26-4- JUSSIEu's OBSERVATIONS 



are the Rhododendra and the Erica ; both adoptee! 

 in the " Genera plantarum," published by me hi 

 1789. The fruit of the former is a capsule, of which 

 each valve, folding itself so as to make its edges 

 approach inwardly, forms a complete loculament, 

 absolutely separate from, and independent of, 

 the loculament formed by the neighbouring valve 

 (valvte introflexce loculum proprium constituentes ) '. 

 The edges of the valves are attached to a central 

 receptacle, bearing numerous minute feeds. 



The valves constituting the capsule of the heathsr 

 are of quite a different structure. Each of them 

 bears along its middle a septum (valva medio septi- 

 fera:)^ which, being applied against the central 

 seed-bearing column, forms a separate loculamenty 

 produced by two valves, conjointly contributing 

 each its half. This distinction, as it affords (even 

 without ether distinctive characters) an invariable 

 line of partition between these two orders, is 

 now adopted by the followers of the natural 

 system, and has enabled M. Ventenat to 

 shew, that the Epigasa, arranged by me among 

 the Erica?, must be transferred to the Rhodo- 



