surprise if difficulty is experienced in growing them equally well. But be the climate what it may, there is one thing that must always be seen to, and that is drainage, to which I have so frequently referred, for damp is more fatal than anything else in this country. If the subsoil is of a very open and porous nature, few drains will be required ; but if, on the other hand, it is heavy and retentive, the drainage will need to be most thorough. In the artificial rock garden, where the subsoil during primary operations is exposed, it should be left in a slope, so that there may be no hollows for the water to lodge in. A few inches of broken stones over it will greatly assist the drainage. At the foot of the bank it 94 ROCK GARDENS is well to make a drain of loose stones which can run along at the back of the rocks bound- ing the pathway ; this will act as a main drain, and other short ones, if thought necessary, can be made to lead into it. This main drain should be made in the earlier stages and before the good soil is filled in. Although drainage is so very important, it is a mistake to imagine that it cannot be overdone, for this is quite possible. The object to be attained is to direct all moisture to the roots of the plants, and then to remove whatever the soil cannot readily absorb. If this is done the excess of moisture will not lodge in the ground and prove fatal to the plants. CHAPTER VII THE BOG GARDEN Position—Natural bog garden—Artificial bog garden— The stream. A BoG garden, on any extensive scale at least, should not be amongst the rockwork, but should rather adjoin it in such a way that the approach to the rock garden may lead through it. But marshy patches of, say, from 5 to 1o square yards in area, here and there through- out the garden, will be found an additional attraction, and give scope for a wider range of plants. : | These spots should have a variety of aspects, some in shade and others in the sun, to suit the requirements of the different plants ; they should not have any formal or well-defined outline, but should merge imperceptibly from the rockwork. To those who are 95 96 ROCK GARDENS fortunate enough to have a stream through their garden the task will be easy, for then all that is necessary to do is to remove the soil to a depth of about 3 feet, where these boggy spots are required, and fill the hollow so made with about 6 inches of coarse stones, a little charcoal, and a compost made up of about equal parts of peat and fibrous loam, with a little sand and broken stones added. If there is difficulty in procuring peat, leaf-mould will do, but peat is to be preferred. The course of the stream should then be directed in such a way that it flows through the spots so prepared ; and in order that the water may be evenly distributed, it is well to divide up the stream into several smaller channels, thereby ensuring the thorough percolation of the soil, and the maintenance of constant moisture all over these marshy spots. Sink the bog garden slightly below the level of the surrounding ground, in order that it may catch all the surface moisture from the rockwork adjoining. As many of these spots as desired can be made by directing the course of the stream from one to the other. It is, however, a THE BOG GARDEN 97 troublesome matter when the stream is so small that there is not sufficient volume of water to supply all the boggy patches with the required amount of moisture, or when it is liable to dry up in summer ; and still more - difficult when there is no stream at all to work with. In either of these cases it will be necessary to prepare these marshy beds in such a way that the moisture will be retained. To accomplish this it will be necessary to make the bottom of the bed fairly water-tight with either concrete or yellow clay. Of the two, concrete gives better and more permanent results. The directions are fairly simple. Dig out a hollow about 2 feet 6 inches deep of the required size and shape. Over the bottom of it put a layer 4 to 6 inches thick of concrete, about 6 parts of coarse sand to 1 of cement. At the lower end, about }-inch above the level of the bottom, put a pipe of 1-inch or 14-inch bore; this should lead into some drain or other means of carrying off the water, in case it at any time be thought expedient to run it off from the bog- 7 GBi ts), ROCK GARDENS bed. There will have to be a plug or tap at the lower end of the pipe, in such a position that it can be easily got at, and the end which is in the hollow should be covered with per- forated zinc to prevent it getting choked. Having fixed this pipe, build a wall of concrete, or bricks if preferred, round the sides of the hollow, about 4 to 6 inches thick, and about 1o inches high, making sure that the top of this wall is approximately on the same level all round. On the lower side, and about 8 inches from the bottom, fix a pipe of about 14-inch bore. This will act as an overflow, and also conduct the water to the next marshy spot and keep the water in the hollow at a constant level, and should have its end covered with per- forated zinc. ‘This being in position, put a layer of about }-inch thick of pure cement all over the bottom and sides to ensure their being water-tight. It is advisable to put on the cement before the concrete has set very hard, say on the following day, for it will be found to bind together better and not be liable to chip off, [ To face page 98. ‘NaGUYVS MOOU SMOHLAV: NI NUCGUVS HOF THE BOG GARDEN 99 It should then be left for several days to thoroughly set, when it can be tested by filling up with water, which should rise to within about 2 inches of the top of the wall and give a depth of about 8 inches. If all is found satis- factory, empty by means of the bottom pipe. Now put in a layer about 6 inches deep of fairly large stones, brick-bats, or any coarse rubbish, to act as drainage; over these place a layer of sods, with the grass sides next the stones. When doing this make sure that the end of the overflow pipe is quite clear and that the water can easily get to it. Having made all secure, now fill in with a compost of peat, leaf-mould, fibrous loam, sand, and broken stones in the same proportions that were given in the beginning of this chapter. If these directions have been carried out, you will have a small (underground) pond with the water kept at a constant level just over the sods covering the drainage. The capillary attraction of the soil will draw the moisture up and keep it constantly damp. If the water to supply _these marshy