PLANTS JAVANICiE RARIORES. G19
most remarkable circumstance in his character of Sterculiece, is his describing the embryo as erect, which, if I rightly interpret his meaning, although applicable to a few species, is directly contrary to the more usual structure. Sterculia as a genus he distinguishes from Heritiera by the existence of albumen, which, however, is not universally present, and dehiscence of the follicles ; the direction of embryo is not noticed in his characters of either of these two genera.
In 1827 M. Auaruste de St. Hilaire, in the < Plantes Usuelles des Brasiliens,' gives, appended to his account of a supposed new species of Sterculia, an improved character of the genus, though in this character he describes the albu- men as a coat of the seed ; the " embryon antitrope ' is considered as common to the whole genus. In the remarks that follow he supports Linnseus's account of his classifi- cation of the genus, in opposition to the observations of Jussieu which he refers to, but which he does not seem to be aware were adopted from Cavanilles, and perhaps also from Forskael, with whom a similar account originated.
In the same year he published a Livraison of his ' Flora Brasilia^ Meridionalis/ in which work, as in the 'Plantes Usuelles/ owing to the state of his health, he was assisted by MM. Adrien de Jussieu and Cambessedes. In the ' Flora Brasilise ' 1 an enlarged, and in many respects improved character is given of Sterculia, of which the principal dif- ference from that in the ' Plantes Usuelles ' is, his admit- ting the existence of albumen divided into two equal seg- ments, which, he correctly states, frequently cohere with the corresponding cotyledons, the first distinct notice I believe of that remarkable economy : he states also, that where the seeds are ascendent, the radicle of the embryo points to the hilum, and where they are transverse, in the diametrically opposite direction. His character, therefore, of the genus is so framed as to include every species of Sterculia of DeCandolle, as far as the author was acquainted with their structure. It will hereafter appear, however, that neither is the existence of albumen universal, nor is
1 Vol. i. p. 277.
�� �
