Contr. Mineral. and Petrol. 15, 67--92 (1967) Optical Properties and Cell Parameters in the Glaucophane-Riebeckite Series* I. Y. Bo~ Lawrence Radiation Laboratory, University of California, Livermore, California Received February 2, 1967 Abstract. A complete set of new optical and x-ray data is given for eleven analyzed alkali amphiboles [Na2(Mg , Fe")3(Al, Fe'")~SisO22(OH)2 ]. Nine new wet chemical analyses are report- ed. Using additional selected data from the literature, variation in refractive indices, extinc- tion angles (y--s), optic angles, density, lattice constants and cell volume are expressed graphically as a function of composition in the glaucophane-riebeckite and magnesiorie- beckite-ferroglaucophane series. Four orientations (G, C, O, and R) of the optical indicatrix within the structure are described and shown to be characteristic of the chemical species glaucophane (G), crossite (C), magnesioriebeckite (0), riebeckite (0), and riebeckite-arfved- smfite (R and 0). Optical properties of the pure end members by extrapolation are: fl y (r--~) c A n e Glaueophane 1.594 1.612 1.618 0.025 b = fl c Ay = 6~ 3.03 Riebeckite 1.702 1.712 1.719 0.015 b=~ cA~= 6~ 3.40 Magnesioriebeekite 1.655 1.671 1.67~ 0.02 b = y c A ~ = 32~ 3.15 X-ray parameters of the end members referred to the C 2/m space group are: a0(A) b0(A ) c0(A ) fl(o) v(A 3) Glaueophane 9.50 17.67 5.29 103.7 864 Riebeckite 9.78 18.08 5.34 103.5 918 Magnesioriebeckite 9.76 ]7.97 5.31 103.9 904 These show very good agreement with comparable measurements on synthetic counterparts. There is some indication that the two proposed synthetic polymorphs of glaucophane (E~sT, 1963) are both more disordered than the natural end member. Introduction The chemical and physical properties of the glaucophane-riebeckite-magnesio- riebeckite group of alkali amphiboles are imperfectly known, despite a large pertinent literature and numerous surveys beginning with Mu~GocI (1906). I~]~C]~TLu 1VIIYASHmO (1957) and D]~]~ et al. (1963) have written excellent summaries of the properties of the group from data in the literature. The reason for the imperfect knowledge lies in the incompleteness of most amphibole de- scriptions. Good chemical analyses, e.g., KU~ITZ' (1930) are available over most of the composition span, but optical data are sparse and in some instances un- reliable. Contributing factors to the paucity of optical data are the difficulties * Work performed under the auspices of the U.S. Atomic Energy Commission. 5*
Transcript
Contr. Mineral. and Petrol. 15, 67--92 (1967)
Optical Properties and Cell Parameters in the Glaucophane-Riebeckite Series* I. Y. B o ~
Lawrence Radiation Laboratory, University of California, Livermore, California
Received February 2, 1967
Abstract. A complete set of new optical and x-ray data is given for eleven analyzed alkali amphiboles [Na2(Mg , Fe")3(Al, Fe'")~SisO22(OH)2 ]. Nine new wet chemical analyses are report- ed. Using additional selected data from the literature, variation in refractive indices, extinc- tion angles (y--s), optic angles, density, lattice constants and cell volume are expressed graphically as a function of composition in the glaucophane-riebeckite and magnesiorie- beckite-ferroglaucophane series. Four orientations (G, C, O, and R) of the optical indicatrix within the structure are described and shown to be characteristic of the chemical species glaucophane (G), crossite (C), magnesioriebeckite (0), riebeckite (0), and riebeckite-arfved- smfite (R and 0). Optical properties of the pure end members by extrapolation are:
fl y ( r - -~ ) c A n e
Glaueophane 1.594 1 .612 1 .618 0.025 b = fl c A y = 6 ~ 3.03 Riebeckite 1.702 1 .712 1 .719 0.015 b = ~ c A ~ = 6 ~ 3.40 Magnesioriebeekite 1.655 1.671 1.67~ 0.02 b = y c A ~ = 32 ~ 3.15
X-ray parameters of the end members referred to the C 2/m space group are:
These show very good agreement with comparable measurements on synthetic counterparts. There is some indication that the two proposed synthetic polymorphs of glaucophane (E~sT, 1963) are both more disordered than the natural end member.
I n t r o d u c t i o n
The chemical and physical properties of the glaucophane-riebeckite-magnesio- riebeckite group of alkali amphiboles are imperfect ly known, despite a large pe r t inen t l i terature and numerous surveys beginning with Mu~GocI (1906). I~]~C]~TLu 1VIIYASHmO (1957) and D]~]~ et al. (1963) have wr i t ten excellent summaries of the properties of the group from data in the l i terature. The reason for the imperfect knowledge lies in the incompleteness of most amphibole de- scriptions. Good chemical analyses, e.g., KU~ITZ' (1930) are available over most of the composit ion span, bu t optical da ta are sparse and in some instances un- reliable. Cont r ibut ing factors to the pauc i ty of optical da ta are the difficulties
* Work performed under the auspices of the U.S. Atomic Energy Commission.
5*
68 I .Y. BorG:
in making measurements on highly opaque iron-rich minerals which character- istically show extreme dispersion of optic directions as well as optic axes. Fibrosity of some members (Mg-croeidolites and erocidolites proper) has also hindered investigation. There are, for example, forty or more chemical analyses of fibrous types in the literature, but fewer than six are accompanied by even a partial description of their optical properties. The large number of analyses of fibrous varieties is doubtless related to the relatively pure state in which they occur naturally, as well as to their commercial importance as asbestiform minerals. The present study was undertaken to establish the elusive relations between optical, chemical, and x-ray parameters. To this end the author has selected eleven representative samples, which were chosen so as to span the composition range of the glancophane-riebeckite series, and she has supplied a complete set of data for each. Included in the group are the type crossite, first described by PALAC~E (1894) from Berkeley, California; the type osannite from pegmatites from Alter Pedroso, Portugal, named by HLAWATSCH (1906) for its distinctive optic orientation; and the riebeckite from St. Peters Dome, Colorado, early descriptions of which [optic plane parallel to (010), LAc~oIx, 1889; Mu~GocI, 1906; JOHNSE~, 1910], strongly influenced optical criteria for recognition of the species. By current chemical and structural criteria, the latter two amphiboles are riebeckite-arfvedsonites. True riebeckites, meaning amphiboles near to the composition of the end member Na2Fe'3'Fe'~"SisO~2(OH)2 are rare; however, one was located and included in the study. Finally, the new data have been combined with selected values chosen from previous investigations to provide graphical descriptions of minerals representing solid solutions between either glaucophane and riebeckite or magnesioriebeckite and ferroglaucophane. The optical indicatrix can have one of four orientations with respect to the crystallography of these alkali amphiboles. Very small variations in refractive indices within grains of a single thin section can produce a dramatic variation in optical properties (orientation). Since the different orientations have been a source of confusion, an indication of which compositions they are associated with is given, as well as an explanation of how they come about. Although some new optical data are presented here for the fibrous varieties, they have not been the subject of systematic study. No at tempt has been made to explain their fibrous habit except to note that both Mg-crocidolites and croeidolites proper are included within very limited ranges of composition. The clearly merit more study, as do the members of the magnesioriebeckite-crossite-ferroglaucophane series, whose properties are only generally outlined here.
Nomenclature
Substitutions into the general amphibole formula AX2Y5ZsO22(OH)2 are the basis of the nomenclature proposed by MIYASHIaO (1957) and adopted here with minor modifications. The formula units (X, Y and Z) for one-half the unit cell content and corresponding cation sites (A, M1, M2, M s, M 4 and Si) (WAR~]~, 1930; PmLLn)S, 1963; GHOSE, 1965): X----Na and Ca in M 4 cation site (6--8 fold coordination) with K and excess Na when (Na~-Ca)~ 2.00 accommodated at A (10 fold coordination).
Optical Properties and Cell Parameters in the Glaucophane-Riebeckite Series 69
Y = R ' " § where I ~ " ' = A 1 w, Fe ' " , Ti in the M S site and R " = M g , Fe" , Li, Zn and Cu in the M 1 and M 3 sites (all six-fold coordination). Z = Si and te t rahedral ly coordinated A1 Iv. ( o H ) = O H § 2 4 7 e l .
FERROGLAUCOPHAN E RIEBECKITE
NQ 2 Fe3"Ar2 Si8022 (OH)2 Na2 Fe 5"Fe 2"'Si8 ~)2z (OH) z
I 0 , ~ . - - . . - |
, o ~
�9 o o!
.8
Fe"+ Mn / e a P CROSSITE " 4
Fe" +Mn 4-Mg ~ �9 i. / 6 "
.4 *" o8 g 5 . " " . . I
'L 4 e ' ' / / ~ _ ' - . ~ . ' �9 �9 o ~ �9
.2 .4 .6 .8 1.0
No 2 Mg3 AI2Si8022 (OH)2 Fe'" + Ti No 2 Mg 3 Fe 2'" Si8022 (OH)2 GLAUCOPHANE Fe"' + Ti +AI 3 ~ ' - MAGNESIORIEBECKITE
Fig. 1. Chemical variation in the glaucophane-riebeckite-magnesioriebeckite-ferroglaucophane group. Formulas of 109 amphiboles are calculated and plotted from chemical analyses on the basis of 24 (O, OH, F, CI) per half-unit cell. Open circles numbered 1--11 are amphiboles for which new chemical, optical and/or x-ray data are presented; filled circles lettered a- -o are from the literature and are identified in the captions to Figs. 3, 4, 5 and 6. Crosses indicate fibrous varieties
The end members in the series (Fig. 1) are represented by the following formulae :
The formula for the related arfvedsonite end member is (Na~.~Ca0.5)3.o(Fe",Mg , Fe'",A1)5(Si, A1)sOe2(OH, F)2 after DEER et al. (1963), where R " --~3.5 and 1~'" _~ 1.5. t~iebeckite-arfvedsonites, (Na, Ca)2.0-3.oFe~:5-3.0Fe~:~-2.0Sis022(O H , F)2, are far more common among natura l amphiboles t han either pure end member. Riebeckite- arfvedsonite typical ly contains appreciable amounts of 1~ and unusual elements such as Zn, Cu and Pb, which suggests t ha t they behave as "scavengers" during crystall ization [see analyses No. 9, riebeckite and No. 10, No. 11, riebeckite- arfvedsonite, Table 1 this work; Bo~L~Y (1963)].
Tab
le 1
. Che
mic
al a
naly
ses
o/gl
auco
pahn
e-ri
ebec
kite
s
1 2
3 4
5 6
SiO
2 57
.73
57.4
8 57
.93
56.7
2 56
.38
50.4
1 T
i02
n.d.
0.
15
0.26
0.
08
0.11
1.
66
Alc
Oa
12.0
4 12
.39
11.9
2 9.
52
8.45
7.
82
Fe20
a 1.
16
2.25
1.
31
4.17
4.
98
8.73
Fe
O
5.41
4.
91
10.7
8 8.
61
9.40
10
.81
MnO
n.
d.
0.02
0.
11
0.19
0.
19
0.14
M
gO
13.0
2 12
.95
8.05
10
.56
9.89
7.
39
Li2
0 n.
d.
n.d.
tr
ace
n.d.
tr
ace
n.d.
