iMALPlWHIACEJS. 447 



petals ; the disk is generally entirely wanting, and it never forms 

 a regular or irregular enclosure outside tlie androceura. The latter 

 is nearly always formed of ten parts arranged on two equal and com- 

 plete verticels. The gynpeceum of the Malpighiacece is always 

 exactly central. In its ovary cells, two or three m number, there 

 is never more than one ovule, always descendent, with the micro- 

 pyle superior, nearly always carried back, in consequence of torsions 

 on one side of the point of attachment. Moreover, the leaves of the 

 Malpighiaecoi are nearly always opposite, an exceptional character 

 in the Sapindacew^ and they are never compoimd. On another side, 

 the Melimea^ in consequence of their great analogies to the regular 

 Hapindacecc^ are also brought near the Malpighiacea' ; they have an 

 hypogynous disk, usually well developed, ovary cells often biovulate, 

 occasionally pluriovulate, descendent ovules, with micropyle turned 

 directly upwards and outwards. The leaves are alternate, frequently 

 compound-pinnate. 



I 



The characters varying in this group and serving, consequently, to 

 distinguish the series and genera, are, in the first place, as we have 

 seen, those drawn fi'om the configuration of the fruit and the abso- 

 lute number of the stamens. A. de Jussieu named the Banisteriece 

 Notopferi/giccc^ the Hireete Pleuro'pterygiece^ the Mulpigldew Aptcry- 

 gicce, the three collectively diplostcmonous, and the Gaudichuudiem 

 meiostemonous. Then come, in a lower rank, the presence or absence 

 of the calycine glands ; ^ the equality or inequality of the petals, 

 entire or dentate ; the glabrous or hairy surface of the parts of the 

 androccum ; the independence or union of the ovary cells ; the con- 

 figuration of the styline branches ; the details of the configuration 

 of the fruit, the wings, sides, ridges, or hairs ; the variations of the 

 inflorescence ; the existence or not of glands or hairs on the surface 

 of the leaves and organs of vegetation. The latter frequently 

 assume, in this family, a form fi-om which they derive their name of 

 MulpigJdaceous, or shuttle-shaped hairs. They are to a certain 



' Thoso glands have Ijooii studied principally very often they are lateral with reference to the 



by A. DE JuHSiEU {Mulpiijh. WA, 92, t. 2), who exterior face of the sepals. To us, they 



has analysed the tissue and secreted product, appear analogous to the stipular glands so 



and has shown its slight value for classification. frequently observed in the family of £'«a/io)'- 



Their evolution on the calyx is usually tardy ; biaccai. 



, 



