30? Iv through a South African summer was not prepared to endure an 

 ■ Snfriish summer in addition without first replenishing? its store of 

 food. 



I have no\f'j given details concerning the vegetative and other 

 characteristics o'^ some of the very remarkable plants at present 

 nTaced under the p-enus ^'esemhryanthemum, in order to explain and 

 1ustif3'- ray reason for senarnting some of them from that genus and 

 raising- them to the rank o"^ distinct genera, ^or the more I study 

 T-esem"^r-"Tnthemum the more do ^ become; convinced that, as at present 

 constituted, it is rather of the nature of a natural order than of 

 a sinp-le ^enus. The great difference in the vegetative characters 

 o"*" various o-roups as above described, is evident to ever^'^one v/hen 

 the Plants ere seen together in a collection. J-t is the general 

 similarity of the flowers of variou^s groups that has caused them 

 to be united into one genus. Yet this similarity, as I have pre- 

 viously stated, is often more apparent than real, for when the 

 structure is carefully examined small distincti^ve characters may 

 o-"ten be found, in fact, I think we may liken ^■%serabryanthemum 

 to such orders as ^puciferse, Cgryophyllaceae end Umbelliferee, 

 where there is often little evident difference to be found in the 

 a'f^pe'^rance of the flo'-'ers of different genera of those orders, 

 there:^ore sm??ll structural differences in the flower are combined 

 wit^ other characters ■^or separating those plants into genera. I 

 propose to do the same with some of the groups at present placed 

 under I.'esembryanthemum, deetling only with the few groups I' have 

 cultivated, as the total number is too vast for me to attempt 

 to classi*^'- without a r-rolonp-ed stnd;"- o"^ all the groups in a living 

 state, ^or the vh-O^e genus requires thorough revision, and that 

 revision, I think, could be bett'r done in South Africc than here, 

 proviflec: that the nomenclature be first carefully corrected by a 

 competent ^erson from a comparison of the plants with the drawings 

 et Kpit of Kpvrorth's type plants; the nomenclature in the nionographs 

 of Salm ^yck, Merger, and the ^lor? ^apensis is b*/ no means 

 trust-i'^orthy in many instances. 



It may be ask^d, what characters are^ to be taken to disting- 

 uish the Plants that are to remain under --esembryanthemum since 

 linne evidently included several distinct types under that genus. 

 Taking hia generic characters in combination v.'ith the species pla- 

 ced in the genus by him, it ^ 'Ould seem that the only definite 

 char^'cters that can be found are that the plants should have J 

 (l) on ench branch or gro^A^th two or more pr irs of distinct leaves, 

 connate only at their base; (2) a '^raonophyllous half five-lobed" 

 calyx, by V7hich I understand him to mean that the calyx is lobed 

 doi'Ti to the to^ of th^ ovary and not produced into a tube above 

 the ovary; (3) a corolla with the claivs of the metals but slight- 

 ly united at the base, i.e., free nearly to the base, not united 

 into a dis"^inct tube; and (4) stigmas 4-5, subulate, erectly re- 

 curved. Linne and others describe the stigmas as styles, but I 

 think that none of the species kno-'-n to Linne have a true style, 

 as the so-cflled styles of all his species are stigr.atic all along 

 to their base. 



'^^hich of the species best correspond to those characters is 

 the pro^'^ler. that has to be solved, pnd when the genus is revised 

 it v.'il? become a nuestion whether plafits differing so much in 

 habit and apT)eBrance from the v/ell-knov/n "Ice plant" U'^, 

 cr'""strllinumJ or from such shrubby/" species as -^-. coccineur.. and 

 -'. inclaudens, as do the t'-'o ne'"' s-^ecies here figured, ■'-. Orpenii 



