56 NEILGHERRY PLANTS. 
stamens inserted on the tube of the corolla. The stamens are either 4 or 5, the filaments 
sometimes, though rarely, dilated and cohering at the base. The ovary is one-celled with 
parietal placentas. In the genus Ewacum the margins of the 2 valves, of which the 
ovary and capsule is composed, are inflexed and somewhat free within the cell of the 
ovary, but more commonly they meet and coalesce, forming a parietal placenta, as shown 
in Gentiana. In Halenia, on the other hand, they enlarge at the point of junction, 
forming a thick fleshy mass, filling the whole of the centre of the cell and bearing two 
rows of seed on each side. This part of the structure is unfortunately very badly brought 
out in the accompanying plate, but is well shown in one subsequently prepared for my 
Illustrations of Indian Botany. And, lastly, in Ophelia there are 4 parietal placentz, a 
very unusual structure in the family, but which may be accounted for on the supposition 
of their being sub-marginal, that is, the ovuliferous margins, in place of being so deeply 
inflexed as to become loose like those of Ewacum, are simply folded in and adhere to the 
face of the valves, thus forming two placentas on each. This peculiarity seems not to 
have been noticed by writers on this family; Grisebach indeed describes the ovules as 
inserted on the suture, which is not quite correct. 
As already mentioned, this family has a very wide geographical range, its species ex- 
tending from either polar circle to the equator, thus at the same time luxuriating within 
a few feet of perpetual congelation and on the most arid plains of the tropics. But while 
this is true to the letter, we must not overlook the fact that, of those found within the 
tropics, the majority inhabit the cool mountain tops, a few only occurring on the plains, 
and most of these only arriving at maturity during the cooler seasons of the year. 
As regards properties, bitterness is the distinguishing characteristic of the family, hence 
many of them are highly esteemed as tonics and restoratives in domestic medicine, and 
several are in daily use for the same purpose in medical practice, the root of the officinal 
gentian (G. lutea) being among the most esteemed. 
In regard to the plant here named, Halenia Perrottetii, I may remark, that it is not 
that species but apparently a variety of H. elliptica—a species also found in Nepaul. 
The true H. Perrottetit has longish, ovate acute leaves, not short elliptic, obtuse ones, as 
in the plant I have erroneously so called. They are very nearly allied, if really distinct. 
EXACUM. 
» Calyx 4-5 parted, segments keeled or winged on the back. Corolla rotate, withering, tube becoming 
globose, limb 4-5-parted. Stamens 4-5, inserted on the throat; anthers remaining unchanged, opening by 
a pore-like slit, at the apex. Ovary spuriously 2-celled, the free inflexed margins bearing the ovules; style 
distinct, declinate, deciduous ; stigma capitate, undivided, or with a slight transverse furrow. Capsule spu- 
riously 2-celled, dehiscing between the inflexed margins (septicidal) which sometimes adhere in the centre, 
sometimes remain distinct, seed minute, immersed in the placenta. Usually annual, erect, straight, herbs, 
very smooth, with terminal cymes; flowers usually blue, but sometimes nearly white. 
The species of this genus are for the most part beautiful flowering plants, the flower, of all I know, 
except E. bicolor, are some shade of, often deep, blue. Grisebach in his generic character describes the 
ovary as 2-celled, with the ovules attached on both sides to the central suture. This part of his character 
I have altered as not being quite consonant with fact. The real structure is not well shown in either of 
the plates, but when carefully examined, there is no difficulty in discovering that they do not cohere in the 
centre, though they meet, the placentiferous margins being covered with ovules which touch, but not the 
placenta. This is well shown in my Illustrations, the dissections of which were executed with more care, 
after — received his monograph, which I had not seen, when the drawings for the accompanying plates 
were % : 
