614 PLANTS JAVANlOffi RARIORES.
more limited than Ventenat appears to have done; con- sidering it as sufficiently distinct from B/dtneriacece, to great part of which, however, Ventenat's two principal characters of Sterculiacece equally belong.
221] In endeavouring to establish the characters andaffini- tiesof Pterocymbium, it became necessary to examine those of the established genera most nearly related to it, and as this examination has led me to adopt considerable alterations in arrangement, and enabled me to make some additions to the number of species, I shall here give the characters of the genera and species of that tribe to which the name of Stercidiece has been applied by DeCandolle and Endlicher, prefacing that account with a history of the tribe, and espe- cially of the genus Sterculia, from its formation to the present time.
In the botanical history of Sferculia, it is not necessary to go farther back than 1747, when Linnaeus first published his character of the genus, 1 founded on the specimens and unpublished figures of Hermannus, whose Ceylon herbarium, of which the Museum Zeylanicum is a catalogue, had been sent from Copenhagen for his inspection. The result of a careful examination of this herbarium was the publication in 1748 of the 'Flora Zeylanica/
Hermann's herbarium (purchased by Sir Joseph Banks and now in the British Museum) contained flowering specimens of Sterculia fcetida and Balanglias, and the fruits of both species were figured by Hermann himself in the volume of drawings which accompanied the herbarium. From these materials, confirmed no clonbt by the figures of the fruits in 'Hortus Malabaricus' and 'Herbarium Amboinense, 5 the Linnean genus was entirely established. The character given, however, is strictly applicable to Sterculia fcetida only ; and Linnaeus was probably induced to refer Bcdavglias to the same genus, either from not having distinctly seen the remarkable form of the flower in that species, or, which is more likely, disregarding that difference, was determined by the exact resemblance in its
1 'Nova Genera Plantarum, respondents Basso?*:? p. 13.
�� �
