158 Rhodora [SEPTEMBER 
well-defined group, of the same grade as Irideae, Burmanniaceae, 
Orchideae, Scitamineae and Hydrocharideae, all of them clearly marked 
out by definite and important characters. It is generally admitted 
that the above suborders, here united under the Amaryllideae, agree 
in the most important characters derived from the flower and seed, 
differing from Hydrocharideae, Orchideae and Burmanniaceae in 
their albuminous seeds, from Scitamineae and Orchideae in their 
regular (or only oblique) flowers, from Irideae and Burmanniaceae in 
their centripetal (not centrifugal) inflorescence and in their stamens, 
from Taccaceae and the majority of Orchideae and Burmanniaceae in 
their axile placentum, from Dioscorideae in their hermaphrodite 
flower, and in all cases there are other characters either less constant. 
or of minor importance . . . Taking therefore the Amaryllideae 
as a whole as one Order, it would include besides the five tribes or 
suborders here enumerated . . . the Vellozieae . . . and 
the Alstroemierieae . . . in which however the secondary in- 
florescence appears to be centrifugal." Under the tribe Hypoxideae 
he placed Hypoxis and Curculigo. 
Bentham and Hooker in their Genera Plantarum (1883) made some 
changes in the five tribes of Bentham. These tribes they called 
Hypoxideae, Amarylleae, Alstroemierieae, Agaveae and Vellosicae. 
Under the Hypoxideae they placed the genera ? Campynema,? Pauri- 
dia, Hypoxis and Curculigo. 
Pax, writing in Engler and Prantl’s Die nattirlichen Pflanzenfamilien 
(1887), placed under the family Amaryllidaceae the subfamily Hypoxi- 
doideae, and under the latter he placed the tribe Hypoxideae, con- 
taining the genera Curculigo and Hyporis. He placed Campynema 
in a subfamily by itself parallel to the Hypoxidoideae; and Pauridia 
under the Haemodoraceae with the comment, that while Bentham 
and Hooker are not certain that it belongs in the Amaryllidaceae, 
neither is it certain that it is any better placed in the Haemodoraceae. 
'The reason for all this uncertainty lies in the fact that while the plant 
has the habit and seeds of a tiny Hypoxis it has only three stamens. 
Baker, in his Synopsis of the Hypoxidaceae (1878), followed the 
plan of Bentham and made the Hypoxidaceae a tribe of the Amarylli- 
daceae. Here he collected four genera Hypoxis, Curculigo, Moli- 
neria and Pauridia, which he characterized in the following manner:— 
