408 Dr. Planchon on Meliantbeae, 



flowers bearing the strictest analogy to the normal or usual structure of those 

 of Melianthus, we shall be inclined to take so striking a conformity as a sign 

 of some close connexion ; and perhaps the very conflicting points, which are 

 distinctive generic marks of the plants under consideration, will throw a new 

 and unexpected light upon their general as well as mutual affinities. 



The plants I have been just alluding to are no ideal types, but two remark- 

 able genera lately discovered in several parts of tropical and subtropical 

 Africa, one of them, Natalia, Hochst., being represented by one species in 

 Port Natal, on the coast of Cafferland, and by another species in Sierra Leone ; 

 while the two known species of Bersama, Fresen. (the second genus), 

 seem to be confined within the limits of the Abyssinian flora. Both genera 

 agree in their habit and structure to such a degree as to leave no doubt of 

 their being rightly approximated by M. Hochstetter, although M. Endlicher, 

 probably through a pardonable inadvertence, leaves Bersama among Ampeli- 

 dece, whose characters are totally at variance with it ; while, at the suggestion 

 of M. Hochstetter, he assigns to Natalia a place among Sapindacece. On the 

 other hand, M. Ach. Richard, who in his 'Tentamen Florae Abyssiniae' has 

 recently given a correct figure of Bersama, seems to make no doubt of its 

 being a true Meliacea ; although the position of the disc, outside instead of 

 inside of the stamens, would suffice to exclude the genus from that very clearly 

 defined order, and point out its affinity to Sapindacece. But if Natalia and 

 Bersama are in fact closely allied to this last group, it is no less certain that 

 they do not come within its limits : for the want of albumen, and the greater 

 or less curvature of the embryo, are characters of primary, and one may say, 

 necessary importance in Sapindacece ; and, contrary to this, Bersama and 

 Natalia have like Melianthus a narrow, straight embryo in the axis of a 

 copious albumen. 



But, to revert to the analogies or differences between those plants and 

 Melianthus, let us take a short notice of the flower of Natalia. There the 

 calyx is comparatively small, and its irregularity not very striking. How- 

 ever the two inferior of its five segments are connected into one, with only a 

 slight emarginature to mark their limits. Five unguiculate and thick petals 

 alternate with the calycine segments, above which their nearly uniform bor- 

 ders are almost equally spread. But a closer examination shows that the 



