tioned, and in this country the Bergia amrnannoides has, in Tamul, received that of Neer-mel-neripoo, or water-fire, a curious coincidence of names in countries so remote.
Remarks on the Genera.
Dr. Arnott and myself following Delile, referred the genus Bergia to Elatine whether judiciously or not may be doubted, now that I find the dehiscence of Bergia is septicidal while that of Elatine, as appears from the statements of those who have examined it with care, is loculicidal. This distinction combined with the difference of habit, of the two sets of plants, might I think with propriety be employed as a distinction to keep them generically separate, notwithstanding the similarity of their flowers, on which account, I, in this work, retain the old name of Bergia for the Elatine verticellata and E. ammanioides of our Prodromus, to which work however, I refer for the distinguishing characters. The accompanying plate represents a species of each genus.
EXPLANATION OF PLATE 25.
A. 1. Plant of Bergia ammannoides. Natural size. 2. Portion of a branch slightly magnified to show the stipules and aggregated axillary flowers. 3. Flowers opened, showing the sepals, petals, stamens, ovary, and stigmas. 4. Stamens separate, back and front views. 5. Mature fruit. 6. The same after dehiscence, the persistent calyx removed. 7- Capsule cut transversely. 8. A seed. 9. The same cut transversely. All more or less magnified.
These figures show the strong tendency that exists in this species, to variation in the number of the parts of the verticels of the flower, 3, 4 and 5 pieces occurring indiscriminately in different flowers, picked from the same stalk.
B. 1. Elatine ambigua. Natural size. 2. A plant removed from the soil and slightly magnified, showing its repent habit. 3. Portion of a branch more magnified. 4. Flower opened. 5. Capsule. 6. The same after dehiscence. 7. A seed. All more or lees magnified.
XXI.—CARYOPHYLLACEAE.
Since the publication of DeCandolle's Prodromus, in which this rather complex order occupies a large space, considerable changes have been made on it by different writers. These principally consist in raising his sub-orders to the rank of distinct orders, the removal of Elatineae, and referring some of the genera included by him to other allied, though in the linear series remotely situated, orders. This being for the most part an extra tropical order, these changes so far as they affect the few Indian genera can be easily pointed out without the necessity of changing the name of the order adopted, after DeCandolle, in our Prodromus. Following Dr. Lindley's arrangement, it being the most recent and perhaps the best, the Peninsular genera ranked under Caryophyllaceae by us are thus distributed. Gypsophila and Silene are referred to Silenaceae; Stellaria, Cerastium, and Arenaria, to Alsinaceae : and Mollugo to Illecebrece the Paronychiacece of our Prodromus. These three orders, along with PortulacecB and Tamariscinece are combined into one " Alliance" Silenales. distinguished by having the "Embryo rolled round mealy albumen; or if this is not the case, herbaceous plants with the joints of the stem tumid, or with scales replacing the leaves upon rod-like branches; almost all herbaceous, or small shrubs." The clause " with scales replacing the leaves" of this character refers to Tamariscineae. The orders are thus summarily distinguished—Portulaceae has two sepals : Silenaceae four or five united into a tube : Alsinaceae four or five distinct : Tamariscineae the dehiscence of the fruit loculicidal, seed hairy : and Illecebreae have leaves with stipules. In this last the dehiscence is also loculicidal. Silenaceae and Alsinaceae are represented in the accompanying plate, and Portulaceae and Illecebreae will be in a subsequent number.
The Caryophyllaceae, are distinguished by having a calyx of 4-5 sepals either united or free. Petals four or five, sometimes unguiculate, sometimes wanting. Stamens equal to, or double the number of petals ; when equal alternate with them. A single ovary of 2-5 united carpels either sessile or stipatate, with 2-5 filiform stigmas. Capsule 2-5 valved, one-celled or imperfectly 2-5 celled, the valves usually partially splitting at the apex, forming twice as many
