ORGANS OF EEPBODUCTION. 



269 



examples are to be considered, therefore, as transitional states 

 between apocarpous and syncarpous pistils. 



It is far more common to find the carpels united by their 

 lower portions or ovaries, and this union may also take place to 

 various extents. Thus, in the Rue {fig. 597, ov) the union only 

 takes place by the base of the ovaries, the upper parts remain- 



Fig. 594. Fig. 595. Fig. 596. Fig. 597. 



Fig. 594. Pistil of Xanthoxylon fraxineum supportpd by a gynophore, g. 

 The ovaries, o, and styles are distinct, but the stigmas, s, are united. 



Flg.h^h. Pistil of korehound {Marrubium vulgare), a Labiate plant. 



Its ovaries, ov, are distinct, tlie styles, s, being united. Fig. 596. 



Pistil of Jfi/oso«i«,aBoraginaceousPlant. ov. Distinct ovaries, d. Styles 

 united. Fig. 59". Flower of Rue iRidagraveolensj, showing the ova- 

 ries, ov, united by their base. 



ing distinct, in which case the ovary is commonly described as 

 lobed. In the Dianthus {fig. 588) the ovaries are completely 

 united, while the styles are distinct ; while in the Primrose 

 {fig. 5G7), the ovaries, styles, and stigmas are all united. When 

 two or more ovaries are thus completely united so as to form one 

 body, the organ thus resulting from their union is called a 

 com'pound ovary. 



When two or more ovaries are completely united so as to 

 form a compound ovary, the compound body formed may either 

 have as many caA-ities separated by partitions as there are com- 

 ponent ovaries, or it may only have one cavity. These differences 

 have an important infiuence upon the attachment of the ovules, 

 as will be afterwards seen when speaking of plaeentation. It 

 will be necessary for us, therefore, to explain at once the causes 

 which lead to these differences. Thus suppose we have three 

 carpels placed side by side {fig. 598, a); each of these possesses 

 a single cavity corresponding to its ovary, so that if we were to 

 make a transverse section of the whole {fig. 598, b) we should 

 necessarily have three cavities, each of which would be sepa- 

 rated from those adjoining by two walls, one being formed by 

 the side of its own ovary and the other by that of the one next 



