AM ARY LLIDACEiE. 
55 
foramen, but is never truly approached to the umbilicus, 
except because the umbilicus is near the foramen ; the um- 
bilicus is situate between the foramen and summit of the 
seed ; the breadth of the umbilical cord reaches to the fora- 
men, wherefore, if the foramen is near to the umbilicus, as 
in Hippeastrum, the cord is slender ; if it is remote, as in 
Curculigo, wider, that is, occupying the intermediate space 
(which, together with the umbilicus, constitutes properly the 
hilum), more extensive than the mere umbilicus. The 
ovules of the testaceous seeds appear to me to proceed at 
right angles from the axis, the umbilical channel making a 
circuit by the line of the raphe to the summit, which is the 
region of the chalaza, and there furnishing nourishment to 
the cotyledon ; and after fertilization the ovule appears to 
slope, the summit being lowered and the foramen propor- 
tionally elevated. The ovules of the fleshy seeds seem to be 
either attached by a short thick cord, or to adhere with a 
wider hilum to the placenta. 
My own observations would lead me to believe that there 
is a very minute communication with the foramen, soon be- 
coming obsolete, along the margin of the umbilical cord, but 
I have not had the use of a powerful microscope, and I can- 
not venture to dissent from the received opinions concerning 
the fertilization of ovules, though from their extreme impro- 
bability my understanding does not willingly yield assent to 
them. 
First Division. — BRANCHING. 
Suborder 1 . Xerophytea:. — Leaves rigid. 
1. Xerophyta. Jussieu . — Flowers terminal. 
% Flowers not terminal. 
2. Vellosia. Vandelli. ( Campderia . Kth. Raddia. Rich.) 
— Perianth scarcely tubed ; filaments some- 
times numerous, superfluous. 
3. Barbacenia. Vandelli . — Perianth tubed. 
Second Division.— CAULESCENT. 
Suborder 2. HypoxidiLe. — Schistandrous, i. e. anthers open- 
ing their whole length. Operculate, i. e. the 
point of the germen superior, and forming a base 
to the style. 
