SPECIES OF ZANNICHELLIA. 265 
The fruit, which, according to the figures of M, Reichenbach (Icon. 
Crit.), appears to furnish such excellent characters, is by no means 
constant in nature, and nothing is more variable than the size of its 
dorsal membrane, which is also either very denticulate or nearly en- 
tire, the same specimen frequently presenting the greatest differences. 
Of the value of its modifications we may judge from our figures of 
the more remarkable forms. Fig. 1 belongs to a Z. vepens from St. 
Cloud; fig. 5 to a repens from the environs of Carlsruhe, sent by 
M. A. Braun; fig. 6 to another repens from the environs of Paris. 
We perceive that, in the first instance, the dorsal margin presents 
three strongly denticulate membranes, and the anterior margin two 
much smaller membranes ; in the second, the margin is simply crenu- 
late ; in the third, it has only a single denticulate membrane. The 
figs. 2, 3, and 4; belong to Z. major; the specimen from which the 
carpel No. 4 was taken, and which approaches most nearly the figure 
of Reichenbach, was found in the department of the Lower Rhine by 
M. Buchinger ; 2 and 3 belong to the same form, but from the envi- 
rons of Paris. Figs. 8 and 9 are the fruit of the same specimen of 
Z. palustris, W. brought from Algiers by M. Bove. If 9 had more 
teeth, it would correspond exactly with Z. gibberosa, Reich. which ~ 
name we might be induced to apply always when the anterior margin 
is furnished also with a denticulate membrane ; but I conceive that it 
must be reserved for plants which possess at the same time a very 
long style, because that author partly characterizes his species by the 
words fructus macrostyli, the length of the style being, in reality, 
the only character at all certain in the fruit; and, although this may 
vary slightly, it may be remarked that in the varieties with four-celled 
anthers it is always nearly equal to the carpel, whilst in those with two- 
celled anthers it only slightly exceeds the half of the carpel ; it being 
clearly understood that our examination must be made when the fruit 
is mature, at which time it is the only character remaining to distin~ 
guish Z. dentata from Z. palustris, inasmuch as the stamens are 
then generally fallen and the stigmas withered. 
Although the length of the style furnishes a sufficiently good cha- 
racter, it is not so with its direction ; generally it is curved backwards, 
that is, its concavity is towards the back of the carpel, but in the same 
specimen it is sometimes found quite straight, or even curved for- 
wards. Nevertheless, in all the species, it possesses one common 
character, that of being enlarged at the base, and bearing on one, at 
least, a projecting line, which is continued on the carpel sometimes 
along the middle, sometimes a little nearer to the back. 
VOL, IX., NO, XXVII. 34 
