IN CEYLON. 87 



varying length. I have not met with these multicellular 

 glands on the calyx of any other species, though they occur 

 abundantly over the carpellary wall of ^. Embryopteris 

 (pi. XIX., figs. 10-11). 



The calyx usually persists on the ripe fruits, but in many 

 of those which form deep cups the fruits seem to become 

 detached and the calyx is left on the twig. 



The general characters of the calyx segments and the 

 ferrugineous hairs with which they are very often coloured 

 are, according to Hallier,* of systematic importance and 

 comparable to similar structures in the Pittosporacefe. 



The same botanistf also calls attention to the fact that in 

 Scytopetalum the segments of the small calyx cannot be 

 distinguished, and suggests an affinity with the Ebenaceas. 

 We have seen that in^X oocarpa particularly, and also to a/ 

 less marked extent in D. Toposia, the calyx segments are 

 often undistinguishable. 



Corolla. The segments are united and usually, though 

 by no means always, isomerous with those of the calyx. 

 The colour is yellow, white, or greenish. The tube and 

 segments are often covered with silky or brown hairs, 

 sometimes on the outside only, at other times on both sides. 

 The margin of the segments in species like JD. hirsute is 

 very hairy. The segments are usually equal, and according 

 to HiernJ are contorted sinistrorsely in aestivation as 

 regarded from inside, except JX oocarpa, in which the 

 aestivation is variously imbricated. The corolla is occa- 

 sionally very conspicuous, as in D. Embryopteris and 

 D. Ebenum, but most usually is small and unattractive, as 

 in JD. Toposia and IX attenuata. The segments may be 

 relatively long, as in D. pruriens, or very short as in 

 D. Toposia. 



For convenience we will distinguish two types of corolla. 

 The first type comprises all those in which the corolla 



* Hallier, I.e., p. 33. | Hallier. I.e., p. 36. J Hiern, I.e., p. 60. 



