PHANEROGAMIA : FLOWERS. 441 



layers are the more tough and solid, and the outer the more fleshy 

 and juicy stone berries (drupes) ; or the reverse true berries 

 (baccce) ; or, lastly, all the layers appear thin and dry, or leathery 

 (achcRnia). All these forms may, with the germen from which 

 they arise, be superior or inferior, one- or many-celled or one- or 

 many-seeded ; which only require to be noticed when deviations in 

 the structure of the germen have arisen through abortion, being 

 otherwise self-evident. 



a. The capsular fruits occur in the most diverse families. The 

 mode of bursting (dehiscence) is especially to be observed. The 

 simplest process is an apparent wholly irregular tearing open at 

 any place (as in Nicandra) : usually, however, the form of this 

 dehiscence is very regular, even though it may be confined to a 

 small part of the fruit (pericarpium poro dehiscens), as in Papaver, 

 Antirrhinum 9 &c. 



The solution of continuity is either vertical or horizontal: in 

 the latter case, the upper part forms a kind of cover upon the 

 under, and the capsule is termed circumscissile. In the first case, 

 the pericarp, &c. falls away in more or fewer separate pieces, which 

 are termed valves.* In many-celled fruits (a) the valves may 



separate entirely from the persistent septa, as in Cobaa scandens 

 (dehiscentia septifraga) ; or (b) the septa may split into two la- 

 mellae, and each valve may bear one of these lamella? on each of its 

 margins (dehiscentia septicida, valvulce margine septiferce) ; or (c) the 

 septa may remain undivided, adherent to the middle of the valves 

 (dehiscentia loculicida, valvulce medio septiferce). If in any of these 

 kinds of dehiscence a stalk-like mass of cellular tissue remains 

 standing in the axis of the fruit, it is called the columella. 



From what has been said, it is sufficiently evident that these 

 solutions in the continuity are not at all dependant upon the ori- 

 ginal composition. Common Botany, however, assumes such a 

 relation, and therefore applies to the line in the external circum- 

 ference of the pericarp, where the edges of real or pretended 

 carpels have become blended, the term of " dorsal suture," which 

 is only half correct according to this very hypothesis, while the 

 term " ventral suture " designates merely the line where the mar- 

 gins of one and the same carpel or similar part have become blended. 



* Sometimes tough cords of cellular tissue, persistent between each two valves, 

 nected in the stigma above (as in Argemone). I do not find that any special nami 



con- 



name has 



yet been made for these. 