C
aO
1.04
0.
43
0.29
1.
16
1.29
3.
99
1Wa2
0 6.
98
6.66
6.
70
6.55
6.
77
7.04
K
20
0.68
0.
05
0.11
0.
06
0.08
0.
57
I-I~O
+ 2.
27
2.17
2.
24
2.19
1.
86
1.17
H
20-
--
0.00
0.
00
0.03
0.
04
0.10
F
n.d.
0.
02
0.05
0.
03
0.01
n.
d.
C1
n.d.
0.
04
n.d.
0.
02
0.01
n.
d.
Les
s 10
0.33
99
.52
99.7
5 99
.89
99.4
6 99
.83
0 =
F an
d C1
0.
02
0.02
0.
02
0.01
99
.50
99.7
3 99
.87
99.4
5
Ana
lyst
W
. Ku~
Tz
D. T
ItAEM
LITZ
C
.O.
ING
AM
ELLS
C
.O.
ING
AN
IELL
S C
.O.
ING
AM
ELLS
•.
HO
LGA
TE
7.78
9 7.
780
Si
0.21
1/8.
000
0.22
0t 8
.000
7.
975~
7.
861~
7.
937
7.44
3 A
IIV
~
. 0.
139
0.02
5J 8
.000
8.
000
0.06
3}8.
000
0.55
7} 8
.000
Alv
i --}
1"7~
s2
3 1.
757/
1.
909/
1.
416/
1.
340/
0.
805/
T
i 1
/ M
nFe"F
e'" --
}0
"610/3
"2270
"118/"
J | 5.
050
0.5561
0"2291
0"015}
2"001 |[] 5.
1710
.1361
0"02
7}2"
0721
.241/]
4.97
8 0.9
97/0"
4351
0"008
}1"85
9 |~]
5.05
8 0.5
28/0"
012}
1"880
1.107
/ [[I
5.08
4 0.9
69j0"
184}
1"958
1.335
/ |~]
4.93
8
0.00
2}3.
170
J 0.
013}
2.90
6 J
0.02
2}3.
199
J 0.
023}
3.20
4 )
0.01
8}2.
979
J M
g 2.
617J
2.
612J
1.
652J
2.
180J
0.
075J
1.
626J
L
i --
--
0.
000
--
0.00
0]
--
Ca
0.15
0)
0.06
2)
0.04
3)
0.17
2 /
0.19
5~
0.63
1 /
Na
1.82
5} 2
.092
1.
748}
1.8
19
1.78
8~ 1
.851
1.
759~
1.9
42
1.84
8[ 2
.057
2.
016~
2.7
54
K
0.11
7/
0.00
9J
0.02
01
o.01
0l
0.01
4/
0.10
7/
0It
2.04
2 2.
042
1.95
9]
2.05
6[
2.02
1 /
1.74
6 /
1.15
3 1.
153
F --
0.
009}
1.9
77
0.02
2/2.
078
0.01
3} 2
.039
0.
004}
1.7
53
C1
--
0.00
9J
o.00
5J
0.00
3J
0 21
.958
22
.023
21
.922
21
.961
22
.247
22
.847
O
7 8
9 10
11
SiO
~ 55
.38
55.1
0 51
.17
49.8
7 48
.30
TiO
~ 0.
36
0.68
0.
63
0.34
0.
38
A12
0 a
5.29
4.
27
1.11
1.
04
2.66
F
%0
a
9.74
10
.61
15.1
8 14
.25
13.4
1 F
eO
13.0
7 9.
78
18.4
8 20
.19
20.4
0 M
nO
0.18
0.
48
2.85
1.
22
1.33
M
gO
6.31
8.
86
0.32
0.
03
1.17
L
i20
n.d.
n.
d.
0.14
0.
44
0.14
Z
n0
n.d.
tr
ace
0.29
0.
67
0.25
C
uO
n.d.
n.
d.
0.13
0.
01
0.02
C
aO
1.10
1.
43
0.80
0.
44
1.40
N
a20
6.40
6.
38
6.28
7.
75
6.91
K
20
0.05
0.
09
0.72
1.
15
1.41
H
~O +
1.
99
1.96
1.
57
0.84
1.
24
tt~
O-
0.09
0.
06
0.04
0.
16
0.03
F
0.02
0.
02
0.60
2.
46
1.12
C
I 0.
01
0.03
0.
04
trac
e n.
d.
Les
s 99
.99
99.7
5 10
0.35
10
0.86
10
0.17
O
:Fan
dC
1
O.0
1 0.
02
0.26
1.
03
0.47
99.9
8 99
.73
100.
09
99.8
3 99
.70
Ana
lyst
C
.O.
ING
AM
ELLS
D
. TH
AEM
T~IT
Z C
.O.
ING
AM
ELLS
C
.O.
ING
AM
ELLS
E
.I-I
. O
SLV
ND
�9
.r r
Si
7.99
7~
7.92
5~
7.92
6 A
1 Iv
0.00
3S 8
.000
0.
075J
8.0
00
0.07
4
AlV
~ 0.
897]
0.
648)
0.
129
Ti
0.03
9~ 1
.994
) 0.
073}
1.8
68~
0.07
4 ~e
'"
1.o5
8J
[4 9
51
1.14
7;
[ 1.
768
Fe
1.57
8)
| "
1.17
6)
| 5.
001
2.39
3 M
n 0.
022}
2.95
7)
0.05
9} 3
.133
) 0.
374
Mg
1.35
7;
1.89
8J
0.07
4 L
i --
--
0.
087
Zn
--
--
0.03
3 C
u --
--
0.
015
Ca
0.17
0]
0.22
0)
0.13
3 N
a 1.
790}
1.9
70
1.77
9} 2
.015
1.
887
K
0.01
0J
o.o1
6J
0.14
1 O
H
1.91
7)
1.88
0)
1.62
2 F
0.00
9} 1
.929
0.
009}
1.8
97
0.29
4 C
1 o.
o03J
o.o
osJ
O
.OlO
0
22
.071
22
.103
22
.074
8.00
0
1"97
1 /
2.9
76
J
2.16
1
1.92
6
4.94
7
7.82
61
7.60
5~
0.17
4i 8
.000
0.
395S
8.0
00
0.01
8]
0.09
8 /
0.04
0} 1
.741
~ 0.
045}
1.7
32~
1.68
3j
[ 1.
589;
|
2.65
01
~4.
917
2.68
6]
}4.9
89
0.16
2l
| 0.
177/
1
0.00
7[ 3
176
J 0.
274(
~ ~
] 0.
2781
"
0.08
9 / .
....
0.
078[
0.
029|
0.
001;
0.
002;
0.
074 /
0.
236 /
2.
358}
2.6
62
2.10
9}2.
629
0.23
0;
0.28
4j
0.87
9]
1.30
1)
1.22
1} 2
.100
0.
558}
1.8
59
o.oo
o;
o.oo
oj
21.9
00
22.1
41
8 N"
Exp
lana
tion
of
Tab
le 1
see
p.
72.
72 I .Y. BORG:
Explanation of Table 1. See Table 2 for description of samples. All analyses new except 1 and 6. 1 ~ Glancophane (KvNITZ, 1930). 2 ~ Glaucophane. 3 ~ Glaucophane. 4 = Glaucophane. 5 = Glaucophane. 6 ~ Crossite (HoLGATE, 1951). 7 = Crossite. 8 ~ Crossite; Cr20 a and V203 absent. 9=Riebeckite; Cr203, trace, PbO, V20 ~ absent; A1203 has been reduced by the wt % ZnO/2 on assumption that part of ZnO was included in gravimetrically determined Al~O 3, 10 = Riebeckite-arfvedsonite; Zr ~ 450 ppm (ZrO~ = 0.05); Ni,V, Cr and Pb ~ 10 ppm; reported A1,Oa treated as in 9. 11 ~ Riebeckite-arfvedsonite; Cr~O trace; PbO, V208 absent reported A120 a treated as in 9. I-I,O(~-) may be low (S. GoLmc~, personal communication). - - Key: n. d. ~ not determined; trace ~ less than 0.02 %.
W i d e l y differing opt ica l o r ien ta t ions and c rys ta l hab i t s have encouraged a p le thora of minera l names in th is group of a lka l i amphiboles . I n t e r m e d i a t e members of the series, whose names have ga ined currency, are l i s ted below toge ther wi th the usage in the presen t t ex t . H i s t o r y of these names and o thers t h a t have been sugges ted is given b y HINTZ~ (1897), HINrZ~ (1938) ; and MxYAS~mO (1957).
Crossite - - Na~(Mg, Fe")a(A1, Fe '")2SisO22(Oi)2 - - initially intended to set apart those members of the glaucophane-riebeckite series having crossed axial planes, viz. optic planes normal to c, (7 = b). This restriction is no longer useful, but the many commonly oceuring intermediate members are conveniently described by the name.
Mg-crocidoli te - - ( including rhodusi te) - - Na2(Mg 2.sFe:~)3.2 Fe~:~ (Si, A1)sO22(OH)2
- - fibrous variety whose composition is closely limited near the magnesioriebeckite end member (• of Fig. 1).
Crocidoli te - - Na2(Fe'2:6Mg.4)3.0Fe'2:0SisO~(OH)~ - - fibrous variety of riebeckite which closely approaches theoretical composition when fresh (tto])Gso~, 1965; HODGSON et al., 1965). Most published analyses when calculated ( • of Fig. 1)
t v v H �9 give a composition near Na2.0(Fe2.0Mg.7)2.TFe2.sSisO22(OH)2 presumably due to oxidation of Fe". (Si+A1 Iv) in tetrahedral coordination is characteristically low in these, i.e., ~ 8.00.
Osanni te - - N a 2.o_3 .0Fe~5-s.0Fe~5'-~.o Sis022( OH , F)2
- - a riebeckite-arfvedsonite named by HLAWATSCH (1906) who demonstrated that its optical orientation [optic plane ~_ to (010), b =7 , (--) sign of elongation] was distinct from what was then called riebeckite. The latter are now also considered to be riebeckite-arfvedsonites (e.g., No. 10, Table 2). The name osannite is used here to designate an orientation of the indicatrix rather thancomposition.
A to t a l of 109 analyses of a lkal i amphiboles are p lo t t ed in Fig. 1. The grouping of F e " ~ - M n and F e " ' - ~ T i on the o rd ina te and abscissa, respect ively , is in considera t ion of the i r s imilar ionic rad i i and comparab le influence on opt ica l pro- pert ies. Amphiboles for which new d a t a are presented (Tables 1, 2 and 3) are ind ica ted b y n u m b e r e d open circles; f ibrous amphiboles are m a r k e d b y crosses. Amphibo les whose publ i shed descr ipt ions are included in plots of va r i a t ion of phys ica l p roper t ies wi th composi t ion (Figs. 3, 4, 5 and 6) are ind ica ted b y le t t e red dots ; the l i t e ra ture reference for these analyses are given in the tex t . Whi le Fig. 1 expresses the gross va r ia t ion found in na tu r a l amphiboles , the me thod of normal- izing all va r i a t ion to 1.00 obscures large depar tu res f rom the usual R'"----3.0 and R'"----2.0 tota ls . Many of the so-ca l ledr iebecki tes p lo t t e d are in fac t r iebecki te- ar fvedsoni tes since X is > 2 . 0 0 and R ' " ~ 2 . 0 0 . Many also conta in large quant i t ies of F~-C1. No. 9 (Fig. 1 and Table 1) closely approaches the theore t ica l compo- s i t ion of r iebeeki te . Pu re g laucophane and fer roglaucophane have no na tu r a l representa t ives .
