488 PROFESSOR GREGORY ON 
slightly expanded at the ends, the angles being bevelled. From the opposite 
ends the margin inclines very slightly, but visibly, to the middle. Nodules very 
conspicuous on the F.V., in which the striation also extends, on each side, to 
rather more than }th of the width of the frustule, which arises from the convexity 
of the S.V. 
I first described this species in my first paper on the Glenshira sand, in which 
the figure was not characteristic. I figured it again in the second paper; giving, 
however, a shorter, broader, and constricted form as the type, and the present 
one as a variety. I have since found it frequent in all the dredgings, but espe- 
cially abundant in one from Loch Fine, and am now satisfied that the linear form 
is typical and the broad constricted form a variety. I give the peculiar and 
characteristic F.V. for the first time. The S.V. in fig. 18 0, is that of a broad in- 
dividual of the linear type. It is generally narrower, and often even no more 
than half this width. The shortest specimens are often still narrower. The 
broad, incurved form, at first regarded as the type, is very scarce in the dredgings, 
compared to the linear form. 
This form has been supposed to be identical with WV. firma, Kiitz, var. 8; but 
its marine habitat at once negatives this supposition ; and, besides, its aspect and 
colour are quite different. NV. firma is brown, while N. maxima is of a pale straw- 
colour. The striation in NV. firma is coarser and more conspicuous; and, lastly, 
N. firma is broader, has acute extremities, and yields several marked varieties, 
such as Enrenserc’s VV. dilatata and NV. Amphigomphus ; while the only observable 
variety of N. maxima is the shorter, broader, incurved one, represented in fig. 2 
of my second paper on the Glenshira sand. 
19. Pinnularia (2) subtilis, n. sp. Pl. UX., fig. 19. Form linear rhombic, very 
narrow, with elongated apices. Length about 00035"; greatest breadth about 
0:00025”. Nodule definite. Costze about 28 or 30 in 0:001"; transverse, slightly 
inclined towards the apices. 
This form occurs in Lamlash Bay. I do not feel quite sure about its genus. 
It may be a Navicula. The whole form is delicate and translucent, and it is far 
from conspicuous. 
20. Pinnularia rostellata, n. sp. Pl. TX., fig. 20. Form linear, broad, with 
acuminate ends, terminating in short, acute apiculi. Length from 0-002” to 
00027”; breadth about 0.0007’. Central nodule definite. Costze strong, subdis- 
tant, inclined near the ends, reaching the median line, about 14 in 0-001”. 
This pretty form occurs both in Lamlash Bay and in Loch Fine. It is not 
frequent, but I have been able to examine a considerable number of specimens, 
which are quite constant in their character. 
21. Pinnularia Allmaniana. P1.1X., fig. 21. Form elliptic-lanceolate, broad, 
extremities subacute. Valve highly convex on one side, concave on the other. 
Length from 0:0016” to 0:0026”; breadth from 0-001” to 0:0014". Coste appa- 
