92 
Hooker . — On Hydrothrix , 
be grouped under well-limited tribes. Returning to Hydro - 
thrix , it differs from other Pontederiaceae quite as much or 
more than Clematis does from other Ranunculaceae, but, 
having regard to its habit and characters, and to its being 
monotypic, I prefer to regard it as an aberrant genus, rather 
than as constituting a tribe of the order. 
Owing to the minuteness and extreme tenuity of the floral 
organs of this plant, and the difficulty of analysing them in 
herbarium-specimens, it is probable that errors in detail may 
be found in both the above description and in the drawing. 
I have, however, no reason to doubt their general accuracy. 
Mr. Gardner, in a note appended to the specimens sent to 
Sir W. Hooker, described the flowers as yellow, placed in 
pairs in a two-leaved membranous sheath, the perianth as 
6 -lobed, with the three lobes that are next the other flower 
more than half narrower than the other three. He further 
states that the plant is submerged and only flowers when the 
water has nearly left it. Having regard to the minute size and 
inconspicuous nature of the corolla in comparison with that 
of other Pontederiaceae, to the fact that I find pollen emitted 
when the flowers are still enclosed in the bracts, and that in 
this state the stigma is applied to the anther, it may well be 
that the plant is self-fertilised 1 , and these flowers cleistoga- 
mous. The figures 4, 5 and 6 represent the inflorescence in 
this stage : it will be observed in these that the anthers 
appear to be opposite the ventral face of the ovary, which 
is, I suspect, due to distortion of the organs under com- 
pression. 
Hydrothrix was elaborately studied by Mr. Benjamin 
Clarke, F.L.S., in 1858, who prepared analyses of it for 
Sir W. Hooker. Of these I have availed myself in preparing 
the accompanying drawing for the Annals, which is the result 
1 In Kerguelen’s Land, in 1840, I found Limosella in flower in a lake under two 
feet of water and several inches of ice, with the corolla closely folded over the 
sexual organs, and containing a bubble of air. The anthers were full of well- 
developed pollen, and the ovules apparently fertilized (Flora Antarctica, vol. ii. 
P- 334)- 
