ILLUSTRATIONS (23). ACICTTLAR-LEAVED THEES. 317 



terrupted dissemination of the same species of tree from the 

 southernmost part of Tierra del Fuego and Hermit Island, 

 where it was discovered as early as 1577 by Drake's expedition, 

 up to the northern Highlands of Mexico, over a meridian ex- 

 tent of 86 of latitude or 5160 miles. Where the acicular or 

 needle-leaved trees, as in the Swiss Alps and the Pyrenees, and 

 not the birch as in the extreme north, form the boundary of 

 arborescent vegetation on the loftiest mountains, which they 

 picturesquely encircle, they are immediately followed in their 

 ascent towards the snow-crowned summits, in Europe and 

 Western Asia by the Alpine roses, Rhododendra, and at the 

 Silla de Caracas, and the Peruvian Paramo de Saraguru, by 

 the purplish-red blossoms of the graceful Befarias. In Lapland 

 the Rhododendron laponicum immediately follows the Coni- 

 ferous trees; in the Swiss Alps, the Rhododendron ferrugi- 

 neum and R. hirsutum, and in the Pyrenees the R. ferrugineum 

 alone ; and in the Caucasus the R. caucasicum. But R. cau- 

 casicum has also been found isolated by De Candolle in the Jura 

 mountains (in the Creux de Vent), 5968 feet lower down, at 

 the inconsiderable height of from 3303 to 3730 feet. If we 

 would trace out the last zone of vegetation near the snow - 

 line we must name, according to our personal observation, in 

 tropical Mexico, Cnicus nivalis and Chelone gentianoides ; 

 in the cold mountainous tracts of New Granada, the woolly 

 Espeletia grandiflora, E. corymbosa, and E. argentea; in the 

 Andes chain of Quito, Culcitium rufescens, C. ledifolium, and 

 C. nivale ; yellow-blossomed Composite, which replace the 

 somewhat more northerly lanose herbs of New Granada, and the 

 Epeletiae, with which they have so much physiognomical re- 

 semblance. This substitution or repetition of similar and 

 almost identical forms in regions that are separated from each 

 other by seas or wide intervening tracts, is a wonderful law of 

 nature. It prevails even in the rarest forms of the floras. In 

 Robert Brown's family of the Rafflesise, separated from the 

 Cytineae, the two Hydnorae in Southern Africa (H. Africana 

 and H. Triceps), described by Thunberg and Drege, have, in 

 South America, their counterpart in the H. Americana of 

 Hooker. . 



Far above the regions of Alpine herbs, of the grasses and 

 the lichens, nay, beyond the boundary of perpetual snow, there 

 occasionally appears a phanerogamic plant, growing sporad- 

 ically, and as it were isolated, to the astonishment of bo- 



