266 OBSERVATIONS ON THE 
The dried stigma forms at the summits of the style a small hook, 
which frequently disappears ; in the flower (see fig. 11) it is peltate, 
slightly concave, papillose, and crenulate, in some of the varieties, 
glabrous, entire, and somewhat narrower, in the others, which charac- 
ters coincide with those derived from the length of the style and the 
number of the cells of the anther. 
As to the length of the peduncles, we should not have thought it 
worthy of mention had not previous writers made use of it in forming 
species ; for it furnishes only very unimportant characters, although 
it be sufficiently constant in the same specimens, appearing to depend 
on the state of the water, at least, the plant in which we have ob- 
served the longest peduncles, grew in a small brook of running water 
at Bone. The partial peduncle exists much oftener than the general 
one. It must be remarked that the employment of this character has 
led to confusion by forming a bad species out of a variety allied to the 
two primitive species; the Z. pedunculata, Reich. presenting, ac- 
cording to that author, two varieties, one a, stagnalis, the other }, 
which is the Z. maritima, Nolte. I quote the latter after Reichen- 
bach, but as I have observed that Z. palustris, W. is almost always 
found near the coasts, whilst dentata grows in inland situations, I be- 
lieve the quotation to be correct ; and as the herbarium of the museum 
contains specimens of the two varieties, communicated by the author 
himself, I have satisfied myself that the variety stagnalis corresponds 
with dentata, and the variety maritima with palustris. 
The number of the carpels is also very variable in the same speci- 
mens, appearing to be most constant when the number is small. 
They are frequently four in number, and groups of five are likewise 
found on the same stem. 
I must caution the student against an appearance which the carpel 
presents when, by maceration, it has lost” its epidermis; the dried 
cellular tissue is then exposed, and exhibits a tomentose aspect; the 
style appears larger and slenderer ; and the back of the carpel, in- 
stead of being bordered with a denticulate membrane, is furnished 
with a range of minute isolated points, as represented in fig. 7. 
The anther (see fig. 14) is oblong, two-celled, and surmounted by 
a minute point, or four-celled and biapiculate. This character ap- 
pears very constant, though M. Gay has remarked that his Z. digy- 
na, which has four-celled anthers, has them also, sometimes, two- 
celled.* 
* This diminution in the number of the cells of the anther must not, 
