2 Rev. W. Smith on the Diatomaceae, with descriptions 



figures, sometimes fail in enabling the student to arrive at the 

 satisfactory determination of species. 



It appears to me that somewhat of this difficulty might be re- 

 moved by a subdivision of this extensive genus, and by adopting 

 as generic distinctions those minuter diversities of structure 

 which the latest improvements in our microscopes have enabled 

 us to detect, and which seem to me not less important than those 

 upon which former observers have relied. 



In pursuing this course I shall restore the genus Pinnularia of 

 Ehrenberg, rejected by Kutzing, and adopt the term Pleurosigma, 

 as descriptive of another group. 



The genus Navicula of Kutzing and other writers will then be 

 resolved into three, whose characters may be given as follows : 



Pleurosigma, mini. 

 Valves convex, sigmoid, striated ; striae resolvable into dots. 



Navicula, Bory. 



Valves convex, lanceolate or elliptical, smooth or striated ; stria? 

 resolvable into dots. 



Pinnularia, Ehren. 



Valves convex, oblong or elliptical, ribbed or pinnated with 

 distinct costae, not resolvable into dots. 



The first of these genera is well distinguished by its sigmoid 

 form ; the two latter contain species which approach each other 

 in outline, but may easily be discriminated by the character of 

 their strise, which invariably present in Pinnularia the appearance 

 of continuous ridges or costse, and which no increase of magnify- 

 ing power enables us to resolve into separate markings ; while 

 in Pleurosigma and Navicula, the strise which are apparent with 

 a low magnifying power as simple lines, become on the applica- 

 tion of more powerful glasses, or the careful adjustment of the 

 light under which the valves are viewed, distinct series of dots or 

 beads probably arising from elevations on the surface of the valves. 

 Jn Navicula it occasionally happens that the elevations are com- 

 paratively distant, and may be recognized with a power of 200 

 diameters or under ; but in Pleurosigma the resolution of the stria? 

 into their constituent beads becomes a task of extreme difficulty, 

 and has from this circumstance been very generally adopted by 

 microscopists as a means of testing the denning power of the 

 object-glass. It now seems to be generally admitted, that 

 although the detection of the striae on Pleurosigma requires the 

 object-glass to be of accurate construction and considerable power, 

 yet the resolution of these striae into beads depends chiefly on 