beds has to be brought in pipes, they should have a tap, so that the flow may 100 ROCK GARDENS be regulated. It is scarcely necessary to mention that every trace of cement and all pipes or taps should be hidden from view by either soil, stones, or plants. When circumstances will permit, it is advisable to have a constant flow of water through these beds; and although this be but a mere dribble, say 10 to 15 gallons per diem, it will be sufficient to keep the water from becoming sour or stagnant. If the supply pipe is arranged so that it reaches the bottom of the bed, it will ensure the water being constantly changed. If it is desired to make the bottom of the bed of yellow clay, the same directions will have to be followed, except that it will be necessary to puddle in about 6 inches of yellow clay instead of the concrete; but neither so water-tight nor lasting a result may be expected. If it is wished to make a bog garden on a scale larger than that just described, and the water supply for it is limited, a concrete basin should be made, following the directions already given. Owing, however, to the expense and the difficulty of getting the levels correct, it is THE BOG GARDEN IOI not advisable to attempt a bog garden of this description of over 150 or 200 square yards in area. But where there is a stream, provided the volume of water is sufficient, there is nothing to limit the size of the garden. It will be necessary, of course, to spread the water from the stream over the bog garden in such a way that it can reach all parts; and as in all proba- bility the proposed site is not level, the simplest and most efficacious method is to cut small channels running from the stream on either side, in the way the farmer irrigates his fields. This method has also the advantage that the moisture can be regulated at will. Another plan, if the channels are considered unsightly, is to lay pipes with a number of holes bored in them a couple of inches below the surface of the ground and connected with the stream. The great disadvantage of this method is that the pipes are so liable to become choked with mud or leaves. The surface of the bog garden should not be flat; undulating or, more correctly speaking, uneven ground is much to be preferred. This 102 ROCK GARDENS will give different degrees of moisture, which will be found most useful, for all plants do not need the same amount. In the hollows can be planted those kinds that require a wet soil, while the higher-lying ground can be reserved for those preferring a drier position. If desired, a pool here and there can be made, and a very charming and attractive addition to the garden these will be, if judiciously placed. They can be made of either concrete or yellow clay. In either case it will be necessary to edge these patches of water with stones or plants, in order to avoid the possibility of any formal outline, and, needless to say, all concrete must be carefully hidden from view. Should the garden be of any considerable size, it will not be feasible to fill it up with any specially prepared compost, such as was recom- mended for the smaller bog beds ; nor indeed is it necessary, for the majority of bog plants, and certainly the coarser-growing ones, are quite happy in any cool, rich, and moist soil ; but in places the ground should be prepared for those which are more delicate or fastidious in their tastes. These spots may be of peat or other THE BOG GARDEN 103 compost, according to the requirements of the future inhabitants. As the soil of the bog garden in general needs to be richer and heavier and to contain more vegetable matter than does that of the rock garden proper, some old manure, leaf-mould, and peat can be dug into it with advantage, after the surface has been thoroughly cleaned. The reader should note most particularly the necessity of getting the ground perfectly clean, for nowhere do weeds grow so rampant or increase and multiply with the same extra- ordinary rapidity as they do in the bog garden. If the ground be not clean and free from weeds to start with, the rightful denizens will have but a poor chance in the struggle for existence, so let no trouble be spared nor time grudged in the preparation of the garden, for if properly done the extra trouble expended will be more than repaid. A few stones half buried may be placed in the bog garden, more especially on the higher ~ and drier spots, for many bog plants like to get their roots under and around them, just as do their neighbours in the rock garden. 104 ROCK GARDENS Whatever paths are required should be made of flat stones in the form of stepping stones. The bog garden should be made in such a position that part of it is exposed to the full sun, while part is in shade, to suit the require- ments of the different plants. This can be done, either by planting the edge with bamboos or similar suitable and fairly tall-growing plants, or by making part of the garden in the shade of some higher-lying ground. As has been mentioned earlier in this chapter, a bog garden of any size should not appear amongst the rockwork, but may adjoin it. A very suitable place is at the foot of the rock- work in the hollow or bank types. Or, in the case of the valley, the approach to the main walk may be through the bog garden. These are merely suggestions, and individual taste and skill will need to be exercised as to where the bog garden would look most natural and work best into the general scheme of the rock garden. Another point that must be taken into con- sideration before fixing on the site is the water THE BOG GARDEN 105 supply, which should be brought by the easiest and most direct route possible, and care must be taken that the overflow can be readily disposed of. Fortunate indeed are those people who have a stream, no matter how small, run- ning through their garden, or are able to conduct one to it. My advice is to have one, if it is at all feasible, for it is the means of adding enormously to the charm of the garden and of considerably extending the range of plants possible to cultivate. If the rock garden is of either the hollow or hill types, the stream may be brought down the face of the rock- work in a series of little cascades into a pool at the foot. The ground on either side should slope gently towards its course. The banks of the stream, at least so much of them as are affected by the water, should be made of fairly large rocks to prevent its en- croaching or washing them away. ‘These rocks, it is scarcely necessary to say, should not be