is the more interesting, because its petals, distinct from the 

 sepals, shew we were right in considering the tendency ot 

 Begonia to be towards the production of a corolla, and that 

 it is consequently correctly associated in the Nixus plantarum 

 ^^'ith Polypetalous orders ; and secondly, that two or four 

 being the number of the floral envelopes, when distinctly 

 formed, the relationship of Se^omacece to OnagraricBis almost 

 demonstrated. 



