Dr. Grcville on some British Diatomaceee. 259 



sides not in the slightest degree curved. Two internal septa 

 run through the frustule, as in Grammatophora, and ujjou these 

 the so-called vittaj are develojicd, but which, from the causes 

 above mentioned, are so much abbreviated as to resemble elon- 

 gated coloured nodules ; for iu the shorter frustules their form 

 becomes merely oblong or even oval. Unquestionably there is 

 an evident affinity with Grammatophora; but upon the whole, 

 seeing that, besides the differences already referred to, all the 

 known species of the last-named genus are marine, I hoj)e it 

 will not be considered as a rash proceeding if 1 propose to 

 establish a new genus for the reception of this curious fresh- 

 water diatom. 



DiATOMELLA, nob. 



Frustules quadrangular (forming at first a piano-compressed 

 filament, at length sei)arating). Coloured vittre two, straight, 

 interrupted in the middle and at each end. Length -0004" 

 to -0010". 



Diatomella Balfuuriana. PI. IX. fig. 10-13. 

 Granimatophoral Balfouriana, W. Sm. MSS. 



The general characters of this minute species are visible with- 

 out difficulty under a magnifying power of 400 or 500 dia- 

 meters. The frustule is surrounded by a thickened border, and 

 is divided into three more or less equal parts by two straight 

 internal septa or bars not clearly defined externally, on each of 

 which, at about equal distances between the middle and ends, 

 are situated two dark-coloured short vitt?e, while at the ends 

 themselves the septa terminate in minute nodules. In the most 

 elongated frustules the coloured vitta^ are linear, but they con- 

 tract in proportion as the frustules diminish in length, until 

 they lose the character of vittaj and resemble nodules. The 

 middle portion of the frustule is blank, while the spaces between 

 the septa and the margin are transversely striated, but it requires 

 a power of at least 600 diameters to bring this character out. 

 The frustules of this diatom are not unfrequently seen in the 

 process of self-division, and one of these I have represented in 

 the plate. It will be perceived that a narrow separation has 

 already taken place, and that in each portion the lateral striae 

 are apparent, while as yet there is only one septum. The next 

 stage in the process would probably be the division of the single 

 septum into two, followed by the development of the blank 

 middle space. 



Melosira nivalis, W. Sm. 



A new species previously determined by Professor Smith, and 

 which will appear in the forthcoming second volume of his work. 



17* 



