378 



University of California Publications. 



[Geology 



believe that they must be very rich in Fe 2 3 . As a rule, all 

 analyses of crocidolite show a large amount of Fe 2 3 (16 to 20 

 per cent. ) ; unfortunately we lack optical determinations of these, 

 and no analysis has been made of Lacroix's positive crocidolite 

 (axial plane parallel to (010) ). Fortunately the question in the 

 riebeckite series is quite clear. There are many riebeckites 

 known; some poor in Fe 2 3 but very rich in FeO, the end mem- 

 ber being arf vedsonite ; some others very rich in Fe,0 3 but poor 

 in FeO, including in the latter osannite, the new amphibole of 

 Hlawatsch. (See the analyses in Table II.) 



The variation within wide limits of the optic constants of 

 those arfvedsonites which have been studied, as Avell as of the 

 riebeckites, is well known ; nevertheless we see a great difference 

 in their chemical constitution. We can, however, in general state 

 that the analyses of some arfvedsonites (of San Piedro III, Sao 

 Miguel XI, Hochenkrochen XIV), and of some riebeckites (Colo- 

 rado, Cape Ann) give a formula of the type of glaucophane 

 (erossite or rhodusite) where Mg is replaced almost entirely by 

 Fe (and Mn) and Ca by Na 2 . 



In the hope of bringing some order into this interesting group 

 of amphiboles, I propose the following classification : 



. ■ , Optic 



Fe'>0 ;! 3 : C A . xuu 2 V char- Formula 

 plane aeter 



Arfvedsonite very — 7U°— 80° 11(010) very ± From 5FeSiO;i, NaoSiO.?. (H 2 SiO;i) to 



poor large 4-5 FeSiOa 1-2 Na 2 SiOj ± FeAlSi:;0 9 



Riebeckite rich ca— 85° ||(010) small — From 3FeSiO:i, 2N»2SiOs. FeoSisOu to 

 (Colorado) 2FeSiOs. Na^SiO:;, 2Fe2Si»Oo ± 



Osannite* rich ca— 80 J-(OIO) very 



1 arge 



H 2 SiO:j 



id). Both the literature and my own observations show that 

 the dispersion of the optic axes and bisectrices stands in close 

 relation to the intensity of the blue color and to the intensity of 

 the absorption. As the latter are a function of FeoSigO,,, as has 

 been stated above, the dispersion must also be a function of Fe 2 

 Si 3 0.j. I believe that the above is sufficient to demonstrate the 

 influence of Fe 2 Si 3 O on the optic properties in general in the 

 AlFe amphiboles of metasilicate type. 



* I owe the information about the optical properties and the chemical 

 composition of osannite to the courtesy of Dr. C. Hlawatch. 



