Cocconeis, Cymbella, Amphora and Amphiphora. 7 



represents two alate, hyaline appendages on each side, thus 

 giving the frustule the form of an hour-glass flattened. Valves 

 forming the primary surfaces thick, oblong, truncate, sometimes 

 striated (plicated?) longitudinally, closely approximated through- 

 out, suddenly compressed and narrowed at each end, supporting 

 on their sides the alate appendages mentioned. Endochrome : 

 flat, double lozenge-shaped, connected ; consisting of a single (?) 

 layer, continuous, presenting a transparent area in the centre in 

 which is the nucleus, and from which a few delicate branched 

 threads radiate towards the sides of the frustule ; chiefly confined 

 to the central valves, but occasionally extending for a short 

 distance into the hyaline appendages. 



Movements. — Like those of Diatomese in general, but with a 

 peculiar contortion, which brings one half of the frustule into 

 right angles with the other. Deduplication, through the broad 

 or primary surfaces. Size ygUo *^ 270 ^^ ^^ ^^^^ long. 



Hab. The brackish water in that part of the main drain of 

 Bombay which mixes with the sea at every tide ; abounding in 

 silty clots of Oscillatoria which float on the surface. 



Observations. — A clot containing several hundred specimens 

 of this species was placed in a small wide-mouthed bottle in the 

 middle of December, where they continued to deduplicate up to 

 the middle of the following June, by the vessel having been re- 

 plenished from time to time with fresh water. Long before the 

 last of these frustules were seen the whole of the other organisms 

 had perished, and the remaining contents of the bottle, which 

 might have been supposed to contain a number of them empty, 

 on being examined, were found to contain none ; hence it may 

 be inferred that the frustules are not siliceous or coherent. The 

 most striking feature about this species is its contortion (fig. 34), 

 which from the thinness and flexibility of the alate expansions 

 is effected to such a degree, that the narrow lateral edges, re- 

 spectively, in one half, are brought almost into right angles with 

 the primary or broad surface of the other ; a phase which this 

 Diatomean is continually assuming, and which, at first, is very 

 difficult to understand. For some time I mistook it for an Am- 

 phora with one half of the frustule split open, and the edges 

 turned back. The central valves are frequently marked with 

 longitudinal lines (folds ?), and fine lines may occasionally be 

 seen cutting each other at acute angles across them; while a 

 linear appearance also presents itself sometimes in the alate 

 appendages parallel to their borders Fig. 37 shows all this ; 

 and sometimes their hyaline transparency is interrupted by 

 white specks. 



This species differs from the navicular forms in the general 

 and greater flatness of its frustule, in the greater expansion of 



