160 MR. C. B. CLARKE ON DIMORPHISM IN THE RUBIACE.E. 



styled. In the latter the stamens are very nearly, or quite, free 

 from the stamen-tube and stand on the ovary. The anthers and 

 the ovaries and ovules in the two forms are indistinguishable, as 

 are the fruits. 



Of Adenosacme, Hook, f., in Bth. & Hk. f. Gen. PL i. 69, gives as 

 part of the generic character, u Stamens inserted on the middle or 

 above the middle of the coro]la." On the other hand, Wallich,in his 

 description of A. longifolia (under the name Bondeletia, in Eoxb. 

 PI. Ind. ed. Carey & Wall. ii. 137), states, u Stamens inserted at the 

 base of the corolla-tube ; filaments very short ; anthers oblong, not 

 elevated beyond the height of the calycine lacunas." The Deccan 

 species of Adenosacmeis A. acuminata, figured by Wight under the 

 name Lawia (Wight, Icon. 1070) ; Wight thereunder states the 

 stamens are at the very bottom of the corolla-tube, but gives two 

 dissections in his plate showing the stamens quite free from the 

 corolla, standing directly on the top of the ovary. Wight was 

 clearly surprised at such a position of the stamens in the order 

 Eubiaceae. I suppose that I have explained the extreme contra- 

 diction in the statements of Hooker, Wallich, and Wight regard- 

 ing this genus. 



In examining again the dried flowers, I find that in my long- 

 styled examples of A. longifolia the corolla easily separates, 

 leaving the stamens standing on the ovary : the pollen of these 

 appears perfect, and like that from the short-styled form. The 

 long- and short-styled forms are in all my examples produced on 

 separate individuals. 



I need not occupy the time of this Society by enlarging on the 

 systematic value attached to the distinction between stamens on 

 the corolla-tube and stamens on the disk, which in the genus 

 Adenosacme fails us even as a specific character. Dimorphism is 

 here carried to a degree which I have not seen before noticed. 



The second kind of dimorphism is perhaps yet more worthy of 

 note, as it is carried into the fruit. 



JBandia uliginoea, DC, is a small tree very common in the 

 swamps of Bengal. I at first supposed there were two curiously 

 similar species. My lamented friend Sulpicius Kurz states (in his 

 4 Forest Flora of British Burma,' ii. 45) that some trees produce 

 sessile flowers succeeded by fruits the size of a hen's egg ; other 

 trees long and slenderly pedicelled flowers succeeded by fruits 

 about half the size of the former. This is so ; the sessile flowers 

 (fig. 3) have the corolla-tube longer with a long style, the stigmas 



