Chamisso and Schlechtendahl) separates by the splitting of 

 the dissepiment into two cocci, which are locuHcidally split, 

 and by some other minor points. Nearly the same characters 

 distinguish them also from Sipanea, which is quite as nearly 

 related as Rondeletia, though very different in habit. They 

 have also been compared with Bouvardia, and in some respects 

 resemble the large white-flovvered species of that genus, but 

 independently of all minor distinctions the Hindsias have the 

 ovules and probably the seeds of the tribe of Rondeletiese, not 

 of that of the Cinchonese. 



'* The Hindsia violacea differs from the better known 

 H, longiflora, chiefly by having much larger, broader, and 

 more downy leaves, the flowers much larger and more hairy ; 

 and by the calyx, of which one, two, or three divisions are 

 much larger than the rest, and more or less dilated and leaf- 

 like above the middle. 



" Both species vary in the size of the flowers, and in the 

 shade of their colour. In the H. longiflora also, and perhaps 

 in H. violacea the stamens are entirely included in the tube 

 of the corolla in some specimens, and in others the tips of the 

 anthers protrude. In the latter case the style appears to be 

 less prominent than in the former, so that these differences 

 may arise probably from a certain degree of sexuality." 



For the foregoing matter we are indebted to the kindness 

 of Mr, Bentham, who agrees with us in regfardincr the Ronde- 

 letia longiflora wrongly referred to the genus in which it has 

 been placed. 



This Hindsia violacea is one of the finest things obtained 

 from South Brazil. It has been imported by Messrs. Veitch 

 and Son of Exeter, who received for it the large silver medal 

 at the Horticultural Society's Garden Exhibition in May 

 last. It will doubtless prove a very easily cultivated green- 

 house plant ; and is certainly unsurpassed in beauty by blue 

 flowering shrubs. 