placed in the form of a straight wall; the stream should be made to wind about, for it will then look so much better. The bottom of the stream should be paved with flat stones, 106 ROCK GARDENS otherwise the soil will be liable to be washed down, especially in time of heavy rains. If the rock garden is of the valley type, the stream can be brought beside the central path, first on one side and then on the other, in order to provide places for planting, both in sun and also in shade. The stream should wind about considerably, and not follow the exact line of the path. The possibilities of variety are endless, though they must depend greatly on the circumstances of each case. While a stream is a good servant, it is a bad master. If provision is not made, great damage may be done by it in time of flood. It should be possible, therefore, to control the volume of water in some way- This can be simply done. Build a wall across the stream at the point where it enters the garden, and in this wall make a hole of such a size that no more than the maximum amount of water conducive to the safety of the garden can come through at a time. As this hole will in all probability not be large enough to carry off all the water when the stream is in flood, a cutting should be made to take away THE BOG GARDEN 107 the overflow and prevent the risk of flooding the rock garden. The bottom of this control- ling hole should be made a foot or more above the bed of the stream, in such a way that the wall below this hole will act as a dam. The advantage of constructing it in this way is that the sand and mud washed down by heavy rains will be checked at the wall and not carried down into the garden. This accumulation of sand and mud must be cleared out periodically, otherwise the dam will cease to be of any benefit. CHAPTER VIII THE WILD AND WATER GARDENS The wild garden—T reatment—List of plants suitable— The water garden—T reatment—List of plants suitable. Tue wild garden has frequently been re- ferred to in previous chapters, so a few words © upon how best to deal with it may be found useful. As may be supposed, there is no mak- ing in a garden of this description ; all that is needed is to plant with discretion and judg- ment. That the secret of obtaining effect is to plant in bold masses has already been pointed out; and important as this is in the artificial garden, it is many times more so in the case of the wild garden. Whereas in the former type the available space is more or less limited, in the latter case the ground is not so curtailed ; therefore, in order to be in keeping with the surround- 108 THE WILD AND WATER GARDENS 109 ings, the planting should be done on very bold lines. It is a great mistake to imagine that only so-called wild-flowers are suitable for the wild garden; many of those species now found in the herbaceous border or enclosed garden would not only look better, but also grow more luxuriantly when introduced into wilder surroundings. Another advantage to be derived from a garden of this sort is that one is enabled to cultivate in it a number of lovely plants which, on account of either their undue vigour or encroaching habit, cannot be associated with choicer and more delicate species. For these the wild garden is the fitting home, in which the full value of their beauty can be obtained. In short, any species of plant can be used, so long as it is strong enough to take care of itself without the attention of the gardener. Every spot should be planted. The ground under trees may be carpeted with the lovely Wood Anemones, Trilliums, Cypripediums, Cyclamens, Snowdrops, Dielytras, Ferns, and a host of others ; while by the margins of streams or in damp meadows the Caltha, Trollius, many IIO ROCK GARDENS species of Primula, Ranunculus, etc., should find a congenial home, and in any grassy spot may be planted Narcissus, Daffodils, Crocus, and other bulbous plants. The Genistas, Cistus, Helianthemums can be used to cover dry and stony banks. In short, whether it be under trees or in the open glade, on margins of a stream or on sunburnt slopes, there is no spot that should not be fully planted. When the approach to the rock garden is through an open glade or fairly exposed bank, no more suitable plants can be found than those of the Heath family, which should be grown in bold masses, and will indeed make an ideal setting for a home of alpine treasures. The best kinds for this purpose are Erica arborea, Codonoides mediterranea, which grow from 3 to 8 feet high. The dwarfer kinds are Erica carnea, carnea alba, ciliaris, cinerea, tetralix, vagans, vagans alba, vulgaris. The large rocks and trees may be covered with many kinds of Roses, the numerous hybrids of the wichuriana type, the Polyantha and Wild Roses, also Clematis, Vitis, Lonicera (Honeysuckle), and such-like. Whenrearrang- [Zo face page 110. WATER-LILY POND, MOUNT USHER. (Photo. by Mzs Delves Broughton.) THE WILD AND WATER GARDENS 111 ing the herbaceous border in the spring or autumn and dividing up the clumps of plants, instead of assigning the pieces taken off to the rubbish-heap, let them be planted in the adjacent woodland, which will shortly and with no expense be converted into a garden that will vie with, if not excel in beauty its more formal prototype. ‘The following is a selection of plants specially adapted for the wild garden :— Achillea Eupatorium. A. millefolium rosea. A. Ptarmica. Aconitums, in variety. Anemone apennina. A. blanda. A. nemorosa. A. japonica, in liek A. sylvestris. Aquilegias (Columbines). Asphodelus luteus. Asters, in variety ; any of the taller-growing species are excellent. Astilbe, in variety. Caltha. Campanula celtidifolia. 