close, surmounted by a very short lacerate ligule. Flowers in Aug. ust, and is very common on salt marshes and sandy sea beaches. It may be distinguished from both the preceding by its awnless glumes and the obtuse rounded tips of the paleæ, and by the want of a woolly beard at the ligule. It forms a considerable portion of the salt meadow hay of the coast.

4. Spartina stricta.-Salt marsh grass. Spikelets loosely imbricated, laterally compressed. Glumes hairy, the keel spinulose, the upper one from one-half to one-third longer than the lower one, one-nerved. Paleæ unequal; the outer palea the shorter, hairy and without lateral nerves; the inner palea longer than the glumes, with two delicate nerves. Spikes from two to four in number, sessile, erect, soft, yellowish. Rachis angular, projecting slightly beyond the imbrication of the spikelets. Culm ten to thirty inches in length, leafy to the top. Leaves convolute, narrow, easily sepa rated from the sheaths. Ligule very short, obtuse, ragged; the upper leaf shorter than its sheath. Root perennial, with strong, creeping fibres. Fig. 69, enlarged spikelet showing the glumes; Fig. 70, paleæ and included floral organs lifted out of the glumes; Fig. 71, the entire plant; Fig. 71a shows the leaf and ligule, natural size. Salt marshes Pennsylvania, &c. There is a variety of this species, which, by some authors, is put as a separate species, known as alterniflora. It has a greater number of spikes, from five to fourteen; they are more slender, from three to five inches long; the rachis is produced considerably beyond the spikelets, having an awnlike appearance; the glumes indistinctly five-nerved. Its odor is rancid and disagreeable. Cattle are very fond of it, eating it greedily; but it is said that the odor is imbibed by the milk, and even the flesh of the animal; though this needs confirmation. It flowers in August and September. It is found in muddy places and the borders of salt marshes; generally in places that are overflowed by the tide. It is used a good deal for thatching, for which purpose it is superior to wheat straw, and lasts much longer. Fig. 72 shows the floret of Var, alterniflora and the glumes; Fig. 73, the paleæ; Fig. 74, the ligule; Fig. 75, the entire plant.

17. CTENIUM.-Toothache grass. Name from the Greek Ktenion, a small comb, from the pectinate appearance of the spike. Spikelets densely imbricated in two rows on one side of a flat arctuate curved (i. e., bent in the shape of a bow) rachis, forming a solitary terminal spike. Glumes persistent; the lower one (interior) much smaller; the other concave below, bearing a stout,recurved awn, like a horn on the

middle of the back. Flowers four to six, all but one neutra; the one or two lower consisting of empty awned paleæ; the one or two uppermost of empty awnless paleæ; the perfect flower intermediate in position, its paleæ membranaceous, the lower awned or mucronate below the apex, and densely ciliate toward the base, three-nerved. Squamule two. Stamens three. Stigmas plumose. In describing this gnus, we, or rather Prof. Gray from whom we borrow it, have introduced the terms neutral flowers and perfect flowers; we must now describe exactly what we mean by these terms. No flower can produce seeds that is not furnished with two separate and distinct organs, viz. The stamens and the pistils; the former produces the pollen or vivifying agent; the later receives it and works those transformations in it which are essential to the perfection of the seed. By analogy, we call the stamens the male organs, and the pistils the female organs; the two together constitute the sexual system of the flowers. When these organs are wanting, the flower is said to be neutral, as it is of no sex and therefore infertile. When a flower has both stamens and pistils in one floral envelope, it is said to be perfect or fertile. When one of the sexual organs is absent in a flower, whether it be stamen or pistil, it is said to be imperfect. Fig. 75a represents a spike of C. Americanum, of natural size; Fig. 756 a spikelet enlarged; Fig. 75c, the flower raised out of the glumes.

Clenium Americanum.-Large glume, warty-glandular outside and conspicuously awned. Culm three to four feet high, simple, pubescent or roughish. Taste very pungent. Wet pine barrens, south Virginia and southward.

