393 NEW PUBLICATIONS. 
the three forms more than marked varieties of one species; but we 
: 
B 
QR 
z 
B 
[-» 
& 
3.2 
B 
om 
S 
4 
T- 
- 
B. 
: dà 
E 
p 
m 
E 
B 
= 
Oo 
Qu 
e 
TE 
> 
S 
e 
[27] 
& 
et^ 
o 
E. gibba,—species that differ so strikingly in every particular. Hi- 
mantidium undulatum is rightly referred to H. pectinale ; and we pre- 
fer uniting the three Cyméelle (lanceolatum, Cistula, and eymbiforme) 
with our author, to retaining them as distinct species, as Rabenhorst 
does in his recently published ‘ Flora Europzea Algarum.' 
Pinnularia must, it seems, by common consent, be allowed to slip 
from among the genera of Diatomacea, notwithstanding Smith's at- 
tempt, by a careful amended definition, to establish it on a more satis- 
factory basis. It is indeed difficult in the smaller species to determine 
whether the valves are striate or costate, and consequently difficult or 
impossible to refer them certainly to Navicula or Pinnularia ; but we 
confess that we give up the genus reluctantly, as we have found it very 
useful in the practical work of determining species, and we still hope 
that more perfect instruments, and better application of light, may en- 
able us to consider Smith's characters as of specific value. While, 
therefore, we cannot object to M. Heiberg, in the present state of our 
knowledge, referring all the Pinnularie to Navicula, we most decidedly 
protest against his uniting in the same genus plants so diverse in their 
orm and structure as the free and naked Navicule proper, the 
stipitate Doryphora, and the frond-invested genera Berkeleya, Schizo- 
nema, and Colletonema. It is true that if all our characteristics of Dia- 
tomacee are to be derived from the siliceous frustules, and if the 
peculiarities of the perfect plant are to be ignored, these genera cannot 
be distinguished ; but surely we have in natural history got beyond the 
period of artificial classifications based on a single organ. Smith has 
erred less in dividing the whole class into two tribes characterized by 
the possession of naked or enveloped frustules, than M. Heiberg in 
uniting in the same genus plants belonging to both of these well- 
marked tribes. It would be absurd to urge in favour of such a classi- 
fication that we cannot tell whether fossil species were stipitate, en- 
closed in a frond, or fre. We cannot set aside the remarkable cha- 
racters and important information derived from the living plants, and 
make a retrograde step, the only benefit of which would be to hide our 
ignorance. On the same ground we object to Zxcyonema, with its 
filamentous frond, being united to the stipitate Cymbella, and Homeo- 
cladia to Nitzschia. 
