358 
I have never had the opportunity of examining Philydrum. Brown has 
lately proposed to constitute it into a distinct order ; for the character of 
which I have translated his generic description in the Prodromus, with a few 
omissions. 
Geography. New Holland plants. 
Properties. Unknown. 
GENUS, 
Philydrum, Banks. 
Garciana, Lour. 
Group IV. 
Essential Character. — Leaves either with many ribs the intervals between which 
are irregularly netted, or with a mid-rib and netted sides ; foot-stalk taper, articulated with 
the stem. Embryo without a latteral slit. Flowers never arranged in a spadix. Floral 
envelopes complete. Twiners or climbers. 
By the strongly marked peculiarity in the foliage, it seems to me desira- 
ble to cut off the following orders from all others. They have a most peculiar 
aspect, and it is plain that they constitute a most remarkable transition from 
Endogens to Exogens. Nevertheless if the organs of vegetation are not taken 
into account, I know of no verbal character which will divide Smilaceae from 
Liliacese, and Mr. Herbert is equally unable to distinguish Dioscoreaceae from 
Amaryllidaceac. I see no objection to the admission of such a character, pro- 
vided it is constant and uniform, and that it is so I hope to shew : first 
remarking that Alismaceae are separated from Ranunculaceae upon a still more 
slender verbal distinction, the embryo of the former being undivided, and the 
other 2-lobed at the apex ; there is no other positive difference between these 
orders, which in fact represent two lines diverging from a common point ; 
yet they are universally recognized, and with great propriety. The first point 
in the character assigned to this group is that the leaves are like those of 
Exogens, that is to say, irregularly netted between the ribs, or ribless with 
netted sides, and that they disarticulate from the stem, a circumstance which 
is indicated by their tapering footstalks. Tliere are no other Endogens in 
which such a character exists, except certain Aracese, to which I shall advert 
presently. The nearest approach to the structure with which I am acquainted, 
occurs in Lilium giganteum, but the leaves of that plant have a flat foliaceous 
petiole and do not disarticulate. The broad-leaved Amaryllidaceous plants 
like Griffinia, Eurycles, &c. are totally different ; their leaves not only having 
no articulation with the stem, but having no reticulations between the ribs 
further than is produced by the anastomosing of the fine parallel secondary 
veins that connect the ribs. In this circumstance Roxburghia alone among 
Retosee corresponds with them ; but the petioles of that genus being rounded 
and disarticulating from the stem, a sufficiently plain distinction still remains. 
From some Araceae the structure of the leaves will not divide Retosae ; and 
it was for that reason that in my Nixus, in which I first proposed to form 
this combination, I actually regarded the group of Retosae as a mere alliance 
among Spadicosae ; but the lax inflorescence and more perfect floral envelopes as 
an external, and the embryo without a lateral slit for the emission of the plu- 
mule as an internal character, will prevent all possibility of confounding Retosae 
with Spadicosae. 
I regard this group as passing directly into Menispermaceae through 
Ripogonum, but as preserving a close connection with Liliaceae through Rox»- 
burghia, and with Amaryllidaceae through Dioscorea. 
