ME. BENTHAM ON LOGANIACEiE. 61 



Logania and Buddleia. Either Logania must be transferred to 

 Scrophularinece and a new name be given to the free Bubiacece, or 

 Buddleia and its allies must be brought over to Loganiacece. I 

 should prefer the latter course, for I think that better characters 

 can thus be given to the two orders. The irregular more or less 

 personate corolla, sometimes indicated only by its bilabiate aesti- 

 vation or by the absence of one or of three of the stamens, and 

 the absence of any stipular connexion between the petioles, would 

 be indicative of ftcrophularinece, and any trace of stipules accom- 

 panied by a regular corolla isomerous with the stamens and quin- 

 cuncial in aestivation would refer to Loganiacece. It is true, there 

 would even then remain some intermediate genera among Scopa- 

 riece, where the quincuncial passes gradually into the bilabiate 

 aestivation, the regular into the personate corolla, but the absence 

 of the stipular line and the tendency to alternation in the leaves 

 would leave these among ScropTiularinece. So also Microcarpcea 

 and Bryodes, which I had included among Buddleiece, would, not- 

 withstanding the presence of the connecting line and a general 

 resemblance to some Mitrasacmes, be rejected from Loganiacece on 

 account of their diandrous or didynamous flowers. 



The approach to Apocynece is indicated in Geniostonia by the 

 contorted aestivation of the corolla, and in Mitrasacme by the 

 partial separation of the ovaries, whilst the styles are joined at 

 the apex ; but, as in neither instance are the two characters com- 

 bined, there is little hesitation as to their retention within the 

 boundary-line of Loganiacece. 



Several of the Euloganiece had been formerly placed among 

 Gentianece, but have been judiciously rejected by the able mono- 

 graphist of that order on account of their axile placentation, 

 besides that they want the bitter principle so universal in Gen- 

 tianece. It is true that the Bubiacece also include a few genera 

 (such as Gardenia, Amaioua, &c.) where the placentation is pa- 

 rietal, as an exception to the almost universal central placentation 

 of the order; that this exceptional character shows itself occa- 

 sionally also in the corresponding genus Fagrcea, and that it cannot 

 in either case be treated as even of tribual importance ; yet here, 

 among free capsular Ifonopetalce, we must give it a much higher 

 value, in order to find a tangible character for the definition of 

 such an eminently natural family as the Gentianece. 



Although aestivation of the corolla does not in Loganiacece any 

 more than in Bubiacece afford good tribual characters, yet in both 

 instances it is a convenient one for the subdivision of the tribes. 



