74 
Dr. Gregory, on the PosUtertiary 
the markings of fig. 36^ could be seen through them, as well 
as others, fallen in part asunder, which demonstrated the 
connection. I then named it in compliment to Dr. Greville, 
who had thus led me to trace the relation between the two 
forms. 
But although it is certain that the form fig. 36^ is a part of 
the entire frustule, fig. 36, it does not appear to be the half 
of that frustule, nor such a part or single valve as usually 
occurs in Amphorae; for if we suppose two such parts in 
apposition, as we may see in A. ovalis, or in several of the 
Amphorae I have here described, such as A. obtusa, the result 
would not be, as in ^. obtusa or A, ovalis, the entire form, but 
something quite different. I have seen two segments thus in 
apposition, but too late to be figured. The effect is exactly 
that of a section through the middle of the supposed frustule, 
and resembles what is seen when an orange is cut in half. On 
the other hand, the entire frustule, as we see it in fig. 36, which, 
after all, may be but half of the entire form, appears as if it 
were formed of parts like fig. 36^, which are thicker on the 
convex margin than on the concave one, placed next each 
other in the same way as the parts or segments of a 
melon. Of such segments we see the backs of seven or 
eight in the view seen in fig. 36. But as the frustule is not 
cylindrical, though convex for a large part of its periphery, 
but rather prismatic, it is probable that there are two sides, 
like that seen in the figure, opposite to each other, and two 
others at right angles to these, also opposite to each other, and 
perhaps narrower. The former, the broader sides, are convex, 
like the sides of a barrel ; the others probably flat. The 
latter most likely represent the rectangular median spaces in 
the common forms of Amphora, and the line of junction of 
the two half-barrels, as we may call them, will pass through 
these narrower sides. If this be correct, then the view, 
fig. 36, may be either that of the entire frustule or barrel, or 
that of the half-barrel seen from the convex side. 1 think I 
have seen specimens of all these views, namely, of the entire 
frustule, which is darker in colour, of the half seen from the 
convex side, and of the half seen from the flat side, or look- 
ing on the plane of a section through the middle of the 
narrower sides of the frustule. But I have not yet seen any 
specimen lying on the narrower side, and thus showing the 
line of junction. This is probably because the convex side 
is so much broader. 
It will be seen that the bars, which are the thick backs of 
the segments, converge like those of an orange or melon, to 
a point on the terminal surface, which is therefore also con- 
