NEW FORMS OF DIATOMACEZ. 485 
middle, the two halves broad and rounded, with subacute extremities. Median 
line broad; central nodule square, definite. Length about 0:0045”; greatest 
breadth 0:0018’. Striation coarse, strongly moniliform, not reaching the median 
line, but leaving a narrow blank space on each side of it. Strize about 18 in 
0-001, much inclined near the apices. 
This form, which I had found frequent in the Glenshira sand, is also fre- 
quent in the new dredgings. It is regarded by many as a variety of NV. didyma, 
and by others as a variety of V. Crabro. Iam disposed to consider it a distinct 
species, on account of its peculiar and very constant form, and also because it has 
a decided light-brown colour in balsam, under a low power, which NV. didyma has’ 
not. It is much larger than N. didyma. As to N. Crabro, the moniliform structure 
in it is always obscure, and the form is also different. I cannot perceive that J. 
Bombus passes into either of these species by intermediate forms. But whatever 
be the ultimate decision on this point, I give it here.as the form called by Euren- 
BERG NV. Bombus ; which, be it species or variety, is at all events conspicuous, and 
very constant in its characters. 
In my last paper on the Glenshira sand, I have figured several Naviculee and 
Pinnularic of the panduriform group, and I have pointed out that this remark- 
able group requires thorough investigation. In order to contribute towards this 
end, I have figured the present form, as well as V. Musca, a new member of the 
same group; and I shall describe, farther on, another, namely a remarkable form 
of Pinnularia Pandura, Bréb. 
13. Navicula Lyra, Ehr. P1. IX., fig. 18 and136. Form oblong-elliptic, broad ; 
often contracted to short produced extremities. Length from 0-002” to 0:0045’ ; 
breadth 0:0007” to 0:0018”. Median line fine, interrupted by a large indefinite no- 
dule, extending transversely. On each side of, and in contact with, the median 
line, is a linear, somewhat broad, striated band; and this is separated from the very 
broad, marginal, striated band by a narrow linear blank space. These linear blank 
spaces are, in each half, united by their base to the extremities of the nodular 
blank. They bend outwards from this point, then inwards, and finally again 
outwards at their extremities, thus forming, in the entire valve, two lyrate shapes 
united by their bases. Hence the name. The lyrate character is often much 
more decided than in the specimens figured. The extremities of these lyrate 
blanks generally reach the margin of the valve near its apices, but sometimes 
fall short of this, as in the figure. Strize about 22 or 24 in 0-001’, somewhat in- 
clined near the apices. 
This species, which occurs in the Glenshira sand, and is scattered through all 
the dredgings here mentioned, has been described, though not as I have described 
it above, in the 2d volume of Professor SuitH’s Synopsis. Professor SmirH seems 
to have seen only a variety, to be presently mentioned, which does not possess the 
lyrate character, and has therefore omitted that character. He refers toa figure 
VOL. XXI. PART IV. 6P 
