176 WALKER-ARNOTT, ON MARINE DIATOMS. 
pure. Doubts are thus thrown on the distinction of the 
species, or on the fresh- and salt-water localities assigned. 
Nav. amphisbena occurs in fresh water, brackish water, and 
in the sea. The first has the extremities capitate, and con- 
spicuously produced ; the second much less so (it is 3 of 
Smith). The third is smaller, with the ends scarcely at all 
produced, and is N. brevis of Greg.; it was named in March, 
1854, by Smith, WN. retusa, n. sp. (before De Brébisson’s 
other species of that name was published), but I afterwards 
satisfied him that it was a mere variety of N. amphisbena, and 
it is therefore not described in the ‘ Appendix.’ 
Pinnularia gracilis has a fresh-water habitat given; but 
a form of it, with a shorter beak, is not uncommon in 
brackish water. This was at one time considered by Smith 
a new species, and called by him P. curta; it is, however, 
omitted in the ‘ Appendix.’ 
Stauroneis pulchella is beyond doubt marine ; but there are 
two forms so different that some think them different species, 
and even refer them to different genera. St. pulchella, 
properly so-called, is got in all our rocky pools, and has the 
stauros (by the way, I wish some more correct name were 
devised, for stauros is not the transverse beam of a cross, as 
is here intended, but the whole cross itself) cuneate, or 
“ dilated towards the margin ;” and the frustule on a F.V. has 
the angles rounded. On the other hand the second form 
occurs in sand-gatherings, has the angles on a front view 
sharp, the ends of the frustule appearing truncate; and the 
stauros (?) is a large circular blank spot around the nodule. 
The striation in both is much the same, and different from 
what is usually observed in this or the allied genera; but St. 
pulchella is of atawny colour ; the other ofa dark bluish green. 
This latter form De Brébisson wrote me (February 21st, 1858), 
he proposed to call Navicula angulata : he had first found it in 
1852, and considered it, with N. pectinalis and retusa, as a 
group very distinct from the other species. As this group is in 
great confusion, the following observations may not be unin- 
teresting. It is not to me quite clear on what De Brébisson 
originally bestowed the name of N. pectinalis ; what Smith 
received from him, and supposed to be so, I have reason to 
think was N. retusa, of which Smith only knew the F.V., 
and in that state it is almost undistinguishable from JN. 
pectinalis of Smith’s ‘ Synopsis.” On February 12th, 1858, De 
Brébisson was disposed to consider the above N. angulata to 
be his N. pectinalis ; but a few days after relinquished that 
view, and formed the opinion that a curious and very distinct 
species (which he since had called Amphora? quadrata) 
