ANNOTATIONS AND ADDITIONS. 147 



i 



tica, has treated and contrasted with great sagacity and 

 clearness two very different phenomena which the distribu- 

 tion of plants presents to us : on the one hand, ' ' unifor- 

 mity of surface accompanied by a similarity of vegetation ;" 

 and on the other hand, " instances of a sudden change in 

 the vegetation unaccompanied by any diversity of geological 

 or other features/' (Joseph Hooker, Botany of the Antarc- 

 tic Yoyage of the Erebus and Terror, 1844, p. 210,) 

 Is there any species of Erica in Central Asia ? The plant 

 spoken of by Saunders in Turner's Travels to Thibet (Phil. 

 Trans. Vol. Ixxix. p. 86), as having been found in the 

 Highlands of Nepaul (together with other European plants, 

 Vaccinium myrtillus and Y. oxycoccus) and described by 

 him as Erica vulgaris, is believed by Robert Brown to have 

 been an Andromeda, probably Andromeda fastigiata of Wal- 

 lich. No less striking is the absence of Calluna vulgaris, 

 and of all the species of Erica throughout all parts of the 

 Continent of America, while the Calluna is found in the 

 Azores and in Iceland. It has not hitherto been seen in 

 Greenland, buo was discovered a few years ago in Newfound- 

 land. The natural family of the Ericacese is also almost 

 entirely wanting in Australia, where it is replaced by Epa- 

 cridea?. Linnasus described only 102 species of the genus 

 Erica; according to Klotzsch's examination, this genus 

 really contains, after a careful exclusion of all mere varieties, 

 440 true species. 



( 20 ) p. 24*."The Cactus form." ' 

 If we take the natural family of the Opuntiacese separated 



