ON ERICA DABOECIA. 265 



dendra, having valva introflexa^ each forming a 

 loculament by itself. 



Dr. Smith,* guided by the same circumstances, 

 has referred to the Rhododendra his Menziesia, a 

 genus, which, with the characters of a fourcleft 

 pitcher-shaped corolla, of eight perigynous sta- 

 mens, of an ovarium liberum furnished with a style 

 and stigma, unites that of a capsule with four 

 valves introflected exactly in the manner of those 

 of Rhododendrum and Epigsea. 



On examining the capsule of Erica Daboecia, 

 the same structure will be perceived as in the 

 Rhododendra, viz. four retroflected valves ; from 

 which circumstance it is naturally to be inferred, 

 that this plant, so far from remaining allied to the 

 genera Erica or Andromeda, must even be entirely 

 detached from the Order of Ericce, and transferred 

 to the neighbouring one. The inspection of its 

 other characters, as derived from form and number 

 of its several. parts, moreover proves that it must 

 belong to the genus Menziesia, j from which it 



* Plantarum iconcs hactcnus incditx. Fa?c. III. p.ig & tab. j6. 

 I Also Mr. Salisbury, in his edition of Thinbug's Distertati* 

 d* Eliot, printed at Featherstone, 1800, mentions the circumstance of 



