XATCHAL CLASSIFU'ATIOX. 125 



Crotalar'ia and Abrus shew, in the empty cleft on the hack 

 of the cvHnder of the filament, this disposition to the separa- 

 tion of the one filament, which, in proper diadelphous plants, 

 we see standing single, whilst in Dipteryx^ Schrcb., the junc- 

 tion of the filaments is incomplete, and in Dalea and Petalos- 

 temon they are less united with each other than with the })arts 

 of the corolla. Are not these instances proof sufficient that the 

 union of the filaments presupposes an original disunion of them? 

 Do we not every day, in many of Caryophylleae, perceive the 

 filaments standing free, whilst in others (Dianthus, Sapona- 

 ritty Silene, and Agrostemma,) they are attached to a ring 

 surrounding the pistil, or are united with the petals ? The 

 tribe of Chenopodeae, to which belong Illecebrum, Herniaria^ 

 Gomphrena, and others, shew the same thing. The fila- 

 ments in Iresine, Paronychia, Tourncf., Anychia, Hern'ia- 

 7'ia, and Bosea, stand free. In Adiyranthes, Illecebrumy 

 Gomphrena, and others, they are united. 



188. 

 What has been said of petals and filaments is also appli- 

 cable to pistils ; that is to say, we are often forced, when we 

 see a simple pistil, to consider it as made up of several : other- 

 wise it would be difficult to understand in what way many 

 plants have several pistils, whilst some nearly related to them 

 have but one. It may here become a question, whether Mespl- 

 lus monogyna be the only original species, or whether there be 

 not many other species, which have from two to three, four 

 and five pistils. When in most of the Grasses we regularly 

 perceive two pistils, it is extremely probable, that the few, as 

 Nardus, Cenchrus, Lygeum, and Spartina, which are mo- 

 nogynous, have their one pistil made up of two that are 

 united. This is still farther confirmed by observing, that the 

 number of the pistils corresponds with the number of lo- 

 culi in the germen. If there be but one pistil, the num- 

 ber of loculi in the germen corresponds with the number of 

 stigmata ; and it is probable that oi-iginally there were as 

 many pistils as stigmata. Lastly, in a cross section of a strong 

 pistil, as for instance in any of the Cactus species, we observe 



