14 for 3r. PhilliDs' book was urgently reouested, it was prepared as 

 raridly as possible, and in the hurry of doing this, that' part of 

 my key containing the characters of the typical species of the genus 

 I'^esembryanthemum was, by sone oversight, omitted from both key and 

 description, and only the characters of the species I had then as- 

 sociated v:ith the typical species v;ere included. For in my earlier 

 keys the genus l-iesembryanthemum was entered in two places J (l) in 

 the group of genera having a placental tubercle in each cell of 

 the fruit and (2) in the group having no placental tubercles. It 

 vjas the entry under the first group that by mischange vf&s omitted 

 from the key and descriptions sent to Dp. Phillips, and as thet key 

 was made the basjs of all the subsequent keys, v/hich v/ere copied 

 and re-copied as nev^^ genera were added, the error was not noticed 

 until long afterv/ards. 



However, it follows from my statement that M, umbellatum, 

 Linn, (an old and correctly understood species) wr- s to be taken as 

 the tjve of the reformed genus, and must be reatained as such, for 

 it would be obvious to any botanist that some mistake has been made, 

 v/hich could easily have been corrected to conform with and include 

 the type by the addition of the italicised words to the description, 

 noted hereafter, or for them to be included in the key, as they were 

 in v/ith or without membranous wings on their outer me.rgin," and 

 "with or v/ithout tubercles at the openings to the cells." 



I find, however, that Dp. Schv;antes has either overlooked that 

 statement or ignored it v;ithnut inquiry, because, instead of correc- 

 ting, he has founded his genus Huschia ut^on the identical type of 

 structure possessed by M. umbellatum, Linn., and has selected another 

 species to be the tyre of Mesembryenthemum. 



"Hiis attemT5t (in consequence of en obvious error) to replace 

 the genus Mesembryanthemum, Linn., as emended b' myself, by another 

 generic name of his ov;n not only does not conform v/ith the code of 

 honour held among scientific v;orkers, but is contr ry to all rule 

 and precedent, and the science of botany would soon become very 

 chaotic if proposed types of genera that have been examined by 

 authors were allowed to be changed in this way at the will of every 

 writer, 



Schwantes if right, however, in separating from ^'^esembryanthemuiii 

 as distinct genera some of the groups I associated v;ith it at the 

 time of revision of the genus, 



"j^e association of the groups alluded to was made because I 

 found that v;hile fruit characters are generally constant for each 

 pronounced vegetative type, yet in some cases they vary very consid- 

 erably among species that unquestionably belong to one genus. I'he 

 most remarkable case of variation noted is in the group separated 

 by Schwantes to form the genus Drosanthemum, wherein there are 

 species in which the expanding-keels of the fruit are contiguous 

 into a central keel, others in which they diverge; some with well- 

 developed cell-wings, others in which they are rudimentary; some 

 in which a placental tubercle is present some large or small or double, 

 other in which it is absent; some with long, others with short 

 stigmas, I will also mention that in ^isphyma austrple the placen- 

 tal tubercle is nresent in some specimens and absent from others. 



It vras this Vfnriation thet caused me to place under ^Mesembryan- 

 themum the plants 3chv/antes has separated from it, and for which I 

 had originally provided by placing that genus in tv/o parts of my 

 key. But I agree vrith Schv/'^ntes th-^t some should be separted. For 

 having again examined all available material, I nov; note that with 