Optical Properties and Cell Parameters in the Glaucophane-l~iebeckite Series 73
T h e gene ra l d i s t r i b u t i o n of p o i n t s a long d i agona l s a t t e s t s t o t h e c o m m o n sol id
so lu t ion b e t w e e n t w o e n d m e m b e r s . T h e s u b s t i t u t i o n s a re Fe'a'Fe'2" for MgaA1 ~
a n d a m o r e l i m i t e d s u b s t i t u t i o n of Fe"A12 for Mg3Fe'~". I t is poss ible t h a t m a n y
of t h e compos i t i ons d i s p l a y e d in F ig . 1 wh ich do n o t fal l e x a c t l y on t h e d i agona l s
w o u l d do so if t h e ana lyses were a c c u r a t e or if t h e samples were free of al l impur i t i e s .
Experimental Method
Separation and Chemical Analysis Separation of the amphiboles was obtained by repeatedly centrifuging samples finer than 100 mesh in heavy liquids (methylene iodide, ~ = 3.32 and tetrabromoethane, ~ = 2.96) and by using the Frantz isodynamic separator on material previously washed with alcohol to remove ultra-fine particles. Separation of severM ultimately had ~o be abandoned because they could not be adequately purified. Single erystMs of riebeckite-arfvedsonite (No. 10 and No. 11) were crushed and put through the Frantz separator to remove adhering oxides and inclusions. The analyses recorded in Table 2 have been corrected for the following contaminants: No. 2, 1% chlorotoid; No. 3, 0.14% C, gravimetrieally determined; No. 4, 0.45 % CO~, gravimetri- tally determined, and 0.72% Ca present as calcite; No. 8, I% albite. These and remaining samples contain impurities less than 0.5 weight per cent (see Table 2). The analyses were done at the Rock Analysis Laboratory, Universi W of Milmesota, by routine gravimetric methods. Spectra of iron-rich members of the series were examined by x-ray fluoresence techniques to detect presence of elements, such as Zn, Ca, Cr, Pb, etc., which are common ill these amphiboles. If zinc and/or copper were present, the amounts were determined colormetrieMly at the University of California, Lawrence Radiation Labora- tory, by methods outlined by t~ADE~ and GtCI~AI~DI (I961). On the assumption that some part of the ZnO was erroneously reported as AI~O 3 in the gravimetrie analysis, the lat~er was reduced by one-half the weight per cent of ZnO (FuR~AI% 1939; p. 9; BOI~LEY, 1963).
Density All densities were measured by using a flotation technique. The powdered sample was centri- fuged in Clerici solutions whose density could be varied by dilution with water until all or most of the powder remained in suspension after the method of BASS (1957) and HEss (1960, p. 19). The density of the solution was measured on a Christian Backer liquid-density balance. The method is inherently very accurate (-t- 0.003 g/cc). I t can be very rapid if a calibration of the A ~ associated with the addition of a drop of water (or Clerici) is made on the centrifuge liquid once the approximate ~ has been reached. The method has the added advantage of allowing an estimate to be made of the variation in ~ within a powder because ~ like other properties of the silicates shows variation within the same specimen (I-IAag~ et al., 1965; WILson, 1960). The data reported here are mean densities. The variation is judged by the densities of the solutions in which 95 % and 5 % of the sample is suspended after centrifuging.
Re/ractive Indices The Becke line method was used to determine the refractive indices of the minerals. The approximate range of the 1%. I. were first determined in white light on grains giving centered or off-centered interference figures by taking advantage of the dispersion of the oils. The final values were determined after appropriately orienting the grains with a universal stage (hemispheres n ~ 1.649) in Nap-light (4 = 590 raft ). The R. I . of the oils were measured with an Abb6 refractometer in Na light under carefully controlled temperature conditions. Before and after each determination of a R.I . of a mineral, the ambient temperature above the stage was measured and the appropriate corrections made to the R.I . of the oils under the tacit assumption that the sample was in thermM equilibrium with the surrounding air. The largest source of error is the uncertainty in the temperature. In the course of a 4-hour deter-
Tab
le 2
. Des
crip
tion
o/sa
mpl
es
g~
Am
- O
rigi
nal
Typ
e ph
i-
No.
(g
iven
bo
le
nam
e)
No.
Loc
Me
Roc
k ty
pe
Impu
riti
es
in a
naly
zed
sam
ple
Sour
ce
Per
tine
nt l
iter
atur
e re
fere
nces
1 A
.M.N
.tt.
Gla
uco-
10
120
phan
e
2 A
.M.N
.H,
Gla
uco-
10
121
phan
e (g
astM
dite
)
3 33
1-M
-121
G
lauc
o-
phan
e
4 SE
-1
Gla
uco-
ph
ane
5 33
1-M
-19B
G
lauc
o-
phan
e
6 47
972
Cro
s~te
7 33
1-M
-56
Cro
ssite
Zer
mat
t, S
witz
.
Pie
dmon
t P
rov.
, It
aly
(Pro
babl
y St
. Mar
cel
or
Val
de
Chi
sone
)
Junc
tion
Sch
ool
Are
a, H
eald
sbur
g Q
uad.
, C
alif.
Seba
stop
ol Q
uad.
, C
alif
.
Junc
tion
Sch
ool
Are
a, H
eald
sbur
g Q
uad.
, C
alif
.
Mon
Hm
ent ~
Iill,
A
ngle
sy, W
ales
Junc
tion
Sch
ool
Are
a, H
eald
sbur
g Q
ua&
, C
alif
.
Gar
net-
chlo
rite
.mus
covi
te
glau
coph
ane
schi
st
Chl
orot
oid-
chlo
rite
-gla
ueo-
ph
ane
rock
(w
ith
min
or
quar
tz)
Qua
rtz-
law
soni
te g
lauc
o-
phan
e sc
hist
wit
h m
inor
ch
lori
te
Gla
ueop
hane
in
alte
red
eclo
gite
Gla
ucop
hane
-chl
orit
e ro
ck
wit
h re
mna
nt g
arne
t,
pyroxene and futile a
nd
minor sphene and
apat
ite
Cro
ssib
e-ep
idot
e qu
artz
- al
bite
sch
ist
wit
h m
inor
m
usco
vite
, ch
lori
te s
phen
e,
hem
atit
e an
d ap
atit
e
Qua
rtz-
eros
site
sch
ist
wit
h m
inor
pyr
hoti
te a
nd h
ema-
ti
te (
?)
Not
rep
orte
d
Non
e
Qua
rtz
incl
u-
sion
s (<
0.5%
)
Chl
orit
e (<
0.5%
)
Chl
orit
e (<
0.5%
)
Epi
dote
(<
1%)
(see
foo
tnot
e p.
84)
Qua
rtz
(<0.
5%)
MA
SON
and
SE
A-
BO
D~W
IO (1
876)
; ~A
~, A
mer
ican
K
UN
ITZ
(193
0)
Mus
eum
Nat
'l H
isto
ry
MA
SON
and S
~.A
- MAN, A
mer
ican
M
useu
m N
at'l
His
tory
U.C
., B
erke
ley
ZAM
~O~Z
NZ (
1906
) ;
II)I
)IN
OS
0911
, p.
468
)
BO
RG
(195
6) i
nclu
ding
ro
ck a
naly
sis
(Tab
le 1
f)
U.C
., :B
erke
ley
TI~
AvI
s (195
2)
U.C
., B
erke
ley
BoR
e (1
956)
S. A
O~L
L,
PIO
LOA
T~ (1
951)
C
ambr
idge
U.C
., B
erke
ley
Bo~
o (1
956)
Optical Properties and Cell Parameters in the Olaucophane-Riebeckite Series 75
0
o.
O9
�9
O9 , ~ ~2 ~ O 9
0 ~-~
O9 ,_,-~
O9
. o9 "~
2
o 0
~ O9
~ . ~ ~ 0 ~ O9
0 0
o
�9 ~ o ~ ~
-~ 0 ,..~ 2,~ .~ NN
o o9 O9
O
O9
5
o9
5~
~ o9
O 9 0
minat ion the ambient tempera ture may change by 3 ~ C, which is equivalent to 0.0018 in some of the high index oils. However, the R. I . are believed accurate to • 0.001. The optical properties of iron-rich members of the series are part icularly troublesome to measure. I n addit ion to the well known in- complete extinction on (010) in monchromatic l ight (V~DL, 1924; ESKOLA and SA~msT~I~, 1930; YAOI, 1953; SHODA, 1954), two Becke lines move in opposite directions when the R. I . of the immersion oil is near to t h a t of the t ransmi t ted ray. An additional compli- cation is extreme opacity of these materials, part icularly to Na-light.
Extinction Angle The ext inct ion angle was determined graphi- cally by measuring the angle between the appropriate optic direction (~, fl or ~) and the c axis on a stereogram. The c axis was located by plott ing the positions of the poles of one or more (110) cleavages and pole of ( 0 1 0 ) = b = f l or ~ as measured with a uni- versal stage. The cleavage angle, ca. 55.5 ~ (see Table 3) and the angle between a, fi and ~, were also graphically determined and used as a check on the accuracy of location. The extinction angle is accurate to 1~ ~ . I t was possible to determine the sign of the extinction angle for three single crystals which were photographed with an x-ray precession camera. SCHUSTER'S rule was followed (WIN- C~ELL and WINCHELL, 1951, p. XV), viz, the sign is considered positive if the ext inct ion direction lies in the acute angle between a and c and negative if i t is the obtuse angle. The signs given in Table 3 are for optical ex- t inctions in a s tructure described as C 2/m. The suggestion t h a t extinction angles on (110) cleavage flakes be used for diagnostic purposes has been recently made by PARKER (1961) and GAzzI (1963). On this account these data (c ^ Y~10, etc.) are presented, together with comparable values calculated from extinct ion angles ( chZ , etc.) and 2V from equations modified from GAzzI (1963). In practice, accu-
! ! ra te measured values of c ^ ~(110) or c A a ( 1 1 0 )
are as hard to obtain as values of c ^ ~ and c A a. Gross errors result if (110) is not strictly parallel to the slide. Yet despite the large scattering of the data, good to very good agreement was had between the average mea- sured and calculated values. This suggests t h a t while the perfect (110) cleavage flake is not common, data measured on flakes t ha t depart
Tab
le 3
. Ph
ysic
al p
rope
rtie
s o/g
lauc
opha
ne-r
iebe
ckite
s
1 2
3 4
5 6
Ple
ochr
oism
c~
co
lorl
ess
colo
rles
s co
lorl
ess
nea
rly
col
orle
ss
nea
rly
col
orle
ss
colo
rles
s fi
v.
lt.
pur
ple
v. l
t. p
urp
le
lt.
pu
rple
lt
. p
urp
le
viol
et
pal
e vi
olet
y
v. l
t. b
lue
v. I
t. b
lue
lt.
blue
lt
. bl
ue
blue
bl
ue
b=fl
b=
fi
b=fl
~
b=fl
O
rien
tati
on
G {T
.O.P.