18. BOUTELOUA.-Muskit grass. Named for Claudius Boutelou, a Spanish writer upon floriculture and agriculture. Spikelets crowded and closely sessile, in two rows on one side of a flattened rachis, comprising one perfect flower below and one or more sterile (mostly neutral) or rudimentary flowers. Glumes concave keeled, the lower one shorter. Perfect flower with the three-nerved lower palea three-toothed or cleft at the apex; the two-nerved upper palea two-toothed; the teeth, at least of the former, pointed or subulate-awned. Spikes solitary, racemed or spiked; the rachis somewhat extended beyond the spikelets. This genus includes three species and one sub-variety.

1. Bouteloua oligostachya.--Glumes sparingly soft-hairy, obscurely papillose along the keel. Lower palea sparingly hairy, the lobes awl-pointed, middle lobe two-cleft at the tip. The sterile flowers, about two in number, are mounted on the summit of a short pedicel, generally an awnless scale, becoming hood-like and coriaceous. The

sterile flower is copiously villous tufted. The spikes are pectinate (i. e., separated into narrow divisions resembling the teeth of a comb) and consist of a great many spikelets, oblong or linear, very dense, solitary and terminal, or few in a raceme, and are from one to five in number. Culms from six to twelve inches high. Leaves very narrow. Rachis glabrous. Found in western Wisconsin and westward.

2. Bouteloua hirsuta.-Lower glume hispid with strong bristles, from dark, warty glands, bristles half as long as the glumes. Lower palea pubescent, three-cleft into awl-pointed lobes. The sterile flower and its pedicel glabrous, the three awns longer than the glumes and the fertile flower. Spikes one to four. Culms tufted from an annual root, from eight to twenty inches high. Leaves flat, lance linear; papillose, hairy or glabrous, very short, edges rough, sheath bearded at the top. It is found on sandy plains in Wisconsin, Illinois and south-westward.

3. Bouteloua curtipendula. Darlington, Torrey and others call it B. racemosa. Spikelets two, rarely three-flowered. Lower glumes very narrow. The upper one hispid on the keel, lanceolate. The perfect flower sessile, lanceolate. The lower palea with three nearly equal awl-pointed teeth; scabrous. The upper one is a little longer; has two keels, and is bifid at the apex. Stamens bright, cinnabar red color. Abortive flower pedicillate and neuter; it has one or two minute scales, and from one to three awns shorter than the fertile flower, the central awn being the longest. The spikes are short, about one-third of an inch long, twenty to fifty in number, in a long and virgate (i. e., wand-like, long, slender and straight) one-sided spike or raceme, rather distant, spreading or reflexed, each of four to twelve spikelets. They are arrayed on two sides of a common rachis, becoming at length secund. The common spike is from eight to fifteen inches long. The culm is tufted, perennial, two to three feet high, geniculate at the base, smooth. Leaves two to three lines wide, tapering to a long slender point, involute when dry, rough on the margin, somewhat hairy on the upper side. Sheath rather loose, the lowest ones often hairy. Ligule short, fimbriate. It flowers in August, and is found on calcareous, dry, hills and plains, from southern New York to Wisconsin and southward. It is found abundantly on rocky lands and hill sides in the counties of Orange and Dutchess. Its brilliant red anthers make it very conspicuous in flowering time. Fig. 75d represents a culm of B. curtipendula, the flowering stalk separated from the culm. Fig. 75e, an enlarged branch Fig. 75f, a pair of glumes, with the flower lifted out of them.

19. GYMNOPOGON.-Naked beard grass. Name from the Greek gymnos, naked and pogon beard. Spikelets of one perfect flower and the rudiment of a second (consisting of an awn-like pedicel, mostly bearing a naked bristle), sessile and remotely alternate on long and filiform rays or spikes, which form a naked, crowded raceme. Glumes lance awl-shaped, keeled, almost equal, rather longer than the somewhat equal membranaceous paleæ, of which the lower is cylindrical involute, with the midrib produced from just below the two cleft apex into a straight and slender bristle-like awn, the upper with the abortive rudiment at its base. Stamens three. Stigmas pencil-form, purple. Leaves short and flat, thickish. includes two species.

