Diatomaceous Sand of Glenshira. 
73 
two valves_, nor have I seen the detached valves of this form. 
As its structure much resembles that of the next species^ I 
shall not here dwell on it. (288.) 
(In the recent deposit so often alluded to^ I find A. crassa 
not unfrequent^ and I have there seen the detached segments, 
which are very remarkable — so much_, that I at first took 
them for an entirely new form.) 
36. Amphora Grevilliana, n. sp. This form, of which the 
entire frustule is represented in fig. 36, and the detached 
valve, or rather segment, in fig. 36"^, is the most remarkable 
of all the forms here described, as Synedra undulata was of 
those figured in the former Paper. 
Form of the entire frustule varies from rectangular, to 
somewhat elliptical, or rather broad elliptical, with broadly 
truncate ends, the ends being convex. Length from 0"003'' 
to 0'0045''. Like the preceding, the whole surface of the 
frustule is made up of vertical, or rather concentric bars, 
the middle one appearing straight, the others slightly curved, 
and converging on the ends. There are blank lines or narrow 
spaces between these bars, of which there are seven or eight, 
not so closely set as in the last species. As in the last, the 
lateral curve lines project half way across each valve in the 
middle, but curve back rapidly to the outer margin, along 
which they run to the ends. The bars are transversely 
striated. Striae conspicuous, but much finer than in A. crassa. 
The frustule has very much the appearance of a cask or 
barrel, but appears to be four-sided or prismatic, as seems 
also to be the case in A. crassa. Colour dark brown. 
The detached segment, as seen in fig. 36"^, is arcuate ; the 
dorsal margin convex, and in some long specimens straight, 
bending forward at the ends to a kind of projecting beak. 
The inner margin, which is faint, is nearly straight, l3ut be- 
tween the central nodule and each end it is slightly incurved. 
Within it is a much more strongly marked and more incurved 
line, the two halves of which meet in a strong nodule. Be- 
yond this again, is another line forming one curve, convex 
towards the inner margin in the middle, and concave at the 
ends. Beyond this is a stiU stronger line, which in the 
middle is concave towards the inner margin, and towards the 
ends is concave in that direction. The whole is marked with 
strong, conspicuous, moniliform, transverse striae. 
For a long time I did not connect the two forms, fig. 36 
and fig. 36"^. The latter I took for a form of A. Arcus, figured 
in my former Paper. But at last, when Dr. Greville pointed 
out some of the diff'erences, I returned to the study of it, and 
soon found specimens of the entire form, fig. 36, in which 
