72 Transactions of the 



not been deemed valid, but the writer is not sure that it has been 



OTerthrown as respects many species. 



An examination of the Pinnulariae on MoUer's " type shde," led 

 to the belief that the costse of such diatoms as P. viridis, ndhilis, 

 &c., had been misunderstood. That instead of broad irresolvable 

 ribs, a truer view exhibited fine Hues of beads springing from the 

 median band, with a furrow between them, and in that furrow 

 another line of beads at a lower level. The curvature of these 

 valves not only renders the exhibition of this structure veiy difficult, 

 but presents it under perspective aspects by no means easy to 

 understand. Occasionally a valve is met with that proves com- 

 paratively easy, and then no one would doubt the complexity of the 

 so-called costae, though much speculation would arise as to the most 

 correct interpretation of the various perspective views that can be 

 obtained. 



The general aspect of valves whose costae can be resolved is that 

 of a number of hues or rows of beads springing from the median 

 band in a series of narrow arches with long sides, in shape some- 

 thing hke ladies' hair-pins, with hollows between them, in which 

 one or more rows of beads may be discerned. 



To investigate these appearances and the various gradations in 

 dififerent species was found impracticable with sohds obtainable in 

 London, and accordingly Herr Moller was applied to through Mr. 

 Baker, and a considerable series, mounted dry, was lately received, 

 and forms the subject of the following remarks : — 



PiNNULAEiA D\TERRUPTA. — The curves of this diatom prevent 

 its coming well into focus with high powers. The beading of 

 the so-called costae is not difficult to show ; but the relation of the 

 " costae" to the median band is not clear. 



P. ATRiDis shows fine ribs composed of rows of dots ; depressions 

 between each pair of beaded ribs, with beaded rows in depressions. 

 The beaded ribs spring from the median band in various planes, 

 and give its edges a serrated aspect. "S\Taen a pair of the hair-pin 

 shaped ribs are reddish, the intermediate bead rows are bluish. The 

 median band composed of beads in various planes. 



P. KOBELis. — Very difficult, except in lucky parts, and valves. 

 Much hke viridis. The headings curve down to a furrow in the 

 median band. Median band beaded in complicated pattern. The 

 beads only visible when the hght comes exactly at the right angle. 

 The rows then seem to go down in a curve, and up again towards 

 the central keel ridge. It is to be expected that the perplexing 

 perspective presented by this object will occasion considerable 

 difierence of opinion as to its real shape, but those who have 

 sufficiently good glasses, and take the requisite pains with the 

 illumination, will see that the so-called costae are very difierent 

 from any published drawings, and that there are beads in the sup- 



