compressis subtricarinatis, dorso deorsum curvato alato-carinato dentibus elevatiusculis cristato, carinis lateralibus rotundatis, lateribus sulco cochleato impressis, stigmate faciali sessili, facie rotundata.

HAB. Rivers; common in the Charles and in the Mystic, and also in the Middlesex canal. Stem slender, compressed, and concave-convex below, much branched, 6 to 12 inches long. Submersed leaves delicate, pale-green, grass-like, connate with the stipules and sheathing, linear, obtuse, 1 in. long, by - lin. wide. Floating leaves coriaceous, oblong lanceolate or linear, the last often a little falcate, furrowed on the under side with 3-7 impressed nerves, half an inch to an inch long and 2-4 lines wide, on channeled petioles 3-5 lines long, which are connate with the stipules, the lower near the middle of the stipules, the upper below it. Stipules of the submersed leaves reduced to short, lacerate ligules. Submersed spikes on very short, erectish, clavate peduncles, 2-6 flowered; spikes of the floating part later, cylindrical, on peduncles half an inch or more in length. Nutlets green, lentiform, and in perfect specimens rendered obovate by the broadly alate-carinate back, which is crested with about 5 prominent, rather distant teeth; the lateral keels rounded: the sides conspicuously cochleate.* Exocarp membranaceous, produced on the back, and forming the wing. Putamen several times thicker, but semitransparent. Seed cochleate. This curious species, which is recognizable by its pale-green color and delicate, grass-like leaves, is more common, in this neighborhood at least, entirely submersed, and perfecting its numerous, few-flowered spikes below the water. The state with floating leaves entirely resembles in habit P. hybridus, from which it differs in its grass-like leaves evidently connate with the stipules, in the attachment of the petioles of its floating leaves to the stipules, in its larger, different nutlets, &c. The following description of the former species, from living specimens, will afford a fuller comparison of these nearly allied plants. P. heterophyllus? Ell. Sk. i, p. 222, seems to accord in many respects with the species above characterized.

P. hybridus, (Michx.): caule compresso striato, ramoso; foliis submersis membranaceis tenuissimis cum stipulis connatis vaginantibusque setaceo-linearibus, apice attenuatis acutis, uninervibus; natantibus subcoriaceis ovalibus lanceolatis linearibusve, subtus nervis 3-7 impressis sulcatis, petiolis planiusculis cum stipulis haud connatis; stipulis nervosis; spicis partis submersæ alaribus capitatis paucifloris pedunculis brevibus clavatis, fructife

The spiral impression of the sides of the fruit is its most striking feature when dry, and reminds one of the outline of a small snail-shell, or of Spirorbis nautiloides.

SECOND SERIES, Vol. VI, No. 17.-Sept., 1848.

30

ris plus minus reversis, natantis cylindricis plurifloris pedunculis longiusculis; fructibus recentibus oblique lunato-lentiformibus compressis tricarinatis, stigmate faciali sessili, dorso deorsum curvato anguste alato-carinato dentato, carinis lateralibus acutiusculis plus minus sinuato-irregularibus, lateribus obscurius cochleatoimpressis, facie acutiuscula. P. hybridus, Michx. Fl. 1, p. 101, e Chamiss. in Linnæa, 2, p. 208. Kth. Enum. 3, p. 132. A. Gr. Man. p. 456. P. setaceus, Pursh, Fl. 1, p. 120, non L. P. capillaceus, Poir. Enc. Suppl. iv, p. 535. Roem. and Schult. Syst. 3, p. 507. P. diversifolius, Barton. Comp. Fl. Philad. 1, p. 96. Torr. Fl. U. S., p. 197. P. filiformis, Pursh herb.!

