784 



HAMAMELIDACE^. 



[Epigynous Exogens. 



Order CCXCIX. HAMAMELIDACE^.— Witch-Hazels. 



Hamamelideae, R. Br. in Abel's Voyage to China, (1818) ; A. Richard Nonv. Eldm. 532. (1828) ; DC. 

 Prodr. 4. 267. (1830); Endl. Gen. clxvii. ; Griffith in Asiatic Researches, (1836;, xix. p. 94. 



Diagnosis. — Umhellal Exogens, with a 2-celled not didymous fruit ^oithout a double 



epigynous disk^ an imbricated corolla, alternate leaves loith stipules, and anthers with 



deciduous valves. 



Small trees or shrubs. Their woody tubes, in some cases, marked by circular disks. 



Leaves alternate, deciduous, toothed, with veins running from the midrib straight to 



the margin. Stipules de- 

 ciduous. Flowers small, 

 axillary, sometimes uni- 

 sexual by abortion. Calyx 

 adherent, in 4 or 5 pieces. 

 Petals 4 or 5, or ; if 

 present, with an imbrica- 

 ted aestivation. Stamens 

 8, of which 4 are alternate 

 with the petals ; their an- 

 thers tm'ned inwards, 2- 

 celled, and 4 are sterile, 

 and placed at the base of 

 the petals ; theu' dehis- 

 cence variable. Ovary 

 2-celled, inferior ; ovules 

 soUtary or several, pendu- 

 lous or suspended ; styles 

 2. Fruit half inferior, 

 capsular, usually opening 

 with 2 septiferous valves. 

 Seeds pendulous ; embryo 

 in the midst of fleshy horny 

 albumen ; radicle superior. 

 According to Brown, the affinity of Witch-hazels is on the one hand with Bininiads, 

 from which they are distinguished by the insertion and dehiscence of the anthers, the 

 monospermous cells of the ovary, the dehiscence of the capsule, the quadrifid calyx, 

 and by habit ; and on the other with Cornus, Marlea, and the neighbom-ing genera ; in 

 some repects also with Ivyworts, but differing in theu' capsular fruit, the structm'e of 

 the anthers, and other marks. — See Abel's Voyage, Appendix. Du Petit Thouars looks 

 upon them as allied to Rhamnads, and Jussieu to Hippurids. Others consider them 

 akin to Mastworts and Spurges, tlu'ough Fothergilla. There can be no doubt that they 

 must be arranged in the Epigynous series, and upon the whole, Bruniads on the one 

 hand, and Cornels on the other, seem to claim the nearest kindi'ed ; from the latter 

 they are known by their stipules and imbricated petals, if any are present. It was 

 formerly supposed that they might be allied to Saxifrages, on account in part of their 

 albuminous seeds and half inferior ovary ; but that resemblance was but a remote 

 one. Mr. Griffith observed in Bucklandia and Sedgwickia that the woody tissue is 

 marked with circular dots something like those of Conifers ; the same sharp-sighted 

 Botanist observed that in Bucklandia the second membrane of the ovule protrudes 

 beyond the foramen in the ripe seed ! 



The species come from North America, Japan, China, and the central parts of Asia, 

 Madagascar, and South Africa. 



The kernels of Hamamelis virginica are oily and eatable. The leaves and bark are 

 very astringent, and also contain a peculiar acrid essential oil. 



GENERA. 



Fig. DXIX. 



I.HAMAMELEiE.— 0%TlleS 



solitary. 



Dicoryphe, Thouars. 

 Dicorypha, Spreng. 

 Corylopsis, Sieb.et Zucc. 



Trichocladus , Pers 

 Dahlia, Thunb. 



Hamamelis, Linn. 

 Trilopus, Mitch. 



Loropetalum, R. Br. 

 Parrotia, C. A. Mey. 

 Fothergilla, Linn. f. 

 Distylium, Zucc. 



II. BUCKLANDEJE. — 



0\Tiles several in each 

 cell. 



Bucklandia, R. Br. 

 Sedgwickia, Griff. 



Numbers. Gen. 10. Sp. 15. 

 Position. — Bruniacese. — Hamamelidace^. — Cornacese. 

 SaxifragacecB. 



Fig. DXIX. — Corylopsis. 1. flowers; 2. branch in fruit ; 3. a flower separate; 4. a stamen; 5. a 

 perpendicular section of the oA'ary ; 6. a section of a seed removed from the capsule and placed with 

 the hilum dowTiwards. 