112 ROCK GARDENS C. persiciflora Moorheimi. - Cardamine trifolia. Crocus, Cypripedium spectabile and calceolus. Delphiniums, Daffodils. Dictamnus, in variety. Dielytra spectabilis. Digitalis (Foxglove). Doronicum excelsium. Dracocephalum virginicum. Echinacea purpurea. Echinops exaltatus and giganteus. Epilobium angustifolium, and other varieties. Epimedium, in variety. Erodium Manescavi, and other varieties. Eryngium giganteum, planum, and other varieties. Euphorbia polychroma and Lathyrus. Geranium, armenium, sanguineum. Heleniums, in variety. Helianthus: Miss Mellish, giganteus, multi- florus, and others. Helleborus (Christmas Rose). Iris, in variety. Lobelia cardinalis, Queen Victoria, and others. Mimulus. Monarda. Montbretia, in variety. Myosotis (Forget-me-not) palustris, sylvatica. THE WILD AND WATER GARDENS 113 CEnothera Lamarckiana, macrocarpa and speciosa. Papaver orientale, in variety. Pceonies, in variety ; these look very well planted in grass. Polemonium, in variety. Polygonatum (Solomon’s Seal). Primula cashmiriana, Japonica, in variety ; sikkimensis. Ranunculus aconitifolius, speciosus. Sanguinaria canadensis. Saxifraga peltata and megasea. Senecio, in variety. Spireas in variety are excellent subjects for shady or damp positions. Symphytum caucasicum, Bohemicum. _ Telekia speciosa. Tradescantia. Trillium grandiflorum. Trollius, in variety. _ Veratrum album and luteo-virides. Verbascum olympicum (Mullein), and other varieties. THe WaTER GARDEN | Fortunate indeed are those people who possess some pond or lake on the margins 8 114 ROCK GARDENS of which they can carry out some water gardening. These margins should be always left in their natural state, provided, of course, the pond has not been made artificially, in which case it may be necessary to break the outline, for anything formal should be carefully guarded against. It will be necessary to have access to the plants which are grown on the margins, and as these latter, in all probability, are soft and boggy for some distance from the water- line, some sort of path will have to be made. Where a firm bottom can be found, this may be done by laying down some rough ballast, ramming it well in and then placing flat stones over it. But where the ground is too soft, a good foundation can be obtained by laying brushwood tied in bundles or by driving piles into the ground. It is scarcely necessary to emphasise the fact that these paths must be in no way formal ; they should meander through rushes, sedges, and other plants growing on the margins, | As before, plant in bold masses to get the [ To face page 114. (‘uoyysnorg saajaq say Aq ‘NAGUVS WHLVM ‘ojo d ) mare : Phase (rae roa - — “. — THE WILD AND WATER GARDENS 115 best effect. The following are a few of the best species for this purpose :— Acorus calamus and gramineus variegatus. Alisma natans, and plantago; will grow in shallow water. Aponogeton distachyon ; will grow in shallow water. Brasenia peltata; will grow in_ shallow water. Calla palustris. These Arum Lilies can be grown in water 3 to 4 feet deep, but are not very hardy. Caltha, in variety ; boggy, moist ground or shallow water. Cyperus longus; shallow water or boggy ground. Gunnera manicata ; grows very large, moist ground. Gunnera scabra ; moist ground. ‘Heracleum giganteum and mantegazzianeum ; for wet margins. | Hottonia palustris (Water Violet) ; for shallow water. Hydrocharis Morsus-rane. Iris Pseudacorus, sibirica ; levigata. Menyanthes trifoliata (Bog Bean). Nymphia (Water Lilies). 116 ROCK GARDENS Polygonum amphibium and Sieboldi; sacha- linense. Pontederia cordata. Sagittaria japonica, fl. pl., sagittifolia variabilis. Thalictrum flavum. Bamboos, in variety. Zizania latifolia. CHAPTER IX THE WALL GARDEN Position and T ypes—Building—Beautifying the old wall. Wuere shall I make my wall garden? This may appear a difficult problem at first, but perhaps the answer may be easy. Make it just where it is most needed. Wherever there has been levelling for lawn or garden, banks will be found of, perhaps, some height. These may not have been laid in grass, owing to the trouble and expense of keeping them constantly mown, and they may be easily clothed in a way more attractive than the usual carpeting of ivy or periwinkle. This is just the place for a wall garden, and if it be made here a double purpose will be achieved, for the wall will give a support to the bank, and in a short time, when covered with masses 117 118 ROCK GARDENS of rock plants, will prove a very attractive addition to the landscape. But it is not only as a support, or retaining wall, as it might more correctly be called, that a wall garden can be made ; for by build- ing what is known as a dry wall (which is in fact a wall without mortar), a very effective division can be made in a garden or elsewhere, which can soon be turned into a home for rock and alpine plants. Nor is it only in walls built specially for the purpose of growing this type of plant that they can be cultivated. Many old walls—the older the better—which are now but an eyesore, can with very little care and trouble be converted into a veritable garden that for many months of the year will be a blaze of colour. Unlike most other gardens, these, when once properly established, will practically look after themselves and require but little further attention. It will surprise many people that not only will they be able to grow a great variety of plants in the wall garden, but they will be able to cultivate there, with comparative ease, some kinds acknowledged to be amongst the THE WALL GARDEN 11g most difficult to keep in the rock garden proper. Nor is the reason difficult to find, for in the wall there is the perfect drainage, and there is not the same risk of loss from excessive damp, which more often than anything else proves fatal to alpines in this country. The chief and only difficulty con- nected with wall gardening is that of establish- ing the plants. If planted after the wall is finished, they can only be very small and therefore young; for they will have to be inserted in the narrow crevices between the stones, often no easy task to accomplish, and, having to be so young, they are more liable to be affected by the vicissitudes of temperature than would the older and _ better-matured plants. For these reasons it will be found _ advisable, when at all practicable, to plant as the building proceeds. Larger and stronger plants can then be used, and their roots can be properly spread out and encouraged to pene- trate into the cooler mass of soil behind the stones forming the wall. This will render them far less liable to suffer from drought. Although this method of building the retain- 120 ROCK GARDENS ing wall is very similar to that used in the construction of the division wall, I think it would. be better to deal with each separately, and afterwards I will show what can be done to metamorphose the old wall. The main principle underlying the construc- tion of retaining walls made for the purpose of growing rock plants is to place the stones in such a way that all available moisture be directed back towards their roots. For this reason, and also because additional strength will be obtained, the walls should not be built perpendicular, but should rather slope back from the foot at say an angle of 60 degrees. Fairly flat stones may be used, and it will simplify the building if they be of much