H. S. Washington — Basalts of Kula. 119 



" brown mineral aggregate " with a surrounding augite-opacite 

 zone. 



Two questions present themselves for solution, and answers 

 to them are now attempted. What is this peculiar " brown 

 mineral?" What, is the cause of the alterations, and of the 

 difference in the two " stages ?" 



Brown mineral. — In regard to the first it may be premised 

 that the nature of the mineral is so far not certain. Nearly 

 all observers call it hornblende, with more or less confidence, 

 basing this conclusion chiefly on the usually observed slightly 

 oblique extinction. As has been seen, in the Kula basalts the 

 extinction was always parallel, and, however the discordant 

 results of other observers may* be explained (as they can be to 

 some extent), the writer gives it as his opinion that the mineral 

 in question is hypersthene, basing this conclusion on the fol- 

 lowing facts : 



a. The unanimous results of all experiments made show 

 that melted hornblende on cooling crystallizes invariably as 

 augite.* "It is hard to see how this tendency can be over- 

 come." 



b. The color and pleochroism of the brown crystals agrees 

 perfectly with that of hypersthene, calling the axis of greatest 

 length c\ as is also usually the case with hypersthene. 



c. The constant parallel extinction observed by me indicates 

 an orthorhotnbic mineral and points to hypersthene. 



d. On the theory that the crystals are hornblende the 

 arrangement at angles of 60° with one another is very difficult 

 of explanation. In hypersthene however we find that the 

 dome 101 is a twinning plane, giving an angle between the 

 two g axes of 60° 58', and such twinned needles of hypers- 

 thene have been observedf in andesite from Bukowina. The 

 case of similarly twinned rutile needles (sagenite) may be re- 

 called. 



e. Augite and hornblende tend to crystallize, when together, 

 in parallel position, and when we use the pleochroism as a 

 means of orientation we find that if the crystals are hypers- 

 thene their «(100) corresponds with that of the hornblende 

 when the c axes are parallel. 



f. The alteration of hornblende to hypersthene (along with 

 feldspar, augite and magnetite) has been several times ob- 

 served, as in andesites from Mexico, Arequipa in Peru, and 

 Cabo de Gata, the rhombic pyroxene in the latter case being 

 called bronzite. 



It must be added that the unsatisfactory and indetermina- 

 tive result of a chemical analysis of a dark spot containing 



*Cf Becker, Zeits. d. d. geol. Ges., xxxvii, 1885, p. 10. 

 f Becke, Min. pet. Mitth., vii, 9:i. 



