The most exquisite effect under roses is the lace work made by myriads of minute, starry, white flowers, especially the saxifrages, which make mossy cushions of foliage. At first I despaired of our ever getting this effect in rose beds. Certainly our summers are too hot and dry for saxifrages. But the stonecrops or sedums are obviously resisters of heat and drought, and that whole genus (except the coarse, tall kinds) ought to be studied from this point of view.

But the most useful effect, perhaps, among the carpeters of rose beds is that of evergreen foliage. I say "useful" because evergreen leaves will, presumably, protect rose roots from alternate freezing and thawing better than dead foliage, even when it forms a dense mat, like that of the moss pink (Phlox subulata). But even if that should not prove true, the ground is bare most of the winter over the greater part of the United States, and it is pleasant to catch a glimpse of some brave, fresh greenery amid universal desolation. For such a use England possesses a perfect plant in the London pride (Saxifraga umbrosa), the thick rosettes of which are indescribably cheery. I fear it can never thrive here as there, but we have plenty of other material, for one February, at South Orange, N. J., I walked over many perfect evergreen carpets of thyme, thrift, sandwort, hardy pinks, etc.

"But how about the practical side?" you may ask. "Can we do it? And how can we protect the roses?"

All such questions, I believe, can eventually be answered with satisfaction. At present the idea is only in the experimental stage here. However, one can see something of it at Elizabeth Park, Hartford, Conn. It is impossible that any one system of rose culture will do well for all varieties and in all parts of the United States. We cannot expect that the new system will keep tender roses from freezing at the root in zero weather. But wherever hybrid teas are hardy without more than two inches of winter mulch, the new plan ought to be successful, because these carpeting plants constitute a living mulch. This will be a great gain for the most populous portions of the North, because we can get rid of the everlasting digging up of rose beds every spring and the heaping of them in autumn with manure, which is often done near the most important windows of the house. Mr. Robinson's rose beds remain in excellent condition for six or eight years without adding any manure to the surface, even liquid manure. It is simply a case of digging deeply (say, three feet) and enriching the soil once for all.

But for those who cannot adapt this idea to their own gardens can anything else of an encouraging nature be said? Heaven knows we need encouragement, for we spend more money on roses than on all other plants put together, and have mighty little to show for it. The Pacific Coast, of course, can grow roses to perfection, and parts of the South can do wonders, but the North finds rose culture the costliest and least satisfactory department of gardening. No beginner will believe this. Every year we hear of immense new rose gardens, and of amateurs who are going to make roses "a life study." But ninety-nine out of a hundred drop out after a few years. They love the rose just as much as ever, but it costs too much to get large double roses. There are dozens of other flowers that will give a better show for the money. So these incipient rose specialists become plain lovers of gardens and hardy plants, which is probably the best thing that could happen for all concerned.

It may be only a negative and Pharisaic satisfaction, but we certainly do not sacrifice as much for mere size of flower as the English do. They say we are constantly talking about "big" things, but at least we do not have in our flower shows thousands of individual roses, each with a stem three inches long, and exhibited on "boards" in serried ranks. Of course, they have informal and artistic arrangements, too, but for many years the dominant idea in English flower shows has been the false and degrading practice of exhibiting big individual blooms on short stems. I admit that this system is absolutely necessary for the study of technical perfection, but it is grossly misleading because, in most cases, such blooms are the results of peculiar and costly methods which the ordinary public cannot use. Therefore, when people crowd to the front to note a new variety at the shows they are only preparing for disappointment. Moreover this horrid method of exhibiting flowers degrades the public taste, because it makes English amateurs ignore garden effects and encourages them to grow roses in the vegetable garden or in places coated with manure simply to get big flowers to cut for the shows. It is an open question whether English shows have not done more harm than good. A better way to spread the love of flowers is for every one to visit one another's gardens, because the cast-iron standards of the florists cannot then tyrannize over us. However, this small comfort may be a fleeting one, since fashions change.

A surer basis for encouragement can be found in our climatic superiority with respect to climbing roses. True, we cannot yet produce the effect (shown on plate 43) of large double roses blooming all summer on house walls up to the second story, but even now we can make a bigger show in June with small roses in large clusters (the sort of thing pictured on plate 44). For the three great Japanese roses seem better adapted to our climate than to that of Europe. They give us more and better foliage than any European species ever can; they are less troubled with insects, and require less care than large, double roses. The many-flowered rose (Rosa multiflora) has already given rise to the Crimson Rambler, of which more plants have probably been sold, it is said, than of any other rose ever introduced. The memorial rose (R. Wich-uraiana) gives us foliage that retains its glossy beauty a good part of the winter. The Ramanas rose (R. rugosa) gives us the best bush and the best foliage of any rose in the world.