486 DESCBIPTION ov {Polygastrtca. 



species these additions are in the form of mammillae, and a good deal 

 resemble the mucous prolongations of some of the Pahnellece, a frua- 

 tule being situated towards the extremity of each. A mammillose and 

 somewhat areola te appearance is thus given to the indefinite frond, 

 whereas, in D. ulvoides, the newly developed additions to the gelatine 

 cohere to fonn a compact, even membrane. 



This species is named after its discoverer, Mr. Dansey, who finds 

 it in small quantity upon rocks on the tidal shore of the river 

 Tamar. 



In a paper published by Mr. Ralfs, subsequently to that by 

 Mr. Thwaites, with the foregoing details, the former gentleman 

 writes : — 



" DicKiEiA Damiseii (Thwaites) does not belong to this genus, since 

 its gelatinous matrix forms an irregular mass, and not a plane frond. 

 Its frustules also diifer, being decidedly siliceous, striated, and 

 having a longitudinal pellucid line and central pimctum (aperture, 

 Kiitz.) in the lateral view." 



Genus Ekctonema (Kiitz.) — Cymbellse (frustules) disposed in 

 longitudinal rows, included in very soft, simple, gelatinous, tubular 

 filaments, (colourless under the microscope.) 



This genus belongs to the family Cymhellece, ofKiitzing; it was con- 

 founded by Ehrenberg with Gloeonema. It(B-alfsAnn. N^at. Hist. v. 16, 

 1843, p. 11) "differs fi"om Sehkonetna, and other frondose genera of 

 Diafomacea, in the form of its frustules, as a single frustule resembles 

 one of CymlclJa or of Cocconema ; therefore, notwithstanding the simi- 

 larity of habit, it belongs to a different series." 



" Certain bodies of a totally different kind have been mistaken for 

 examples of this genus, especially congeries of ova of different insects ; 

 but these ova, although cymbiform and aiTanged in longitudinal 

 series, are neither siliceous nor striated- 



" The lateral surfaces of the frustules being convex, are observed 

 in the front view, in which also the frustules are quadiilateral, with 

 two puncta at each end. These pimcta are less easily discerned in 

 the dorsal view, as the dorsum is longitudinally convex. The 

 lateral view is semi-elliptic, with numerous transverse striae, which 

 are interrupted, as in Cocconema, by a longitudinal pellucid line." 



Kiitzing describes two species: "but (says Mr. Ralfs,) I doubt 



