552 On some new or oh scui^e species of Plants. 



pear distinct from the latter except by the presence of a true co- 

 rolla, or from the former except by the cohering calyx. Perhaps, 

 then, notwithstanding the ditference of structure of the ovary, 

 Schwpjia and Mackaya may be linked together, and form a small 

 (artificial?) group among the .Calyciflorse, connecting Santalaceae 

 with them, as Olax does with the Thalamiflorse ; a view which 

 would be strengthened, if we Avere to consider with Ur Lindley, 

 while speaking of Anthobolese, Nysseae,* and Santalacese, that the 

 superior or inferior fruit was a character of more importance than 

 the position of the ovules. Without, however, granting this, for 

 which I am not prepared, the two may be placed near each other, 

 till other and better affinities be discovered. 



I may here add with regard to Schoepjia, that although Vahl, 

 (under Codonium,') and some other botanists, describe the ovarium 

 as half-superior, that portion which is elevated above the margin of 

 the calyx is perfectly solid and fleshy, and appears to me to be more 

 an epigynous disk, similar to what is observed in many Rubiaceae, 

 than a part of the ovarium itself: the ovary is thus truly inferior, 

 and this constitutes the principal objection to uniting the genus to 

 Olacineae. The ovules are narrow oblong, attenuated at their point 

 of attachment, and reach from the apex of the columella nearly to 

 its base, with the angles of which they alternate. In S. arborescens 

 the style is cylindrical without any trace of furrows, and reaches to 

 about the middle of the corolla ; the stigma is capitate, and not at 

 all lobed ; it is, however, slightly flattened : the lobes of the corolla 

 are recurved and acute, nor do I see how S. Jlexuosa is to be dis- 

 tinguished except by its rather narrower leaves. Vahl mentions 

 that there is a small tuft of hairs on the corolla at the back of the 

 anthers, which I perceive also in my specimen (from Bahia:) there 

 thus seems to be no distinction in this respect between the Ameri- 

 can and Indian species ; but in one of those from India, and pro- 

 bably in both, the stigma is clavate and 3-lobed, and the corolla in- 

 fundibuliform, by which Decandolle's two sections may be charac- 

 terized. 



* The genus Ni/ssa differs from Santalacese not only in habit, but by the very 

 large embryo nearly as long as the albumen, which is thin ; the whole structure 

 of the seed is not very unlike that of Fagonia cretica, and some other genera of 

 the class Rutacese, as also of Euphorbiaceae : In those species I have examined 

 I have found only one ovule in the ovarium 



