One of the less known and, perhaps, somewhat rare saxifrages; it is a curious, distinct, and beautiful form, being of that class which the lover of the ornamental kinds most admires, for not only is it attractive all the year round, but additionally so when there cannot be seen any part of a growing or decaying flower stem upon it, and when its silvery, but lax rosettes, with their encrustments and glistening leaf dots, are perfectly matured, which is the case during mid-winter. I fear the illustration (Fig. 89), can give but a poor idea of the pleasing silvery-grey colour, which, when the specimen is dry, overlays foliage of a dark and glossy green, to say nothing of the numerous and regular spots which so charmingly enliven the specimens. I am unable to learn to what species it is most nearly related; its name, which doubtless has reference to its peculiar form and habit, would seem to isolate it even from its parents, if such are known; it, however, belongs to that section having thick leathery leaves, ligulate, encrusted, arranged in rosette form, and having excavated dots. Saxifraga lingulata, S. crustata, S. Australis, S. longifolia, and S. carinthiaca belong to the same section; but S. paradoxa differs much in general appearance from them all, and remarkably so in one or two respects, as, indeed, it does from the whole genus, thus justifying its name. The uneven length and arrangement of leaves, the casting off of the encrustments as a skin or in flakes, exposing to view a finely-polished surface, and the general web-like appearance of the tufts, are all peculiar to it. Of all the varieties of its section it most resembles S. carinthiaca and S. Australis; these forms, however, grow in compact rosette form, having leaves of more even size and shape. Our subject is irregular in every way, many of the leaves pushing out to double the length of others, and becoming attenuated at their junction, or club-shaped.

Its flowers are insignificant and similar to those of S. Aizoon, but more dwarf in the stem. The leaves are ½in. to 3in. long, very narrow and tongue-shaped, sometimes obtuse and club-shaped; stout, dark green, with a greyish crust-like covering, and deeply dotted with bright spots. The leaves are arranged in lax rosettes and are reflexed or pressed flat to the earth nearly all their length. The habit is very pretty in established and fair-sized specimens, which accommodate themselves to the form of surface, and the longer or erratic leaves become so interlaced with the other parts as to appear woven; this habit and the bright bead-like dots go to make the plant more than ordinarily attractive. It should be in every collection of choice Saxifrages; it is charming as a pot specimen, plunged and grown out of doors the year round.

On rockwork it should have a place, too, among the gems, being a neat and slow grower; its position should be near dark-coloured stones, where it will prove most telling. In damp weather its silvery parts are obliterated, but a breeze of half-an-hour or a beam of sunshine soon brings it into full beauty again. Gritty peat and a little loam suits it well; I have it doing nicely in ordinary garden soil; but if the more carefully prepared composts are employed, the results well repay the pains so taken. Its propagation is easily carried out by root divisions; early spring is a good time for the operation.

Flowering period, May and June.

Saxifraga Pectinata.

This belongs to the encrusted section, being most distinctly toothed; from this it takes its name; the teeth are large for such small leaves. Specimens of this Saxifrage, though small, are exceedingly pretty. Excepting when there is fog or rain, it is nearly white; and the rosettes, of various sizes, from ¼in. to 1in. across, are not only neat in themselves, but are densely and pleasingly arranged in a hard flat mass. It is never more beautiful, not even in May and June, when it flowers, than in November, when the growth is both complete and ripened, and the scaly substance which is spread over the leaves and the silvery teeth combine to render it attractive.

The flowers are of the usual form, and are produced on stems 4in. to 6in. high; they are white. The leaves seldom exceed ½in. in length and 1/8in. in width; they are spathulate in form, stout, and rigid. The rosettes are somewhat flattened and numerous, and give the idea of greenish-white flowers.

S. p. hybrida is a variety of the foregoing species, and without pretending to say what the type has been crossed with to produce this handsome form, I may, for the purpose of conveying an idea of what it is like, say that it approaches S. aizoon, which also flowers in May and June. In all its parts it is larger than the type; the leaves are greener and more strap-shaped, and are more erect, but not so rigid; the habit, too, differs—it forms more rounded tufts. In all these respects it will be seen to resemble S. aizoon. It is a lovely form; the sparkling teeth are relieved by the fine dark green ground of the foliage.