II(Ol
O)
G {T
.O.p.
H(OI
O ) G
{T.O
.P.II(
010)
b fl
G _
../T
.O.P
.II (
010)
G
G
_ fT
.O.p
.H(0
10 )
b =
fi
T.O.
P.[I(0
10)
Ab
sorp
tio
n
y =
fl>
c~
y
=fl
>
c~
fi =
y>
a
fi =
Y>
~
Y =
fl>
~
Y =
fl>
c~
Ref
ract
ive
~ 1.
611
1.61
0 1.
613
1.62
6 1.
632
1.64
6 in
dice
s fi
1.
630
1.63
0 1.
638
1.64
4 1.
6455
1.
658
(#t=
590
mix
) y
1.63
3 1.
632
1.64
2 1.
646
1.64
6 1.
658
Bir
efri
ngen
ce y
--~
0.
022
0.02
2 0.
029
0.02
0 0.
014
0.01
2
Dis
pers
ion
fin
elin
ed
]in
clin
ed
Jin
clin
ed
Jinc
line
d ]i
ncl
ined
Ji
ncli
ned
[r <
v
wea
k
[r <
v
wea
k
[r <
v
wea
k
[r <
v
wea
k
[r <
v
wea
k
[r <
v
wea
k
Ex
tin
ctio
n
c A
~
6 ~
6.25
~
4 ~
6 ~
7 ~
7.5
~ an
gle
c A
r~1
0(c
A y
~10)
cale
7.
5 ~
(8.5
~
8 ~
(9 ~
7
~ (7
~
10 ~
(10
~
13 ~
(13
~
11 ~
(15
~
(~t=
590
mix
)
Sig
n of
ex
tin
ctio
n a
ngle
(@
)
Opt
ic a
ngle
2
V
--43
.6 ~
--
41.8
~
--4
8.9
~
--38
.1 ~
--
21
.5 ~
--
5 ~
(2
=5
90
m~
) 2V
(cal
c.)
--4
2.4
~
--3
6.5
~
--43
.1 ~
--
35
.0 ~
--
21.3
~
0 ~
Den
sity
~
(med
ian)
3.
092
3.08
0 3.
129
3.13
9 3.
149
3.20
6*
(obs
. va
riat
ion)
=J
=0.0
30
~_0.
009
=L0.
016
-4-0
.011
=
~0.0
05
--
0 (c
ale)
3.
088
3.07
7 3.
138
3.13
2 3.
161
--*
*
Cle
avag
e (1
10)
A (
li0)
(cal
c.)
55 ~
13'
55
~ 1
3'
55 ~
10'
55
~ 2
6'
55 ~
20'
55
~ 1
0'
angl
e
Un
it c
ell
a o (
A)
9.55
0 ~
0.00
1 9.
547
• 0.
001
9.53
5 •
0.00
1 9.
595
:k 0
.002
9.
609
:~ 0
.002
9.
614
:E 0
.002
C
2/m
b o
(A
) 17
.842
:k
0.00
3 17
.738
•
0.00
2 17
.745
-t-
0.0
02
17.7
98 •
0.
003
17.8
13 :
k 0.
004
17.8
82 :
J: 0
.003
Z
= 2
c o
(A
) 5.
296
• 0.
001
5.29
8 :E
0.0
01
5.30
6 ~
0.00
1 5.
307
• 0.
001
5.31
1 •
0.00
2 5.
312
-k 0
.001
fl
(~
10
3.70
-4
-0.0
1 10
3.68
•
103.
54
• 10
3.66
•
103.
61
~0
.02
10
3.71
:t
:0.0
1 a o
sin
fl
(A_)
9.
278
9.27
6 9.
270
9.32
4 9.
339
9.33
9 V
(A
a)
871.
9 87
1.7
872.
7 88
0.6
883.
5 88
7.2
O
* H
OL
GA
TE
(19
51).
**
Acc
ura
te c
alcu
lati
on n
ot
poss
ible
.
�9
7 8
9 10
11
Ple
ochr
oism
c~
#
Ori
enta
tion
Abs
orpt
ion
l~ef
ract
ive
indi
ces
fl
(~=
59
0m
b)
y
Bir
efri
ngen
ce 7
--cr
Dis
pers
ion
pale
yel
low
bl
ue
purp
le
b=
7
C {
T.O
.P.
i(0
10
)
fl=
y>~
1,
659
1.67
0 1.
674
0.01
5
hori
zont
al
r <
v m
od
erat
e
Ex
tin
ctio
n
c A
y,
etc.
c
A f
l 2
~ an
gle
c A
7~1
o(C
h ri
~0)(
calc
), 3~
(1~
(2
= 5
90m
~)
etc.
Sig
n of
ext
inct
ion
angl
e O
ptic
ang
le
2V
-59
.8 ~
(A=
590
mM
J 2V
(cal
c)
--61
.5 ~
Den
sity
@
(med
ian)
3.
211
@ (o
bs.
vari
atio
n)
~:0
.020
@
(cal
c)
3.22
3
Cle
avag
e (1
10 A
(11
0) (
tale
) 55
~ 1
7'
angl
e
pale
yel
low
de
ep b
lue
dk.
sky
blue
de
ep b
lue
yell
ow
It,
grey
blu
e de
ep p
urpl
e d
ark
gre
y bl
ue
It.
gree
nish
yel
low
b =
~
b =
#
c b
o {T.
o.,.a(
OlO)
{T
.o.e.l
l(01o
) fl
=~
>~
~
>7
>#
~
>#
>;"
1.66
3 1.
693
1.68
6 1.
666
1.70
1 1.
694
1.66
9 1.
709
1,70
05
0.00
6 0.
016
0.01
45
cros
sed
or
~cr
osse
d fi
neli
ned
hori
zont
al
(r <
v
stro
ng
[r X
v e
xtre
me
r ~
v st
rong
cArl
11
~ cA
~.
4 ~
cAa
6 ~
10 ~
(11 ~
) 4
~ (2
~ )
7 ~
(5 ~
)
(+)
(-)
--91
~
f--9
4 ~
at
591
m~x
--
87 ~
[-
-80
~ a
t 48
7 m
~z
--9
0 ~
90
~ --
84 ~
3.18
9 3.
396
3.41
4 :t
:0.0
22
• s
3.19
6 3.
380
3.40
6
55 ~
20'
55
~ 3
0'
55 ~
48'
deep
blu
e ye
llow
dk
. gr
ey b
lue
b=
y 0
{T.O
.P.
5_(0
10)
1.69
3 1.
700
1.70
9
0.01
5
cros
sed
extr
eme
r<v
cAa
11 ~
2
o (2
o)
--10
2 at
591
m b
-
-
93 ~
at
48
7m
~
_9
8 ~
3.40
6 :~
o.o2
o 3.
405
55 ~
46'
Un
it c
ell
a o
(A)
9.64
7 =L
0.0
01
9.67
3 •
0.00
2 9.
769
• 0.
002
9.82
3 •
0.00
2 9.
846
~ 0.
001
C 2
/m
b o (
A)
17.9
05 =
]= 0.
003
17.9
23 •
0.
003
18.0
48 •
0.
003
18.0
21 i
0.
004
18.0
68 •
0.
002
Z=
2
co (
A)
5.31
6 •
0.00
1 5.
316
• 0.
001
5.33
5 ~
0.00
1 5.
328
• 0.
002
5.33
4 =E
0.0
01
fl
(~
103.
60
:kO
.O1
103.
68
4-0.
01
103.
59
~0
.01
10
3.70
i
O.O
1 10
3.81
:~
0.01
ao
sin
fi (
A)
9.37
7 9.
398
9.49
5 9.
544
9.56
2 V
(A
8)
892.
5 89
4.4
914,
3 91
6.4
921.
4
~a 9 9~
B 8 m"
~a g
78 I. u BoRa:
from this position form an approximately gaussian distribution. Additional complications in measuring extinction occur in amphiboles with fibrous habit, small grain size and/or low birefringence; a universal stage must be used to orient the grains so as to obtain acceptable results.
Optic Angle Standard orthoseopic procedures were used with sodium light to measure 2V (or actually 2 H before appropriate corrections are made for the difference between R.I . of the mineral and that of the hemispheres). Thin sections were mounted on a standard Leitz-4-axis universal stage with hemispheres of n = 1.649 and a central plate of n : 1.520. Choice of objective (U.M.2) and a fully opened substage diaphragm minimized the potentially large errors in the method described by MvNno (1963 and 1964). I t is likely that optic angles smaller than 65 ~ are accurate to within a degree while angles between 80~ ~ which are readily deter- mined to within 2 ~ under good circumstances, are q- 30--5 ~ These large angles are associated with iron-rich members in which optical anomalies and strong absorption are the rule. The optic angle given for each amphibole analyzed is the average of the results from six or more grains, each measured four times. The angle varies from 3o--8 ~ (average 4 ~ ) in any sample. Especially thin sections were cut normal to (001) and hence subnormal to the acute or obtuse bisectrix in the single crystals of riebeckite-arfvedsonite. In these the optic angle was deter- mined by using the method of KAMB (1958) for slightly off-centered figures viewed with an objective of N.A. 0.85. The precision of these determinations is not high because of the diffuse nature of the isogyres, even in monochromatic light. 2V's were checked against values calculated from refractive indices with the equation
cos~ V~ = ~2 (~2__~) f12 (~__ ~2)
or approximate values read from the charts of McAND~w (1963, Figs. 1 and 2).
X-Ray Data Cell parameters were measured by the powder method with a camera of 114.6 mm diameter, 0.02 mm capillaries made from low-absorbing glass, and manganese filtered, Fe radiation. Eccentricity and absorption errors were corrected by BACON'S (1948) extrapolation method used in conjunction with a silicon internal standard. All data had been previously corrected for film shrinkage. On the average, 50 diffraction lines from 2 0 ~ 1 2 ~ ~ from each amphibole were measured; of these an average of 39 could be unambiguously indexed and used in a least-squares code (MAST), which was run on an IBM 7094 computer. Indexing was accomplished with visually estimated intensities from a series of precession and Weissenberg films of glaucophane, crossite and riebeckite-arfvedsonite (No. 4, No. 8, No. 10, Table 2, respectively) taken with Me and Fe radiations, respectively. Strong high-index diffraction lines not included in the Weissenberg and precession films (e.g., ~55, ~65, and ~04 of iron-rich members in a C 2/m cell) were indexed on the powder films by comparison with theoretical patterns computed from the chemical analyses, likely site occupation, and atomic coordinates from a refinement of the riebeekite structure (COL- WLL~ and GIB~S, 1964; GIB~S, personal communication) with SMITh'S program (SMIT~, 1963). Use of the Wm~TAK~R (1949, p. 315) atomic coordinates for "crocidolite" transformed to a C 2/m cell (Wu 1953, pp. 49--50) were less successful in this connection. The chief source of error in the cell dimensions is incorrect determination of the d-spacing of the planes in the sample and internal standard because of overlap, which is unavoidable in complex patterns. For this reason data from the generally more accurate back reflection region could not be used to the exclusion of data from the front reflection region. Cell parameters and indexing are for the (001) centered cell (C 2/m) which was chosen following DONNAu rules (DefrAY and Now~ct~x, 1954, p. 150). The data are given with pertinent probable errors. Volumes calculated therefrom have a maximum probable error of ~ 0.56 A a (0.05 %).
