including a new Arrangement of Phanerogamous Plants. 191 



Differences in the Position of the Single Carpel. 



Ovaries in which the carpel is single are for the most part the 

 result of the abortion or non-development of one of the carpels 

 of a dicarpous ovary ; and the position of the carpel depends 

 consequently upon whether it is an anterior, or a posterior, or 

 one of two lateral carpels which is absent. 



1. When the single carpel is anterior. When the ovai'y in 

 Myrtacefe, Bruniacese, Onagracese, Polygalacese and Acanthacese 

 is one-celled, it is the posterior cell which has disappeared, and 

 to these may be added Tetragoniacese and Ulmacese among ape- 

 talous plants*. 



2. When posterior. A single carpel posterior may be ex- 

 plained in an analogous manner. In Houttuynia for example, 

 the carpels when reduced to two are for the most part anterior 

 and posterior, and in the instances of single carpels they are all 

 directly posterior. This offers an explanation for the unusual 

 position of the carpel in Piperacese, where when single it is con- 

 stantly posterior. 



3. JVhen lateral or oblique. The position of the single carpel 

 on one side of the flower, either directly lateral or more or less 

 oblique, is frequently owing to the same circumstance, as in 

 Elatostemma and Morus, where the carpels are frequently lateral. 

 The stigmas here being two, one of them is continuous with the 

 dorsal rib of the fertile carpel, and the other corresponds with 

 the placenta, being a part of the rudimentary carpel. 



From the causes particularly noticed as occasioning the dif- 

 ferences of the position of the two carpels in dicarpoiis ovaries, 

 it necessarily follows that in such cases a single carpel anterior 

 and a single carpel lateral would be the same carpel in different 

 positions tj and consistently with this inference, those Orders 

 which have the carpel always lateral are also very nearly allied 

 to those in which it is always anterior, with the exception of 

 Nyctaginese (and some others in which the carpel is lateral only 

 in part) as afterwards explained. Such Orders are therefore 

 placed in the Proterocarpous Division. 



Position of the Carpels when Three. 



Prom the foregoing and other analogous examples it seems to 

 follow as a theoretical inference, that the regular number of car- 

 pels in all ovaries where they are definite is three, or a multiple 



* The term anterior is used as synonymous with inferior, and posterior 

 with superior. 



t Those cases must be excepted where two carpels right and left occur 

 in an ovary in which the third carpel is anterior, as in Menispermum lauri- 

 folium and Maranta dichotoma. 