These comb-leaved Saxifrages belong to the more neat and effective rock plants; the type, at least, is of alpine origin, and under cultivation it seems most happy amongst the stones. I have grown these kinds as pot specimens, on nearly flat beds, and as edging plants; and in every position they prove attractive. It is very strange that such pretty forms are not more generally seen in gardens; they will grow well on walls and the tops of outhouses, and are good subjects for town gardens. Any kind of sandy soil will do for them; that of a vegetable character is, however, the best; they may be planted with choicer things, for, unlike many of the genus, they are not rampant growers. Practically, they need no propagating; for as the specimens spread they make new roots, and at any time one or half a dozen rosettes may be slipped off for planting elsewhere. It is better, though, to avoid this with small plants, as their full beauty is not realised until they become of considerable size.

Flowering period, May and June.

Saxifraga Peltata.

A new species to English gardens, hardy, herbaceous, and perennial, imported from North America; it is a truly noble plant. The illustration (Fig. 90) will convey some idea of its fine form, but the reader must rely on the description for its size when fully developed. When the flowers of this Saxifrage are in their best form, the noble foliage is scarcely half developed; a drawing, therefore (though it could hardly be made at a stage when the plant is more interesting), must necessarily fail, in this case, to give any more than an approximate idea of the parts undeveloped. Not only is this the largest species of the extensive genus at present grown in this country, but its form is both distinct and noble.

The flowers are produced on stems 18in. high and ¾in. thick at the base, being covered with long stiff white hairs, which are very conspicuous on the reddish stems. The flowers are similar to those of most of the genus, as may be seen by the one given in the drawing; they are arranged in massive heads, 3in. to 6in. in diameter, and rose-coloured. The leaves at the flowering time are 6in. or 9in. across, having stout, round, ruddy stems, 8in. long, covered with stiff hairs; they form a junction with the leaves in an unusual way, viz., near the centre, whence the specific name peltata, or umbrella shape; but the form of the leaves at the flowering period, which is funnel-shape, is, a little later on, reversed, the edges bending downwards. The younger leaves are folded and hooked downward, having the appearance of stout fern fronds just out of the ground, and their stalks are much contorted. The more advanced leaves are seen to be seven-cut, each lobe divided and sub-divided by cuts less deep, the whole leaf being richly toothed and veined. The under side is covered with hairs, the upper surface being smooth, shining, and of a pleasing bronze-green colour. Later, the foliage in every way increases very much in size, reaching a height of 2ft., and each leaf measuring nearly a foot across. The root or rhizoma is horizontal, progressive, jointed, and fibrous at the joints, and nearly 2in. in diameter; it may be clearly traced on the surface, but the fibrous parts go very deep.

It is said to be a bog subject; fortunately, however, this fine plant may be grown otherwise than in a bog, but it should not want for depth of rich soil. This I believe to be a more important condition than a boggy situation, inasmuch as I have grown my specimen for three years on the top of a dry mound; but the soil is good rich loam, and fully 5ft. deep; and to show that this strong-growing subject needs a good depth of soil, I may mention that I had occasion to dig up a piece, when it was found, for the operation, to require both the strength and tools that trees demand, the fibrous parts being deep and tough. When fairly established it makes rapid growth, and when in full leaf it proves very effective. Its propagation is easy with healthy plants; a length of the creeping root, with a crown to it, should be cut from the parent stock just before growth commences in early March. If planted as indicated in the foregoing remarks, and kept shaded with a leafy branch for a month or two, there need not be any fear about young plants becoming established the first season.

Flowering period, June.

Saxifraga Purpurascens.

The flowers are produced on stout stems, 8in. high, arranged in branched heads, of a rose or rosy-purple colour, and bell-shaped. They are, however, soon damaged by unfavourable weather, and there is little about the plant at that period to render it more attractive than its fellows; its finer qualities are developed as more genial weather prevails. When the stout foliage grows glossy, waved, and of a deep clear green colour, the edges of the leaves become lined with red as if hemmed with red silk; the leaves also have the edges irregular in form, the outline broadly oval, 4in. to 6in. long, and they are veined and slightly wrinkled; during the autumn a yellow tint starts from the edge, and in time becomes a vermilion, which is all the more effective from the leaf being of leather-like substance.

It enjoys a deep rich loam; and, evidently, to place its roots in contact with pieces of limestone is beneficial. Rare as the plant is, this is all that I do for it, and not only does it remain healthy, but it has increased greatly in size during the last year. I have not as yet tried to propagate it, but so far as I can judge there will be no difficulty in forming young stock by root division. It has hitherto enjoyed a happy immunity from all garden pests, not excepting slugs.

