48 THE VICTORIAN NATURALIST 



leaves and flower, but different in fruit, requires yet to be cleared up. 

 Only Mr. Clialmeri^ specimens are in fruit, but devoid of flowers. 



Our plant is evidently allied to S. guttata, but tbe foliage and 

 panicle:? are not liairy, the lobes of tlie calyces are not so long, and 

 the fruitlets are not turgid. More distant it is yet from P. Roxhurghii 

 on account of the shorter stalklets, color of flowers and form of 

 fruitlets. The plant shows also some afiinity to*?. Blumei, S. cuspidata 

 and a few otliers as yet imperfectly described. 



The collections, formed during the Australian Geographical 

 Society's Exi^edition, contains from the tStrickland-River a Sterculia, 

 wlrich may be referable to S. rubiginosa; but in absence of fruit the 

 identification remains doubtful. 



In seeds of a Sterculia, sent from India as belonging to S.foetida^ 

 nearly the same proportion of the cotyledons to the albumen is 

 observable, as in those of Edelfeltii, but the cotyledons being not 

 readily separable from the albumen; these Indian seeds have a 

 yeiy small yellowish carnulent strophiole of oblique ovate-renate 

 shape; there is no substance between the brittle thinly coriaceous 

 testa and the cartilaginous inner tegument. The carpolo^ic 

 characteristics are evidently among the best for the demarkation of 

 the species of Sterculia also. 



In all probability the bast of these Papuan congeners, like that of 

 some otlur species, could be used for textile fabrics, and the seeds 

 may prove edible as those of several other kinds. The macerated 

 flo\\ers exude much gelatinous mucilage, thus these trees are sure to 

 yield a Bassorin- Gum. 



In using also here plainly the English term fruitlets (instead of 

 carpels from Latin or better carpids from Greek), for the disjoined 

 portions of Sterculia-fruits, developed each from a distinct ovary, 

 no ol'jection can be taken to this expression merely on account of 

 size, in as much as also the leaflets of so very many kinds of plants 

 are very large indeed, and as in the same manner stalklets may very 

 often be thick and elongated, and therefore of large dimensions. 



It ?eems to me, that we would obtain a very acceptable generic 

 group, if in restoring Fzriin'aiut we added to it Hildegardia, Scaphium 

 and Pterocymbium. the fruitlets of all being membranous and the 

 radicle inferior, — Tarrietia being very closely oognato, and Courtenia 

 having by Bentham and Masters already been joined to Cola, a 

 genus, which received so far back as 1023 its name from Caspar 

 Banhin, as shown by Pfeiffer. — CarpoplyJlium is likely also referable 

 to Firmiana, in as much as Miqnel does not positively speak of the 

 direction of the embryo. Pterocymbium would form a separate 

 section in the genus Ftrviiana, on account of tlie disposition of the 

 anthers, uniseriate as in Cola and Heritiera. The fruitlets of iT. 

 literalis occasionally ripen two seeds. Sterculia paUens is referable 

 to Firmiana, and nearest allied to F. diversifolia. 



