Hypericum.] hypericace^ . 87 



centre, wilh 2 or even 4 opposite, very narrow, membranous ridges or wings on 

 alternate sides between each pair of steta-leaves, usually naked below and simple, 

 greenish or purplish, with a few scattered black dots, much and corymbosely rami- 

 fied above, the branches opposite, axillary and decussate, the lower ones short and 

 barren, gradually elongating as they ascend and at length becoming floriferous. 

 Leaves numerous, opposite, sessile, decussate, dark (sometimes bright), elliptical, 

 oblong-elliptical or obovate-elliptical, in 13. very narrow and sublinear, those of the 

 main stem about 1 or IJ inch long, and above ^ an inch wide, of the branches 

 smaller, prominently 7-ribbed beneath, the midrib very acute, and all pellucid 

 when viewed against the light, usually very obtuse and rounded, sometimes a lit- 

 tle pointed, quite entire, with a row of little black dots or glands along their 

 slightly deflexed margins, and a few occasionally sprinkled over the under sur- 

 face, in addition to the more or less numerous pellucid points, which are occasion- 

 ally nearly wanting as in H. dubium, a species not yet detected in the Isle of 

 Wight, but which, though much resembling the present, besides having the calyx- 

 segments obtuse, has the leaves covered with a network of anastomosing pellucid 

 veins, by which it may at once be distinguished from every variety of our common 

 plant.* Flowers very numerous, showy, of a bright golden almost orange-yellow, 

 in repeatedly forked clusters terminating the superior branches, constituting toge- 

 ther a large paniculate corymb. Pedicels unequal, mostly with a lanceolate, leafy, 

 toothed hract at the base of each. Calyx divided nearly throughout into 5 lan- 

 ceolate, somewhat unequal, acute or rarely rather obtuse segments, which are 

 3-ribbed, pellucidly striate, remotely serrato-dentate at top, mucronato-acuminate, 

 sprinkled here and there with a few black dots. Petals above twice the length of 

 the calyx, irregularly and obliquely oblong, sulcalo-striate, bluntly crenulale 

 towards the apex on one side chiefly, with a black dot in the sinus of each notch, 

 and a few more marginal ones at the back and scattered at random over the sur- 

 face, occasionally, too, streaked with purplish black. Stamens numerous, in 3 

 usually distinct sets, which are alternate with the styles, shorter than the petals ; 

 anthers of 2 round lobes, with a purplish black dot between them. Styles 3, 

 straight, divergent, with simple crimson stigmas. Germen ovoid, somewhat fur- 

 rowed or rugose. CapsM/es about J of an inch long, ovoid-conical, reddish and mem- 

 branous, 3-cleft or tricarpellary at the summit, 3-celled, furrowed and wrinkled when 

 fresh with translucent warts and ridges. Seeds numerous, deep chocolate-brown, 

 oblong-cylindrical, obtuse at both ends, reticulato-punctate, in close parallel rows, 

 the areola shallow and angular. 



The var. ^. is, as Gaudin observes, remarkable for the more erect branches and 

 much narrower leaves, the pellucid dots on which are fewer but considerably larger 

 than in the common form. His other characters I do not observe in my speci- 

 mens, or they are at least liable to considerable variation, as indeed is the breadth 

 of the leaves, which approach those of the ordinary state of the species by sensible 

 gradations in different examples, and even on the same stem. 



The var. y. approaches H. guadrangulum, L. {H. dubium of Leers and British 

 authors), and H. delphinense of Villars, in having the sepals comparatively broad 

 and obtuse, and I was at first inclined to suppose it might be that species, which 

 from the descriptions of authors seems to be a very variable if not a doubtful one, 

 as its name implies. But on comparing our variety with specimens of the true 

 quadrangulum, kindly communicated to me by Mr. Leighton from Shropshire, I 

 am convinced his plant and mine are not the same, the latter having neither the 

 4-angled stem, the imperforate mucronate leaves, nor the black streaks on the petals ; 

 even the breadth of the sepals is less considerable in ours than in the Shropshire 

 specimens, the upper leaves in which are however sparingly pellucido-punctate, a 

 character said to obtain in the true H. quadrangulum sometimes in as great a 



* Such is the case in all the specimens I have received from various parts of 

 Britain, and others gathered by myself at Killarney. I do not find any notice 

 taken of this character by any author I have consulted. 