Flowering period, April to June.

Saxifraga Pyramidalis.

This is a very handsome form or variety of S. Cotyledon, and belongs to the alpine regions of Europe. As a decorative subject for our gardens, it is highly and deservedly esteemed; its attractiveness consists more in the numbers and arrangement of the flowers than in any beauty which belongs to them individually, though they are not devoid of that quality.

Of the many hundreds of species and varieties of Saxifrages which bloom during the month of June, this is one of the most distinct and useful as a decorative flower, and where the Saxifrages are grown in large collections, as they often are, giving more than an ordinary amount of pleasure compared with collections of other genera, the kind now under consideration always asserts itself as one of the first order of merit. Not only in its blooming state, but all the year round, it is very effective and striking; it is a free grower, having handsome, large rosetted foliage.

The flowers, as will be seen by the one given, natural size, in the illustration (Fig. 92), are of the common Saxifrage form, but rather more highly coloured in the central markings than the general run. They are produced on stout stems, 2ft. high, well and evenly branched in the form of a pyramid, whence the specific name. Each flower will be ½in. or more across; they are very numerous, and, partly from the fact that they remain perfect for a very long while, and partly because of the habit of the plant being to open all its flowers about or near the same time, the large panicle of bloom is very fine. The leaves, as already hinted, are formed into lax rosettes, which are 5in. to 7in. across; they are strap-shaped, narrowing slightly at the connection, half an inch wide, the outer ones being reflexed; the edges are finely serrated, and irregularly lined with a silver colouring.

This is a capital plant for rockwork, where it shows itself to much advantage; but specimens are much finer grown in beds or borders, where the moisture and temperature at the roots are likely to be more equable; besides, I find that, owing to its small quantity of roots, all of which are very near the surface, when grown on rockwork they may often be seen bare on inclined surfaces, and the weight of the flowers drags them entirely out of the soil on one side. They may be planted as an edging to a shrubbery, in bold groups, or as ordinary border flowers. So useful has this variety been found by professional gardeners that it is now largely grown in pots in single rosettes, which, after becoming well established, send up their rich plumes of blossom, all the finer for having been kept clean under glass. So grown, nothing can better repay the small amount of trouble which they give in order to place them in the conservatory as showy specimens; all they require being a 4in. pot, well drained, a compost of half-rotted leaves, and fat loam and sand. Put in one rooted offset any time from June to the end of July, the earlier the better; plunge the pot to its rim in sand or ashes until next spring, when it may be taken under glass if desired. To have fine flowers, the offsets should be pinched off as they appear. I may also mention that a somewhat shady situation has proved conducive to large and better coloured flowers; between irises 4ft. high and shrubs 6ft. high, the opening being not more than 3ft., running north and south. The specimen from which the drawing is taken was grown along with many others. A baking or dry treatment is often not only given to plants of this genus, but believed to be of advantage to them; it may be to some, but there are exceptions, and this is one without doubt. All the sections of Saxifraga to which it belongs are fond of good loam, well enriched. It is propagated from offsets taken as soon as they are from an inch to two inches across; they may either be put into nursery beds or be planted in their blooming quarters.

Flowering period, June and July.

Saxifraga Rocheliana.

Saxifraga Umbrosa.

This common flower is well known, and is only mentioned here as the typical form, and by way of introducing a beautiful variety called S. u. variegata, broad cushions of which, from their verdant condition, good habit, and pleasing variations of leaf colour, are amongst the more attractive objects of the garden in January. It hardly need be said that the plant is not valued for its flowers, which are similar to those of the parent form and borne at a corresponding date. The leaves, however, are much less in size and more flatly arranged in rosette form, they are also recurved at the edges. The markings are of two colours, creamy-white and pink, and there are many shades of green. The forms of the markings are most irregular, as striped, flecked, marbled, dotted, and edged; the various shades of green blended with pink and white, although figured on one of the commonest plants we know, render such plant worthy of a place in every garden, and more especially on rockwork.

