KONGL. SV. VET. AKADKMIKNS HANDLINGAH. BAND 26. N:(> 2, 29 



Marine: Gullmarsfjord, Sweden! Atlantic coast of N. America (Lewis.)- 



This species, which I liave found at Fiskebiickskil (West-coast of Sweden) nioiig Zostera, is 

 scarcely silicious, the frustnles disappearing almost completely, if burnt on a glasscover. 



This very interesting species, according to Lkwis lives in colonies, and forms curved fila- 

 ments of 12 or more frustules. He figures such a colony in top-view, which shows that the 

 frustules keep in contact by means of the wings. 



Other species, which perhaps belong to Amphoropsis are Amphipr.? jmradoxa Grev. (Edinb. 

 N. Ph. J. XVIII p. 41 f. 21, 1863) remarkable for its coarsely moniliform stria-, and Ampli. 

 Thicaitcsiana Grev. (1. c. p. 183 f. D) notably distinguished by its strongly marked wings, which are 

 striate, other parts of the valve not showing any striation. 



Dictyoneis Cl. (1890). 



Valve elongated, panduriform or lanceolate. Median line straight with the terminal fissures 

 usiially in contrary directions. Valve with double structure; the upper stratum finely punctate, 

 with puncta disposed in decussating rows; ihe interior stratum irregularly reticulate with rounded 

 cellules. Marginal cellules frequently larger than the others, forming a row of false loculi. Con- 

 necting zone not complex. 



The first known species of this genus is Navicitla marijinata Lewis. The larger marginal 

 cellules of several species give to the valves some appearance of Mastogloia, in which genus seve- 

 ral forms of Dictyoneis have by various authors been placed. In the year 1877 Grunow expresses 

 the opinion that the marginal cellules seem not to belong to a separate plate as in Mastogloia, 

 but to the valve itself and proposes a new genus for Mast.? reticulata and the allied species. In 

 the year 1890 I proposed (Diatomiste I p. 14) for this genus the name Dictyoneis, founded on the 

 peculiar structure of the valve. Unfertunately I have in that paper committed an error, as I then 

 supposed the stratum with coarse reticulations to be the upper instead of the lower. 



The median line of most species of Dictyoneis is straight and ends in terminal fissures, turned 

 in contrary directions. The central pores are moderately distant, except in D. naviculoides, where 

 they are very approximate. On both sides of the median line are narrow structureless zones, which 

 are more silicious than the other parts of the valve and are united to the central nodule. 



The genus Dict3^oneis is well distinguished from all the other naviculoid genera. Its syste- 

 matical place is doubtful, as there are no intermediate forms connecting it with other diatoms. 

 The structure of the outer stratum of the valve as well as the terminal fissures ])oint perhaps 

 to some relation to Pleurosigma. 



The species are all marine and belong to warmer seas. Fossil forms occur at Oamaru and in 

 Hungary. 



This genus comprises forms, which are so closely connected, that the distinction of well 

 defined species is very difiicult, as is frequently the case with truly natural groups. Such charac- 

 teristics as the outline of the valve, the coarseness of the reticulation etc. are very variable. 



2. 



Artificial hey. 



rgiiial cellules larger than the otl: 

 — — not — — 



f Va: 



Marginal cellules larger than the others 3. 



— — not — — 2. 



Valves panduriform D. jamaicensis Grev. 



lanceolate D. naviculacea Cl. 



( Valve not constricted D. Thumii Cl 



< — slightly constricted 4. 



\ — strongly 6. 



