198 p.cters cen be derived, only en obconic, globose, obcordpte or two- 

 lobed fleshy mass and e flovrer. And ss the sixe of the e:rov;ths 

 of the ver^r seme rlent in different yeers under cultivation diff- 

 erence in colour end nerkings, and nerhf^^s every slight difference 

 in form by which to distinguish different closely allied species 

 v>^en out of flower; end these ere not characters that cen be 

 clearly defined by words, end ^'.re sometiraes verieble. Itie flov/- 

 ers often differ likev/ise in small but constant characters thpt 

 cannot be described, for colour cannot always be relied upon to 

 ser^erate allied s^^ecies, since different plents of the SKrae spe- 

 cies nia3'- vrry from white to rosy in the colour of their flowers. 

 Havrorth and others have thought that the exsertion of the ovary 

 fron the body of the growth or its inclusion within it is a good 

 specific character, but I find it the exertion or inclusion of 

 the ovary and calj^-x-tube varies in different years upon the very 

 sane plant, so that it cannot be implicitly relied upon. -here- 

 fore, only good coloured drav.-ings with dissections accompanying the 

 descrintions can really enable anyone to correctly discriminate 

 betv/een the closely allied species of this genus, and such illus- 

 trations I have been for some years prepering for a future work. 

 Meanv/hile, I have done the best I can to classify the species by 

 c-rou^in'? them in accordance with the character of their surface, 

 form end markings. 



There is one point that it is very important to note, and that 

 is the difference in size and appearance between newly imported 

 rlpnts and those \''ery seme plants after being in cultivetion a 

 few years. The behaviour of different species in this respect 

 varies, however; some do not alter at all, v;hile others alter 

 gre-'tly. ?or example, the grovrths of the besutiful little 2. 

 m^nusculum are no larger to-day, after 13 years cultivrtion, than 

 the-"- v-ere ■'-hen rf'irst sent to Kev in 1908, while the growths of 

 C, globosum, V7hen introduced at Kew, were about half an inch in 

 di^- meter, yet a portion of the plant given to myself developed in 

 tv-o yeers PTOwths 1-li inch in diameter, and in subsequent years 

 veried from i-l:l inch in diameter, a size that it probably never 

 attains in nature. The most remarkable chenge that I have notic- 

 ed hes been in the cese of the very distinct ^. pilosulum, which has 

 altered its form from obcordately obovold to an elongated ovoid 

 shape, pointed at the top. N.E. Bpov/n 



(To be continued, ) 



--esembrvanthemum end some new .^enera separated from it. 

 Gsrd. Chron. HI. 715 214. 1922. 

 (Continued from page 198.) 



''K: In the descriptions which follow it will al^.&ys be necess-_ 

 ary to take the smaller dimensions given as more nearly indicc^ting 

 the size of the growths as they occur in South Africa. The time 

 at which the f lexers are stated to expand is "^eenwich time, not the 

 so-celled ''summer-time. " 



As in the ce.se o~" the gener--- previouslj'" dealt vdth, the name 

 Hesembrypnthemiim is indlceted by the letter ^^ in the synonymy. 



The groups under v'hich I have arranged the species in most 

 ceses ere surely ertificiel, end do not admit of a proper se- 

 nuence of the nearest allied species. 



In the descriptions the breadth of the growths is the greater 

 diameter, and the thickness the lesser diameter across the top. 