It has this drawback—it is not constant. In some gardens the markings die out. This, however, need not be, for a rather dry situation and rich soil will produce rosettes of large size and good figuring. Still, there will be fully half of the rosettes entirely green in a large patch; this is more desirable than otherwise. The marked ones have a more starry effect in such a green setting; it is only when all become green that disappointment is felt. Sometimes I have noticed rosettes, about the size of a penny-piece, all one colour—creamy-white—which, when cut from the plant, very much resembled a carnation. Such abnormal forms are of no moment to the botanist, but if nine out of every ten persons who see this plant are interested, not to say pleased with it, it ought not to be entirely neglected. It is most effective in patches 1ft. to 2ft. broad. In propagating it the more finely marked pieces only should be taken.

Flowering period, May to July.

Saxifraga Wallacei.

This plant may be grown in pots or borders, as edging, or on rockwork, and in any kind of soil; but to have fine specimens and large flowers it should be planted in calcareous loam, and be top dressed in early spring with well rotted manure. I have it as an edging to a small bed of roses; the position is bleak, but the soil is good; it furnishes large quantities of cut bloom, and otherwise, from its rich hawthorn-like scent, it proves a great treat. So freely is its handsome foliage produced that it, too, may be cut in quantities for table decoration. If the flowers, or some of them, be left on, the tufts will form a pretty setting for a few other small flowers of decided colours.

To increase this Saxifrage is a simple matter during the warm season: The twiggy tufts should be pulled asunder, no matter whether they have roots or no roots; if dibbled into fine soil, deeply dug, and shaded for a week or two, they will form strong plants before the winter sets in.

Flowering period, April to August.

Scilla Campanulata.

A hardy bulbous perennial, introduced from Spain 200 years ago. It very much resembles the English hyacinth—H. nutans, or Scilla non-scripta—better known as the wood hyacinth. Handsome as this simple flower is, it might have been omitted from these notes as a plant too well known, but for the fact that there are several varieties of the species which are less known, very beautiful, and deliciously fragrant, entitling them to a place amongst other choice flowers, both in books and gardens.

Of the typical form little need be said by way of description. The flowers are bell-shaped, pendent, blue, and produced in racemes of many flowers. The leaves are lance-shaped, prostrate, and of a dark shining green colour.

S. c. alba differs from the type in having its white flowers arranged more evenly round the scape, being shorter in the divisions of petals and wider at the corolla; the habit of the plant, too, as may be seen by the illustration (Fig. 94), is more rigid and neat. In a cut state the flowers are not only very lasting, but if gathered clean, they are suitable for the most delicate wreath or bouquet.

S. c. carnea has pink flowers.

All the forms of S. campanulata are cheerful and effective spring flowers. They should be grown in bold clumps, and if under slight shade, where many other things cannot be well grown, all the better; still, they are in no way particular—any aspect, position, or soil will answer for these robust flowers. Such being the case, few gardens should be without at least the finer forms of the large Bluebell. So fast do these varieties increase by seed and otherwise, that any remarks on their propagation are unnecessary.

Flowering period, April to June.

Sedum Sieboldi.

Sedum Spectabile.

Hardy and herbaceous. This is one of our finest autumn bloomers. During September, the broad massive heads of small rosy flowers, which are arranged in cymes 6in. across, are very attractive, and will, with average weather, keep in good form for a month. This species is somewhat mixed up with another called S. Fabarium; by many they are said to be identical, but such is not the case. I grow them side by side, and I may say that they are as "like as two peas" up to midsummer, when they begin to diverge. S. Fabarium continues to grow to the height, or rather length, of 2ft., and tumbles over; the foliage has a lax appearance, and the flowers are very pale. Concurrently S. spectabile has grown its stems and glaucous leaves to stouter proportions, and crowned them with more massive heads of bright rose-coloured flowers, at the height of 15in. It is larger in all its parts, with the exception of length of stem, and by September it is nearly twice the size of S. Fabarium; it also stands erect, so that then the two species suggest a contrast rather than a comparison, S. spectabile being by far the more desirable.

I find, however, that it is much slower in increasing itself; the best way to propagate it is by cuttings dibbled into sand in early summer. The commoner one increases rapidly and often bears the wrong name; care should therefore be taken to obtain the true species, after which it will not give much further trouble, thriving in any kind of soil, but it should be planted in the full sunshine, when its habit and flowers will be greatly improved. It will bear any amount of drought—indeed, it seems to enjoy it. My finest clump is on a very dry part of rockwork, where it has always flowered well. These two Stonecrops and a variegated variety are some of the very few hardy plants which slugs do not graze; at any rate, it is so with me; neither do other pests attack them, but the humble bees literally cover their flowers the whole day long at times.