the same thickness, for then the ‘‘ courses” will work in evenly. These stones should be firmly fixed and should all slope down towards the back. The lower stone should slightly protrude beyond that immediately above it, in order to catch the rain falling from it. No mortar should be used, but soil only, and not very much of that—just a sprinkling between each course of from 4 to 1 inch thick; if [Zo face page 120. (-uojy8norg saataq s1yq Aq *oj0y) WHHSA LNOOW ‘NAGUYVS TIVM AO Luvd THE WALL GARDEN 121 more soil were used it would be liable to be washed out by heavy rain. Pieces of slate may be inserted between the stones, and provide a little room to plant in, and also to take off undue pressure on the plants. As has been said earlier in this chapter, it is advisable, when possible, to insert the plants as the building proceeds, for larger plants can thereby be used and their roots properly spread out, and also the chances of leaving vacant spaces behind them, which is so often the cause of plants unaccountably dying, will be reduced to a minimum. The advantages of planting at this stage can only be fully realised by those who have spent many and weary hours, both trying to the temper and not unfrequently most painful and damaging to the skin of the knuckles, endeavouring to insert the delicate and obstinate roots of alpines into chinks where they did not wish to go. So, having laid a course of stones, put a sprinkling of soil over them ; then the plants, spreading the roots out well; then a little more soil, making sure that all the crevices are well 122 ROCK GARDENS filled, and on top another course of stones, and so on till the desired height is reached. Where a dry wall, which is also to act as a dividing wall, is required, it must be built hollow in the centre, to allow for the soil which is needed for the plants AN a’ 4 ! “i ; / ! ! ) ! 1 aA, BT Sie 2 re = ey a > a or t ’ ’ 1 «=, ARS e U ear ee : ; i r) 1 7 t Y 5 aa ts i feet eis ae ~> Fic, 6 Both sides of the wall should slope back from the foot at an angle of 80 degrees, and there should be at least a foot of soil in the centre. A wall of say 6 feet high and 2 feet thick at the top would require a base of 4 feet 6 inches wide, allowing for the stones being about 12 inches long (fig. 6). This wall THE WALL GARDEN 123 should be built in exactly the same way as described for the retaining wall, and the stones must slope back so as to catch the rain. It will be found a good plan if about 6 inches of stones are laid in the bottom of the wall to act as a drain, and this will also reduce the chances of heavy rain affecting its stability. In this case also it is advisable, when cir- cumstances permit, that the planting be done | simultaneously with the building. The soil in the centre should be filled in as the stone- work progresses, and should be well rammed home and made solid. We now come to the old wall which requires beautifying ; and the older it is the - better, so long as there is no danger of its collapsing and thereby burying our treasures or ourselves in its ruins. The first thing to do is to open up as many holes in its face as is consistent with its stability. In the bottom of these holes it is a good plan to place pieces of slate or flat stones, tipped up in front and slightly protruding, say about 4 to 1 inch, in order to catch the rain ; 124 ROCK GARDENS these may be, if necessary, made secure by a little cement or mortar, which, however, should not be visible. Into the holes thus prepared a compost made up of finely sifted loam, leaf- mould, and a little sand should be packed, every care being taken that all nooks and crannies are well filled and that no empty spaces are left. The top of the wall will most probably be covered with grass and weeds. This should all be carefully removed, and then as many pockets as possible made, using, as described above, pieces of slate or stone to keep the soil in position. It is well, when making the pockets in the face of the wall, to vary their position as much as possible. By this I mean avoid having the protruding stones immediately one above the other, for the rain will be pre- vented from reaching the lower one, a state of things that should be carefully avoided. One of the chief difficulties to contend with in wall gardening is the liability of the plants to suffer from drought when young and tender and be- fore their roots have had time to work back into the cooler regions behind. As the space THE WALL GARDEN 125 available for planting is generally so small and cramped that it is necessary to use very tiny plants, which will often be found difficult to establish, it is advisable to use seed. There is then a better chance of the seedlings establish- ing themselves, for they will not at any time suffer from disturbance. One method which will be found particularly useful when dealing _ with some very tiny crevice is to mix the seed and soil into a sort of paste, which can be worked into the required spot. As the aspect of the wall must unavoidably be governed by the circumstances of the case, all that can be done is to adapt the plants to suit the situation ; if in shade, plant with those preferring a shady home, and vice versa. A list will be found in Part II. giving the names of plants best suited either for sun or shade. CHAPTER X PROPAGATION Propagation by seed—By division and cuttings—Methods of raising from seed — Methods of propagation by division—By cuttings—By layers. Many alpines when in cultivation enjoy at the most but a short life; in some cases not more than five to eight years. Whether it is the same in nature is for obvious reasons difficult to say, but the fact remains that plants will have to be frequently renewed. It is therefore advisable to keep a stock at hand to meet the deficiencies. There are four ways of propagating alpines, namely, by seed, division, cuttings, and layers. Reproduction from seed is in some ways the best. It ensures healthy and vigorous plants, which, being more acclimatised, are therefore 126 PROPAGATION 127 more likely to thrive than would be the case if propagated by any of the other methods. Seed also provides the only means of raising a new variety and of improving a freshly im- ported strain. But, on the other hand, unless artificially fertilised, the results from seeds are often disappointing, for, in place of being an improvement, they frequently turn out to be only very inferior copies of the parent plant. This is due to cross-fertilisation, especially noticeable in the case of