1893.] G. King — Materials for a Flora of the Malayan Peninsula. 237 



cell ; style various, or stigma subsessile, capitate. Drupe usually 

 ellipsoidal, more or less distinctly trigonous, with a 1-3-celled, 1-3- 

 seeded stone ; cotyledons often partite, contortuplicate. Distrib. Tropi- 

 cal Asia, chiefly in Malaya. Species probably about 90. 



The genera Santiria and Ganarium are not separated from each 

 other by any bold, well-marked distinction. As regards the Indian 

 species of both genera, I find the following characters the most diag- 

 nostic : — 



Canakium. Santiria. 



Calyx cupular or campanulate. Calyx cupular, never campanulate, 



Petals longer than broad, slightly sometimes quite flat. 



narrowed to the base. Petals rotund, with very broad 



Disc small, annular, or thin and truncate bases. 



cupular, or represented only by Disc large, cupular, fleshy, often 



the tube formed by the united corrugated. 



filaments. Filaments dilated at the base, but 



Filaments dilated at the base and never united into a tube. 



sometimes united into a tube. Flowers all hermaphrodite and uni- 



The flowers with fertile anthers form, as are the inflorescences. 



(practically the male flowers) 



smaller than those with fertile 



ovaries, but in larger inflores- 

 cences. 



In the majority of the species of Canarium in which the filaments 

 are united into a tube, I can find no other disc than that tube. 



The genus Trigonochlamys is also closely allied to Canarium and 

 Santiria, being distinguished from both by its much larger calyx, 

 depressed-globose ovary, and spherical drupe. In fact the characters of 

 the three genera so overlap each other, that, in my opinion, they must 

 either be kept distinct by characters more or less minute, as Messrs. 

 Bentham and Hooker have done, or united into a single genus. Dr. Engler 

 steers a middle course ; he keeps up Canarium and Santiria, but 

 unites Trigonochlamys with the latter, dividing Santiria into three 

 sections: (1) Icicopsis, with one species (S. Planchoni) ; (2) Trigono- 

 chlamys, with four species ; and (3) Fu-santiria with 22 species. The 

 section Icicopsis contains one 3-androus plant, which, as it has the 

 dimorphous flowers and inflorescence of Canarium, I have transferred 

 to that genus. I admit that the drupes of the plant in question more 

 resemble those of Santiria than of Canarium : but, in the majority of its 

 characters, it appears to me to agree better with the latter genus. As 

 regards Dr. Engler's section Trigonochlamys, it is difficult to see how 

 Canarium and Santiria are to be kept distinct as genera, if the four 



