A OKOGRAPHICAL SKETCH OF THE JUAN FERNANDEZ ISLANDS I43 



simplifies matters for the surveyor wiiere, in contradistinction to Masatierra, 

 boats cannot be used to move from one place to another. 



The longitudinal axis is directed N by W — S by VI. The statements regarding 

 the si/e of the island vary a great deal: 9 miles (14.4 km) by 4 miles (6.4 km) with 

 a circumference of about 46 km and an area of CS5 sq. km, all according to Lopez; 

 1 1.7 by 7.4 km (Johow's map); Branchi, 10 by 6 km, area 84 sq.km; Gunther, 

 9.25 by 6 km, and Guzman, 17(1) by 7 km, a circumference of 55 km and an 

 area of 85 sc]. km. The figures taken from my map are 10.3 by 6.2 km and the 

 area 57.6 scj. km. 



Geology and morphology. 



In 1908 Ouensel, the only geologist who ever visited Masafuera, made a 

 hasty survey of the rocks along the coast from O. Casas to Loberia Vieja and 

 north of Casas from the shore to perhaps 1300 m altitude. His field notes and 

 specimens served him for a description of the principal geological features. In 

 structure the two islands differ considerably. Both are built up of innumerable 

 volcanic beds of varying thickness and tilted about 20°, both exhibit two more 

 or less distinct horizons, both are traversed by vertical dikes, but here the 

 resemblance ends. On Masatierra the lower strata consist of lavas very rich in 

 olivine, interbedded with tufifaceous deposits, deeply weathered and denuded, 

 resulting in the formation of V-shaped valleys. The lower horizon of Masafuera 

 consists of hard, dark gray to black vesicular feldspar basalts (Ouensel 2 p. 58), less 

 rich in olivine and very resistant to the denuding forces. These rocks prevail 

 up to approximately 1000 m. My notes from 191 7 contain a statement that W 

 of Vicente Porras, where Playa Larga begins, a bed of ''conglomerate" was 

 observed, traversed by harder dikes, but unfortunately no specimen was pre- 

 served. In a paper not quoted by Ouensel, who did not mention the occurrence 

 of agglomerate beds on Masafuera, R. A. Philippi described 14 rock samples col- 

 lected by Germain, most of them "losen, von der Hohe heruntergerollten Blocken 

 entnommen". I doubt that anybody will be able to identify these rocks from 

 Philippi's unscientific description, but two of the samples suggest the occurrence 

 of tufifaceous deposits: they consist of "Rapilli" cemented together. 



At about 1000 m a lava bed of a different aspect was met with, a light gray, 

 "iddingsite-bearing phyric olivine basalt" (Quensel 2 pp. 58 — 63). Between 1000 

 and 1 1 00 m a rock of quite another type occurred, described as a light yellow- 

 ish green soda trachyte. Apparently a closely related type is found also at 

 lower levels; boulders of this rock are so common in the talus along the 

 south shore that the place was called Tierras Blancas. Ouensel arrived at the 

 conclusion that "lavas of soda-trachytic composition have been emitted at different 

 times and that they have alternated with lavas of more normal basaltic com- 

 position" (2 p. 66). 



The upper horizon, from about 1 100 to 1400 m. is formed by dense, ash 

 gray feldspar basalts, which probably form the whole upper complex of lava 

 beds with exception of the summit of Mt. Inocentes ; these rocks very much 



