144 
Wallich^ on the Diatom-valve. 
Were the structure hexagonal^ these changes could not 
occur in the foregoing order or number ; for, inasmuch as 
in any hexagonal arrangement only three series of lines are 
present_, being disposed on the same plane, the changes to 
the predominant series would take place six times only, 
namely, at every 60^ in a complete revolution, each series 
being twice repeated. 
It has been pointed out by Mr. Hunt {' Journal of Microsco- 
pical Science,^ vol. i, p. 175) that the boundaries of a portion of a 
valve, belonging to one of the diagonally marked group, to 
which moisture had accidentally gained access, were in strict 
accordance with the view of diagonal lineation ; whereas they 
were not reconcilable with any other view of the structure. 
A similar fact may, at any time, be witnessed in balsam- 
mounted specimens to which air has gained partial access, or 
in dry mounted slides, affected by the ordinary atmospheric 
moisture. 
Lastly, the lines of fracture, as before stated, invariably 
tally with the thinnest portions of the valves in the two groups, 
that is, with the diagonal series in the one, and with the 
longitudinal and transverse series in the other ; a result at 
variance with the indefinite lines of fracture observable in true 
hexagonal structures. 
Without reverting, then, to theoretical points, I would sum 
up the general conclusions for Avhich I conceive sufficient 
evidence has been adduced. They are as follows : — 
That the growth of the diatom valve ceases entirely, either 
at the period of its liberation from the connecting zone of the 
parent valve, or immediately afterwards. 
That, subsequently to this period, no change of configura- 
tion takes place in the siliceous valve, except along its 
margin, where fresh siliceous secretion may, under certain 
conditions, be produced. 
That the normal figure of all markings whatever is circular, 
or approaching thereto. 
That these markings are arranged on the surface of the 
diatom valve in a determinate order, according to the inhe- 
rent tendency of the species ; but that the ultimate figure of 
those markings is due to forces exerted upon the young valve, 
whilst in a plastic condition, and retained within the connect- 
ing zone of the parent frustule. 
And lastly, that variation in size, and in the degree of fine- 
reasons already given, they may be left out of the proof, if the experimenter 
desires, without in any degree vitiating the result; for, in this case, the 
change to the predominant series only, would occur four times, instead of 
eight, in a complete revolution, namely, at each 90". 
