Vol. 4] Murgoci— Classification of the Amphiboles. 381 



sonite, and other such amphiboles contain some Fe'" which re- 

 places Al. Perhaps the darker varieties correspond, as in the 

 glaucamphiboles, to those richer in Fe'". I have found karin- 

 thinedike amphiboles in many rocks, generally alkali rocks, other 

 than schists, an occurrence which is also put on record in the 

 literature of the subject. As I possess no analyses, the identifi- 

 cation with karinthine or soretite is doubtful. 



In fact, if we consider the analyses of pargasite, karinthine, 

 soretite, philipstadite, hudsonite, hastingsite, noralite, etc., we 

 notice that while alkali, CaO, and the sum of ALO ; , and Fe 2 3 

 remain almost constant, FeO and MgO vary in opposite direc- 

 tions ; karinthine has 7 FeO : 43 MgO, noralite 40 FeO : 5 MgO, 

 and the others have intermediate proportions. As far as the 

 literature is known to me, I do not know of a continuous varia- 

 tion in optical properties which could be explained by this vari- 

 ation of chemical composition, and I may state again that the 

 amount of iron as FeO (not Fe 2 3 ) has no influence upon the 

 physical properties of the amphiboles. This does not hold for 

 Fe 2 -0 :! . The sum of the coefficients of A1 2 3 and Fe 2 3 in the 

 above named series is almost constant, = 14 (the extremes being 

 12 and 18, karinthine and noralite being almost free from Fe'", 

 while the analyses of the other members give more or less Fe'" in 

 the proportion of % to % ) . I have tried to prove above that the 

 size of the angle of extinction in this series also is a function of 

 this proportion. 



Further, in the case of karinthine we have seen, and in the 

 case of barkevikite, philipstadite,* hudsonite, hastingsite, etc., 

 the literature supports the suggestion, that the angle of extinc- 

 tion is proportional to the intensity of color and absorption ; the 

 phenomenon is quite clear if these amphiboles show a zonary 

 structure, as frecpiently happens. 



Unfortunately we have not always good and complete deter- 

 minations of the optical properties corresponding to the analyzed 

 individuals, and accordingly we cannot, in general, verify the 

 dependence of the angle of extinction and the intensity of ab- 

 sorption on the proportion Fe 2 O g : A1 2 3 , but the few precise 



* E. Daly, On a New Variety of Hornblende, Proc. American Academy 

 of Arts and Science, XXXIV. 16, 1899. 



