This mode of grovv-th is, of course, onlj'- a modification of 

 that of the genus ^heiridopsis, to which ^«'i trophy Hum seems to be 

 in some vay allied, for the genus comprises two groups, one in v;hich 

 the vegetative and flowering growths are both alike, are permanent 

 and have distinct internodes, and another group in which the per- 

 manent vegetative part of the plant is practically stemless and 

 sessile on the ground, without internodes, but produces tall flow- 

 ering stems v;ith distinct internodes, which die after fruiting, 

 A vegetative growth of this latter group with its free leaves spread- 

 ing close to the ground, and the cone of united leaves arising 

 from between them, in habit is so very similar to Cheiridopsis pe- 

 culiaris in the same condition of grov/th, that without seeing flow- 

 ers and fruits most peorle would J)roabably consider them to belong 

 to the same genus. 



This remarkable group is one of those I alluded to in The 

 Gardeners' Chronicle, 1925, Vol, I^XXVIII, p. 412, as being left 

 under Mesembryantheraum when my key to the genera was constructed, 

 because I had no fruits or adequate material by which to separate 

 them, Schwantes, hpwever, has rightly separated the group and 

 founded the genus I-itrophyllum for its reception at the place above 

 quoted, under the following structural characters J — ^lov/er solitary, 

 terminal, brsctless, sometimes breaking through the body of the 

 cone, with shorter or longer pedicels. Cgiyx 4-5-lobed. i^etals in 

 several series, linear, white or pink, Stpiaens numerous, spreading. 

 Stigmas 7, subulate, as long as or shorter than the stamens. Cap- 

 sule 7-celled; cells v.'ith cell-covers; rlscentes parietal, without 

 placental tubercles ;^expending-keels diverging upwards, viith moder- 

 ately broad wings. Seeds unknovm. 



It is not stated from which species these characters were de- 

 rived nor which is the tvj)e of the genus, ^t in the very same vol- 

 ume, p. 321, he divides i-itrophyllura into tv;o genera, and states 

 that --. mltratum is the type of that genus, which he there defines 

 as follows: — "-^edicel vdth two short, thick bracteoles at the base, 

 Petals and stamens numerous. S^igmgs 5^ filiform. Capsule 5- 

 celled, vdthout cell-covers; ex^^anding-keels parallel, not touching 

 one another, with small, marginal wings. Seeds Pear-shaped, slightly 

 papulose," — Now, these characters, stated to be derived from M. 

 mitratum, are so different from those of his original description 

 (quoted above), that it is obvious that both sets of characters 

 could not possibly have been made from the same species, and, have 

 been made from the same species, and, therefore, M. mitratum can- 

 not be the tyT)e of the genus as originally established, ^or it is 

 those original characters that must be maintained for the genus, as 

 it is not Permissible and is against all rule and precedent to de- 

 liberately change the characters of a genus in this way, unless it 

 can be demonstrated that a mistake has bee made, ^ut here is no 

 mistake, for the characters of his first description do apply to 

 certain species, and, moreover, we also find that ^^chwantes actually 

 uses them (at the same place quoted) to found his genus Conophyllum, 

 which must, therefore, rank as a synonym of Mtrophyllum. 



'•Vith reference to i-. mitratum, the only claim to generic dis- 

 tinction is that the cells of its fruit have no cell-Vv-ings and the 

 expand ing-keels are parallel instead od divergent, for there isno 

 other difference. I regard this as merely a structural variation 

 among species belonging to the seme genus, because among the large 

 number of fruits I have examined I find that although fruit-structure 

 is fairly constant and reliable (in conjunction with vegetative char- 



