Mr. B. Clarke on a new Arrangement of Phanerogamous Plants. 189 



\-flori, aut simpliciter pauciflori, breves, in paniculam foliosam 



vel subnudam dispositi. 



I have nothing to add to the excellent description and arrange- 

 ment of the species, as determined by Mr. Bentham*, in D.C. 

 Prodr. X. 193. 



XVII. — Observations on Relative Position ; including a new Ar- 

 rangement of Phanerogamous Plants. By B. Clarke, F.L.S. 

 &c. 



[Continued from p. 90.] 



Part II. 



On the Position of Carpels. 



As the progress of discovery shows that Jussieu's system in its 

 primary divisions, viz. Monopetalous, Polypetalous, and Ape- 

 talous, leaves unassociated plants between which there is the 

 closest resemblance in both structure and habit, and that in 

 numerous instances, — it has become desirable to form primary 

 divisions depending on different characters, but retaining as far 

 as possible those of Jussieu as of subordinate value. How far 

 the present attempt is successful the Tables will show, and the 

 researches connected with the relative position of carpels to the 

 axis will I trust prove of interest, and may also assist in deter- 

 mining questions of affinity which at present remain unsettled. 



It being so common for the ovary to consist of two or only 

 one carpel, in either case having a variable relation to the axis, it 

 becomes interesting to trace the cause of this reduction, and 

 more especially the causes of the variations in their position ; 

 and of these inquiries, the varying position of the two carpels of 

 dicarpous ovaries affords the most satisfactory explanation. Thus, 

 the cause of the difference of the position of the carpels when 

 reduced to two, is explained by the mode in which the changes 

 of position occur when a tricarpous ovary becomes dicarpous. 



Differences in the Position of the Carpels when Two. 



1. When the two carpels are right and left with respect to the 

 axis. In the genus Carex the three carpels are ordinarily two of 

 them right and left and one posterior, and when reduced to two, 

 they are (in the species examined) uniformly right and left ; in 

 Malpighia coccifera the three carpels have also the same relation 

 to the axis, the posterior one being smaller ; and in Banisteria of 

 the same family, the carpels when only two are right and left. 



* Analytical details of the aestivation and structure of the five different 

 sections of this genus will be given in plate 63 of the 111. So. Amer. Plants. 



