204 WALKER-ARNOTT, ON ARACHNOIDISCUS, ETC. 
side of the central nodule being smaller and differently 
curved from that on the other, whereas, in most genera of 
the group, these two portions are alike. 
The incurvation of the valves is at its minimum when they 
are in the incipient state, at its maximum when the frustules 
have separated after self-division. The proximate cause is 
the contraction of that half of the valves which in the disk of 
paper is next the eye, and its consequent thickening, for the 
quantity of silex is probably the same on both sides; but the 
cause of the contraction I cannot explain. By this bending 
in of the valves the central nodule, which was apparently 
marginal, now assumes a position sometimes nearly vertical 
to the line of fission, but sometimes reaching to only one 
fourth of that distance. The maximum incurvation and also 
the projection of the median line on the plane of the field of 
view (a line of double curvature) vary in the same species, 
although generally within certain limits. The true form of 
the valve would be nearly got by completing the apparent 
outline by the addition of the incurved part; but the apparent 
outline of the frustule itself, at the maximum incurvation of 
the valves, is found to be more easily observed and con- 
venient for description ; and to this, there is no objection, if 
those frustules only be used in which the connecting mem- 
brane is at its minimum breadth. The strize on the valve are 
precisely alike on each side of the median line, but from 
some being seen through the medium of the valves, and 
others presenting themselves directly to the eye, the former 
appear faint and hazy; the latter, between the median 
line and the outline, are therefore alone referred to, in 
their definitions, by the more eminent diatomists. From 
the minuteness of the object, its inequality of surface, the 
necessity of using high powers and these with a large angle 
of aperture, it is impossible to bring all the parts into view, 
without altering the focus, or to give a perfectly accurate 
representation of the structure except by models. 
The portion separating the two valves, or the connecting 
zone, varies much in all diatoms, but more so in Amphora than 
in most other genera. At first it is narrow, then it becomes 
broader and broader, but is in most species striated very 
differently from the valves, if striated at all.* The new or 
intermediate valves, which it projects, are at first destitute of 
* The circumstance of the structure of the connecting zone being the 
same, or very different from that of the valves, indicates two groups or 
sections of the genus. As this is quite independent of self-division, it is 
of more importance than the splitting of the connecting zone into several 
plates, at least until the cause of this be ascertained. 