Aubrietias, Aquilegias, and many of the Saxifraga family. Neverthe- less, many species do come quite true, a great deal, of course, depending on whether there are varieties of the same species grown sufh- ciently close for bees to carry their pollen from one to the other. Another disadvantage in raising plants from seed is that they often take two or even three years before they flower ; indeed, the seeds of some kinds take as long as eighteen months, or even more, to germinate at all. During this time the seed-pans need to be carefully looked after and kept free from | weeds and that arch-enemy, Marchantia; to say nothing of the trouble and difficulty of 128 ROCK GARDENS judicious watering, neither letting soil get too dry nor keeping it too moist. Propagation by division or cuttings, when possible, is on the whole preferable, for the plants are then sure to be true and reach maturity much sooner. As this method, how- ever, can only apply when one either possesses or has access to a plant from which to take the pieces or cuttings, and when it is possible to pot them up soon after being removed from the parent plant, it would not be feasible when collecting abroad, in which case try to get seeds and raise plants from them. There are some kinds, indeed, though not many, which will neither admit of division nor grow readily from cuttings: for example, many of the Aquilegias family ; these can therefore only be raised from seed. In grow- ing alpines from seed, the secret of success lies in procuring fresh seed that will germinate readily, and also yield a much larger per- centage of plants. In order to ensure its being fresh, when possible save the seed your- self. Therefore look out carefully for seed, especially in the case of the rarer kinds. Often PROPAGATION 129 the first intimation that is given of fertile seed having ripened is the finding of small seedlings round the parent plant. It isa great mistake to be in a hurry to remove the flower- heads as soon as they are withered, and, more- over, the seed-pods are in many cases nearly as ornamental as the flowers themselves. At the same time it is well to exercise discretion, for the seedlings from some plants may prove as troublesome as any weed. Having gathered the ripe seed, sow at once. Nature sows her seed as soon as it is ripe; why should we imagine we know better and lay it aside for several months? Seeds may be sown in the open or in pots, but many species of plants do not ripen their seeds till late in the autumn, and these should be sown in pots and wintered in a frame or cold-house. In any case I should recommend this treatment for any of the rarer species. The best time to sow in the open is in April, but seeds may be put in any time from April to August. So, should seeds come to hand during the summer months, it is better to sow at once rather than to keep them over for 9 130 ROCK GARDENS the following spring, for then a whole season’s growth is gained. Even if sown as late as the beginning of August, the seedling ought to be strong enough to take care of itself during the winter. Yet it is not advisable to sow in the open so late in the season except in the case of seed from plants of a hardy nature, which germinates quickly. The seed-bed, which should not be more than 4 feet wide, will require to be carefully prepared, in some warm but sheltered position. Dig the soil well to a depth of 8 inches, making it as friable as possible; clean the ground well, removing all weeds and large stones, and, if not naturally light, add plenty of sand and leaf-mould. Over this put a layer about 4 inches thick of a good light compost made up of loam, leaf-mould, sand, and refuse from the potting bench. This mixture should be all put through a 4- inch riddle. Now level the surface of the bed, which will be ready for the seeds. Choose a day for sowing in calm, mild, and open weather, when the ground is rather dry and in good friable condition, Make PROPAGATION 131 little drills across the bed about 6 inches apart. ‘These should not be made with a hoe or similar tool, but get a piece of straight stick about 14 inch wide, and, laying it on the ground, press it gently down till it leaves an even and smooth hollow about an inch deep. This will give a good firm bed for the seeds. Therein sow the seeds as evenly and thinly as possible, and cover sparingly with some good light soil put through a ;4,-inch riddle. In many cases the merest sprinkling will suffice, for seeds should only be covered to a depth about equal to their diameter. Nothing more need be done except to keep down weeds and see that the beds do not get too dry; but watering will have to be done very carefully and with a very fine rose, otherwise the ground will become baked and hard. Watering should be limited as much as possible, and, in order to pre- vent undue evaporation, the beds should be protected from hot sun. It is a good plan to put a little cocoanut fibre between the drills, which will help to keep the 132 ROCK GARDENS ground damp and lessen the chances of its getting baked. The seedlings may be left where sown until they are fit to transplant to their final habitat. In the stronger- growing kinds, if they come up too quickly, a little thinning can be done with advantage ; but care should be taken not to disturb the roots of those that are to remain. In this way many of the best rock plants may be raised. They have this advantage over those sown in pots, that they are less liable to suffer from the vicissitudes of temperature, nor do they receive any check when being pricked out. But do not imagine that your bed of seedlings is free from all dangers, for sooner or later hungry slugs will find them out during their midnight forays, and then woe betide them, for untold — havoc will be wrought before you visit in the morning what was the night before the pride and joy of your heart. There is one thing, however, no slug or snail will cross, and that is zinc. . So -get a sheet//otiaime and cut it into strips about 3 inches wide, and after the seeds are sown put this all PROPAGATION 133 round the bed, thereby making a rampart which, from the slug’s point of view at least, is impregnable. I never sow, nor would I recommend other people to risk sowing, the rarest or finest seeds in the open air when they come to hand at a time unsuitable for sowing out of doors, such as the late autumn. But as they should be sown at once, it is advisable to use pots which can be put in a frame or cold-house. In this treatment there are two great difficulties to contend