100 p. T. CLEVE, SYNOPSIS OF THE NAVICULOID DIATOMS. 



nensis and D. prisca. In others, as in Amphora egregia, we meet with transverse costse, alternating 

 with double rows of puncta or alveoli, exactly as in Diploneis gemmata, and many others. The 

 space , between the longitudinal and median lines correspond to the furrows of Diploneis. The 

 ventral side of the valve, which has been more modified than the dorsal side, retains still in many 

 cases a longitudinal crest, although all other structure has disappeared. In some forms the longi- 

 tudinal line of the dorsal side fades away, and there is left of it only a blank space, as in forms 

 of Amphora ovalis; in others no trace of it is perceptible, although other characteristics remain 

 marking the affinity with forms which possess distinct longitudinal lines. The forms with longi- 

 tudinal lines on the dorsal side may be classed in two groups, viz. with complex, and not complex, 

 connecting zone. I propose for the former the name Diplamphora, founded on their supposed re- 

 lation to Diploneis, and for the latter I reserve the name Amphora, sensu strictiori. 



There is another group of forms, which have a longitudinal crest not on the dorsal, but on 

 the ventral side. These forms, which I include in the group Calamphora, are of doubful rela- 

 tionship. 



Akin to Calamphora is a group of forms with a row of costse on the ventral side but 

 without the longitudinal crest. All of these, known at present, are fossil, and I propose for them 

 the name Archiamphora. 



A large group of Amphora is of the type of A. coffceiformis or A. salina. They are 

 without longitudinal lines, but have a complex zone, protracted, and frequently, capitate ends. 

 Their strise are more or less distinctly punctate. I include these forms in the group Halamphora, 

 but am unable to trace their connection with any of the divisions of Navicula. 



Another group of Amphora, characterized by a complex zone, absence of longitudinal lines, 

 and distinctly punctate strige, constitutes the the group Oxyamphora. There can scarcely be any 

 doubt that this group of forms, many of which have a stauros, is akin to the Microstigmatica 

 among the Naviculee, particularly the subdivision Stauroneis. 



The same may be the case with the two groups Amhlyamphora and Psammamphora, both 

 characterized by the direction of the median line, absence of longitudinal lines, by their finely 

 punctate strise, but differing in the simple or complex nature of the zone. 



As Cymhamphora I regard forms, which have great resemblance to Cymbellce, but not di- 

 stinctly punctate strise. The zone is simple, not complex. I am unable to trace their connection 

 with other groups of naviculoid diatoms. 



There remain some forms, which I cannot comprise under groups above named, and which 

 I treat of in an appendix. 



If the above named large groups of Amphorse were admitted as distinct genera, which I 

 believe they ought to be, the synonymy would be still more intricate than it is at present. I 

 propose for this reason, that the species of the different groups should retain their generic name 

 Amphora, which in all cases signifies that they are asymmetrical Naviculse. This will also afford 

 an opportunity of testing my views, which are entirely new. before admitting the proposed 

 new genera. 



In many Amphorse, belonging to different groups, a peculiar, structureless, very hyaline 

 limbus occurs, which seems to be a flat plate projecting from the dorsal side of the valve. I am 

 in doubt how to regard this peculiar organ, which perhaps may correspond to the wing in the 

 genus Tropidoneis. As specimens of the same species are found with and without this limbus, it 

 seems not to be of great importance for the distinction of species. 



Subgenus Amphora Cl. 



Frustules in outline usually elliptical, with truncate ends. Connecting zone broader on the 

 dorsal than on the ventral part, without longitudinal divisions and not transversely striate or 

 costate. Valve asymmetrical, more or less lunate. Median line biarcuate. Dorsal part of the 