It

1. Gymnopogon racemosus. The awn of the abortive flower shorter than its stalk, equalling the pointed glumes, not more than half the length of the awn of the fertile flower. Spikes flower bearing to the base, five to eight inches long, soon divergent. Culms clustered from a short root stock one foot high, wiry, leafy. Leaves oblong, lanceolate. Perennial. Flowers in August and September. Grows in sandy, pine barrens, New Jersey to Virginia and southward.

2. Gymnopogon brevifolius.-Lower palea ciliate near the base, short awned; awn of the abortive flower obsolete or minute; glumes acute. Filiform spikes long, peduncled i. e., flower bearing only above the middle. Sussex county, Delaware and southward. Fig. 75g represents G. racemosus, reduced in size. Fig. 75h is a magnified spikelet, showing the glumes and floral organs.

20. CYNODON.-Bermuda grass. Scutch grass. Name from kuon, a dog, and odous, a tooth. In this genus the spikelets are one-flowered, with a mere naked, short, pediceled rudiment of a second flower, imbricate-spiked on one side of a flattish rachis. The spikes usually digitate at the naked summits of the flowering culms. Glumes keeled, pointless, rather unequal. Paleæ pointless and awnless, the lower larger, boat-shaped. Stamens three. Low, diffusely branched and extensively creeping perennials, with short, flattish leaves. It has but one species.

1. Cynodon Dactylon.-Glumes linear, compressed, somewhat hispid on the keel. Professor Gray, in the definition of the genus given above, says that the glumes are pointless. They are distinctly pointed in all the specimens in our possession. Paleæ pointless, awnless, the lowest, chartaceous, longest and largest nerved on the edges, boat-shaped, smooth, longer than the blunt rudiment. Spikes

three to five, about one inch long, generally purplish. Culms six to twelve inches long. Sheaths smooth, longer than their leaves, woolly at the top. Leaves filiform, one to two inches long, rough on the edges, joints near the base covered by the sheaths. Ligule wanting. Fig. 76 is a magnified spikelet, showing the glumes and floret. Fig. 77 is a part of the rachis and two spikelets magnified. Fig. 78 is a magnified view of the paleæ opened with the blunt rudiment of a second flower. Fig. 79 is the upper part of a sheath and a leaf, showing the hairs which replace the wanting ligule. Fig. 80 shows the appearance of the whole plant, of the natural size, though we have seen it twice as large from the Island of Jamaica. It is of no agricultural use, as far as we know, where the better species will grow. Yet it has a very wide geographical range. It is well known at the south as Bermuda grass and forms a very considerable part of their scanty pastures. The most common pasture grasses of Jamaica, West Indies, were forwarded to us by a friend residing there, and fully one-third of these belonged to this genus, and we are assured that it holds the same numerical rank in most of those islands. It grows on the sandy shores of the southwest coast of Cornwall, but we believe that it is found in no other portion of Great Britain. It is found in Spain, Portugal, Italy, Turkey, Greece, islands of the Mediterranean, north Africa and western Asia. This wide geographical range would seem to indicate that it was intended to serve some useful purpose, if we could only ascertain what that purpose was. It flowers in July and August. It is very useful in the same way that Calamagrostis arenaria is, binding loose sands by its long roots, which send out long cord-like stolons that fix themselves firmly in the loose sands by producing at every joint a cluster of strong radical fibres, which cover the surface by a rapidly forming net work, as well as by interlacing below. It is only found growing spontaneously in warm latitudes, and is quite incapable of withstanding the effects of excessive cold and moisture. It suffered in London in the cold winters of 1860-61 quite as much as the Arundo Dorax, Tripsacum dactyloides, and other semi-tropical plants of the order. We are assured that it is identical with the Durva or Doorba grass of Hindostan, where it is the most common and valuable fodder grass of the region. We have never seen this grass, but it is probable that something in the soil or climate of that region increases its size and adds to its nutritive. value. If these special qualities could be artificially communicated to the soil of our southern States, where it grows so naturally, it would be a priceless boon to the planters. T. Afflec reports that