WALKER-ARNOTT, ON MARINE DIATOMS. 173 
valve only which is undulated, a character which often dis- 
appears after the valves have been separated from the con- 
necting zone by beimg boiled in acid; the margin and the 
connecting zone itself is not undulated, as in several genuine 
species of Cyclotella. 
Amphiprora paludosa is said by Smith to be found “in 
fresh or slightly brackish water ;’ and here I shall suppose 
his figure to indicate the form intended ; this has a peculiar 
indentation on each half of the valve at the place where the 
ala or keel meets the body of the valve; but this indentation 
is not noticed in the text, and does not seem to have been 
considered of any importance by Smith. I have no slide 
from him; but I have examined the same form from Wool- 
wich, Hull, the coast of Norfolk, Ayrshire, &c., and invaria- 
bly accompanied by marine Diatoms. I therefore consider 
it, like all others of the genus, as peculiar to salt or at most 
brackish water. I have before me also a form without the 
indentation, and with striz (sometimes very obscure) on the 
body of the valve; which is thus intermediate between the 
figured A. paludosa and A. alata on the one hand, and between 
it and A. duplex on the other: so that I cannot avoid drawing 
the conclusion that these four forms belong to one natural spe- 
cies, recognised by the alate carina, and tendency to twist when 
the valves are separated and either dried or put into balsam. A. 
Ralfsit has the same tendency to twist, but the keel is not alate. 
In the same way A. lepidoptera represents those that have an 
ala but do not twist, while A. vitrea and elegans conjoined form 
a fourth species, without an ala, and with no tendency to 
twist. All other species of the genus may, as appears to me, 
be reduced to one or other of these. 
Amphora minutissima is said to be always parasitic and from 
fresh water ; but Mr. Okeden has found, in South Wales, two 
forms of apparently the same species: one in a salt-water 
marsh, parasitic on Nitzschia sigma ; the other in fresh water, 
and not attached to anything. 
Coscinodiscus minor of Smith is stated to be from fresh 
water; but what he had described he afterwards satisfied 
himself was only the detached valves of Melosira nivalis. 
Kiitzing’s (and I believe Ehrenberg’s) species of the same 
name is from the sea or brackish water, and is not uncommon 
at the mouth of the Clyde, and on many parts of the east 
coast of England: it is figured in Gregory’s paper on ‘ Clyde 
Diatoms,’ as the 8.V. of his Orthosira angulata ; but certainly 
does not belong to it, for Orth. angulata in no respect differs 
from Orth. marina. I have one or two gatherings of Cose. 
minor from Cumbrae (from Mr. Hennedy), with the valves 
copious, but no trace of any Orthosira. 