HAB. Small ponds and plashes; Cambridge, Plymouth, Tewksbury; and extending through the Middle and Southern States. Stem very slender, compressed, striate, shortly and rather distantly branched, about a foot long. Submersed leaves membranaceous, setaceous-linear, counate with the stipules and sheathing at the base, attenuate and acute at the tip, channeled on the upper side, one-nerved, 1-3 inches long. Floating leaves subcoriaceous, a little more oval than in the last, furrowed on the under side with 3-7 impressed nerves, 6-10 lines long by 3-4 wide, on rather flat petioles which are free from the strongly nerved stipules. Peduncles longer than in the last, more or less reversed in fruit. Nutlets minute, half the size of those of the last, very oblique, rounded, compressed, tricarinate, the back narrowly alate, with 6-8 approximate teeth, the lateral keels acute, and often sinnate-toothed when dry, the sides obsoletely cochleate-sulcate. Exocarp very thin, a little produced on the back. Putamen semi-transparent. Seed cochleate. This species, P. Spirillus, and P. Claytonii are peculiar to this country, and, at first sight, are not easily referred to either of the commonly received sections of the genus. They differ from the Heterophylli or Diversifolii of authors in characters much more important than the solitary one (the floating leaves) in which they are considered to agree with them, and shew, as it seems to me, very clearly, that this group cannot be retained in arranging the North American species. P. Claytonii belongs to the Graminifolii, irrespectively of its coriaceous leaves, and the two others mentioned, if they accord with the Heterophylli in their floating leaves, accord at the same time with the Graminifolii in their submersed leaves and inflorescence, and with the Vaginiferi or Coleophylli in the attachment of their leaves to the stipules. This last character is not easily noticeable in P. hybridus, and seems to have escaped attention, but it is conspicuous in the nearly allied P. Spirillus. The arrangement proposed by Fries (Novitiæ, edit. 2, p. 27) for the Swedish Potamogetons, admitting two principal sections, I, of those with submersed leaves broader in the middle (Plantaginifolii; in which the Natantes or species with simple stems constitute the

first tribe, and the Lucentes or those with branched stems the second); and II, of those with linear submersed leaves ( Graminifolii; in which the Compressi or species with free stipules are the first tribe, and the Pectinati or those with stipules and leaves connate, the second), best includes our American species, and, as is always the case with a natural disposition, sheds light on their difficulties.

AGROSTIS. In the article in a former volume of this journal to which reference has already been made, an account was attempted, from considerable and authentic materials, of the species of Agrostis formerly considered to constitute the genus Trichodium. The Agrostis scabra there described, is, according to Dr. Gray, (Man. p. 577,) not the A. scabra of Willdenow, and is thus without name. As it seems to me distinct from the more southern A. perennans, I shall venture to re-instate it, and subjoin, for fuller comparison, a revised description of the other species.

Agrostis campyla, (mihi): culmis e basi geniculato ramoso erectis glabris; foliis lanceolato-linearibus planis striatis scabris vaginis glabris; panicula diffusa ramosa ramis 4-6-verticillatis breviusculis flexuosis patentibus divaricatisve scabris; floribus oblongis acutis glabriusculis, glumis inæqualibus acutis s. cuspidatis carina inferioris sæpius scabra superioris glabra margine scariosislin. longis; palea ovata longiuscula glumam superiorem vix haud æquante acuta glabra. Agrostis scabra, Tuckerm. Obs. in Sill. Jour., xlv, p. 44, non Willd. e Gray. Trichodium scabrum, Muhl. Gram. p. 61. Torr. Fl. U. S., p. 83.

HAB. New England. New York, Torrey. Pennsylvania, Muhl. A stout, erect grass, with broad, lanceolate-linear leaves, a diffuse panicle with more or less flexuous branches, and oblong smoothish florets, the glumes being considerably unequal, and the upper one mostly smooth.

A. perennans, (Tuckerm.): culmis e basi gracili geniculato ramoso erectiusculis procumbentibusque glabris: foliis patulis longiusculis linearibus planis striatis scabris vaginis lævibus; panicula tenui-elliptica demum oblonga laxiuscula ramis verticillatis erectiusculis scabris; floribus lineari-lanceolatis acuminatis, glumis angustatis acutissimis s. cuspidatis carinis scabris circiter lineam longis subæqualibus; palea lineari-lanceolata glumis breviori acuta glabra. Sill. Journ., xlv, p. 44. Cornucopia perennans, Walt. Fl. Carol., p. 74. Trichodium perennans, Ell. Sk. 1, p. 99, tab. v, f. 2. Agrostis Cornucopia, Fraser! in Gent. Mag. 59, p. 873, cum Icone. Agrostis anomala, Willd. Sp. 1, p. 370. Trichodium decumbens, Michx. Fl. 1. p. 42. Muhl. Gram., p. 60. T. scabrum, Darlingt. Cest., p. 54, nec Muhl.

