OF INDIA AND CEYLON. 237 
a^ree with H. Griffithii in the characters of the thallns and 
of the floral shoots, also in those of the spathe and its mode 
of opening, differing only in the fruit. I am inclined to lay 
very great stress, as mentioned above, on these broad points 
of agreement in thallns and spathe characters, and think 
that this group of species forms a very natural genus so far 
as our present knowledge of the order goes. The large well- 
developed primary axis of H. olivaceum also, which in many 
respects recalls the tall shoots of Sphærothylax, and is far 
more developed than the primary axis of our Podostemons 
and Dicræas. forms another argument in favour of generic 
separation, at least until we know the characters of the 
various species of the Eupodostemeæ better. It is true that 
Hydrobryum now forms a genus in which there is great 
variety in the fruit characters, usually considered of generic 
importance, but this is nothing unusual, and the same is the 
-Case in the next genus to be considered, Farmeria. I there- 
fore propose, as in several other cases, to reinstate Tulasne’s 
genus with the same sub-genera. 
(1) Euhydrpbryum, Tul. Fruit 
with 12 ribs, isolobous. 
Stigmas usually cuneate or 
obcuneate, toothed. Thal- 
lus crustaceous 
(2) Zeylanidiuni, Tul. Fruit 
with 8 ribs or none, aniso- 
lobous, one lobe deciduous. 
Stigmas deltoid, cuneate, 
or subulate, sometimes 
toothed. Thallus crusta- 
ceous, lobed, or ribbon-like, 
regularly branched. 
Fruit sessile, smooth 
Fruit stalked, 8-ribbed. 
Thallus crustaceous. Sta- 
mens short 
H. Griffithii, Tul. 
H. sessile, Willis. 
H. olivaceum, Tul. 
