GREVILLE; ON NEW DIATOMS. 233 



Hub. Cliimborazo, Barbadoes deposit ; C. Johnson, Esq. 



There can be no question but that this magnificent species, 

 for beauty and interest, stands at the very head of the genus, 

 distinguished, as it is, for many fine forms. The only two 

 examples as yet knoAvn were both discovered by my acute and 

 indefatigable friend, Mr. Johnson. The first is a fragment ; 

 neverthelesb, as it possesses a perfect angle and centre, and 

 the greater part of two sides, admits of being correctly described. 

 The other is entire, and, although considerably smaller, is a 

 splendid object. Both enrich my cabinet — the first by the 

 kindness of ]Mrs. Bury, the second by the no less generous 

 donation of the discoverer. It is at once evident that this 

 unrivalled species belongs to the small group of Triceratia, of 

 which T. marginatum of Brightwell has been considered the 

 type, the valve having a central triangular space, and a broad 

 border, divided into compartments by transverse lines or costal. 

 It further belongs to the section of that group containing the 

 species just named and T. pulchen'imuin, which have the 

 central triangular space filled up (not blank), suggesting the 

 idea of a small Triceratium being laid (after the manner of 

 patchwork) in the middle of a larger one. It is, however, 

 essentially separated from those previously described by the 

 conspicuously cellulate, lace-like structure of the Avhole valve. 

 The pseudo-nodules are very large, more or less ovate, filling 

 up the rounded angles, densely punctate, the puncta becoming 

 larger and passing into evident cellules at the base. Beneath 

 the pseudo-nodules is a broadly obconical vacant space, equiva- 

 lent to the part which has been called a second pseudo-nodule 

 in T. marginatum, as well as to the unpunctate portion of the 

 pseudo-nodule in T. pulcJierrimum, but has so little in common 

 Avith the pseudo-nodule that it will be more correctly described 

 (when present) as the pseudo-nodular blank space. The cen- 

 tral triangular compartment is very similar in outline to the 

 same part in the two species above mentioned ; but, instead of 

 being filled with radiating costse, is largely cellulate, with the 

 addition of a few costse radiating from the centre. The two 

 specimens now before me indicate a considerable range of 

 variation. In the valve first discovered (the largest) the 

 angles are much less rounded, the sides straighter, and the 

 cellulation more oval. The pseudo-nodules are ovate, and the 

 somewhat undulating costaj are eleven or twelve on each side. 

 In the specimen figured, the pseudo-nodules are obovate, the 

 cellulation rounder, and the lateral costse only five or six. In 

 this example the cellulation at the base of the pseudo-aiodule 

 extends into the blank space, and has the appearance of being 

 torn at the edge as if it had been originally continued through- 



VOL. II. NEW SER. K 



