They also approach Myrtaceae through Fuschia, hut differ in the ahsence of pellucid dots and in having definite stamens. This last character equally distinguishes tliem from Philadelp/ieae. The filiform style and ahsence of albumen separate them from Halorageae, from which
they also differ in habit. They have also been compared with Loaseae but are readily distinguished by their binary, not quinary, arrangement of the parts of the flower which mark that
order, leaving altogether out of consideration the acrid stinging properties possessed by Loaseae.
Geographical Distribution. The species of this order are numerous, amounting according to DeCandolle's I'rodromus to 24S. including Trapa and Circcea, the greater p or . tion of which are American, though, to some extent, found in all countries from the extreme north, through every degree of latitude, to the equator, ami thence nearly 50 degrees south, but most abound in the temperate regions. In India, within the tropics, the species are few and except Ludwigia paroiflora and Jussitea repens, of rather rare occurrence. In the more temper- ate regions of the Himalayas thirteen species of Eptlobium have been found, one of which has been figured by Dr. Royle.
Properties and Uses. Of these almost nothing is known. The roots of the evening primrose jEnotkera biennis are edible, and it is cultivated for their sake.
Remarks on Genera and Species. In an order containing only two genera within the limits to which this work extends, there is not much room for remark on this division of the subject, yet I cannot altogether pass it over, as I think we have more genera than we can find good characters for. Jussicea is readily distinguished by having twice as many stamens as petals, but in both Ludwigia and Isnardia they are the same, and in truth, so far as I can discover, there is no good difference between these genera, the plant here figured, which is unquestionably a Ludwigia, seems to agree equally well with the character of Isnardia.
EXPLANATION OF PLATE 101.
Ludwigia parviflora.
1. Small plant — natural size.
2. Expanded flower.
3. Stamens.
4. Capsule cut transversely.
5. Ovary cut longitudinally.
6. Mature capsule splitting, showing the ripe seed.
7. Ovary cut transversely, 4-celled, with several rows of ovules in each — all more or less magnified. Nos. 4 and 7, are misplaced.
SUB-ORDER CIRCÆEÆ.
This sub-order consists of but one genus, essentially distinguished from all the true Onagrariae on the one side, by the cells of its ovary having only one ovule, and from Hydrocaryes on the other, by its ovules being erect not pendulous.
From Halorageae it is equally distinguished by its erect, not pendulous, ovules, and still further by its ex-albuminous seed. The following is Lindley's character of both the sub-order and genus.
" Calyx superior, deciduous, tubular, with a two- parted limb. Petals 2, alternate with the lobes of the calyx. Stamens 2, alternate with the petals, inserted into the calyx. Disk large, cup-shaped, filling up the whole of the tube of the calyx, and projecting beyond it. Ovary 2 (or 1) celled, with an erect ovule in each cell ; style simple, arising out of the disk ; stigma emarginate. Fruit 2 (or 1) celled, 2-valved, 2 (or 1) seeded. Seeds solitary, erect ; albumen none ; embryo erect ; radicle short, inferior. Herbaceous plants. Leaves opposite, toothed, stalked. Flowers in terminal and lateral racemes, covered with uncinate hairs."
Affinities. These have been already pointed out under the preceding order.
Geographical Distribution. They have only been found in the northern hemisphere. The species though few in number, have a wide range over the northern regions of Europe, America and Asia— Royle has figured one from the Himalayas and I have one from both the
