would be a Rose-house. This should be a light span-roofed structure, glazed nearly to the ground at the sides, with ample facilities for ventilation, and, if possible, the power of heating it occasionally in winter. The Tea-scented and other tender kinds of Rose might be planted out in beds in such a house, and pillars or light iron arches would afford the means of supporting the more valuable climbing sorts. Houses of this class, in fact, are not now unfamiliar to cultivators, and are in the highest degree remunerative in regard to yielding enjoyment. But they want to be applied to a decided rosery, where they would be singularly harmonious and apposite.

A plan of a rosery, embracing the appurtenance just named and other peculiar accompaniments, has already been casually given at p. 180. Two additional designs will be found incorporated in figs. 214 and 232, where their situation and general arrangement will be duly indicated. I shall therefore simply present here two further sketches, in which there is a marked dissimilarity of treatment. The first, fig. 188, is of a rosery which I made lately in the neighbourhood of Dulwich, near London. It lies in a sheltered and partially detached corner of the grounds, and is connected with the kitchen-garden, on the north side, by the walk at the top of the engraving, the walk to the right leading eastwards into the general pleasure-grounds, through some wire arches, covered with climbing Roses; that to the left being finished by a handsome summer-house; and the southern walk, which quickly turns westwards, being conducted through a small wood to another part of the estate. The whole is nicely open to the south, south-east, and south-west, on which sides only shrubs exist. On the other margins, larger trees mingle with the plantations.

Great simplicity and roundness of form will be observed in the beds, and the grass openings, with the grass verge round the edge of the walk, are varied and ample. The references will make the details quite intelligible:

It will be seen that each important tribe is brought together in beds by itself, and if the sorts be nicely selected and mixed, such an arrangement will be found usually more productive of

harmony of character and tone, than any merely promiscuous mixture of all the groups.

The other plan, fig. 189, was prepared for Harman Grisewood, Esq., at Daylesford. In this case, the top of the engraving represents the west, while the north is to the right, and the south to the left. The site of the rosery-the only suitable and convenient one that could be found-is in a rough appendage to the pleasure-grounds, which is annexed to the lawn on the east side of the house, and has a walk carried round it, and a collection of Coniferous plants scattered along its sides. The rosery forms a break in that walk, and is on a somewhat elevated

spot, level in itself, but with a slope to the west beyond the range of the circular walk.

A very large extent of grass lying around the plot, the beds are

And because

to have box-edgings, with gravel walks between. it would have appeared like a sudden break in the principal lawn, and would have looked exceedingly bare in the winter, as viewed from a distance, and nothing in the way of denser plantations or masses of shrubs would have been admissible, on account of the open character of this part, it is proposed to environ the plot, irregularly, with a collection of Hollies. These, as specimens, and as thus brought together, would be very interesting in themselves, while they would supply the needful framing in, shelter, and cover, without producing too much shade, or interfering with the leading purposes and characteristics of the spot.

The beds are to be filled with Roses of one class or tribe in each, the figures near them referring rather to the plants represented by a small cross (×) in the centre or other parts of them. Thus, 1 indicates that the cross in the middle of this bed is for a pillar Rose, on a pole 8 feet high. The figure 2 likewise means that the crosses in these beds are for standard Roses, 3 feet 6 inches high; and the figures 3 denote that specimens of standard Cotoneaster microphylla, about 3 feet high, should be placed where the crosses occur in the adjoining beds. Of the specimens, 4 are pillar Roses, on poles 6 feet high, 5 are half-standard Roses, 2 feet high, and 6 are Irish Yews, to be kept to a uniform height of about 3 feet. The remaining references are to Hollies,

This list, including all the best and most recognised kinds of Holly, may further help to guide those who wish to make a collection of them. I have occasionally met with other very distinct varieties in provincial nurseries, and having local names. And to any person who can appreciate the beauty and value of the tribe, every really different variety will be a desideratum. In the Bagshot nurseries, too, I have noticed the variegated hedgehog Holly pruned and trained into a standard, and thus making a highly useful plant for a formal garden.

But ere I pass from the consideration of the Rosery, it may be well to mention that certain places afford facilities for growing a collection of Roses by the sides of a walk, rather than in a regular garden. Such a walk I remember to have designed for

William Wailes, Esq., of Saltwell, near Gateshead, where regular oblong beds were cut out in a band of grass on either side of the walk, and specimen standards occurred between the beds. The walk itself along the front of the kitchen-garden-was entered through a wire arch mantled with climbing Roses, and ended in an arbour or bower, over which similar Roses were intended to ramble.

5. Although there are very few places of sufficient magnitude to admit of the formation of what has been termed an Arboretum, or complete collection of trees and shrubs, classified according to their natural affinities; and where there is actually room for it, such a gathering, according to the received notion of it, would be by no means ornamental, while it would necessarily comprise many species and varieties that are quite unworthy of cultivation; there is the greatest propriety in selecting the most distinct or interesting members of certain tribes, and allotting a separate space to them within the general compass of the pleasure-grounds. And one of the most pleasing of such departments would be the Pinetum.

By the term Pinetum, however, I do not profess to describe a spot that necessarily accommodates all the known or hardy species and varieties of Coniferous plants. Much less do I seek to advocate the common method of dotting these about, as single specimens, at nearly regular intervals, by the sides of a walk made on purpose to exhibit them. I merely wish to recommend the introduction, where practicable, in some remoter and wilder part of the pleasure-grounds or woods, and particularly where there are natural sloping banks, of varied aspect, with an ordinarily sandy or rocky substratum, of a careful selection of the most peculiar or most ornamental kinds, and to distribute these about very irregularly, in broken groups or as single specimens, according to the conformation of the ground, the character or value of the plants, and their fitness for entering into combination with others or for standing alone. In short, the object of a Pinetum should be to produce a new and unique, but always picturesque scene or succession of scenes in a place, with the occasional exhibition of a very perfect specimen, and not, as is usually the case, a merely monotonous succession of specimens.

A remarkably eligible site for such a Pinetum would be a small winding valley, in an old wood, towards the outside of