} 
7 
SPECIES OF ZANNICHELLIA. 271 
north, where one species, Z. polycarpa, apparently quite distinct, 
appears. 
M. Fries appears to have found this species, as well as our two, in 
Sweden (Nov. Flor. Suec. mant. prim.) ; for he mentions the 
major, Benngh. under the name of palustris, and the repens, al- 
though with doubt. His pedicellata evidently appears to be the pa- 
lustris, Willd. ; and, on the other hand, if the Z. contorta is a good 
species, the south will also possess its species. The dentata, W. has 
not yet been found in Barbary, but this may be accounted for by rea- 
son of little more than the coasts having been yet explored; and, as 
we have remarked, the varieties of dentata are generally found in 
inland situations, and those of palustris near the sea. Besides, these 
plants are of a nature easily to escape the researches of travellers, and 
consequently we are prevented from tracing with exactness the limits 
of their vegetation. 
The difference in number of the cells of the anthers seems to us a 
remarkable fact, more especially as the number of these cells appears 
susceptible of variation in the same species, according to the observa- 
tion of M. Gay. When there are four cells the anther is biapiculate, 
which renders it probable that, in this case, there are two stamens ag- 
glutinated. It is true that, according to Micheli, the anthers are 
sometimes found with three cells ; but it appears to us that this diffi- 
culty may be easily overcome by supposing that in one of the anthers 
there is one cell abortive. A new genus from Madagascar ( Diplan- 
thera, Du Petit Thouars), published by Du Petit Thouars, entirely 
confirms our opinion, its stamens being formed of two bilocular an- 
thers, situated on the same filament, but at unequal heights (see figs. 
15 and 16). 
If we reflect on this variation in number of the anthers and carpels 
in Zannichellia, and on the presence of the number four in Ruppia 
and Potamogeton, a number altogether abnormal in monocoytledons, 
we shall be induced to consider it as probable that the flowers of the 
Potamez are spadixes (spadices), and that each carpel or each sta- 
men constitutes a flower ; which will approximate this family to the 
Aroidex, and remove it from the Juncaginez, with which it has such 
affinity. 
