ADDITIONAL COMMENTS ON THE GEOLOGY OF THE JUAN FERNANDEZ ISLANDS 59 



Fig. 16. Vesicular feldspar basalt, x 8. Normal development under i 000 m elevation. Masafuera. 



sition. The analysis (Table I, No. 7), however, shows that the lava must be 

 assigned to the basaltic group, though in many respects it obviously differs from 

 all the dark basaltic rocks of lower elevations on the island. With some hesitation, 

 in my former publication I referred the rock to the basanite group, assuming 

 that a content of virtual nepheline might be concealed in the groundmass, which 

 was readily affected when treated with HCl. This circumstance Lacroix assumes 

 to be characteristic for the basanitoid lavas of the Pacific, where he says the 

 nepheline often "n'est pas exprimee; elle est restee a I'etat potential dans la verre. 

 La poudre des ces roches, traitee par HCl s'attaque plus au moins facilement" 

 (C. R. 169, 1919, p. 402). Bartii has, in his paper 'Pacificite, an anemousite 

 basalt', suggested that the virtual nepheline may be concealed in the plagioclase 

 in the form of carnegieite. For such basalts he proposes the name pacificite, or 

 if olivine is present, olivine pacificite and on account of my former description, 

 refers the lava in question from Masafuera to the pacificite group (21, p. 60; 

 22, p. 380, 510). 



In publishing the description of the rocks from the Juan Fernandez Islands 

 in my earlier publication the analyses were not recalculated according to modern 

 methods. This has been done now and the results recorded in the appendixed 

 table of analyses. With regard to the rock, then tentatively named basanilc, the 

 calculated norm shows no nepheline and an excess of 1.99 % of aluminium. Ihis, 

 together with a content of 3.36 % H2O, gives the impression tiiat the rock is de- 

 composed. On this account Lacroix, in reproducing my anal jos ■:: ilie Juan 

 Fernandez rocks, omitted this analysis, "car les resultats ne curresi- nclcr.t pas 