Flowering period, August to October.

Sempervivum Laggeri.

The flowers are more than an inch across, of a bright rose colour, and very beautiful; the central flower is invariably the largest, and the number of petals varies from six to twelve. The leaves are in rosette form, the rosettes being sometimes 2in. across, nearly flat, and slightly dipped in the centre; a downy web, as fine as a cobweb, covers the rosette, it being attached to the tips of the leaves, and in the middle it is so dense that it has a matted appearance. The leaves are very fleshy, glandular, and of a pale green colour. Slow in growth, habit very compact; it has a tender appearance, but I never saw its web damaged by rain or hail.

Many grow it in pots for indoor use; it finds a happy home on rockwork or old walls; it should have a dry and sunny situation, and, with these conditions, it will prove attractive all the year round. It thrives well in gritty loam; a little peat rubbed in with the grit will be an improvement and also more resemble its native soil. To preserve it from the bad effects of our damp winters, it need not be taken indoors, but sheets of glass should be tilted over the specimens during the short days, when they are dormant; the glass should not touch the plant. This seems to be the nearest condition we can afford it as a substitute for the snows of its mountain home, and I may add, for years it has proved effective; in fact, for several years I have left specimens in the open without any shelter whatever, and the percentage of loss has been very low, though the seasons were trying. It propagates itself freely by offsets; if it is intended to remove them from the parent plant, it should be done early in summer, so that they may become established before winter, otherwise the frosts will lift them out of position.

Flowering period, June to August.

Senecio Pulcher.

Autumn is the heyday of Composite flowers. The one now under notice has the merit of being of an unusual and beautiful colour, viz., purplish crimson. It is, in fact, a new plant in English gardens, and has been justly described as one of the finest imports of recent years; it has only to be seen in order to commend itself to all lovers of hardy flowers (see Fig. 96). It is a robust grower, ranking with the more noble subjects suitable for the borders. Its hardiness is doubted by many, and a few have suspected its perennial quality; but notwithstanding the warm climate of South America (whence it hails), it has proved both hardy and perennial in this country. Excessive moisture is its greatest enemy.

Its bright purplish-crimson flowers are daisy-shaped and large, the centre being a fine golden yellow—on strong young plants the flowers will be 3in. across. Moreover, they are numerously produced on stems 3ft. high, in branching cymes, and last a long time in perfection; with favourable weather an individual bloom will stand above a week, and the plant provides itself with abundance of buds for succession. I never yet saw a specimen that developed half its buds, but this brings me to notice one of its faults (for it has more than one), viz., it is too late in blooming; at any rate, in Yorkshire we rarely get more than three weeks' enjoyment of its flowers, when, but for severe frosts, it appears capable of blooming for two months. To some extent this may be remedied, as will be shown when I refer to its culture. The radical leaves are over a foot long, stem leaves much smaller, very dark holly green of leather-like substance, the edges very unevenly shaped, the general form of the leaf being something like the cos lettuce.

The cut blooms are indeed fine and cannot well be inappropriately used. This brings me to fault No. 2. The flower stems are very hollow and dry, nearly as much so as the hemlock or kex, and I have found that when flowers have been cut, either from the moisture collecting in the stem, or some such cause, rot sets in lower down, and soon the branches of bloom head over. I tried cutting to a joint where the cavity was stopped, but the pith when so exposed soon gave way, so that latterly I have ceased to cut the flowers, unless the occasion was worth the risk. A specimen not cut from did not suffer from stem rot. I, therefore, blamed the cutting. There may, however, be other causes; at any rate, there is the fact of fine flowers in their prime falling over, and it is worth one's while to try to find out from what cause it happens, and if my theory is not the true one, it may prove useful as a hint.

It likes a deep and rich soil, and well deserves to have it; if left out all the winter, a piece of glass should be put over the crown, because it has the fault (No. 3) of rotting in the centre, as I believe from water being conducted down its spout-like stems; but even under the most neglected conditions it stands our winters, and the rootlets send up a number of small growths in spring. These may make plants, but will not be reliable for bloom the following autumn; the damage should be prevented if possible. Another plan, by which two points are gained, is to grow young plants in good-sized pots and winter them, plunged in cold frames, not failing to give plenty of air. In April these, if compared with others in the open garden, will be found to be much more forward, and the first gain will be that, if planted out then, they will flower much more vigorously, and, secondly, they will start earlier by two weeks at least. To propagate this fine border plant, the very long and fleshy roots may be cut into pieces 6in. long and dibbled into fine soil; they are somewhat slow, but pretty sure to "go"; they should be protected from slugs, which are very fond of the young leaves. On young stuff, grown apart from the flower beds and borders, quicklime may be used, which would otherwise be unsightly.