Optical Properties and Cell Parameters in the Glaucophane-l~iebeckite Series 79
The cleavage angle (110)^ (li0) was calculated from cell parameters using the relation
a o sin fl [(110) ^ (1i0)] = 2 tan -1 bo
Optical Properties
Orientation o/ the Indicatrix
Within the amphibole group studied, there are four distinct orientations of the optical indicatrix. They have been called the "glaucophane" , "crossite", "osanni te" and "riebeckite" orientations (T~SG~R, 1956). To emphasize tha t they specify
ORIENTATION OPTIC tZ' ( COLORLESS ) SYMBOL PLANE
( ' ) ~ B=b IOLET) G (OIO1 Cx)
a (YELLOW) ( . )
IOLET ) C L (OIO) {x)
~ (YELLOW) (o)
7=b ~E~-IBLUE 1 o • ( o o )
~" (YELLOW) (o)
~E /~=b Y-BLUE) R (010) (•
Fig. 2. Orientation of the indicatrix with respect to the basal section of the amphiboles. Dashed lines are traces of the optic plane (T.O.P.). The cleavage angle (ll0)A (110) and pleochroism are indicated on the section normal to the e axis. Symbols (o, • and A) are associated with the refractive indices plotted in Figs. 3 and 4
part icular optic orientations ra ther than chemical species, they are abbreviated to G, C, O, and t~ orientations respectively in subsequent discussion. A p]ot of composition versus refractive indices such as Fig. 3 shows tha t the orientation changes where substi tutions cause any two of the three refractive indices to cross. 1% would also be represented ff ~, fi and y were extended until fl and crossed e.g. beyond the bounds of either Fig. 3 on the riebeckite-rich side or Fig. 4 on the magnesioriebeckite-rich side. For the sake of explanat ion it is convenient to assume tha t the absorption axes are parallel to the vibration directions within the crystal, although this is not necessarily true (SIIoDA, 1957). Then we may associate the vibration direction nearest to the c axis with a blue pleochroie color (A in Figs. 2, 3, and 4), that nearest to the a axis with colorless, yellow or pale brown pleoehroism (.), and that parallel to b with violet, purple or dark grey blue (x). Thus if the relative
O~
Tab
le 4
. O
ptic
al o
rien
tati
on
Nam
e of
ind
i-
eatr
ix o
rien
tati
on
(sym
bol)
Che
mic
al t
yp
e P
leoc
hroi
sm *
In
dic
atri
x
Ex
tin
ctio
n
ori
enta
tio
n
angl
e E
xam
ple
s
"Gla
ucop
h~ne
" ff
!.?.
~fPh
~ .........
G
som
e cr
ossi
tes
Cro
ssite
"Cro
ssit
e" C
so
me
Mg-
rieb
ecki
tes
Rie
beck
ite
som
e ri
ebec
kite
- "0
san
nit
e" 0
ar
fved
soni
te
yell
ow
viol
et
blue
�9
•
A
b =
~ I1
(o10
) c
A
7
yell
ow
blue
vi
olet
b
: y
c A
fl
�9
A
• •
(010
) an
d c
a. •
c
blue
to
yell
ow-
grey
blu
e b
~ ~
c A
M
g-ri
ebec
kite
s in
digo
b
row
n
J_(0
10)
and
ca.
IIc
A
�9
•
incl
udin
g M
g-
croe
idol
ites
an
d
rhod
usit
es
"l~
iebe
ckit
e" 1
~ R
iebe
clci
te-
dar
k
dar
k g
rey-
ye
llow
- b
= fl
c
A
arfv
edso
nite
s bl
ue
blue
b
row
n
II (01
0) a
nd
lie
A
• �9
@ 1
, 2,
thi
s st
ud
y
6, t
his
stu
dy
; H
OL
GA
TE
(19
51)
@ 7
an
d 8
, th
is s
tud
y;
NIK
ITIN
and
K
LE
ME
N (
1938
)
•IY
AS
I-II
RO
a
nd
IW
AS
AK
I (1
957)
:~ 9
, th
is s
tud
y;
AL
NB
~RG
(193
0);
I)H
EM
IST
E~
et a
l. (1
950)
PA
LA
C~
an
d W
AR
R~
(19
11) ;
V
~D
L
(192
4);
:~ 1
1 th
is s
tud
y
WY
lVIO
~D a
nd
WIL
SO~
(195
1);
MC
LA
CH
LA
N (
1951
)
])~.
VIL
LI]
~S (
1948
) ;
KIn
G (
1961
)
=~ 1
0, t
his
stu
dy
; IV
[ACK
Au e
t al
. (1
949)
; :B
AIN
(193
4) ;
CO
LE
MA
~ (1
951)
; L
AC
RO
IX (
1889
); B
OR
LE
Y (
1963
)
* G
iven
for
0.0
3 m
m s
ecti
ons;
gre
enis
h an
d b
lue-
gree
n ti
nts
may
be
cons
pico
us i
n th
inn
er s
ecti
ons.
A
Vib
rati
on d
irec
tion
nea
rest
a a
xis.
•
Vib
rati
on d
irec
tion
par
alle
l to
b a
xis.
�9
Vib
rati
on d
irec
tion
nea
rest
c a
xis.
�9
Optical Properties and Cell Parameters in the Glaucophane-Riebeckite Series 81
magnitudes of the refractive indices associated with these three absorption axes can be determined, the orientation of the indicatrix can be determined uniquely. In other words, the pleochroie scheme is distinct for each of the four orientations. Descriptions of the four orientations are given in Table 4 and in Fig. 2 ; the latter shows the basal section of amphiboles having each of the four orientations. Lacking measurements of R.I . , the orientation of the optic plane can be used in conjunction with the sign of elongation to recognize the type of indicatrix orientation. If one restricts the term riebeckite to alkali amphiboles in which A-~ X ~ 2.0--2.3, i.e., the A position is either empty or only partially filled, most riebeckites have the O orientation. Riebeekite-arfvedsonites in which A~-X-~2.5--3.0 can have either I~ or O orientations, possibly depending on whether (F~-C1) is high (-~0.7--1.5 formula units) or low (-~0--0.7 formula units) respectively. I t is likely that fl and y indices do in fact intersect within the diagram (Fig. 4) and that the pure end member has the R orientation. However, most naturally occurring magnesioriebeekites including the fibrous varieties typically have the O orientation. A C orientation also is possible for intermediate types (m, Fig. 1; from M~AsHmo and IwAsA~, 1957) because substitution of Fe'3'A1 ~ for Mg3Fe~" is associated with a more rapid increase in n associated with the yellow and violet absorption axes than the blue (Fig. 4). Optical orientation of crocidolites is not well known because of their tendency to occur in bundles of fibers whose individual members are less than 1000 A across and because coatings of iron oxide and quartz impurities which decorate external faces make them difficult to measure. MIL~S (1942) and SVZVKI (1939) describe the R orientation, while P~ACOCK'S (1928) and the author's measurements below on erocidolite (from MalipsdrLft, S .E . of Pietersburg, Transvaal, South Africa) indicate the orientation is 0 :
= deep sky blue fl ---- pale yellow y ~ purple b --~ y, T.O.P. _L (010), c ^ ~ ---- 0~ ~ 2 V = (--) large.
I t is possible that the exact orientation of the indicatrix depends on the degree of weathering and hence on oxidation of Fe", which commonly occurs in crocido- lites (HonasoN, 1965). The interpretation of the hollow fibers of Mg-croeidolites and crocidolites photo- graphed by KIN~ (1961) with an electron microscope remains elusive. Electron micrographs taken by the author of the crocidolite from South Africa described above also suggests that some fibers are hollow. These observations are reconciled with difficulty to WmTTAKWU'S (1949) structural analysis of Mg-croeidolite. In all other respects the crocidolites resemble amphiboles, and the possibility that the tubular appearance is an artifact, e.g., a focusing effect, has not been eliminated.
Variation in Optical Parameters Glaucophane-Riebeclcite Series. Trends shown in Figs. 3 and 4 represent variation in optical properties of amphiboles which closely represent binary solid solutions. Ideally, the abscissa represents compositions along the diagonals of Fig. 1. As few amphiboles of this composition have been reported in the literature, Figs. 3 and 4
6 Contr. Mineral, and Petrol., Vol. ]5
82 I.Y. BorG:
are constructed from amphiboles whose compositions only approach this ideal. Examples that deviate, especially iron-rich members (e.g., No. 6, Fig. 1) also depart from general trends in optical properties. The type crossite has been omitted altogether from Fig. 3 for the same reason. In addition to new data (numbered 1--11) Fig. 3 includes data from the literature (lettered). The amount of Ca in M~ is on the order of 0.2 units per hag-unit cell in ~11 except No. 6, HOLGATE'S erossite, and p, Eg~sw's E4GL, in which it is higher.
9 0 % 80 ~ 60 ~ 4 0 ~ 20 ~
2V 0 ~ 20 ~ 40 @ 60 ~ 80 ~ 90o~
0 2 5
z 0 0 R I E N
TATI 0 N G ORIENTATION I C ORIENTATION : ]~ b = ) " - - b= B O R I I ( O I O ) b=)" O.P.L(OIO) O.R•
Fig. 3. Variation in optical properties in the g]aucophane-riebeckite series. Iqumbered points correspond to new data in Table 3; lettered points are form BANJO, 1959 (a and b), PEEMISTE~ et al., 1950 (e); E~ST, 1964, E4 GL (1o). Dashed lines indicate uncertainties in the curves because of lack of data; the dashed 2 V and (~--~) curves are computed from refractive indices extrapolated across the vacant area. Symbols (e, • and A ) represent n near or parallel to te a, b and c crystallographic axes respectively
Noteworthy features are:
1. The well known increase in l~.I. with increased concentration of transition elements (Fe'", Ti, Fe", and Mn). 2. Non-linear variation of n near to c (A) with composition. Natural riebeckites have lower indices than might be expected from a linear extrapolation from the crossite region of the diagram. This is due to the substitution of 1~ and C1 in the OI-I position (Bo~LEY, 1963 ; COMEFO~O and KoH~, 1954), which is a common and perhaps constant feature of natural riebeckites and riebeckite-arfvedsonites. 3. Insensitivity of extinction angle to change in composition. Higher values ( > 1 0 ~ occur in erossites (e.g., No. 8) which contain more of the magnesio- riebeckite molecule.
Optical Properties and Cell Parameters in the Glaucophane-Riebeckite Series 83
4. Rapid and cyclic variation of 2 V with increased (Fe'", Fe" and Ti) beyond G150Ris0. The dashed line for 2 V in Fig. 3 is constructed from the variation in refractive indices; the very few natural examples tha t have compositions between Gl~oRi60 and Gll0Ri~0 are inadequately described in the literature or are represented by inferior chemical analyses.
5. Close agreement of indices for synthetic glaucopane and extrapolated values for the end member, viz:
~ =1 .596 4- 0.003 7 = 1 . 6 2 0 • C A 7 = 1 0 ~ ~ (av. of 8) (ERnsT, 1961; COLV~L]~ et al., 1966, Table 3)
=1 .594 f l = 1 . 6 1 2 y = 1 . 6 1 8 c^ 7 = 6 ~ (Fig. 3, this work) = 3 . 0 3 2 V = ( - - ) 52 ~ (7 - - e )=0 .025 , b = f l O.P.[](010)
For this composition the degree of disorder in the M 2 site does not appear to effect materially the refractive indices in contrast to its effect on lattice para- meters (see p. 86).
