.330 EUPI10EBIACE.E (Brown). [UupJwrhia. ^ 



"When grown from seed, a large globose body partly buried 'in the ground is 

 developed, from which the numerous branches arise, as in all other species of this 

 group. But when one of the branches is rooted, the globose main body is not 

 iormed, but it becomes a cylindric or clavate stem rising some inches or even up 

 to 2 feet above the ground and produces branches at the top. It is upon branches 

 rooted in this manner that E. CovDnelimi j DC, and E. Caput- Med asm, var. major, 

 Ait., wore founded and supposed to represent distinct species, the habit being 

 different. From the description given by Miller, Alton, and Haworth of E. 

 FractuS'Pitd {E. Caput-Mtdmx, var. geminata, Ait.), I can find no character to 

 separate it from typical E. Caput- MediiScV, and I have not seen any specimen of 

 it. Gardeners of the period seem to have distinguished it as the " Little Medusa s 

 Head," whilst the typical E. Caput- 31 edusie was known as '* Medusa's Head " ; so . 

 that it is probable that it was of a somewhat smaller growth than the type. 

 Although Miller's description seems to have been made from a rooted branch, the 

 plant when originally introduced was probably either only a small specimen of E. 

 Caput-Meduspe or of a sex ditferent from the plants then in cultivation. A young 

 plant in cultivation at Kew in January, 1915, answered exactly to Millers 

 description of E. Fructus-PiuL In the Ist edition of Alton's Ilortus Kcicensis, 

 var, minoi' seems founded upon the plant chiefly intended by Isuard under the 

 reference quoted by Alton, which was certainly E. Caput-Mcdusx, Whilst in the 

 2nd editiL»n, under var. minor, Aiton only refers to the plant figured in De 

 Candolle, Plant, Grass, t. 150, which is another species, see under E. Bergen. 



■ 



119. E. Bergeri (N. E. Br.) ; dwarf, succulent, spineless ; main 

 body of the plant subglobose or obconic, thick, fleshy, bearing a 

 large number of radiating branches on the upper part, glabrous ; 

 branches 3-9 in. long, ^-| in. thick, usually simple, cylindric, 

 tessellately tuberculatCj often curved, glabrous, green ; tubercles 

 rhomboidj mostly 3-4 lin. lon*:^ and about 24 lin. broad, A-l Hn. 

 prominent ; .leaves 3-6 lin. long, |-1 lin. broad, linear-spathulate, 

 acute, channelled down the face, glabrous, deciduous ; peduncles 

 solitary in the axils of the tubercles at the ends of the branches, 

 2-5 lin. long, stout, bearing 1 involucre and about 4 bracts, some- 

 times persisting and thickening after the fall of the flower and 

 becoming clavate or branch-like ; involucre 2^—3^ lin. in diam., 

 cup-shaped, with 5 glands and 5 transversely rectangular or sub- 

 qnadrate finely toothed lobes ; glands spreading, 1-1 J lin. in their 

 greater diam., transversely oblong, green, with 3-7 short or subu- 

 late greenish- white teeth on the outer margin ; ovary included in 

 the involucre, glabrous; styles about 1| lin. long, united for about 

 half their length, with slightly spreading thickened cuneate stigmas, 

 notched at the apex ; capsule and seeds not seen. E, Cffpid' 

 Medusse, Lam. EnajcL ii. 416 ; DC. PI. Gra^s. L 150, not of Linn. 

 E, fructuspina, Sweet, Hart, Suburh. Lond, 107. E. Fructus-Pini, 

 Siceet, Sort. Brit, ed, i, 356, exch var, /3, not of Miller, E. parvi- 

 mcmraa, Berger in MonatsscJir, KaJd. ix. 92, ivith fg., and SiiJd\ 

 Euphorh. 113, /y. 30, vot of Boiss. 



South Africa : without locality, cultivated specimens ! 



The above description i^ made partly from dried material, partly from a living 

 specimen, both taken from the type plant of E. parvimammay Berger, kindly sent 

 to me by Mr. Alwin Berger. But it is certainly not the E . parvlmaramd of Boissier. 



E, BergcH is only known from cultivated plants. It is very similar in appear- 

 ance to Z*. Caput-Medus^, Linn., but the branches are rather more slender and 

 the involucre-glands quite ditferent. It may possibly be a hybrid. 



