including a new Arrangement of Phanerogamous Plants. 197 



and Piperal Alliances^ especially in such Balanophorese as have 

 but one carpel, in which the inflorescence resembles that of 

 Aracese. This opinion perhaps may derive some support from 

 the figures of the species of Balanophora by Mr. Griffith in 

 ' Trans. Linn. Soc' vol. xx., where the carpels are apparently 

 irregular in their position ; and should this prove to be the fact, 

 it will be an additional reason for regarding Ehizanths as the 

 common basis of Endogens and a part of Exogens, which seems 

 to be indicated by his proposed distribution of them. 



Endogens. 



That all the great sections of Exogens in their higher deve- 

 loped forms may become Proterocarpous is evident from the 

 Tables (vide Table II. Derivations), and hence it might almost 

 be anticipated that Endogens would also ; but as it is very rare 

 in those sections of Exogens which approach Endogens, so 

 it may be very rare in Endogens themselves, and hitherto I 

 have only observed it in Pontedera. It seems however not un- 

 likely to occur in the greater part of Orchidese, but as in the 

 lower forms of Endogens the position of the carpels when two is 

 variable, the exceptions may be confined to petaloid forms with 

 irregular flowers, such as Pontederacese. 



Gymnosperms. 



In Pirnis the flattened expanded ovary is always anterior, as is 

 also the succulent carpel of Podocarpus ; and as far as the other 

 genera of this section of Phanerogams show any traces of a car- 

 pel, or an envelope which possibly may be a rudimentary carpel, 

 it is always anterior, as the external tunic of Saxe-Gothcea and 

 Gnetum. 



DiDYNAMOXJs Stamens. 



Prom the fact that as far as the didynamous monopetalous 

 families show any tendency to suppression in the carpels of their 

 dicarpous ovaries, it is for the most part or always in the pos- 

 terior one, it might be supposed that the stamens and carpels 

 follow the same order in this character, as it is the posterior 

 stamen which is deficient, and thus the cause of didynamous 

 stamens might be explained. And possibly this may be Mr. 

 Ralph's reason for considering the anterior as the odd carpel in 

 Scrophulariacese (vide Proceedings of Linn. Soc. vol. i, p. 284). 

 But this coiTcspondence in position between the fertile carpel and 

 stamens is perhaps confined to the Proterocarpous Division, as 

 in repeated instances in which the stamens of Pimelea decussata 

 were reduced to one, it proved to be always on the opposite side 

 of the flower to the carpel, the stamen being anterior and the 