6. There is similar close agreement of the indices for synthetic riebeckite with the extrapolated values for the end member although the latter presume minor F substitution for (OH).
= 1 . 7 0 9 • Y =1-718 4- 0.004 c ^ / ~ = 3 ~ 4-6 ~ (av. of 15) (E~sw, 1962, p. 704; COLVrSLE et al., 1966, Table 3)
=1 .702 /3=1.712 Y = 1 " 7 1 9 c ^ ~ = 6 ~ (Fig. 3, this work) = 3 . 4 0 2 V = ( - - ) 5 0 ~ , ( 7 - - e )=0 .015 , b = y O . P . •
a n d oa. I[ c .
Optical data for both glaucophane and riebeckite end members show poor agreement with comparable data computed from linear regression equations (WINcH~LL, 1963, Table 4).
7. The composition at which fl and 7 cross is approximately G1~61%i4~; ~ and/~ become equal at G117Ries3 where the C orientation gives away to O.
Magnesioriebeckite-Ferroglaucophane Series. Variation in optical properties of the magnesioriebeckite-ferroglaucophane series is indicated in Fig. 4. The detail is tentative, since accurate optical data for compositions which are simple solid solutions of the two end members, magnesioriebeekite and ferroglaucophane, are lacking. There is no reason to believe, for example, that refractive indices change linearly with composition. The composition at which the 0 orientation changes to C is poorly fixed at MgRiToFeG180.
In addition to new analyses (indicated by number) data were compiled from the literature (indicated by letter). SWITZ~'s crossite (1951) (d), Harvard collection No. 12873, was also included after the optical properties were reexamined and refined as follows:
:r = 1.661 fi = 1.666 7 = 1.668 r < v , strong horizontal dispersion = p a l e yellow f l = d e e p blue 7 = d e e p purple f l = Y > ~ b = 7
o . P . i (OLO) 2V (--) ~ 83 ~ 86.5 ~ 90 ~ 91.8 ~ 95~ - - average 89.0 ~
c^f l = 9.5 ~ 9.5 ~ 10 ~ 1 0 . 5 ~ average 10.0.
The magnesioriebeekite from Bizan (MIYASItIRO and IWASAKI, 1957) which has the C orientation could not be satisfactorily included in the group. As its x-ray
6*
84 I .Y. BORG:
pa rame te r s are also unusua l (p. 87), i t is l ike ly t h a t the zoning and colorless cores of the sample descr ibed b y the au thors have compl ica ted corre la t ion of composi t ion wi th phys ica l proper t ies . Ref rac t ive indices for several of the crossites in the group are lower t han migh t be an t i c ipa ted f rom the general t rends of Fig. 4. R e l e v a n t observa t ions are t h a t two (No. 6* and i) conta in more Ca and A1 Iv t h a n is t yp i ca l of o ther members of the series, and the o ther (b) is no t a s imple b i n a r y solid solution.
0 ORIEN- TATION
L. b =7 " I~ C ORIENTATION - - O.P.• b=y O.F?.L (.010)
89o / / \ \ 6 0 ~ 4 0 o //'.~ \\ 2oo "(-} (§ . /
2V 0 ~ 20 ~ I 4 0 ~ , k \ 60 ~ 80 ~ ( + ) ~ - )
40~ " . 0 0 5
can 20 ~ �9 CA ~ ~'~-~ lO ~ .
0 o
~ , 2 5 ";a/c. p (rheas) o o o 5,20 P
G o o 3.15
1.680 -
n 1.660 1.650 I 1.64-0
I I I l I ~ I I l l I
f gh 8 d b i 7c q i i , i i i
1.0 .8 .6 .4 0 CROSSITE
MAGNESIORIEBECKITE ( Fe" '+ Ti ) + ( Mg+L i ) FERROGLAUCOPHAN E
N(~ 2 Mg 3 Fe~" Si 8 022 OH) 2 ]~OTAL Y ~ o 2 Fe~"AI 2 Si s 022 [OH) 2
Fig. 4. Variation in optical properties in the magnesioriebeekitc-ferroglaucophane series. Curves compiled from numbered amphiboles in Table 3 and literature data as follows: (q) IWASAKI (1963); (f) POLOVINKA. 1953), !~o. I I I ; (g) ])E VILLIE~S (1948); (h) WYlYIOND and WILSON (1951); (b) BANNO (1959); (i) BLOXAM and ALLEN (1960); (C) LACROIX (1941); and (d) SWlTZER (1951), a crossite with newly determined optical properties (see text). All chemical analyses are calculated on the basis of 24 (0, OH, C1, F). Symbols (e, x , and ~k) represent n near or parallel to the a, b and c axes, respectively. In order to recognize the trend in 0 with change in composition, the measured data have been augmented with a calculated density for the magnesioriebeckite end member (Fig. 6). Dashed lines indicate uncertain trends; as in Fig. 3, the dashed 2V and (7--~) curves are calculated from the most likely values of ~, fi, and 7
Extrapolated values for the optical parameters of the pure magnesioriebeckite are in good agreement with measurements on synthet ic counterparts, viz . ,
= 1.654 _u 0.003 7 ---- 1.672 ~: 0.003 c ^ ~, /~ or 7 ----- 4 ~ ~ =h 4 ~ (av. of 7) ( E ~ s T , 1960, Table 10, runs with hemat i te -magnet i te buffer only; ERNST, 1963, p. 245-247; COLWLLE et al., 1966, Table 3)
:r = 1.655 fi = 1.671 7 = 1.672 c h ~---- 32 ~ (Fig. 4, also see p. 87 this work) e----3.15 2 V ---- ( + ) 0 - - 2 0 ~ (7 - -~ ) - - 0.02 b = 7 0 . P . J _ ( 0 1 0 ) a n d c a , l[c.
* Microprobe analysis indicates a Na-poor and Ca-rich phase is strongly segregated within the amphibole suggesting presence of minute epidote inclusions.
Optical Properties and Cell Parameters in the Glaucophane-Riebeckite Series 85
b o
{D
There is only fair agreement with predictions from linear regression formulae (WI~cHELL, 1963, Table 4). HonI (1954) and WINCH]~n (1961, Table 4) have shown tha t substitution of equal numbers of actions per unit volume into clinopyroxenes increases 1%. I. in the order of Mg, Mn, Fe" and Ti. An analogous effect is observed in these amphi- boles although the effect of Mn and Ti on g . I . cannot be confidently appraised relative to tha t of the other metals in the list. t~ecall ~hat 1%. I. increases with increased total Fe'", Ti, Fe" and Mn (Fig. 3). Note tha t a slight increase in n
N~2 Mg3AI2 Si8 022 (OH)2 TOTAL Y No 2 Fe3"Fe'~Si s 022 (OH) 2
Fig. 5. Variation in cell dimensions, cell volume and density in the glaucophane-riebeckite series. Solid numbered points correspond to new data in Table 3. Lettered points are En~ST'S measurements (1963, Table 10) on amphiboles described by BANNO (1959) (a and b) and PgmlrSTER et at. (1950) (e); PAPIK~ and CLA~K (1966) (n); and E~gs~ (1964, Table 4, E 4 GL and E 16 GL) (p and s). Open square is measured value rather than the calculated value. Chemical analysis of (s) was corrected for 4.2% epidote impurity and recalculated to 100.28 total. Chemical analyses and 9 are calculated for 24 (0, OH, C1, F) per half-unit cell
associated with the a ( � 9 and b ( • ) axes results from substitution of MgsFe'~" for Fe~'AI s (Fig. 4). Thus, there is a greater increase from the substitution of Fe" ' for A1 vI than Fe" for Mg. The situation is reversed for the index nearest ~he e axis ( �9 The optic angle varies widely with small changes in composition; its magnitude is not readily related to particular substitutions.
Variation in Cell Parameters
Glaucophane-Riebeckite. Cell dimensions, volumes and calculated densities of amphiboles are plotted versus composition in Fig. 5. Cell parameters indicated by open circles were measured by ERNST (1963, Table 10) on previously analyzed amphiboles (a and b), ]~Ax~o (1959); (e) PH~MISTE~ etal. (1950). Densities were calculated from these data; stoichiometry is based on 24 (0, OH, F, C1) computa- tions. The X site of a]l contains nearly 2.00 cations per half-unit cell and Ca is
86 I.Y. :BORG:
a minor constituent except in No. 6*, p and s. Other amphiboles measured by E ~ s T but not included in Fig. 5 either contain large proportions of the other two end members or have analyses whose totals depart sufficiently from 100.0 to be suspect, e.g., E 13 GL (EI~NST, Table 4, 1964), PALAC~E and WAn~EN'S No. 12 (PALACRE and WAn~EN, 1911) and SWITZE~'S riebeckite No. 11 (EaNsT's numbers in Table 10, 1963). Amphiboles No. 10 and No. 11 (Table 3) are not included because AX ~ 2.66 and 2.63 respectively, and because their fluorine and chlorine contents are high. Both of these factors result in smaller cell dimensions, par- ticularly b 0 and c o than would otherwise be measured (F~OST, 1963). Comparison of the parameters of three iron-rich amphiboles included (Nos. 9, 10 and l l , Table 3) also tend to verify F~osT's observation (p. 381) that increased Ca and Fe" is associated with increases in b 0 but not c 0. Note for example that on the basis of F and Na contents (Table 2) it might be expected that amphiboles Nos. 9, 11 and 10 are in order of decreasing b 0 and c o dimensions. This in fact is true except for the b 0 dimension of No. 11, the amphibole which has the highest combined Ca and Fe" content. Two crossites containing somewhat more Fe '" and Ti than required by a mixture of G1 and Ri end members depart slightly from the otherwise linear trends (Fig. 5). Extrapolated parameters for glaucophane and riebeckite end members are given in Table 5, together with E ~ s T ' s data for synthetic counterparts (1962, 1963,
Table 5. Comparison o] cell dimensions o/ synthetic end members and extrapolated values/or glaucophane and riebeckite
End Synthetic glaucophane End Synthetic riebeckite member (ERNST, 1963, Table 8; member (ERNST, 1962, Table 10; glaucophane COLVILLE et al., 1966, riebeckite COLV~LE et al., 1966, C 2/m Table 2) C 2/m Table 2)
C 2/m C 2/m
I II (av. of 15) (av. of 10) (av. of 8)
a 0 (A) 9.50 9.75 9.64 9.78 9.73 b o (A) 17.67 17.91 17.73 18.03 18.06 c o (A) 5.29 5.27 5.28 5.34 5.33 fl 103.7 ~ 102.8 ~ 103.6 ~ 103.5 ~ 103.3 a o sin ~ 9.22 9.518 9.37 9.51 9.47 V (A) a 864 897 877 918 913
(cale) 3.03 - - - - 3.39 s - -
COLVILLE et al., 1966, Table 2). Agreement is very good in the case of riebeckite, but poor for glaueophane irrespective of which of E~ST 'S proposed polymorphs (I and II) is considered. The possibility of gross error in measurement of cell parameters is not likely since ERNST'S x-ray data for natural amphiboles (1963, p. 258) included in Fig. 5 (open circles) and the new data presented here (solid circles) are consistent. The differences in parameters of the G1 end member and synthetic glaucophane lie primarily in the a o dimension and the related cell volume. If occupation of the M s site by A1 to the exclusion of other cations produces minimum a 0 sin fl
* See footnote p. 84.
