184 Transactions of the Royal Microscopical Society. 
In a preparation of Moller from Porto Kico I observed a still 
more interesting abnormity, viz. the multiplication of the valves 
within the frustule. Here the connecting membrane was very 
short, and the enclosed frustules (numbering 1, 2, 3 or more) be- 
came gradually smaller as they approached the centre like a nest of 
pill-boxes. PI. CXCVI., Fig. 4, b. W. Smith has delineated a similar 
growth in Orthosira Dickiei, and I have observed a like abnormity 
in a small variety of Orthosira clistans. Can this be a craticular 
state of the Melosireas, similar to that of Navicula cuspidata, N. 
Perrotettii Meridian circulare, and Odontidium mesodon, and 
which Pfitzer, in his beautiful paper on the Bacillaria, compares to 
the resting spores in other algae, and the thickening of the mem- 
brane for their protection from unfavourable influences? But in 
what manner do these abnormal frustules multiply and reproduce 
a new series of normal forms? Certainly not by conjugation or 
self-division. 
The inner frustules of Cerataulus leevis are sometimes very 
small, and it is impossible to comprehend how these small frustules 
can reproduce by self-division frustules of the average size ; the 
diminution of the frustule is always the result of self-division, and 
not the contrary. Much still remains to be done towards the 
interpretation of these and other facts connected with the repro- 
duction of the Diatomaceae. 
I may here add that many genera of Biddulphieae and Eupo- 
discefe are very doubtful. It is now generally admitted that 
Amphitetras and Amphipentas ought not to be separated from 
Triceratium, and it is even difficult to find a sufficient distinction 
between the latter genus and Biddulphia. 
Triceratium striolatum Ehr. ( - T. membranaceum Bright- 
well, T. Biddulphia Heiberg) is the triangular form of B. rhombus. 
(See Cleve, ‘ Bihang. till. k. Svenska Akad. Handlung,’ Bd. I., tab. 
xiii., fig. 2.) 
T. spinosum is a triangular form of B. yranulata Koper. 
\T. spinosum is often quadrangular and B. reticulata sometimes 
three-sided. — F. K.] 
Hydrosera triquetra Wallich, is also only a triangular variety 
of T. compressa Wallich. And Cerataulus leevis sometimes occurs, 
as we have just seen, with three processes. 
[Cleve (l. c., tab. iv., fig. 3, a, b) figures a frustule of Biddulphia 
aurita (which he considers sporangial), resembling our Fig. 4, b, 
inasmuch that a small frustule has been formed within it,* but in 
neither case can I imagine the larger frustules to be sporangial. 
* la the August part of this Journal, p. 75, I suggested the probability that the 
endoehrome may under certain conditions possess the power of producing (? by 
means of microspures) perfect frustules without conjugation. This would perhaps 
explain the abnormal conditions just described. 
