APPENDIX. 49 



in most cases grows black in drying, always possessing a valvate 

 sestivation and other very distinct characters^ classed with Atropa 

 and Mandragoray genera quite different from them and each 

 other, and possessing a remarkably imbricate sestivation. Thus 

 also in the Lycine(2y there is an equal amount of comphcation, 

 for we see Dunaliay Tochroma^ Pcecilochromay Acnistus, and others 

 with valvate sestivation, associated with several distinct groups 

 that possess an imbricated sestivation : among these we find 

 Juanulloa^ Solandraj &c., and also Marckea^ Thinogeton^ &c.j and 

 all these again congregated with Lycium — groups perfectly di- 

 stinct from one another. We meet with Juanulloa and Marckea^ 

 having almost a straight embryo, placed among a number of 

 genera having a nearly annular embryo. 



The genus ThinogetoUy arranged by M. Dunal among his Ly- 

 cinetey is said to be closely allied to Jahorosay HimeranthuSy Dory- 

 stigma and TrechoTUBteSy genera which he has singularly placed 

 among his Atropinece. This distinguished botanist can never 

 have seen a specimen or drawing of Thinogetouy or he would 

 never have ventured on such a conclusion : its aflfinity, as I have 

 elsewhere shown, is toward Scopolia, Physochhena and CacabuSy 

 genera that I have placed with Hyoscyamus, on account of their 

 many uniform characters, particularly that of the operculiform 



dehiscence of their fruit. 



In M. Dunal's tribe of the Dature^y we meet with a similar 

 degree of irregularity, in the association of perfectly incompatible 

 genera. Thus Didyocalyw is placed here, while Thinogeton is 

 arranged among the Lycineay and yet these two genera are 

 identically the same. This genus with a spiral embryo, and 

 Datura with a nearly annular embryo, are associated with So- 

 landra, where the embryo is nearly straight or but slightly 

 curved ; in this respect M. Dunal has followed the example of 

 older botanists, who, for no other reason that we can imagine, 

 drew this conclusion, because in former times Solandra grandi- 

 flora was the Datura scandens of Plumier. Solandra^ is as totally 

 distinct in habit from Datura as it is in structure ; it is a climb- 

 ing plant, with large coriaceous leaves and orange-coloured 

 flowers of large size, with a thick fleshy corolla, having a ven- 

 tricose funnel-shaped tube, and a border of five large fleshy lobes, 

 which in estivation are so deeply imbricated that they com- 

 pletely overlap one another. In Daturay on the contrary, the 

 corolla is white or of a lurid blue, with a nearly entire or pen- 

 tangular border ; this in sestivation is plicated into five deep folds, 

 which almost meet in the axis, and these folds are torsively and 

 spirally twisted round the common centre, having then- margms 

 thus valvately coherent in juxtaposition : no two cases of more 

 extreme difi'erence could have been selected. On account of the 



H 