Optical Propertie and Cell Parameters in the Glaucophane-Riebeckite Series 87
(the distance between silica chains) and a 0 dimensions (ERNST, 1963), then it is likely tha t the natural end member is a completely ordered form, whereas both the synthetic varieties (glaucophane I and I I ) are disordered counterparts tha t differ from each other in degree. Recent s tudy of a natural glaueophane (Gls0Ri~0) has in fact shown tha t the structure is ordered and tha t M e is occupied chiefly by A1 (PAPIK~ and CLA~K, 1966). A similar but less pronounced order-disorder relation might be expected in riebeekites. The average ionic radius at M2= Fe" ' , Fe" in a disordered structure is greater than tha t in an ordered structure M ~ F e " ' (Fe" ' is 0.64 ~ ; Fe" is 0.74 A); hence a 0 sin fl should vary depending on the occupancy of the site. Magnesioriebec/cite-Ferroglaucophane. Parameters for members of the magnesio- riebeekite-ferroglaucophane group are shown in Fig. 6. In addition to new data (solid circles) ERNST'S x-ray data (1963, Table 10) (open crieles) are plotted against compositions which were calculated on the basis of 24 (0, OH, C1, F) from analyses in the literature. Calculated densities have been excluded for those amphiboles whose reported water content departs appreciably from the ideal. Unaccountably, one magnesioriebeekite for which data are available [MxYAs~ao and IWASA~(I, 1957 (m, see Fig. 1); ERNST, 1963, Table 10, No. 5] does not fit into this scheme although the analysis is good and its composition indicates it is nearly a binary solid solution (also see p. 83). Agreement between parameters measured on synthetic magnesioriebeekite and extrapolated values for the end member (Table 6) is very good.
Table 6. Comparison o] cell dimensions o] synthetic end member with extrapolated values o/ magnesioriebeclcite
End member Synthetic magnesioriebeckite magnesio- (ERnsT, 1963, Table 3; riebeckite COLVILLE et al., 1966, Table 2) C 2/m C 2/m
(av. of 7) a 0 (A) 9.76 9.73 b o (A) 17.97 17.95 c o (A) 5.31 5.30 fl 103.9 ~ 103.3 ~ a o sin fl (A) 9.47 9.47 V (~3) 904 901
(eale) 3.15 (?) - -
E//ect el Cation Substitutions on Cell Parameters
Inspection of Figs. 5 and 6 indicates the following: 1. Cell dimensions increase as per cent of the riebeekite molecule increases, i.e., as Fe'3'Fe'~" substitutes for MgaA12. 2. The increase in parameters associated with Fe ' " substitution for A1 vI is much greater than the increase associated with Fe" substitution for Mg; otherwise the magnitude of slopes in Figs. 5 and 6 would not be similar. [Slopes of %, b0, and c o a 0 sinfi are slightly greater for data in Fig. 5 than in Fig. 6, whereas slope of fi angle is less in Fig. 5 than in Fig. 6. This might be anticipated from consideration
88 I .Y. BORe:
of ionic radi i , viz., the difference betweel AI ' " (0.52 A ) and F e ' " (0.64 .A) is g rea te r t h a n be tween Mg (0.67 A) and F e " (0.74 A).] 3. F e " subs t i tu t ion for Mg increases the b 0 d imension more t h a n a 0 and a 0- sin ft. The effect can be seen in the grea te r difference in slopes associa ted wi th MgaAl 2 ~ Fe~'Fe'~" and Fee"A12 ~ MgsFe~" in the case of b0, t h a n a 0 and a 0 sin ft. Analogous effects on b 0, a o and a 0 sin fl have been observed for e l inopyroxenes in which the Ca conten t is s imilar (B~ow~, 1960, p. 24; W~CH~LL and Tilling, 1960, Table 3 - - 5 ; and especial ly VISWA~TATHA~, 1966, p. 433 435).
3.25 3.20 3.15 3 .10
- p ( c o / c ) o o "
o
V 9 0 0 . (~ .3) 8 9 5 .
8 9 0 . 8 8 5 . �9 o 9 .85
13 o
�9 o 9 .75 " 9 .65
18 .00 - bo 9 .55 bo 17 .95 -
(,~) 17 .90 17.85 Co o 17 .80 o.-.-o o �9 c - o �9
i o 4 . o o - - B o ,8 1 0 3 . 7 5 - -
( ~ Oo Sin'8 9 . 4 0 ~
ooSin,8 9 . 3 0 (~,) 9 . 2 0 I i I I 6 b I I I
(3 o
(~,)
5.35 co 5 . 3 0 (~,) 5 . 2 5
I I I .2 .I 0 1.0 .9 .8 .7 .6 .5 .4 .3
CROSSITE
MAGNESIORIEBECKITE ( Fe ' "+T i ) + ( M g + L i ) FER ROGLAUCOPHAN E
NQ2 Mg3 Fe2'" S i8022 (OH)2 TOTAL Y No2 Fe3"AI2 S i 8 0 2 2 ( 0 H ) 2
Fig. 6. Variation in cell dimensions, cell volume and density in magnesioricbeckite-ferro- glaucophane series. Solid numbered points correspond to new data in Table 3. Lettered points are ER~FST'S (1963) measurements on amphiboles described in the literature as follows: (j) WmT~K~R (1949); (k) ER~TST (1960); (1) ERNST (1963); (d) SWITZER (1951); (b) crossite, BANJO (1959). Formula and ~ calculated from chemical analyses on the basis of 24 (0, OH, C], F)
4. The negligible change in the c o d imension in bo th series is to be expec ted since the length of the silica chain is no t g rea t ly inf luenced b y cat ion subs t i tu t ion in M4, M I + M s + M 2 in e i ther amphiboles or pyroxenes . Subs t i tu t ions of A1 Iv for Si has smal l effect con t r a ry to observa t ions on hornblendes (BI~NS, 1965, p. 314).
Acknowledgments. The cost of the chemical analyses was defrayed from a G.S.A. Project Grant 682--55, which is gratefully acknowledged. A large portion of the work was done at PRI~TC~rON (1955--1959) while the writer was a Visiting Fellow in geology. Use of depart- mental facilities as well as H. HEss' interest and inspiration were greatly appreciated. Most of the x-ray work was carried out at the University of California (Lawrence Radiation Laboratory, Livermore, and Department of Geology, Berkeley). Help from the staff especially with the photography of single crystals was invaluable. Thanks are due Prof. W. G. ERNST for critical comment and for making available the revised optical and x-ray parameters of synthetic amphiboles prior to their publication thereby greatly improving the argument herein. Thanks are also due to those who supplied samples (S. O. AGRELL, R. SIT~GAM, •. HOLGATE, C. FRO~DEL, W. Qc~D~, B. M~so~T, D. S~A~A~T, J. P}IEMISTER, H . M . E . SCIt~Ir J. GLOV]~R, M. BASS, D. FT~INN, E. SA~PSO~T) and made the study possible.
Optical Properties and Cell Parameters in the Glaucophane-1%iebeckite Series 89
References
AL~E~G, L. : The hornblendes from the Mauriupol alkaline body. Bull. Geol. Prospect- Service 49, 75 (1930).
BAcon, G. E. : The determination of the unit cell dimensions of non-cubic substances. Acta Cryst. l, 337 (1948).
B ~ , A. D. N. : The younger intrusives of the Kudau Hills, Nigeria. Quart. J. Geol. Soc. London 99, 201--215 (1934).
BA~o, S. : Notes on rock-forming minerals (10) glaucophanes and garnet from the Kotu district, Sikoku. J. Geol. Soc. Japan 65, 658--663 (1959).
BAss, M. : Effects of gamma irradiation on the physical properties of minerals. Am. Minera- logist 42, 100--104 (1957).
B I ~ s , 1%. A. : The mineralogy of metamorphosed basic rocks from the Willyama Complex, Brooken Hill District, New South Wales. Part I, hornblendes. Mineral. Mag. 35, 306--325 (1965).
BLox~ , T., and J. B. ALLEN: Glaucophane schist, eclogite and associated rocks from Knocknormal in the Girvan-Ballantrae complex, S. Ayrshire. Trans. l~oy. Soc. Edin- burgh 64, 1 (1960).
BODY, WIG, C.: ~ber den Glaueophan yon Zermatt. Poggendorffs Annal. Phys. Chem. 1~8, 224 (1876) [See ZAMBONI~I (1906)].
BORG, I. Y. : Glaucophane schists and eclogites near Healdsburg, California. Bull. Geol. Soc. Am. 67, 1563--1584 (1956).
BOgLEY, G. D. : Amphiboles from the younger granites of Nigeria. Part I. Chemical classifica- tion. Mineral. Mag. ~3, 358--376 (1963).
BnOW~E~S, M. : Glaucophane schists from the North Berkeley Hills, California. Am. J. Sci. 252, 614--626 (1954).
BRowx, G.M.: The effect of ion substitution on the unit cell dimensions of the common clinopyroxenes. Am. Mineralogist 45, 15--38 (1960).
COL~N/~, 1%. G.: 1%iebeckite from St. Peters Dome, Colorado. Bull. Geol. Soc. Am. 62, 1517 (1951) (Abst.).
COLVILLE, ALA~, and G.V. GIBBS: The refinement of the crystal structure of riebeckite. Bull. Geol. So. Am. Progr. Miami Beach Meeting 31 (1964) (Abst.).
COLVmLE, P. A., W. G. Eg~s% and M. C. GrL]3EgT: Relationships between cell parameters and chemical composition of monoclinic amphiboles. Am. Mineralogist 51, 1727--1754 (1966).
Co~]~FoRo, J. E., and J. A. KoH~: Synthetic asbestos investigations. I. Study of synthetic fluor-tremolite. Am. Mineralogist 39, 537--548 (1954).
DE~I~, W.A., 1%. A. HowIE, and J. ZUSSMA~: The rock forming minerals: chain silicates. vol. 2. London: Longman 1963.
DON~AY, J. D. H., and W. NowAcF~I: Crystal data: Classification of substances by space groups and their identification from cell dimensions. Geol. Soc. Am. Mere. 60 (1954).
E~Ns% W. G. : Stability relations of magnesioriebeckite. Geochim. et Cosmochim. Acta 19, 10--40 (1960).
- - Stability relations of glaucophane. Am. J. Sci. 259, 735--765 (1961). - - Synthesis, stability relations and occurrence of riebeckite and riebeckite-arfvedsonite
solid-sointions. J. Geol. 70, 689--736 (1962). - - Polymorphism in alkali amphiboles. Am. Mineralogist 48, 241--260 (1963). - - Petrochemical study of coexisting minerals from low grade Schists, Eastern, Shikoku,
Japan. Geochim. et Cosmochim. Acta 28, 1631--1668 (1964). Es~:oLA, P., u. T~. G. S~T,S~EI~: ~ber die unvollkommene Ausl5schung einiger Amphibole.
Compt. rend. geol. Finlande No. 3, 13--19 (1930). FI~OST, M. T.: Amphiboles from the younger granites of Nigeria; Part II. X-ray data. Mineral.
