T)i\ Grcvillc on some British Diatomacese, 257 



Odontidiuin anomalum, W. Srn. PI. IX. fig. 7-9. 



Odontidium anomidum, W. Sin., Ann. of Xat. Hist. vol. xv. p. 7. 

 pi. 1. fig. y. 



A new and very interesting diatom, di.scovered last summer 

 by Professor Smith in the Ccveimcs at au elevation of about 

 4000 feet, and by Professor Balfour and Mr. Lawson in Braemar 

 immediately afterwards. It was exceedingly scarce, and mixed 

 with Melosira disians, Kiitz. This species is described and figured 

 by Smith in his paper on the Diatoniaceai of the South of 

 Prance, recently published in the ' Annals of Natural History.' 

 Its chief character rests on the linear valves, constricted towards 

 the obtuse extremities; but the very peculiar structure of the 

 valves as seen in the front view affords besides a most conspi- 

 cuous distinction. This structure arises from the presence of 

 " internal cells," which Professor Smith regards as an abnormal 

 condition of the filament, because frustules of the more usual 

 description (as occurring in other species) may be occasionally 

 found side by side with others containing internal cells. Theo- 

 retically, this view may be correct ; but it is certain that in the 

 only specimens hitherto observed in France and Scotland, the 

 filaments are almost exclusively composed of frustules containing 

 the internal cells, and which thence derive a most marked cha- 

 racter; while the normal frustules are, as Professor Smith justly 

 remarks, rarely to be detected. It would be a very curious fact 

 if this diatom should prove to be known everywhere in an 

 abnormal condition — and that that condition should be one of 

 increased development. This so-called abnormal structure — 

 which according to my view differs considerably from that exhi- 

 bited in the figure whicli illustrates Professor Smith's paper — is 

 very beautiful under the microscope. Each frustule in the figure 

 referred to contains an internal cell, elliptical or linear-elliptical 

 (as the filaments happen to vary in breadth from "OOOG" to 

 •0015"), converging to a point at each end, and traversed longi- 

 tudinally by a median line. The following, on the contrary, is 

 the result of my own observations, made, I may add, before I 

 saw Professor Smith's paper. ^Vhen a frustule becomes abnormal 

 it is divided by an elliptically curved line, the centre of the 

 curve appi'oaching very close to the lateral margin, while the ex- 

 tremities of the line terminate nearly in the middle of each end 

 of the frustule. On each side of the termination of the line are 

 situated the minute nodules which are present in every condition 

 of the diatom. A single frustule is thus subdivided into two 

 cells, as will be more readily perceived by referring to the figures. 

 The same arrangement precisely, takes place in the adjoining 

 frustule, only the direction of the line is reversed ; so that when 



Ann. ^- May. N. Hist. Ser. 2. Vol. xv. 17 



