RoEER^ on Biddulphia. 
5 
the outline, size, and arrangement of tlie valve and connecting - 
membrane in the perfect frustales of this genus; and to this 
cause we may trace the long list of synonyms appended to 
nearly every species. The determination of any constant 
characters upon which to fix their limits as a genus is, as 
stated by Professor Smith, a matter of some difficulty, but 
they all agree in having rounded or compressed frustules, the 
terminal valves having submarginal processes always placed 
near the extremities of the long diameter when the valve is 
oval or approaching lanceolate in outline, always more or less 
reticulate in structure, generally spinous, and united by a 
connecting membrane often of considerable breadth, which 
is also reticulated in the majority of the species, and in this 
and the genus Amphitetras frequently projects beyond the 
suture of the valves when undergoing self-divison. 
But though the exact definition of the genus presents some 
difliculty, when we come to consider what is or what is not 
a species, upon what grounds their limits are to be defined, 
and the terms in which any specific characters are to be 
drawn up, the variations that occur in different localities, 
and even in any abundant gathering, are so remarkable, that 
the greatest uncertainty prevails, and the most discordant 
opinions are entertained as to the limits that should be pre- 
scribed, or even upon what grounds the specific characters 
should be based. That Ehrenberg and Kutzing have erred 
in placing any dependence on the presence or absence of 
the spines on the surface of the valves, I think, is now 
generally admitted by all who have carefully studied the 
genus. The number of lobes in those species in which they 
occur, is also subject to great variation, ranging from three 
to seven in B. pulchella ; and if the Denticella polymera, of 
which only a separated valve appears to have been seen by 
either Ehrenberg or Professor Bailey, is correctly referred to 
B. Tuomeyii of the latter author — of wdiich, from the figure 
and description, I think there can be little doubt — we have in 
that species a variation ranging from one to thirteen lobes, 
each of which, if the system of Kiitzing was adopted, would 
have to be made into a distinct and separate species. But 
this ground of specific character has been abandoned by 
Professor Smith in the first-named species, and is equally 
without weight in any other. The connecting membrane, 
again, is immediately after self-division exceedingly narrow, 
but increases in breadth as the frustules approach maturity, 
till from being barely perceptible, it frequently exceeds in 
breadth, and is sometimes nearly double the length of the 
valves themselves, before the act of self- division is again 
completed ; and though in some species more or less areolate, 
