centibus; intermedia perfectd biloculari; lateralibus sepiis dimidiatis. 
Pollen ...... nunquam invent. Ovarium superum, oblongum, triloculare, 
polyspermum, ovulis placente central affixis, horizontalibus. Stylus fili- 
formis. Stigma concavum, capitatum, triangulare, papillosum, (nunc 3-par~ 
titum laciniis bilobis, monstrosum, ut im icone). Capsula ex zcone D. Miers, 
oblonga, pallidé brunnea, torulosa, trilocularss, 3-valvis, polysperma: valvis 
medio septiferis. Semina parva nigra corrugata, funiculo nigro vestcato 
seminis ipsius magnitudine. 
nnn eee 
First discovered in ‘the neighbourhood of Valparaiso, by Mr. James 
M‘Rae, on his voyage to the Sandwich Islands, in the service of the Horti- 
cultural Society, on board His Majesty’s ship the Blonde, commanded by 
Lord Byron. It flowered in the Greenhouse in September 1825, within 
a few weeks of its arrival. An inconspicuous bulbous plant, thriving well in 
a cool greenhouse, planted in sand and loam. 
- The whole structure of this most remarkable plant is so peculiar, that we 
scarcely know whether the definition and description of the parts of fructifi- 
cation above given will not be considered more paradoxical than just: 
and yet if the analogies the various organs bear to those of.other plants be 
carefully considered, their structure will scarcely admit of any other interpre- 
tation. With respect to the five petaloid leaves, which are here described 
as bractese, and which bear a considerable degree of resemblance to a 
perianthium, it may be observed, that this appearance is more apparent than 
real. They neither correspond in insertion nor in number with the segments 
of a monocotyledoneous perianthium, nor do they bear the same relation to 
the parts contained as a perianthium should bear. The three outer are not 
inserted on the same line, but are distinctly imbricated at the base; and the 
two inner do not complete the second series, as would be required in a 
regular monocotyledoneous perianthium. 
But if we were to admit, for a moment, the possibility of these bractec 
being segments of a perianthium, what explanation could be given of the 
setiform processes proceeding from their base, or of the central fleshy slipper- 
like body from within which the stamens proceed? The former bear no 
determinate relation to the other parts of the flower in their insertion; they 
are subject to much ‘diversity of form and number, being sometimes eight, 
consisting of ¢wo unequal subulate bodies proceeding from each edge of each . 
lateral segment, the outermost of the two being wider than the innermost, 
and being, moreover, not unfrequently a manifest process of the margin of 
the segment itself; sometimes having their number reduced to four by the 
suppression of the exterior processes ofeach lateral sezment; and occasionally 
having the outer processes suppressed on one segment, and not suppressed on 
the other. In the many flowers which have been under examination, the pro- 
cesses, moreover, were always constituted of cellular tissue alone, without 
either trachee or tubular vessels. These circumstances being considered, it 
will scarcely be proposed, we presume, to identify them with abortive 
stamina. If they are, notwithstanding what has been advanced, determined 
to be the perianthium itself, what becomes of the outer segments which had 
previously been referred to perianthium? for it would be difficult to trace 
any analogy between the structure of Gilliesia and of those genera in which 
a third series is added to the usual senary division of Monocotyledones. 
