88 iivPEEiCACE.E. [Hypericum. 



degree as in H. perforatum ; " nor are the bl.iok (lots and streaks on the petals of 

 H. quadrangulum less inconstant, being, it is stated, sometimes wholly wanting or 

 nearly so. Kocbf says that in H. quadrangulum the sepals are oWongo-ovate, 

 faintly toothed at the tips (in my plant they are elliptic-lanceolate, toothed and 

 mucronate, as in that from Shropshire), the 3 exterior obtuse, the 2 inner some- 

 what pointed (spilzlich), usually as long as the germen (Fnichtknoten) in the 

 full-blown flowers or somewhat larger, and either sprinkled or not with few or 

 many black dots. In E. B. the sepals are drawn very blunt, with broad white 

 margins, of which I find no mention made by Smith or any other author, and 

 therefore conclude it to be an inaccuracy of the draughtsman. Beside the breadth 

 and bluntness of the calyx-segments, which are very variable in degree, there 

 remains no certain diagnostic of H. quadrangulum but the obscurely 4-angled 

 stem, which is evidently seen in Mr. Leighton's specimens, where the alteniate 

 and faintly projecting pair of wingsj may be traced, though their presence scarcely 

 affect the rotundity of the stem, which on a transverse section appears nearly as 

 cylindrical as H. •perforatum, and our present variety, where the secondary ridges 

 are quite wanting. A certain mark of distinction however between H . perforatum. 

 and H. dubium is to be found in the pellucid reticulations on the under side of 

 the leaves of the latter when held against the light, and which are far less nume- 

 rous and conspicuous in the former, or nearly obsolete. 



This species is extremely common in Canada and throughout the U. States, 

 where it is generally supposed to have been introduced from Europe. I have 

 never in America fallen in with our broad-leaved form y., which so much resem- 

 bles H. dubium; nor do the botanists of that country appear to be acquainted 

 with any other than the narrower more elliptical-leaved variety represented in Fl. 

 Danica and in Fl. Londin., which, being that of most frequent occurrence in 

 Europe also, may be assumed as the typical state of the species. Our broad- 

 leaved form is said by Koch (RohU Deutschl. Fl.) to be found only in southern 

 Switzerland and Upper Italy, and to be absent altogether from Germany. Dr. 

 Darlington (Florula Cest. p. 68) gives this vulnerary herb a very bad character 

 in America for producing troublesome sores on horses and horned cattle, especially 

 those which are white or have white feet and noses, by mere contact with it. As 

 such results are not observable in this country, the charge is most likely unjustly 

 brought against it for effects produced by very different causes. This remark 

 upon the " balm of the warrior's wound " reminds one of what used formerly to be 

 said of the Dutch, that they were good subjects at home, but very bad masters 

 abroad. 



3. H. quadrangulmn, L. Square-stalked St. John' i-wort. St. 

 Peter's-wort. " Stem herbaceous 4-angled somewhat branched, 

 leaves ovate with pellucid dots, sepals erect lanceolate acuminate." 

 —Br. Fl. p. 80. E. B. t. 370. Bab. in Trans, of Bat. Soc. 

 Edinh. vol. i. p. 83. 



In low wet meadows and thickets, moist woods, and along the sides of ditches, 

 rivers, &c. ; very common. Fl. July — September. Fr. October. If. 



E. Med. — In various places about Ryde. 



W. Med. — Woods by Yarmouth and Thorley. 



Herb perfectly glabrous, with a somewhat fetid odour, like that of H. hircinum, 

 when bruised, though Smith calls it a pleasant lemon scent. Root somewhat 

 creeping, of one or more stout, fusiform, woody, and many slender, flexuose 



* Riihling's ' Deutschlands Flora' fortgesetz von W. D. J. Koch, Ver. Band, 

 s. 351. 



■\ L. s. c. 



+ The ridges or wings do not spring from the woody fibre, but are merely 

 formed by a duplication of the epidermis continued into the midrib of the leaf. 