with, namely, the growth of Marchantia, and keeping the pots at an even degree of moisture. With regard to the first, there are several ways of preventing its growth. ‘The best and surest means is to bake the soil thoroughly and use nothing but boiled water. This is an absolute preventative, but entails rather more trouble than most people would care for. Another way is, to prepare the pots for sowing, and then to water them thoroughly with boiling water—but it must be boing 134 ROCK GARDENS -—and then leave them for a day or more before using. This will destroy most of the seeds and spores of weeds and kill all worms and insects, which is no little advantage. For subsequent waterings use nothing but boiled water. If neither of these plans is adopted, the only course left is to pick off every particle of the Marchantia as soon as it appears, which is by no means easy to do without disturbing the seed or young plants, and will entail considerable labour if there are a large number of pots to look after. : It is most essential to keep the soil at the proper degree of moisture, which is not at all easy to accomplish, and will require constant attention. Nothing proves more fatal to seeds than to allow the soil to become dust-dry ; while, on the other hand, if it is kept too moist, the seeds are very liable to rot away. Above all, avoid alternate conditions of wet and dry. For, germina- tion having once commenced, if the soil is subsequently allowed to dry up, the seeds will be irrevocably destroyed. PROPAGATION 135 Watering overhead is all right in the case of seeds that germinate fairly quickly, say in a month or two; but not when, as is the case with some, it will take twelve or even eighteen months before any signs of life are visible. Under such circumstances it is very difficult to prevent the surface getting hard and baked, when the pots will not “take the water.” So, for all seeds which are slow in germinating, I should advise not watering from overhead, but supplying the moisture from below. This may be done by placing the pots in a saucer of water, when the capil- lary attraction of the soil will draw it up, and if there is plenty of drainage in the pots, as there should always be, there is no danger of the soil becoming water-logged. Personally, I am trying another method, which so far seems quite satisfactory, and will, I hope, still further reduce the attention required. As perhaps my readers may care to try it for themselves, I will explain my system. Into one of my small two-light frames, 6 feet by 4 feet, I put a wooden partition down the centre from back to front. In one of these 136 ROCK GARDENS divisions I fixed a lead trough or tray, com- pletely filling the bottom. This tray, which is filled with water, is therefore 4 feet by 3 feet, and about 4 inches deep. An inch pipe connected with the tray goes through a hole cut in the back of the frame, and protrudes about 24 inches beyond the woodwork. By means of this pipe it can be seen at a glance what amount of water is in the trough, and, through it, more can be added if necessary. When I had ascertained that the trough was quite level, which can easily be done by filling with water, I placed perforated zinc all over it, about 2 inches from the bottom. The zinc I kept in position by resting it on some flat tiles about 2 inches thick. Over this I put small stones to a depth of about half an inch, and then about 8 inches of granulated peat. If this description has been carefully followed, it will be seen that about 1 inch of the peat is in water. By this means moisture will be drawn up through the whole mass, which will be kept constantly damp, and all that will be necessary to maintain this condition is occasionally to look at the pipe which protrudes outside the PROPAGATION U7 back of the frame, and, when required, to add water to keep it up to the desired level. Pots containing seeds should be plunged up to the brim in the peat, which will keep them sufficiently damp, and also tend to preserve them at an even temperature. By this device the danger of the pots becoming too dry is reduced to a minimum, and as the -moisture is merely absorbed from the outside, there is no chance of their becoming water- logged. A few minutes two or three times a week is all the attention they will require. Much time may be gained, especially in the case of the slower germinating seeds, by making, in a pit or frame, a very gentle hot- bed, and either sinking in the bed the pots in which the seeds are sown, or else covering it with about 4 inches of fine soil and sowing the seeds on that. Personally, I think the chief objection to this method is that the seed- lings will require to be very carefully treated until they are sufficiently hardened off. There is also the danger of the hotbed cooling down before the seedlings are strong enough to look after themselves. I therefore much prefer the 138 ROCK GARDENS cold frame, for although its action is slower, better and stronger seedlings are thereby obtained. Some people also recommend soaking the seeds for a short time in warm water before sowing. This treatment rather tends to hasten germination, as it softens the outer covering of the seed, which with some is so very hard that it would take, in the ordinary course, a long time before it becomes affected by the moisture of the soil. There is a right and a wrong way of doing most things, and, although it may be thought dificult to make a mistake in sowing seed, it is by no means as simple a process as many people imagine. With the seeds at hand, choose out the requisite number of pots of 5 or 6 inches, according to the amount to be sown, a separate pot for each kind. Into these put the drainage of broken crocks, filling to about a quarter the depth of the pot, and over this put a layer of moss to prevent the soil getting into and choking up the drainage. | Next prepare the compost, which will of course vary according to what is to be sown, PROPAGATION 139 but for general purposes it may be made up of good light fibrous loam, mixed with a fair amount of sharp sand and leaf-mould in the proportion of about three to one. Put this through a ;!;-inch riddle, keeping a sharp look-out for worms, which should be promptly removed. Put the coarse part remaining in the sieve over the drainage in the pots, filling them half full. Then fill up the re- mainder with the fine soil. Now press all down firmly, either with a piece of wood or the bottom of a pot, which will answer just as well, until the surface is about half an inch