HAB. Carolina, Fraser! Pennsylvania, Darlington! Southern Ohio, Sullivant! A delicate grass, sending up many assurgent, slender culms from the procumbent base. Leaves rather

long and lax, narrow, linear. Panicle slender, at length somewhat diffuse. Florets linear-lanceolate, acuminate. Glumes narrowed, nearly equal. Palea lanceolate-linear. A plant first brought into notice by Walter, who seems to have had rather extravagant notions of its agricultural value, and afterwards collected and cultivated by Fraser, who published a figure of it in the Gentleman's Magazine, October, 1789, together with a specific character prepared by Dr. Smith. The name proposed by Fraser should perhaps take the place of Walter's, whose account is very imperfect. The principal characters in the above descriptions have been noticed, more or less, by the different authors who have published the species, and seem to be constant in all my specimens.

CERATOSCHENUS MACROPHYLLUS, (Sp. nov.): cymis compositis; spiculis gracilibus patentibus; nuce oblongo-obovata basi acuta compressa lævi, setis filiformibus duplo-stylo persistente subtriplo-breviore; foliis angustatis rigidis glabris culmum superantibus.

HAB. Plymouth, Mass. New Jersey, Dr. Knieskern! Found by me at Plymouth in 1839, and distributed afterwards under a provisional name which has not since been taken up by the friend who proposed it. C. macrostachys, A. Gr. (Rhynchospora macrostachya, Torr. in Gray Rhynch. n. 14,) of which the present has been considered a state, has closely fascicled, and (especially the axillary ones) somewhat simple cymes; erectish, stout spikelets; broad-obovate (exactly spoon-shaped with the tip truncate) nuts which suddenly taper to the produced base, the bristles more than twice as long, and the style more than four times as long as the nut; and softish leaves which are scabrous on the margins and shorter than the culm. In the species now proposed all the cymes are compound and rather loosely flowered; the spikelets slender and spreading; the nuts oblong-obovate or rather pyriform, tapering evenly to the acutish base, the bristles about twice as long, and the style more than three times as long; the leaves narrowed, rigid, smooth, overtopping the culm. These characters appear to be constant both in the New Jersey and the Plymouth plants,

ART. XXI.-Descriptions of Shells found in Connecticut, collected and named by the late Rev. J. H. Linsley; by AUGUSTUS A. GOULD, M.D.

SOME months since, specimens of the shells indicated and named by the Rev. J. H. Linsley as new, in his "Catalogue of the Shells of Connecticut,"* and of which he intended subsequently to give full descriptions, were put into my hands by his daughter, with the request that I would examine them, and conclude the work which he commenced so well. I have therefore done so. Some of them, I am well satisfied, had been previously described; others are new. The following are the results of my examination.

ASTARTE MACTRACEA. (No. 71.) Testà parvâ, solidâ, subtriangulari, sed anticè rotundatâ et ad basim arcuatâ, concentricè costato-undulatâ, inter undas radiatim striolatâ, fulvo-viridi, fuscoradiatâ, apice acuto; areolâ magnâ, profundâ.

Lat., alt.

poll. Fig. 1.

Fig. 2.

Astarte mactracea.

Shell small and solid, nearly quadrant-shaped, the apex acute, somewhat behind the centre, with a divergence of nearly a right angle, the posterior and basal margins regularly curved while the anterior margin is nearly a right line. The surface is undulated with about fourteen concentric, rib-like waves, and is marked between the ribs with very minute, regular radiating striæ. The color is pale yellowish green, with fine pencillings of dusky radiations. The areola is very large, deep and broad.

There is only one valve of this shell, but its characters differ so widely from any described species as to allow no doubt of its being distinct. The figure accompanying Mr. L.'s catalogue is not characteristic, the posterior side being too much excavated.

CYTHEREA MORRHUANA. (No. 85.) This seems to be quite a young and small specimen of C. convexa, Say.

UNIO PEQUOTTINUS. (No. 105.) This is undoubtedly a valve of U. latus, Raf. (U. rectus, Say). It is said to have been found in connection with Indian bones and other articles. As this shell

* See this Journal, vol. xlviii, p. 271.