NATURAL HISTORY SOCIETY OF GLASGOW. 217 



of puncta. This is apt to mislead, and many have in consequence 

 rested ahnost solely on the presence of puncta, and introduced 

 species into the genus which ought to have been placed elsewhere. 

 The great character of Nitzschia is that each valve has one con- 

 spicuous keel or angle, sometimes near the middle, but often near 

 the margin; but tliere is always a portion of the valve beyond the 

 keel. The puncta are confined to the keel. The keel is best seen 

 when the valve rests on its two margins, the keel being then 

 elevated ; when the angle or keel forms nearly a right angle, and 

 the valve lies on the broader half of the valve, the other half 

 appears to coalesce with the keel. Smith's figures 115 A. and C, 

 in Tab. XIV., will illustrate these positions. It results from this 

 striation that the zone which connects the two valves is not 

 parallel to either half of the valves, but is more or less oblique. 

 Smith has shown this in his ideal sections on Plate XIV., fig. 115 

 and 118, although the two figures he there gives are scarcely cor- 

 rect, exhibiting very convex sides to the valve on each side of the 

 keel — in general, the portions on each side of the keel are nearly 

 flat. Several species of Tnjhlionella have also got puncta at one 

 side of the valve, but these are strictly marginal ; there is no keel 

 or angle there, so that the zone is quite parallel to the valves. 

 This may be seen obscurely in T. acuminata, and more distinctly in 

 T. apiculata of Gregory, a species which, from not attending to this 

 character, Smith referred to his Nitz. duhia B. N. tliermalis, Grun. 

 (or Surirella tliermalis, Kg.), if I may rely on specimens from 

 M. de Br^bisson, is a species of Tryhliondla. I have seen but 

 sparingly what I consider the same, from the Victoria Tank at Kew, 

 collected in Feb., 1856. This Smith considered a strongly striated 

 variety of Nit. linearis. In Nitzschia the part of the valve on each 

 side of the keel has neither a keel nor depression; in Tryhlionella 

 the valve has almost always one or other. The same character 

 will distinguish Nitzschia from Denticula, one species of which — 

 now pretty well known, although not published, as D. clecipiens 

 Arn. and Ryl. — was considered by Smith, Avhen I first met with it 

 at Kew (in the sample already noticed), as a variety of Nitzschia 

 mimitissima, with conspicuous strife. N. curvula of Smith (whether 

 this be the Nav. curvula Ehr., I have not yet been able to ascertain, 

 but it appears to be so, although Kiitzing's figures from French 

 specimens seem different), or N. sigmatella of Gregory, has two 



keels or angles, one near each margin of the valve. It is, there- 



2c 



