THREE species of rubus are popularly called dewberry in the eastern states - Rubus Canadensis, R. hispidus, and R. trivialis. These are also all known as low-blackberry and trailing-blackberry. In the botanies the name dewberry is usually restricted to Rubus Canadensis. These three species are much confused in the popular mind, and it is also probable that comparatively few botanists have a clear understanding of them. This confusion is due in part to the great similarities in the species themselves, and probably fully as much to the imperfect characterizations and descriptions in the books. An instance of an error in the books, which is everywhere copied, is found in the case of Rubus hispidus. It is said to inhabit swampy places and low woods, and in Gray's Manual it is called " running swamp-blackberry," while it is common upon sand banks ! I never saw thriftier specimens of it than I collected this summer upon clean and dry white sand. Neither do the books say that its leaves are glossy above, which is the fact, at least when it grows in the sun, and which affords one of the best distinguishing characters.

Rubus Canadensis is generally distributed throughout the region east of the 100th meridian from Newfoundland to, Virginia, probably. A'. trivialis represents it in the south, extending from Virginia to Florida, and westward to New Mexico. R. hispidus has much the same range as R. Canadensis, except that it reaches southward to Georgia.

R. Canadensis and R. trivialis are closely allied. Both have pointed or acuminate leaflets, which are singly serrate, or, more properly, dentate, with coarse teeth, stems beset with sharp and curved prickles, and both are comparatively stiff and strong growers. The chief contrasting points of the two species are these : R. Canadensis - Main stems rather sparsely and lightly prickly; leaves thin and deciduous, either destitute of prickles or provided with weak ones, and more or less hairy; leaflets ovate and comparatively large; sepals often prolonged and leaf-like and lobed. R. trivi-a/is - main stem mostly thickly beset with stout prickles; leaves firm, and nearly or quite evergreen, with usually stout-prickled petiole and midrib, and smooth, or very nearly 90 ; leaflets oval-oblong or almost lanceolate and small; sepals not prolonged nor cut, so far as I have seen.

From these species Rubus hispidus is distinguished by its very small and weak long stems, which are beset with hair-like bristles, which only rarely become stout enough to scratch the hand seriously ; the obovate and obtuse shining leaflets, the terminal one of which is not stalked ; the fifi form pedicels and spreading inflorescence, and the small flowers and fruit. It is in every way a much more delicate plant than either of the others. A sprig of it is shown in Fig. 2.

Now, what is the cultivated dewberry ? It is always said to be Rubus Canadensis, yet it is by no means certain that it should be referred to that species. The Lucretia dewberry looks like a distinct and new species, yet it had probably better be referred as a variety to R. Canadensis until we know our rubuses better. The frontispiece, Fig. 1, is a good illustration of the Lucretia dewberry. The leaves, especially those on the sprig at the left, are seen to be very broad and straight at the base, and to be doubly serrate and jagged. These are characters which do not belong to the type of Rubus Canadensis. Fuller characters of separation may be given as follows:

Rubus Canadensis, Linn., var. roribaccus, new variety (ros, dew; baccus, berry). Plant larger and stronger ; leaflets broad below, usually triangular-ovate, doubly serrate with small teeth, and more or less notched or jagged; peduncles longer, straighter and stouter, habitually more numerous and more conspicuously overtopping the leaves ; flowers very large (sometimes two inches across !); sepals uniformly larger, some of them much prolonged and leaf-like and conspicuously lobed (sometimes becoming an inch long and wide); fruit much larger.

But the cultivated dewberries are not all the same. I have a specimen of a plant which I once grew, sent me as "Lucretia's Sister," which is, apparently, true Rubus Canadensis. I have no record of the source from which I obtained this variety. I also have specimens of Bartell and Mammoth, which are clearly distinct from the Lucretia. They have the coarsely dentate leaves of R. Canadensis, nearly smooth pedicels and less conspicuous sepals than Lucretia. They are, apparently, robust forms of R. Canadensis, but my specimens are insufficient for positive determination.

The Lucretia dewberry is said to be native to the hills of West Virginia, and is said to have been introduced from there by B. F. Albaugh & Sons, of Covington, Ohio. It has been in cultivation for a number of years. Neither the fruit nor any part of the plant has ever been truthfully figured, so far as I know. The trade cuts and lithographs of the dewberry are unlike it.

Fig. 2. Rubus hispidus.

As a fruit, the Lucretia dewberry has given poor results with me. The plants are unproductive and the berries are seldom well-formed. But others have obtained remarkable results from it, and I am confident that it possesses promise. It is not improbable that the variety is not uniform in character, and that it needs more systematic attention in the way of careful selection ; but it is evident that I have not yet learned how to grow it.

L. H, B.