126 TAXONOMY. 



that there is not merely one canal in its axis to the germen, 

 but that several such passages stand in a circle, and pro- 

 bably represent as many united pistilla. When we obsei'\'e 

 how the four caryopses of the Labiatae and Asperifoliae sur- 

 round the single pistil, we cannot but believe that the latter 

 part is made up of four individual pistils, especially when 

 we attend to the almost complete separation of the stigma, — 

 the distinct coherence of two individual pistils in the Perilla, — 

 the deep disunion of them in the Thymhj'a, in EcJiium and 

 EcMochilon Desfont., — and, above all, when we attend to the 

 four divisions of the stigma in Cleonia and Coldenia. 



189. 



Lastly^ we must apply this idea of an union of parts even 

 to the fruit. What we mean is, that caryopses, in indefinite 

 numbers, are more original than capsules with many loculi ; 

 that the latter must be considered rather as a collection of 

 united simple capsules, and that when, in a simple fruit, we 

 observe a certain irregularity, we may safely suspect that ori- 

 ginally there were several loculi, but that they have become 

 abortive and united. 



We may prove this account by striking instances. Eyery 

 person knows that the fundamental genus of the Ranuncu- 

 lese, from which they take their name, contains caryopses in 

 indefinite numbers, which are also found in Myosurus, Ane- 

 mone, Thalictrum, and so forth, but which pass in Xan- 

 thorrhiza into the single-seeded utriculus, and in Aconitum 

 and Paoniu into capsules which open laterally. On account 

 of the relationship of Nigella to those plants in other respects, 

 we must consider its fruit also as single capsules, which open 

 laterally, and which are only a little united at their lower 

 parts. Now, when we observe a capsule with five loculi 

 in Vallea Mut., which is evidently a Ranunculus, nothing 

 is more natural than the conclusion, that in this case, as 

 well as in Nigella, there have been five simple capsules which 

 have here undergone a union. If we attend to the fun- 

 damental genus, from which the Diosmeae have their name, 

 we find it carrying five pointed caryopses, surrounded by an 



