27^^^ pulling the seeds off their fiinicles or stalks, -and as all the 

 p"ener8 of this group having axile placents tions have open cells 

 to the fruit, without ceM-v/inrs, the seeds are easily wahed out 

 of the cells by the rein. The fruit at Fig. 110 F. has axile pla- 

 centas and open cells, the black spots shown beinf' seeds, -a-ll 

 the other fruits represented >?t ^ig. 110 have their plecentrs on 

 t>ie outer -"-pll or floor o-^ the cells, and the cells are roofed 

 vdth ce''l-win/:'s as indicated on the fruit D at b, v/hile at c of 

 the ^r^e firwve is shown the tubercle thst nearly closes the open- 

 in^ 0-" the cells, and seemr' ps if designed to prevent the escape 

 o^ the seeds. ""'Hist is the purpose of this tubercle? In all the 

 fru:its I have been able to examine that have this tubercle, it is 

 8lwfi7''s accompcnied by stiff (not thin and flexible) cell-wings, 

 "'hich still further tend to prevent the escape of the seeds. 

 Hov do the latter get out of the cells? I have failed to solve 

 tbe '"T-oblem* the ansvrer to ' ' luestion must be v/orked out by 

 some One in South Africa. 



^Vhen the valves are expended some of these fruits are very 

 ^rett""-, st^r-like or "^lover-like objects, as may be seen from 

 their F^'re^.Tence in ^ig. 110. On account of its flower-like app- 

 eprfince, I have even seen the fruit of one srecies sold in London 

 under the erroneous name of "Rose of *^eric^.. , ' name that is also 

 often mise-^nlied to other plants the t expan^;. i/nen ■■•'etted and close 

 nr^ when dry. -he true "Hose of "ericho" v/ill be found referred 

 to later or " note under Oephelophyllum dubium, 



A curious circumstence connected with the cr-psules of ^-esembrj''- 

 snthemum is that in the yeor 1777 the^'' v.rere nistsken for i'ungi \)j 

 Ha gen, who, in the Berlinische ^ammlungen zur Beforderung der 

 arzne-^n^issenschsft, V. ix. , p. 133, founded his genus Hediviva, 

 contsininp' three species, upon the capsules of different species 

 of ITesenbryanthemum. See Konir and Sins, ^i-nnals of Botany, V. i.^ 

 p. 365 (1804). 



This is not the rlace to enter into the details of the re- 

 markable structure of these fruits, but as I have found that cer- 

 tain combinations of structure possessed by them are usually as- 

 societed with certain types of vegetative character, I have util- 

 ized the vegetative and fruit characters in combination in the fol- 

 low:nr- pages to divide the grour. into smaller genera. It must be 

 understood, however, that my knov.'ledge of the fruit of these plants 

 is YBT^r imr>erfect, owing to my inability to obtain an abundant 

 suT^^ly, yet vrhere I have been able to examine the fruit of tv.'O or 

 more s^^ecies of the same grou^^ or genus as I nov; consider, the 

 combination of characters as given seems to indicate that they 

 will be conformable with all the species placed under each genus. 



As the main r^urpose of s plant's life is to produce fruit, it 

 is an accepted conclusion that the fruit is one of the most (if 

 not the most) important of riant structures, and therefore where 

 a general uniformity in its structure is found it indicates close 

 relationshir or corjnon origin of the species have such structure. 

 But where one or more of the organs of which the fruit is composed 

 is undevelOT^ed. or suppressed or otherwise modified, it indicates 

 a diverp-ence from the relationship, and being usually accompanied 

 b''- '^ome dif-^erence in the flower structure or in the vegatitive 

 or^rr^ns is regarded as of generic import. This is practically the 

 basis uron v/hich all large ptoups have been divided into genera, 

 ^nd is adorted here for the division of i^^es embryo- nthemum, vegata- 

 tjve, floral and fruit char-^cters all being taken into considera- 

 tion. 



And, judp-inff from the seedlinf^s of the fev; genera I have been 