Mag. 33, 377--383 (1963). F u ~ , N. H. (ed.): Scott's standard methods of chemical analysis, 5th ed., vol. 1. New
York: D. van Nostrand Co., Inc. 1939.
90 I .Y . BORG:
GAzzr, E. P.: On the microscopic determination of the amphiboles in grains. Am. Minera- logist 48, 422--429 (1963).
G~os~, S. : A scheme of cation distribution in the amphiboles. Mineral. Mag. 35, 46--53 (1965).
GRoss, E. B., and E. W. 1-I~I~RICH: Petrology and mineralogy of the Mount Rosa Area, E1 Paso and Teller Counties, Colorado. II . Pegmatites. Am. Mineralogist 51, 299--323 (1966).
HAOI~ER, A.F. , S. S. LEU~G, and J. M. DEs~ISO~: Optical and chemical variation in minerals from a single rock specimen. Am. Mineralogist 59, 341--355 (1965).
I-I~ss, H. H. : Stillwater igneous complex, Montana. Geol. Soc. Am. Mem. 80 (1960). HI~TZ~, C. : Handbueh der Minera]ogie, 2. Bd. Berlin 1897. - - Handbuch der Mineralogie, Erg.-Bd., Lieferung 1--4, Leipzig 1938. HLAWATSCg, C.: ~ber den Amphibol yon Cervadaes, Portugal. Festschrift ffir H. ROSEN-
~ s c ~ , Stuttgart. S. 68--78. 1906. HODGSON, A.A. : The thermal decomposition of miscellaneous croeidolites. Mineral. Mag.
35, 291--305 (1965). - - A. G. FREeMAn, and t t . F. W. TAYLOg: The thermal decomposition of crocidolite from
Koegas, South Africa. Mineral. Mag. 35, 5--30 (1965). HOLGAT]S, 1~I. : On crossite from Anglesey. Mineral. Mag. ~9, 792--798 (1951). HORI, 1~. : Effects of constituent cations on the optical properties of clinopyroxenes. Sci.
Papers, Coll. Gen. Educ. Univ. Tokyo 14, 71--83 (1954). IDDINGS, J. P.: Rock minerals, 2nd ed. New York 1911. IWASAKI, M. : Metamorphic rocks of the Kotu-Bizan area, eastern Shikoku. J. Fae. Sci.,
Univ. Tokyo, Sect. I I , 15, 1--90 (1963). JAKOB, J., u. E. BRANDENBERGER: Chemische und rSntgenographische Untersuehungen an
Amphibolen: I. Mitteilung: Die Osannite yon Alter Pedroso. Schweiz. mineral, petrog. Mitt. 11, 140--162 (1931).
Jon~ssE~, A.: ~ber den Krokydolith yon Griqualand West. Neues Jahrb. Mineral. Geol. Central B1. S. 353--356 (1910).
KA~B, B. : Isogyres in interference figures. Am. Mineralogist 43, 1029--1067 (1958). FANa, D. : The occurrence and comparative mineralogy of S. Australian magnesian crocidolites
(rhodusites). Trans. Roy. Soc. S. Australia 84, 119--128 (1961). KV~TZ, W. : Die IsomorphieverhEltnisse in der ttornblende-Gruppe. 1NTeues JaMb. Mineral.
Geol. 69, B-B Abt. A, 171--250 (1930). LAcgoIx, A.: Sur une roche a amphibole sodique (riebeckite), astrophyllite, pyroclore et
zircon, Colorado. Compt. rend. 109, 39--41 (1889). - - L e s glaucophanites de 1Niouvelle-Caledonie et les roches qni les accompagnent, ]eur
composition et leur genese, M6m.-Academie des Sci., T. 65, Paris 1941. MACKAY, R.A. , R. GgEE~WOOD, and J. E. ROCKI~GgAM: The geology of the plateau tin-
fields - - resurvey 1945--1948. Appendix: The morphology of the Jos plateau. Bull. Geol. Survey Nigeria No. 19 (1949).
McAsDREW, J. : Relationship of optical axial angle with the three principal refractive indices. Am. Mineralogist 48, 1277--1285 (1963).
McLAc~LA~, G. R. : The aegirine-granulities of Glen Lui, Braemar, Aberdeenshire. Mineral. Mag. 29, 476~495 (1951).
MiLsS, K. R. : The blue asbestos-bearing banded iron formation of the Hammersley range, Western Australia. Australia (West.), Geol. Survey, Bull. 190, Pt. 1 (1942).
MIYASHIRO, A. : The chemistry, optics and genesis of the alkali ~mphiboles. Jour. Fac. Sci. Univ. Tokyo, Sect. II , Pt. 1, 11, 57--73 (1957).
- - , and M. IWASKAI: Magnesioriebeckite in crystalline schists of Bizan in Sikoku, Japan. Jour. Geol. Soc. Japan 63, 698--703 (1957).
M u ~ o , M. : Errors in the measurement of 2 V with the universal-stage. Am. Mineralogist 48, 308--323 (1963).
- - Reply to comment on "errors in the measurement of 2 V with the universal stage". Am. Mineralogist 49, 813--815 (1964).
Optical Properties and Cell Parameters in the Glaucophane-i~iebeckite Series 91
MVRGOCI, G.: Contributions to the classification of the amphiboles. Univ. Calif. (Berkeley) Publs. ]3ull. Dept. Geol. Sci. 4, 359--396 (1906).
NrKITI~, V., u. 1%. KL~E~: Crossite aus Voduo bei Skiplje. Neues Jahrb. Mineral. Monatsh., ]3eil.-]3d. 74, 129--131 (1938).
OsAx~, A., u. 0. U~m~VEg: i)ber einem Osannithornblendit ein Feldspath freies Endglied der Alkalireihe yon Alter Pedroso. Sitz.ber. heidelberg. Acad. Wiss., Abt. A Math.-Phys. Wiss., ]3d. V. A, No. 16, 1--10 (1914).
PALAC~E, C. : A rock from the vicinity of Berkeley containing a new soda amphibole. Univ. Calif. (Berkely) Pbls. Bull. Dept. Geol. Sci. 1, 181--192 (1894).
- - , and ]3. E. Wa~n~N: The chemical composition and crystallography of parisite and new occurrence of it in the granite-pegmatites of Quincy, Mass., U.S.A. Am. J. Sei., Ser. 4, 31, 549 (1911).
P~vIKE, J., and J. CLAn~:: Cation distribution in the crystal structure of glaucophane II, the high pressure polymorph (Abst.). Program Geol. Soc. Am. Annual Meeting 1966, San Francisco, p. 157 (1966).
PAnK~n, 1%. ]3. : l~apid determination of the approximate composition of amphiboles and pyroxenes. Am. Mineralogist 46, 892--900 (1961).
PEXCOCK, M. : The nature and origin of the amphibole-asbestos of South Africa. Am. Minera- logist 13, 241--285 (1928).
P~E~IST~n, J., C. O. H ~ v n u and P. A. SA~I~: The riebeckite-bearing dikes of Shetland. Mineral. Mug. 29, 359~373 (1950).
P~LLrrS, 1%.: The recalculation of amphibole analyses. Mineral Mug. 33, 701--711 (1963). POLO~U~I~A, Y. In.: Cummingtonite and alkali amphiboles from Krivoi 1%og. Mineral.
Sbornik, Lvov Geol. Soc. N. 7, 167--186 (1953) [See Mineral. Mag. 57, Abst. 13 (t956)]. l%A])]~n, L. F., and F. S. G~ALDI: Chemical analysis for selected minor elements in Pierre
Shale. U. S. Geol. Survey, Profess. Papers 391-A (1961). S~ODA, T. : On the anomalous optical properties of heikolite. Mineral. J. (Sapporo) 1, 69--83
(1954). - - On probable elliptic vibration of light in heikolite. Mineral. J. (Sapporo) 2, 114--133
(1957). - - Elliptic vibration of light in some alkali amphiboles. Mineral. J. (Sapporo) 2, 269--278
(1958). SMI~, D. K.: A Fortran program for calculating x-ray powder diffraction patterns. Univ.
of Calif., Lawrence 1%adiation Laboratory UCRL-7196 (1963). SvzvKI, J.: A note on soda amphiboles in crystalline schists from Hokkaido J. Fac. Sci.,
Hokkaido Imp. Univ. 4, Ser. 4, 507--519 (1939). SwITz~n, G. : Mineralogy of the California glaucophane schists. Calif. Dept. Nat. 1%esources,
Div. l~Iines, ]Bull. 161, 5--70 (1951). ThAws, 1%. ]3. : Geology of the Sebastopol Quadrangle, California. Calif. Dept. Nat. 1%esources,
Div. Mines ]3ull. 162 (1952). T n 6 ~ n , W.E. : Tabellen zur optisehen ]3estimmung der gesteinsbildenden Minerale.
Stuttgart: E. Schweizerbart 1956. Vn~D~, A.: ~)ber einen 1%iebeckit. Z. Krist. 60, 135 (1924). VILLI~nS, J. E. n~: Note on unusual amphibole from Zesfontein, S.W. Africa. Trans. Geol.
Soe. S. Africa ~1, 77--80 (1948). V~SWA~T~, K. : Unit cell dimensions and ionic substitutions in common clinopyroxenes.
Am. Mineralogist 51, 4 2 9 ~ 4 2 (1966). W~gg~ , ]3. E.: The crystal structure and chemical composition of the monoclinic amphi-
boles. Z. Krist. 72, 493---517 (1930). W ~ t K E R , E. J. W. : The structure of ]3o]ivian crocidolite. Aeta Cryst. 2, 312--317 (1949). WILSON, A.: Co-existing pyroxenes: Some causes of variation and anomolies in optically
derived compositional tie lines, with particular reference to eharnockitic rocks. Geol. Mug. 97, 1--17 (1960).
WI~C~LL, A.M., and H. W ~ c ~ : Elements of optical mineralogy, part II, 4th ed. New York: John Wiley & Sons 1951.
92 I.Y. BORn: Optical Properties and Cell Parameters in the Glaucophane-Riebeekite Series
WINCIIELL, H. : Regressions of physical properties on the compositions and c]inopyroxenes: II. Optical properties and specific gravity. Am. J. Sci. 2~9, 295--319 (1961).
- - Clinoamphibole regression studies: I. Regressions of optical properties and density on composition. Mineral. Soc. Am. Spl. Paper, 267--277 (1963).
- - , and R. TILLING: Regressions of physical properties on the compositions of clinopyroxenes: I. Lattice constants. Am. J. Sei. 2~8, 529--547 (1960).
Wrcl~OrF, R. W. G. : Crystal structures, vol. I I I , chpt. XII, p. 49--50. Intersci. Publ. Inc. Publ. Inc. 1953.
WY~o~D, A.P., and R. B. WILSON: An occurrence of crocidolite near Robertstown, S. Australia. Trans. Roy. See. S. Australia 74, 44--48 (1951).
YAGI, K. : Petrochemical studies on the alkaline rocks of the Morotu district, Sakhalin- Bull. Geol. See. Am. 64, 769--710 (1953).
ZA~IBOI~n~r, F. : ~ber den metamorphosierten Gabbro der Roeca Bianca in Susa-tal. ~eues Jahrb. Mineral. Geol. ~, 62--104 (1906).
Dr. I. Y. BORG University of California Lawrence Radiation Laboratory Livermore, California
