398 
characterised by the adherence of its frustules to Algae of 
larger growth. This character is true of most of the species, 
and even where not strictly applicable, as in E. gibba and 
others, we detect a disposition to rely on a foreign body for 
support, their frustules being usually embedded in the mucus 
of some member of the family Palmellaceee.” Recently I 
have had an opportunity of ascertaining that Epitliemia gibba 
is strictly an adherent species, and that its supposed excep- 
tional character in this respect has been based on insufficient 
observation. 
In the month of August last, when on a visit to Sir Walter 
Calverly Trevelyan, Bart., at his seat at Wallington, in the 
heart of Northumberland, I observed in a large pond of clear 
water near the mansion an abundance of the greater bladder- 
wort ( Utricularia vulgaris) growing erect in the water. As 
the leaves and stems of this plant presented a cloudy appear- 
ance, characteristic of parasitic growths, I subjected them to 
a careful microscopical examination, and found that the 
appearance alluded to was produced by an abundance of 
adherent Diatomaceae, the species being, Epitliemia gibba, 
Ehrb. ; Epitliemia ventricosa, Kiitz. ; Epitliemia granulata, 
Ehrb. ; Epitliemia Sorex, Kiitz. ; Sgnedra capitata, Ehrb. ; 
Cocconema lanccolatum, Ehrb. All were exceedingly abun- 
dant, and it was an interesting fact to observe that Epitliemia 
gibba and E. ventricosa were attached, by one extremity, 
not by the ventral surface of the frustule, as is the 
case with the other members of the genus. E. gibba thus 
shows an affinity for the genus Sgnedra, being in other 
respects somewhat aberrant in its characters, and by no 
means typical of its own genus. 
Epitliemia ventricosa, Kiitz., I feel satisfied, is a small or 
undeveloped form of E. gibba, with which it agrees in every 
respect except in being very much shorter. In the Walling- 
ton gathering frustules are plentiful of every size between 
the two forms, and show the gradation between them. 
Grunow has expressed his belief in the identity of the 
two. With this view 1 fully concur; and, although Raben- 
horst believes them to be widely different (‘ Europ. Diatom.,’ 
p. 65), I feel assured that E. ventricosa should no longer 
rank as a species. The two forms are generally found asso- 
ciated in the same localities. 
It is a somewhat curious fact that immediately after self- 
division in Epitliemia gibba the newly formed valve of each 
frustule is quite linear, and without the gibbous middle it 
afterwards assumes. Ivutzing has delineated this peculiarity 
both in the larger and smaller frustules (‘ Bac.,’ pi. iv, 22, 
and pi. xxx, 9, b). E. gibba is subject to considerable varia- 
