'^. 



SPECIES OF MESEMBRYAXTTIEMUM. 



G9 



Dyek's figures of this group aro incorrectly namedj as drawings of Hawortli's 

 typical plants at Kew clearly demonstrate. These plants hybridise freelv, 

 but with me fail to produce seeds when fertilised with their own pollen, as 

 they seem to re(juire to be cross-fertilised, so that seeds obtained from a 

 nursery or any source where every care has not been taken to prevent 

 hybridisation are quite likely to produce hybrids for the unsuspecting 

 cultivator instead of the typical species* 



/c 



.n 



Und 



er 



yS 



S, without •i-ivino- them names, all founded upon plants figured and dcscrihed 

 1)}' Dillonius. As those four jilnnts of Dilloiiius certainly do not all belong 

 to one species, Haworth separated them and gave each a distinct specific 

 name ; for they differ from one another in appearance, in the size and shape 

 of their leaves, in details of their flowers, and in their capsules and seeds. 

 In modern monographs the dislinctivc characters of their capsules and seeds 

 are ignored. Of these four species I have oidy seen seeds of the true 



^' 



1 



covered with a minute puhesconco as seen under a microscope — a structure 

 that T have not seen on an}' other sp(!eies in the genus that I have examined. 

 I have examined seeds of some of the garden hybrids that are cultivated 

 under tbe name of AL Jiixjn'iforme^ but they are quite different. 



Unfortunately, when Ilaworth separated these plants specifically he 

 selected the variety /5 of Linne as being the type of M. lingniforme, and gave 

 a new name to the Liniiean type of that species. AVilldenow corrected this 

 error, but other authors have copied and j)erj)etuatcd without invest! iration 

 tfie mistake made by Haworth. The correct synonymy of these four 

 plants as I understand it Is as follows : 



M. LATUM, Haw. Obs. p. .180 (171)4); Misc. p. 32 ; Synop. p. 220 ; & 

 Rev. p. 98 : ^\\M. Sp. PI. vol. ii. p. 1020 : DC. Prodr. vol. iii. p. 422 : 



Perger, Mesemb. p. 241 under M. Ihiyinforme var, ohliquum, Berger. 

 Af. medium, Haw. Suppl. p. 88 (1819) : & Rev. p. 95 : DC. Prodr. vol.'^iii. 

 p. 421. AI. lingDcefornie, Lodd. But. Cab. t. 1307, not of Linnaeus. Af. 

 Ihigviforme var. ^, Linn. Sp. PI. ed. 1, p. 488 (1753). AI. Unguiforme var. 

 lata, Weston, Univ. Bot. vol. i. p. 172 (1770), corrected to var. latum. 



Mese 



§ 8, fig. 



^ 



Sonder in Fl. Cap. vol. ii. p. 104. AI. follu lingui/ormi lailore. Dill. Hort. 

 Eltb. p. 236, t. 184, fig. 22^, not fig. 226 as quoted by Linna^ns. 



SoT'TH Africa. Locality and collector of the tj-pe unknown, but probably 

 from the southern coast region (see under M. medium^ p. 132). 



According to an original coloured drawing at Kew of the typical 

 M. medium, which was described from a plant introduced by Burchell, there 



LINN. JOL'EN. — BOTANY, VOL. XLY. 



G 



