jphysiologically considered. • 533 



in all these cases, the whorl of fruit- leaves situated at their base 

 is attached to the corresponding bundle of vessels of their pla- 

 centae, and connected with their edges. The pith of the axis dis- 

 appears more or less in this form. An intermediate formation 

 between these forms of fructification and the last-mentioned 

 takes place, for example, in Papaver and iS/ymphae'a, in which the 

 middle column has disappeared ; the medullary rays, however, re- 

 main, which have many bundles of fibres, and produce ovula over 

 their whole surface. Here, also, two principal modifications occur. 

 a. The separation of the fruit-axis begins under the point, 

 and continues to the base, so that all the bundles of fibres 

 bearing ovula end in a style, and form, as it were, the skeleton 

 of the fruit. The fruit-leaves are so attached to this skeleton, 

 that the two edges of each cover the two separated bundles of 

 vessels belonging to each placenta, so that two bundles of fibres 

 from different placentee are brought together. These grow more 

 or less firmly to a false placenta, which is produced from the inter- 

 srovvn edfifes of the fruit-leaves,^ and vvhich must therefore alter- 

 nate with the lobes of the scar (Lappen der Narbe) formed by 

 the point of the fruit-leaf. In this manner of fructification, if the 

 original placenta should grow undivided, it would rest on the 

 midrib of the fruit-leaf; but I know of no example of this. If 

 the inter-growth of the two bundles of fibres which form the 

 false placenta is weak, they remain fixed, on the bursting of the 

 capsule, to the separating edges of the fruit-leaves, as in Helian- 

 themum and Parnassia ; but if it is firm and interwoven, on the 

 bursting of the fruit-leaves, the false placenta remains as a skele- 

 ton of the fruit, as is beautifully exemplified in Argemone. The 

 greater number of variations of this class of fruits arises from 

 the failure of several bundles of vessels, as well as of the corre- 

 sponding fruit-leaves of the whorl, so that in Chelidonium only 

 two placentae and fruit-leaves are found. The separated bundles 

 of fibres grow together, as in Argemone, and remain, after the 

 bursting of the fruit-leaves, as two opposite placentae. In Glau- 

 cium, these two placentae are also united by an enlargement of 

 the pith of the fruit-axis, which becomes very large, and is co- 

 vered with a thin epidermis. This growth fills the fruit almost 

 entirely, and divides it into two compartments ; so that the two 

 bundles of fibres of each false placenta are separated by it, and lie 

 in the two different compartments. The fruit of the Cruciferae 

 is formed in the same simple manner; only the enlargement 

 of the pith covered with an epidermis is much more delicate 

 and slight than in Glaucium. Afterwards it diminishes still 

 more, so that, when the fruit is ripe, the pith is only found in 

 patches between the two membranes of the partition, producing 

 the light and dark spots and stripes on which R. Brown pro- 

 posed founding generic characteristics. In some, it is, however, 



M M 4 



