DR. M. T. MASTERS ON THE PASSIFLORACEA. | 599 
nanthe the corona is similarly arranged in two rows: the outer one is in the form of a 
membranous tube or sheath, surmounted by lorig threads; the inner one is also mem- 
branous, but shorter and entire at the top. To the inner surface of this second corona 
the five stamens are congenitally adherent, or, more correctly speaking, not separate 
from it. 
In the genus Passiflora the form and arrangement of the rings of the corona vary in 
the different subgenera, and may, indeed, sometimes be used to distinguish species from 
each other. In Passiflora bilobata, Juss., the flower is described as destitute of corona, on 
which account it was made the type of a genus Astephananthes by Bory de St.-Vincent. 
Mr. Sowerby, in the second volume of the ‘ Transactions of the Linnean Society” *, has 
given accurate figures and descriptions of the arrangement of the corona in several 
species of the genus. In Passiflora quadrangularis, in which the coronal filaments are 
. in greatest abundance, he describes the following series, going from above downwards, or 
from without inwards:—1, perfect rays, consisting of two or three rows of distinct 
threads; 2, imperfect rays, consisting of a variable number of short and, as it were, 
imperfect or rudimentary threads; 3, a membranous ring, the **false operculum ” found 
in P. quadrangularis, and in one or two other species only; 4, a membranous sheath 
called the * operculum," which shuts off the nectary proper from the upper part of the 
flower-tube, and below which is sometimes to be seen, 5, a thickened rim projecting 
from the side of the nectary, and partially dividing it into two compartments, an upper 
and a lower; lastly, at the base of the gynophore is often to be seen, 6, a shallow mem- 
branous or thickened cup. 
Such an arrangement as that just mentioned pertains especially, with certain variations 
in individual species, to the subgenus Granadilla. It may be remarked that the several 
rows are often closely approximate, so that almost the whole inner surface of the flower- 
tube is lined with these outgrowths. 
In the other subgenera the number and arrangement of the parts of the corona are 
different. Mr. Sowerby's divisions are strictly accurate in the case of the species men- 
tioned by him, and apply very well to a large number of other species also; but, from the 
difficulty in some instances of ascertaining the exactly corresponding rows in different 
species, it is simpler to employ such terms as outer or inner, faucial, median or basilar, 
according to their position—or filamentous or membranous, according to the nature of 
the outgrowths in question. 
In the subgenus Astrophea the corona consists of two or three rows of filaments spring- 
ing from the upper portion of the tube; of these the uppermost or outermost are con- 
siderably the longest, and are usually more or less dilated and sickle-shaped at the ex- 
tremity. The innermost corona, the “ operculum,” emerges from near the base of the 
tube, at some distance from the outer ring, and consists either of a membranous tube or 
of a row of short filaments; in the former case the membrane is often inflexed. 
In the subgenus Plectostemma, Mast., including the sections Cieca, Dysosmia, Poly- 
anthea, Distemma, and Decaloba of other writers, the flower-tube is usually short and 
saucer-shaped; the rings of the corona emerge from the mouth of the tube; the outer 
* Trans. Linn. Soc. ii. p. 19, t. 3, 4, 5. 