below the edge of the pot and fairly smooth. This will leave things ready for sowing. Take the seed packet and open carefully, either over the pot or a sheet of paper, if the seeds are very small, and then sow as evenly and thinly as possible. To cover, hold the riddle in one hand at an elevation of about 6 inches over the pot, and on it scatter the soil, keeping the riddle gently moving all the time. By this method the soil will be spread evenly over the surface of the pot. The depth of covering is generally supposed to be equal to 140 ROCK GARDENS the diameter of the seeds ; but these are often so very small, that a better way is just to cover till no seeds are visible on the surface. In some cases the merest dusting will suffice. Over this I find a thin layer of fine silver sand very beneficial in keeping the surface clean and free from moss, which, even if it does grow, can easily be removed without dis- turbing the seeds. Special mixtures of soil will be required for some seeds. For those that like lime, the Saxi- frage family, for example, and especially the encrusted section, it can be added in the form of lime rubbish, and less loam and leaf-mould used ; whilst for those that like peat, such as the Ramondias, Rhododendron, etc., the compost should consist of fibrous loam, peat, and leaf- mould, with a little sharp sand. In the case of the Androsaces, sandstone broken up quite small will be found of service. In short, whatever soil the mature plant does best in should be used for its seeds. ‘The soil for either potting or sowing seeds should not be so dry that it will not take water readily, nor yet so wet that it is liable to cake or stick to the riddle; PROPAGATION 141 it should just be sufficiently damp to work easily. Having sown the seeds, label carefully, and then cover the pot with a piece of glass to prevent undue evaporation. This should not be removed till the seedlings are well up. Now place in a cold frame, close to the glass, and see that each pot is as level as possible. _ The frames should be kept shaded from the direct rays of the sun, whilst admitting as much light as possible. ‘Try to keep the tempera- ture as even as can be. Before the seedlings appear, the frame may be kept closed ; but when they are up more air will be required, or they will be liable to damp off. As some seeds take such a very long time to germinate, do not be in a hurry to throw them away, even should they not appear within twelve months. Presuming the seeds are sown in March and nothing appears up to the autumn, they should be surfaced with a light dusting of soil and kept over the winter, being watered sparingly during the winter months. Except in the case of seeds that are known to germinate quickly, I should 142 ROCK GARDENS not despair of their coming up, even had they been sown eighteen months; but after that they may be thrown away, for the chances of success are by then almost nil. If there is no frame available, the seeds should be sown rather later, say about the beginning of April, and the pots plunged in sand or cinders in a warm and sheltered corner, protected from the hot sun and heavy rains. But I strongly recommend a frame when- ever possible, for the temperature and moisture can then be so much more easily regulated and kept even. After the seedlings are up and strong enough to handle easily, they may be pricked off into 4-inch pots, or, still better, into shallow boxes about 5 inches deep. After pricking out, the pots or boxes should be kept protected from the sun for a few days until the seedlings have recovered from the move. As frost has great effect on pots and is very liable to loosen the roots of plants in them, some protection will be required during the winter months, although this may be obviated to a certain extent by plunging into a bed of sand or fine cinders. ~ PROPAGATION 143 The seedlings may remain in these boxes or pots till the following spring, by which time they should have made good roots and may be transferred to their final home, or, if required for stock, be potted up into 24-inch thumb-pots, which should be plunged in the sand or cinder bed. Another method which may be adopted in the case of stronger-growing plants is to prick out direct from the seed-pot into a bed specially prepared with fine soil. Plants so treated will also require protection for a few days from hot sun. If the instructions just given are carefully followed there is every reason to expect that the result will be a success. But, to give it every chance, it is absolutely essential to obtain fresh and fertile seed, to provide ample drainage in seed-pots, keep them clean and free from all weeds and moss, and, above all, to keep the moisture even, for nothing proves more fatal to germination than alternate conditions of moisture and drought. _ Most alpine and rock plants lend themselves very readily to propagation by division, espe- cially that large and typically alpine family 144 ROCK GARDENS the Saxifrages, as do also the Androsaces, Primulas, and a host of others too numerous to mention. As an example of the method to be adopted, let it be assumed it is wished to increase the stock of Saxifraga apiculatum. Remove from the parent plant some of the shoots next the ground, breaking them off as close to the main stem as possible. A careful examination will disclose tiny rootlets attached to every piece so removed, each of which, when potted up in suitable soil, will grow readily. In a similar way any number of rooted pieces may be obtained from the mossy Saxifrages, without in the least spoiling the appearance of the plant from which they are taken. In deal- ing with the rarer kinds of Saxifrages, especially those of the encrusted section, such as diapen- stoides, Burseriana, Boydi, etc., more care will have to be exercised, and it is advisable, before removing a portion’of the plant, to see if it is rooted. With some Saxifrages, such as “#ifur- cata, Whitlavu, roots will be found springing from joints along the creeping stem, much in the same way as with strawberry runners. In PROPAGATION 145 the case of §. Cotyledon and its varieties, hybrids of S. J/ongifolia, S. Griesbachi, and others of a like habit, the rosette from which the flower-spike springs dies after flowering, but round it will appear a number of offshoots, each of which will have developed roots by the autumn. They can easily be separated from the main stem, and will have formed sturdy plants by the following year. The Androsaces also are easily increased by division. Take a plant of