486 PROFESSOR GREGORY ON 
in vol. i. (fig. 152a@*) given as WN. elliptica, which is not lyrate; but the name 
given by EnrenBere proves that he regarded the lyrate character as a principal 
one. I have therefore figured it, in order to show that it occurs in Britain as 
EsReENBERG described it. 
14. Navicula Lyra, Ehr., var. 8, abrupta. PI. IX., figs. 14 and 146. Form 
usually oval, more or less elongated ; sometimes linear in the middle, broad, with 
parallel sides, and obtusely acuminate at the extremities. I have hardly ever seen 
it with contracted and produced ends, as is so often observed in NV. Lyra. Size and 
striation as in V. Lyra, but the blank spaces, which are linear, as in that species, 
instead of being recurved at the ends, or lyrate, bend inwards at the ends, so as 
to form two narrow ellipses meeting in the central nodule. These linear blanks 
in this variety stop abruptly at some distance from the terminal margin of 
the valve, which, in NV. Lyra, they often, though not always, reach. 
This form, which is frequent in the Glenshira sand, as well as in the dredg- 
ings, is that already referred to as having been figured in vol. i. of the Synopsis 
(fig. 152 a*), as N. elliptica, and since referred to as N. Lyra, in vol. ii. 
I figure it here, both that it may be compared with the NV. Lyra of EarEnserG, 
and that it may be contrasted with WV. spectabilis (fig. 10), which is supposed by 
some to be a form of V. Lyra. The form of the latter, and the fact that the 
blank spaces in it are not linear, but broad, and reach the margin, all which cha- 
racters are very constant; to which may be added the rich brown colour of N. 
spectabilis in balsam, under a low power, seem to me to be sufficient to distinguish 
it from N. Lyra. The reader is requested to compare fig. 10 with figs. 13, 134, 
14, and 140. The latter forms are colourless in balsam. 
15. Navicula Smithii, var. 8, fusca. P1.TX., fig. 15. Form an elongated oval, 
broad, with rounded ends. Length from 0-003” to 0:0063", and even more; 
greatest breadth from 0:0014’ to 0-0028". Median line narrow at the terminal 
nodules, which are a little within the apices; broad, and formed of three parts, 
all ending in expansions, on each side of the central nodule. Nodule large, broad, 
indefinite. Striation very coarse, and coarsely moniliform, not reaching the me- 
dian line, but leaving on each side of it a narrow blank line, terminating in the 
angles of the nodular blank. The whole spaces, taken together, form two very 
acute long triangles, base to base Strize about 10 in 0-001". At about one-third 
of the distance from the median blank lines to the margin, the striz are traversed 
by a strong, dark line, which is often, as in the figure, nearly rhombic, but is 
generally curved, though very slightly. This line is caused by a ridge or eleva- 
tion of the valve, and is very conspicuous. Valve thick, and highly convex, of a 
strong brown colour, in balsam, under the } or 3 of an inch objectives. 
This form, which is very conspicuous, occurs, like the two preceding, both in 
the Glenshira sand and in the dredgings; and in that of Professor ALLMAN from 
Lamlash Bay and one from Loch Fine, it is frequent. I give it as a variety of WN. 
