of the Flower of the Papilionaceae. 159 



Yet as there is no rule without at least an apparent excep- 

 tion, there may be persons who can bring forward a number of 

 facts which appear to speak against the correctness of the 

 theory here advanced ; but these exceptions serve, as far as I 

 have hitherto become acquainted with them, only to confirm 

 and extend the above positions, which I only maintain for 

 the true Papilionacecs suflficiently well characterized by their 

 (Bstivatio veooillaris. 



One might mention, for instance, the large groups of the 

 Ccesalpinece and Mimosea, which can scarcely be separated 

 from the family of the Leguminosce, in w^hich the almost re- 

 gular five petalled corolla now and then occurs together with 

 the characteristic pod, as not being in harmony with the law 

 above stated for the PapilionacecBy although the forms of flower 

 which here occur are nothing more than modifications pro- 

 duced by that law. 



The Caesalpinece are distinguished in addition to the erect 

 embryo, which is of no importance in our inquiries, from the 

 Papilionacece by the imbricate, the Mimosece by the valvate, 

 aestivation. 



The former appears to be produced by the calyx in the 

 C(Ssalpine(B being generally quinquepartite to the base ; it is 

 therefore not able to inclose the floral parts so tightly and to 

 press them on one another, as a gamosepalous calyx; the petals 

 can consequently develope more freely and adopt that aestiva- 

 tion originally peculiar to them. 



In this case almost all the petals are of like size and form ; 

 they expand freely, not being prevented by the calyx, and 

 approach in their outer appearance more to the rosaceous co- 

 rolla than to the papilionaceous : the stamina likewise rarely 

 cohere inter se, and we here find them arranged in two circles. 

 If on the contrary the calyx is cohering {Coulteria, Hb. B. 

 Kunth, Ccesalpinia^ L., &c.) we immediately find the papilio- 

 naceous corolla make its appearance. 



Further, when the ovarium in the Casalpinece is spherical 

 or cylindrical, then it will be less eccentric than the usually 

 occurring compressed ovarium ; its axis will approach nearer 

 to the imaginary floral axis than is otherwise the case, for it 



