PERUVIAN REGIONS. 



209 



with the corolla spreading and more or less rotate, usually blue, but 

 in one exceptional species bright red. An allied genus was further 

 remarkable for the spreading corolla having its numerous segments 

 alternately blue and white. 



4. Garyoplvjllacem. Consisting principally of inconspicuous-flowered 

 and rigid-leaved Saginm? and Cherleria-s?; growing in alpine tufts or 

 masses, each based on a thick woody subterraneous stem. 



5. Malvacece. Including perhaps as many as three genera: the 

 species mostly blue-flowered, and one of them much resembling the 

 so-called Sida? Pichincensis. 



6. Valerianacece. Some of the species excessively congested. 



7. Umhelliferoi. Less frequent than in the alpine ground of the 

 Austral Andes, South of the Tropic; and consisting chiefly of con- 

 gested Fragi^osas. 



In addition to the above seven prevailing Tribes, Paronychiacece 

 were frequent ; including however but a single species, its congested 

 tufts rendered conspicuous by projecting membranaceous stipules. 



Of Chilian and Puegian analogies, some were rather striking: as 

 the presence of congested UmhellifercB ; of a congener of the Fuegian 

 Caltha? appendiculata; and of a one-floivered Plantago, with resisting 

 leaves, stellately dividing the alpine turf. 



On the other hand, the connection with the alpine ground to the 

 Northward seemed more intimate; and recalling specimens obtained 

 by Jameson on the Andes of Quito, I recognized Equatorial alpine 

 analogies, in the divided-leaved Saxifmga, rotate-flowered Gentianas, 

 and the blue-flowered Malvaceous genus that includes Sida? Pichin- 

 censis. 



Negative characters. One point of difference from the Equatorial 

 alpine ground was however remarked; the absence of large-flowered 

 Culcitiums. 



The rarity of Lichens in such a moist alpine district did not seem 

 readily explicable. {3Iusci., or mosses were not noted, and are proba- 

 bly also rare). 



Contrary to all expectation, annual plants ^\ ere not altogether want- 

 ing; at least, the slight roots of two diminutive species of Compo- 

 sites would under a different climate be regarded as annual. 



The absent tribes of plants continued generally the same as in the 

 Paramera below : but there were further deficiencies, in the disappear- 

 ance of 



53 