Flowering period, August to October.

Sisyrinchium Grandiflorum.

The generic name of this flower is in reference to the grubbing of swine for its roots, and means "pig-snout." The common names may be seen, by a glance at the cut (Fig. 97), to be most appropriate; that of Satin-flower is of American origin the plant being a native of Oregon, and is in reference to its rich satiny blossom; that of Rush-lily, which is, perhaps, an even more suitable name, has been recently applied to it, I believe, in this country. It is applicable alike to the rush-like form and habit of foliage, and the lily-like purity and style of flowers. It was sent to this country in 1826, and yet it is rarely met with in English gardens. Some think it scarcely hardy in our climate in certain soils. I happen to have grown it for six years, which period includes the recent severe winters, and it has not only survived but increased in a moderate degree. This took place on rockwork facing south; in the autumn of 1881 I divided the specimen, and planted a part of it in the coldest part of my garden, which is not without clay, though far from all clay; that division is now a strong plant, and has made an extra crown; it forms the subject of the present illustration. Let me state, in passing, that it is naturally a slow grower. The very severe weather of the week previous to my writing this note, in March, 1883, when 23deg. of frost was registered, which cut down the bloom stems of Hellebores and many other well-known hardy things, did not hurt this subject very much; I am, therefore, confident of its hardiness from six years of such experience.

The flowers are 1in. to 1½in. long, and about as much across when open, of a fine purple colour, with a shining satiny appearance; the six transparent petal-like divisions are of uneven form, having short bluntish points; from the openness of the corolla the stamens and style are well exposed, and they are very beautiful. The flowers are produced when the plant is about 6in. or 9in. high, the buds being developed on a rush-like stem, and enfolded in an almost invisible sheath 2in. or 3in. from the apex. Gradually the sheath, from becoming swollen, attracts notice, and during sunshine it will suddenly burst and let fall its precious contents—a pair of beautiful flowers—which dangle on slender arching pedicels, springing from the sheath-socket. They seem to enjoy their new-born freedom, and flutter in the March wind like tethered butterflies. Their happy day, however, is soon over; their fugacious petals shrivel in three or four days. The leaves are rush-like, ribbed, and sheathed.

I have found it to thrive in loam, both light and moderately stiff, also in vegetable soil and sand; it likes moisture, but not of a stagnant character; between large stones, at the base of rockwork, suits it in every way; it may also be grown by the side of the larger kinds of snowdrops for contrast and effect. Impatient of being disturbed, it is not wisdom to lift it for any purpose, provided it is making progress, or until it has formed strong tufts; when, if it is desirable to increase it, and during early autumn, the long roots should be got well under, and taken out of the ground as entire as possible; from their wiry nature they are then both easily cleared of earth and divided into single crowns; these should be replanted in positions deeply dug, and where they are intended to remain, being carefully arranged without any doubling up. After such pains have been taken with so well-deserving a plant, there will be little to fear for its future, no matter how severe the winter may prove.

S. g. album is a white-flowered variety, of which, however, I have had no experience. Since these lines appeared in serial form, a lady, cultivating a good collection of choice hardy flowers, has informed me that this variety is very fine, and in every way commendable.

Flowering period, March to May, according to positions or climatic conditions.

Soldanellas.

All the Soldanellas love a vegetable soil, as peat or leaf mould, to which, when under cultivation, a liberal quantity of sand should be added. If grown in pots, they make lovely specimens, and should be plunged in sand and kept moist; but I find my specimens to grow much more vigorously when planted out, as they are at the base of a small rockery, rather below the level of the neighbouring walk, which forms a miniature watershed for the supply of moisture. I also fancy the liverwort, which surrounds them, rather helps them than otherwise. Certain I am, however, that moisture is the great desideratum in the culture of this genus. My difficulty with the planted-out specimens is to keep them from being grazed off by the slugs; a dash of silver sand every day or two has sometimes proved of use. When the Soldanellas once get into proper quarters they make rapid growth; I have divided them most successfully in April and May.