NATURAL HISTORY SOCIETY OF GLASGOW. 221 



collected the sample, he had better opportunities of examining it. 

 My friend Mr Eylands was (at least at one time) of opinion that 

 N. birostmta was a mere variety of N. cusindata, and that S. 

 craUcida was a diseased state of the same. I am not quite dis- 

 posed to adopt these views; but, from the frequency of N. 

 Urostrata, where S. cmtkula is also got rather copiously, not only 

 in the Marton deposit, but in a slide I possess from Mr Gr. 

 Norman, from the ruins of Persepolis, I am inclined to think that 

 N. Urostrata and S. craticula are the two valves. of the same species 

 — the former being probably the upper valve. At one time I 

 considered that N. Urostrata was the lower layer of the valve of 

 S. craticula; while the portion with ribs formed the upper layer, 

 as I have mentioned already when alluding to Cocconeis Grevillei; 

 and the frequency of valves with costse and no striae confirm this 

 view — the chief objection being the usually much more radiating 

 striae of the one than of the other. It is to be hoped that ere 

 long 5^. craticula may be got in abundance, in a living state, in its 

 breeding ground ; for by obtaining entire frustules the mystery at 

 present attending it can only be solved. 



I have already alluded to Denticula, as allied in some points to 

 Nitzschia, and in others to TryUionella, with which it is more 

 closely connected. In the autumn of 1864 Mr Rylands, of War- 

 rington, and I endeavoured to clear up the British species; and I 

 will conclude this paper with a synopsis of them which we then 

 made out. But first let me state that Mr Smith in his observa- 

 tions on D. ohfusa and D. tenuis has rested the specific distinction 

 on what is of no value (the shape of the frustule and valve), and 

 omitted that from the costae on the valve. In his figure, indeed, 

 he has represented the side view of the frustule, not the valves, 

 and probal^ly this led him to misunderstand the structure. 



In some species the costae are much ddated at the base, and 

 these dilitations form ocelli, when the front view (F. V.) of the 

 frustule is examined, similar to those seen in Epifhemia argus, E. 

 ocellata, and a few others. The costae are attached to one margin 

 only of each valve, and this margin, in the entire frustule, may, in 

 the lower valve, lie either directly under the sunilar margin of the 

 upper valve or under the opposite margin; in other words, the 

 costae, in the one case, are attached to the adjacent angles of the 

 same frustule, in the other, to the alternate angles. By striae or 

 costae pervious, is meant that they extend from the one margin of 



