214 PROCEEDINGS OF THE 



valve only. For those wlio are interested about the productions 

 of the Clyde, I may add that G. splendida, of Gregory {f. 29), is 

 not different from C. jmnctatissima, Greville, from the West Indies, 

 nor is either a really distinct species from C. mousii, Sm., and 

 perhaps all of them Ijelong to C. punctata, oi Lobarzewsky. So 

 far as my observations go — but I have obtained no specimens from 

 the weed — the upper valve has no true median line or central 

 nodule; and, on the other hand, G. crebrestriata, of Greville, 

 appears to me to be the lower valve of his G. jmndatissima. In C. 

 mousii there are 10-12 striae in -001, the dots forming the striae 

 being 15 in -001. In the same from West Australia, the striae are 

 12-13 in -001, and the dots also 15 in -001. In G. pimdatissima 

 the striae are 15 in -001, but the dots are likewise 15 in -001; so 

 that although the striae do vary a little in number, even in the 

 same sample, the dots comprising them are isometrical. 



Some species referred to Cocconeis may be here noticed. 

 One is C. GrevilUi; Smith's figures and description are apt to 

 mislead. He says (Syn. i., p. 22), that his figure, tab. iii., f. 35 

 a is the lower valve, f. 35 (V^ the upper valve, and f. 35 d 

 the upper valve from within: this is not the case; f. 35 a is pro- 

 bably the npper valve; but if so, the nodule and median line are a 

 pictorial embellishment only, the upper valve possessing neither; 

 f. 35 a* exhibits both valves combined, and, if the artist was not 

 deceived, f. 35 a is the same again; f. 35 a is the lower valve, but 

 not quite correct, in so far as the horizontal costae or ribs are 

 represented as interrupted in the middle by a longitudinal 

 rugged band, which I have never seen, and must have been 

 accidental. .This lower valve consists of two layers; when both 

 are united, the median line and nodule of the lowest plate are very 

 obscurely seen; this has the striae fine and very unlike those on 

 the upper valve: sometimes the upper layer separates; on this are 

 the costa attached to the margin, and no striae, but these costae 

 leave a few granules on the lower plate, marking where they 

 adhered to it. If there be two layers to the upper valve, I have 

 seen nothins: which I would consider the inner one. The lower 

 layer of the lower valve has the striation, median line, and nodule 

 of C. major of Gregory, but I have never seen it quite destitute of 

 the granules I have noticed above : if such specimens are found, it 

 would render it difficult to say whether C. major ought to be 

 referred to G. GrevilUi; or, as I have done, to G. pseudomarciimta. 



