170 ONAGRACE,E. [EpiloUuin. 



A few specimens in Alverstone lynch. In considerable abundance under trees in 

 a plantation at the foot of Hatchet-close Cliff, but does not flower there. In a 

 boggy pasture between Fullford and Perieton farms, in great plenty, 1840. On 

 a moist bant near the church at Bonchurch, probably an escape from the grounds 

 close by. In profusion on the Wilderness. In several parts of Bordwood lynch, 

 1841. In the wood near Cook's Castle, very near the Castle, on the left hand of 

 the road winding up the hill from the side facing Appuldurcombe, H. C. Watson, 

 Esq., in litt., 1840. Between Byde and Sandown, in woods ? Rev. G. E. Smith. 

 Between Byde and Brading, but the spot not noted, N. B. Ward, Esq., probably 

 the same station as the last. 



p. Abundant amongst the furze on the S.W. slope of Yarbury Hill, near Niton, 

 but does not always flower there. In a small wet thicket by a moory meadow a 

 little W. from Blackwater mill, in plenty (probably this form). 



Moot creeping extensively, white, cylindrical and soft. Stem erect, 3 — 6 feet 

 high or even more, rounded, smooth, leafy, red or purple. Leaves alternate or 

 scattered, various in size and breadth, lower ones lanceolate, 6 or 7 inches long 

 and 2 inches wide, those higher up narrower and smaller ; all sessile, glabrous, 

 dark green above, glaucous beneath, entire or with a few obscure glandular sevra- 

 tUres, their surface waved or undulated, with thejmargins a little deflexed, veined 

 at right angles to the very prominent often reddish midrib. Flowers in a long, 

 leafless, almost spiked raceme, stalked, bright purplish pink, smaller and deeper- 

 coloured in my wild specimens than is usual in gardens, sometimes white, pro- 

 duced in succession throughout the latter part of summer and autumn. Sepals 

 linear-lanceolate, acute, coloured, forming a cross, the 3 upper ones equal, approxi- 

 mate, nearly straight, the points of the 2 lateral curving a little upwards, the lower 

 segment longer than the rest and bent backwards, like the spur of some Orchis. 

 Petals broadly obovate, about as long as the calyx, slightly emarginate, veined 

 and wavy, suddenly contracted into the very narrow claw. Stamens spreading, 

 bent downwards ; their filaments enlarged and furrowed at the base ; anthers 

 oblong, bursting along the front of each cell ; the pollen of copious pale blnish or 

 greenish angular granules. Style strongly deflexed, placed on a green nectarife- 

 rous gland, glabrous, excepting just beneath the deflexion, where it is surrounded 

 by a belt of white hairs ; stigma 4-cleft, bright purple, the segments obtuse and 

 twisted together. Capsules about 2^3 inches long, erect, linear, furrowed and 

 truncate, reddish, a little canescent with close-pressed down, sometimes warty 

 along the angles. Seeds, — 2 ranked in each cell, yellowish, oblong, wrinkled, 

 glabrous, crowned with the very fine, white, simple, sessile pappus. 



The Yarbury-hill plant I have determined to be our var. /3., the smaller form 

 with long capsule, or E. macrocarpum of Stephens.* On this station the plant is 

 very dwarf, the soil being extremely dry, and the situation much exposed. The 

 leaves are very glaucous beneath, and dark above, but more spreading and flatter 

 than usual in this foim, though on some of the specimens they are as much curled 

 and corrugated as they ever are. The flowers are rather paler in colour, approach- 

 ing in this respect though not in size the larger garden form, the E. angustifolium 

 p. brachycai-pum of Babington. The form and colour of the anthers are those of 

 the normal wild type. It is evident that the two forms of E. angustifolium do 

 not depend on soil or situation, since our long-fruited plant occurs in the driest 

 and most exposed situations, as well as in swampy thickets. 



* Mr. H. O. Stephens, in vol. viii. of the 'Annals and Magazine of Natural His- 

 tory,' p. 170, endeavours to establish a new species of Epilobium, allied to the 

 present, which he calls E. macrocarpum, distinguishing the latter by its very long 

 linear capsules, paler leaves and less deeply coloured flowers, from what he consi- 

 ders to be the true E. angustifolium, the capsules of which he describes as short 

 and turgid, the flowers and leaves darker in colour. I think I have remarked a 

 short-fruited var. of E. angustifolium in gardens, but the co-existence of the linear 

 form in the seed-vessels, with a deep colour in the leaves and inflorescence, is 

 proved in the description of the species given above, and is, I apprehend, purely 

 accidental. The figures in Engl. Bot, Fl. Lon. and Fl. Dan. do indeed all 



