1. _Meloseira._--Valve punctate, with a constriction or furrow between edge of valve and girdle. 2. _Gaillonella._--Valve punctate, with a circular collar or crest near edge of valve. 3. _Lysigonium._--Valve punctate, neither keeled nor constricted. 4. _Hyalodiscus._--Valve punctate in the centre; border with decussating radial lines. 5. _Stephanopyxis._--Border of valve with a crown of thorns; valve areolate. 6. _Pyxidicula._--Valve areolate, with a border of spines. {14}MELOSEIRA AG. (1824), em. DE TONI (1892) (melos, a limb or member, and seira, a chain) Frustules globose, ellipsoidal or cylindrical, concatenate, closely joined together. Valve either simply punctate or punctate and areolate. A constriction of the cell-wall, forming a furrow between the edge of the valve and the girdle, is more or less evident. The genus Meloseira constituted by Agardh has been variously modified by Kuetzing, Thwaites, Wm. Smith, Van Heurck, De Toni, and others. In Systema Algarum Agardh included certain species of Conferva, of Lyngbye, Dillwyn and others, and limited his genus to frustules more or less globose (fila articulata ad genicula constricta), although in his Conspectus Criticus (p. 64), he modifies the description (fila teretia articulata, articulis diametro æqualibus vel longioribus) to include M. varians. As, however, Lysigonium Link, Gaillonella Bory, and other genera enlarged by Ehrenberg and Kuetzing, came to be included under Meloseira, Thwaites suggested the division of the genus into two: Orthosira, in which the frustules are not convex at the ends and Aulacosira in which no central line is apparent but with two distinct sulci. Wm. Smith adopts the genus Orthosira but rejects Aulacosira, including all forms under the former genus and Meloseira, suggesting that differences "exist in the formation of the sporangia" of the two genera. M. varians and M. crenulata appear to form auxospores or sporangial frustules in different ways, as will be noticed hereafter. As, however, the present state of our knowledge is so limited and as much confusion would result in further changing the nomenclature, I shall adopt, for the most part, the division made by De Toni, separating Gaillonella and Lysigonium and employing the name Meloseira as emendated in Sylloge Algarum, although, as stated, it omits the species of Agardh. That a further division may be necessary is indicated by the differences existing between the Orthosira forms and the others. ANALYSIS OF SPECIES Frustules cylindrical and lengthened: Valves with two distinct furrows; granules small distans Valves with coarse granules granulata Valves denticulate on the margin crenulata Valves denticulate and constricted roeseana Valves with row of large puncta on the girdle side undulata Frustules cylindrical and compressed: