Topaz-bearing rocks from Mount Gibson, North Queensland ...

11
American Mineralogist, Volume 77, pages303-313, 1992 Topaz-bearing rocks from Mount Gibson, North Queensland, Australia C. JonNsroN Department of Geology, JamesCook University of North Queensland, Townsville, Queensland, Australia B. W. Cruppnr,r, Department of Geology, Australian National University, Canberra,A.C.T., Australia Ansrn-lcr Dikes, sills, and irregular pods of topaz-bearing aplite, lithium fluorine granite, and topazite represent F-rich extreme differentiates of a suite of mineralized alkali feldspar granites near Mount Garnet, northeasternAustralia. Topaz is a magmatic phase in two aplites and the topazite and a hydrothermal mineral in the third aplite, lithium fluorine granite, and the alkali feldspar granite. The granites plot close to the 100 MPa (l kbar) HrO-saturated ternary minimum in the haplogranite system(consistent with their history of extended fractionation), but the aplites are significantly displacedfrom that minimum, suggesting either vapor phaselosses of qtartz and feldspar components (into pegmatites) or crystallization from hydrous fluid. Trace elements(e.g.,Rb, Ga, Nb) retained in feldspars, micas, and opaque minerals are enriched in the granites and aplites but not in the topazite. Other trace elements are strongly depleted, having been partitioned into halogen-rich hydrous fluids and lost into the country rocks. The strong enrichment ofBa, Sr, and Eu in the highly fractionatedtopaz aplites is unusual and is thought to indicate crystallization from or reequilibration with hydrous fluids rather than silicate melts. Topaz aplites of this area are chemically similar to ongonitesof central Asia and central Arizona, U.S.A., and small differences in major element chemistry reflect the differing F content of the provinces. Topazites are similar to those reported from the U.S.A., and the granites are chemically similar to topaz rhyolites of the western U.S.A. INrnooucrroN The topaz-bearing rocks of the Mount Gibson area,12 km east-northeast of Mount Garnet, North Queensland, form part of the CoolgarraBatholith, a mineralized com- positebody ofCarboniferous age(Richards,1980; John- ston and Black, 1986). The topaz-bearing rocks are as- sociated with alkali feldspar granites and microgranites. Chemical analyses of the rocks, which include topaz ap- lites, lithium fluorine granite, andtopazite, are discussed in this paper and compared with similar rocks reported from the western U.S.A. and Moneolia. Rrcrout cEor,ocy AND TECToNTc SETTTNG The North Queensland region comprises two major structural elements:(l) the Precambrian shield compris- ing multiply deformed metasedimentary rocks and (2) to the east,mid-Paleozoic sediments of the Hodgkinson Ba- sin, possibly underlain at depth by Precambrian base- ment (De Keyser and Lucas, 1968).These two structural elements are separated by the Palmerville Fault (De Key- ser, 1963). This major crustal discontinuity, which is thought to have developed mainly during the Devonian period, has been the locus for Permo-Carboniferous vol- canism and plutonism including the Coolgarra Batholith The tectonic development of the region has been dis- cussed by Arnold and Fawckner (1980). They consider that before the late Silurian an active magmatic arc oc- cupied the eastern margin of the Precambrian craton. During the late Silurian, the Hodgkinson Basin formed by rifting and crustal extension as subduction gave way to upwelling of hot, low-density material from the man- tle. Sedimentation within the basin beganduring the late Silurian and continued until the late Devonian. Conti- nental margin subduction recommenced between late Devonian and the early Carboniferous,causingthe basin to be uplifted, tectonized, and cratonized. Isotopic studiesby Black and McCulloch (1990) sug- gest that the voluminous felsic magmas of Permo-Car- boniferous age (including the Coolgarra Batholith) were derived from mantle material by a two-stage partial melt- ing process. New crustal material was successively added from below by underplating or emplacement into the lower crust. Remelting of the underplatesand fractional crystallization of the resultant magmasproduced the ob- served felsic volcanics and intrusives. Gnor,ocv oF THE MouNr GrnsoN lnre The topaz-bearing rocks intrude and brecciate tightly folded quartzo-feldspathicsedimentsof the Hodgkinson Formation, a thick sequence of Siluro-Devonian flysch. (Oversby et al., 1980). 0003-o04x/92l0304-o303$02.00 303

Transcript of Topaz-bearing rocks from Mount Gibson, North Queensland ...

Page 1: Topaz-bearing rocks from Mount Gibson, North Queensland ...

American Mineralogist, Volume 77, pages 303-313, 1992

Topaz-bearing rocks from Mount Gibson, North Queensland, Australia

C. JonNsroNDepartment of Geology, James Cook University of North Queensland, Townsville, Queensland, Australia

B. W. Cruppnr,r,Department of Geology, Australian National University, Canberra, A.C.T., Australia

Ansrn-lcr

Dikes, sills, and irregular pods of topaz-bearing aplite, lithium fluorine granite, andtopazite represent F-rich extreme differentiates of a suite of mineralized alkali feldspargranites near Mount Garnet, northeastern Australia. Topaz is a magmatic phase in twoaplites and the topazite and a hydrothermal mineral in the third aplite, lithium fluorinegranite, and the alkali feldspar granite. The granites plot close to the 100 MPa (l kbar)HrO-saturated ternary minimum in the haplogranite system (consistent with their historyof extended fractionation), but the aplites are significantly displaced from that minimum,suggesting either vapor phase losses of qtartz and feldspar components (into pegmatites)or crystallization from hydrous fluid.

Trace elements (e.g., Rb, Ga, Nb) retained in feldspars, micas, and opaque minerals areenriched in the granites and aplites but not in the topazite. Other trace elements arestrongly depleted, having been partitioned into halogen-rich hydrous fluids and lost intothe country rocks. The strong enrichment ofBa, Sr, and Eu in the highly fractionated topazaplites is unusual and is thought to indicate crystallization from or reequilibration withhydrous fluids rather than silicate melts.

Topaz aplites of this area are chemically similar to ongonites of central Asia and centralArizona, U.S.A., and small differences in major element chemistry reflect the differing Fcontent of the provinces. Topazites are similar to those reported from the U.S.A., and thegranites are chemically similar to topaz rhyolites of the western U.S.A.

INrnooucrroN

The topaz-bearing rocks of the Mount Gibson area, 12km east-northeast of Mount Garnet, North Queensland,form part of the Coolgarra Batholith, a mineralized com-posite body ofCarboniferous age (Richards, 1980; John-ston and Black, 1986). The topaz-bearing rocks are as-sociated with alkali feldspar granites and microgranites.Chemical analyses of the rocks, which include topaz ap-lites, lithium fluorine granite, andtopazite, are discussedin this paper and compared with similar rocks reportedfrom the western U.S.A. and Moneolia.

Rrcrout cEor,ocy AND TECToNTc SETTTNGThe North Queensland region comprises two major

structural elements: (l) the Precambrian shield compris-ing multiply deformed metasedimentary rocks and (2) tothe east, mid-Paleozoic sediments of the Hodgkinson Ba-sin, possibly underlain at depth by Precambrian base-ment (De Keyser and Lucas, 1968). These two structuralelements are separated by the Palmerville Fault (De Key-ser, 1963). This major crustal discontinuity, which isthought to have developed mainly during the Devonianperiod, has been the locus for Permo-Carboniferous vol-canism and plutonism including the Coolgarra Batholith

The tectonic development of the region has been dis-cussed by Arnold and Fawckner (1980). They considerthat before the late Silurian an active magmatic arc oc-cupied the eastern margin of the Precambrian craton.During the late Silurian, the Hodgkinson Basin formedby rifting and crustal extension as subduction gave wayto upwelling of hot, low-density material from the man-tle. Sedimentation within the basin began during the lateSilurian and continued until the late Devonian. Conti-nental margin subduction recommenced between lateDevonian and the early Carboniferous, causing the basinto be uplifted, tectonized, and cratonized.

Isotopic studies by Black and McCulloch (1990) sug-gest that the voluminous felsic magmas of Permo-Car-boniferous age (including the Coolgarra Batholith) werederived from mantle material by a two-stage partial melt-ing process. New crustal material was successively addedfrom below by underplating or emplacement into thelower crust. Remelting of the underplates and fractionalcrystallization of the resultant magmas produced the ob-served felsic volcanics and intrusives.

Gnor,ocv oF THE MouNr GrnsoN lnre

The topaz-bearing rocks intrude and brecciate tightlyfolded quartzo-feldspathic sediments of the HodgkinsonFormation, a thick sequence of Siluro-Devonian flysch.(Oversby et al., 1980).

0003-o04x/92l0304-o303$02.00 303

Page 2: Topaz-bearing rocks from Mount Gibson, North Queensland ...

304 JOHNSTON AND CHAPPELL: TOP AZ-BEARING ROCKS

Treue 1. Modes

Sample noRock type

631 632 633 637 638 640Alkali

Potas- Lithium feld-sium Soda fluorine Topa- spar

Aplite aplite aplite granite zite granite

Modal mineralsQuartzMicroclinePlagioclaseBiotiteSericiteFluoriteTopazVugsAccessories

40.9 23.324.1 44.130.4 2s.91 .0 4 .30.1 0.00.0 0.03.5 2.10.0 0.0

<0.1 >o.2

39.0 29jl14.9 33.739.6 32.53.0 4.30.6 0.00.3 0.12.6 0.30.0 0.0

<0.1 <0 .1

76.50.00.00.00.00.0

14.59.0

<0.1

36.546.412.54.60.00.00.00.0

<0.1

f f i " " ' 0 n , " , ' " b i o r i , . e r . . i , e f f i a " r , ' . - o . s . "

IT+lT:"";Ll?; i ; r : : , : : - : i " t ; t : ; "" ' 7," ," ." ,E

r " . s r . i n . d . r k . r r . r d s p a ' s ' . n i t 6 l - _ _ - l " " o " . , . " " " . . , . " , t " "

652 "" ' " ' ' " ' " "" '

Fig. l. Location diagram and interpretative geological mapof the Mount Gibson area.

This is the youngest formation within the HodgkinsonBasin (Fig. l). These sedimentary rocks occur as a pen-dant within the roof zone of a composite granitic batho-lith that, in this area, comprises a medium-grained alkalifeldspar granite (the Pinnacles Granite) and smaller bod-ies of alkali feldspar microgranite, mixed pegmatites, andaplites. These rocks display a close spatial associationwith the topaz-bearing rocks, and studies by Johnston(1984) suggest that the latter are extreme differentiates ofthe Pinnacles Granite or related phases. Dating by John-ston and Black (1986) has shown that the age ofthe Pin-nacles Granite (31 I + 5 Ma) is indistinguishable fromthat of aplite 632 (312 + 4 Ma), which is consistent withthat relationship. Similarly the initial 875r/865r ratios areidentical (0.74 + 0.04). Contact metamorphism is welldeveloped within the roofpendant (Blake, 1972), andbi-otite- and monazite-rich selvages occur immediately ad-jacent to the topaz-bearing intrusions. The area is strong-ly mineralized, and many of the creeks in the area havebeen worked for alluvial cassiterite.

PrNNlcr,ns Gn-c.Nrrn

The Pinnacles Granite is a medium-grained alkali feld-spar granite that outcrops over some l2km2 in the MountGibson area. Principal constituents are quartz and beigepotassium feldspar, with accessory albitic plagioclase andprotolithionite (sample 640, Table l). Traces of hydro-thermal fluorite, topaz, and sericite are present. Other

accessories include ilmenite, a bastniisite group mineral,columbite, thorite, zircon, and sulfides.

In thin section (Fig. 2a) this granite shows considerabledeuteric alteration, viz., sericitization of plagioclase feld-spar, kaolinization (?) of potassium feldspar, chloritiza-tion of micas, replacement of micas by fluorite, devel-opment of swapped albite rims (Smith, 1974; vara dePijpekamp, 1982), and replacement of potassium feld-spar by hydrothermal topaz. Relatively euhedral potas-sium feldspar is characterized by extreme developmentof perthitic albite; clearly its original composition wassodic. The lithium iron mica present is protolithionite(brown, orange-brown, pale yellow, Z > Y > X, respec-tively), which is commonly interstitial and invariablypartly chloritized.

APIITTS, LITHIUM FLUORINE GRANITE'

AND TOPAZITE

These highly evolved rocks are volumetrically very mi-nor components of the Coolgarra Batholith. For example,the aplite (sample 631) occurs as a thin sill (<200 mm)with a pegmatitic selvage 30-40 mm wide on its loweredge. This fine-grained white rock consisting ofquartz +topaz + potassium feldspar + albite + zinnwaldite isporphyritic in both topaz and quartz, with the latter oc-curring as distinctive small (l -mm), round grains. Its tex-ture is saccharoidal due, in part, to the abundant ground-mass plagioclase.

The aplite rich in potassium feldspar (sample 632) wascollected from a thicker sill (750 mm) and is generallyfine grained with some large (5-mm) grains of quartz andtopaz. Modal layering is defined by varying concentra-tions of zinnwaldite (black) and potassium feldspar,qvartz, andtopaz (white), presumably resulting from flow-sorting processes during intrusion. Small pegmatitic

schlieren may be present at the contacts of the aplite andcountry rocks.

The soda aplite (sample 633) is a more heterogeneousrock than samples 631 or 632. \t occurs as a pale, fawn-colored, irregular mass closely associated spatially withsample 632. The texture is granular, with average grainsize I mm. Occasional quartz phenocrysts and aggregatesup to 20 mm and phenocrysts of topaz (10 mm), albite

L E G E N D

Page 3: Topaz-bearing rocks from Mount Gibson, North Queensland ...

JOHNSTON AND CHAPPELL: TOP AZ-BEARING ROCKS

Fig. 2. Photomicrographs. (a) Alkali feldspar granite sample 640; strong development of exsolution and grain boundary albiteclearly shown. (b) Lithium fluorine granite sample 637; euhedral plagioclase (Abn*) is surrounded by interstitial zinnwaldite andperthite-free potassium feldspar (K0. (c) Aplite sample 631; "porphyroblasts" of quartz andtopaz ?resorb early formed feldspars.Note unaltered appearance. (d) Topazite sample 638; euhedral topaz, anhedral qluartz, single lath ofalbite and kaolin-filled vugsclearly shown.

305

" . o

et

(5 mm), and potassium feldspar (15 mm) are also present.This aplite is plagioclase and quartz rich (Table l).

The lithium fluorine granite (sample 637) occurs as asill (>2 m), its upper surface studded with large (50-100-mm) euhedral potassium feldspar crystals. This white,speckled, fine-grained granular intrusive contains approx-imately equal proportions of quartz, potassium feldspar,and plagioclase but little topaz (Table 1). The most dis-tinctive features in hand specimens are the presence of

small (l-mm), rounded quartz phenocrysts and a weakalignment of mica crystals.

The topazite (sample 638) was collected from an irreg-ular, podlike body (<3 m in thickness), plunging north-east at 30'. The material sampled fllls the voids betweengianl quarlz and topaz crystals radiating inward from thefloor and ceiling of what must have been a large cavity.It is a pinkish white, fine-grained, equigranular, friablevuggy rock comprising quartz, accessory topaz, rare albite

t':is;,r,hi;

/s \ .

Page 4: Topaz-bearing rocks from Mount Gibson, North Queensland ...

306

TABLE 2. Mineral chemistry

JOHNSTON AND CHAPPELL: TOPAZ-BEARING ROCKS

632 637Plag Plag

2 7

631 632 637 631 632 637Mica Mica Mica Kf Kf Kf

4 3 5 3 1 8

631Plag

1

Sample no.MineralAnalyses

12.00.04.0

0.00.00.00.00.13.80.0

32.O

16.0

4.0

0.00.00.00.00.43.60.0

sio,Tio,AlrosFeOMnOMgoCaONaroKrO

RbroLirO

Total

siTiAIrrlAlrerAlFeMnMg

NAKclRbLio

39.81 . 1

20.018.00.54.40.10.49.20 .10.80.6

95.0

6.60.1

1 . 42.5z .c0.11 . 10.00 .12.O0.00-10.4

24.0

10.53.82.10.7

39.11 . 1

19.619.80.52.90.00.39.40.1

92.8

6.60 .1

1 . 42.52.80.1o.70.00.12.00.0

24.O

41.7o.4

21.418 .80.5o.20.00.49.70 .10.94.3

98.4

1 . 13 .12.60 .10 .10.00 .12.00.00.12.9

24.0

1 1 . 12.82.21 . 0

64.10 .1

1 7 0

0.00 .10.00.00.3

15.70.0

98.2

65_10.0

18.00.10.00.00.01 . 1

15.30.0

99.6

12.00.03.9

0.00.00.00.00.43.60.0

32.0

16.0

0.49.7

89.94.0

oo.c0.0

19.50.10.00.00.7

10.50.10.0

97.4

1 1 . 90.04.1

0.00.00.00.13.70.00.0

32.0

16.0

3.795.70.63.8

68.30.0

19.80.00.00.10.5

11.20.20.0

100.1

1 1 . 90.04.1

0.00.00.00 .13.80.10.0

32.O

16.0

3.395.6'I .14.0

68.10.0

19.40.00.00.0o.2

11.40.10.0

99.2

't2.00.04.0

0.00.00.00.03.90.00.0

32.O

16.0

1.397.90.83.9

64.70.0

18.20.10.10.10.01 . 2

15.00.0

99.4Formula proportions

6.9 12.00.1 0.0

A I + S iT i + F e + M n + M gC a + N a + KFel(Fe + Mg)AnAbOrTotal cations

o, t,10.7 2.788.5 96.64.O 3.9

10.53.72 .10.8

32.0

16.0

Note.' Mica : zinnwaldite, Kf : potassium teldspar, Plag : albite, dash : not measured or not quoted. Li and Rb measured by AAS on mineralseparates.

laths, muscovite shreds, and accessory cassiterite (Tablel). Pockets of gem topaz are not uncommon. Some band-ed coarse-grained variants are also present. The F contentof topaz from this unit is very high at l9.lo/o (Johnston,1984). This implies a magmatic rather than hydrothermalorigin for the rock (Burnham and Ohmoto, 1980). Wetherefore do not consider it to be a greisen.

In thin section, the aplites, lithium fluorine granite, andIopazite display a number of distinctive features. In con-trast to the Pinnacles Granite, they display only minordeuteric alteration. Principal rock-forming minerals in-clude euhedral, randomly oriented, twinned albite (Abnu,grain size 0.2- I .0 mm) and interstitial potassium feldspar(Orrr-ro) with prominent microcline twinning (Tables Iand 2). Interstitial zinnwaldite (red-brown, orange, verypale yellow with Z > Y >> -l') occurs in variable concen-trations and contains small grains of zircon, monazite,and thorite surrounded by pleochroic halos (Fig. 2b); rarefilm perthite is only present in the lithium fluorine gran-ite, and the lack ofperthitic albite exsolutions in the po-tassium feldspars of the aplites indicates crystallization

under subsolvus conditions, with almost pure end-mem-bers (Ab, Or) crystallizing (see Table 2). Aplite samples631 and 632 contain abundant subhedra ofquartz andtopaz with rounded resorbed inclusions ofalbite and po-tassium feldspar (Fig. 2c). These are thought to have growntoward the end of crystallization in response to risingconcentrations of F. The vuggy nature of the topazite isshown in Figure 2d.

The appearance of topaz as a possible magmatic phasein aplite samples 63 I and 632 indicates a higher F con-tent during crystallization for those aplites than for thePinnacles Granite. Measurements by Congdon and Nash(1988) on a topaz-bearing vitrophyre from west-centralUtah. U.S.A.. indicate that an F content between 2 and3o/o may be required before topaz will crystallize. Also,experimental work by Webster et al. (1987) has shownthat HrO saturation and high F contents in residual meltswill stabilize topaz near the solidus. In contrast, sample633 has small amounts of clear, anhedral hydrothermaltopaz that replaces potassium feldspar and sometimes al-bite.

Page 5: Topaz-bearing rocks from Mount Gibson, North Queensland ...

JOHNSTON AND CHAPPELL: TOPAZ-BEARING ROCKS

TABLE 3. Chemical analyses

Sample 631no

RocktyPe

Aptite

307

S,q,NIpI,o PREPARATIoN AND ANALYTICAL

TECHNIQUES

Material was collected by drilling and blasting. Sampleweights ranged from 5.0 kg for medium-grained samplesto 2.5 kg for fine-grained material. Each sample wascrushed and milled using tungsten carbide faced equip-ment.

Major elements (except NarO, Fe2*, and F) were mea-sured by wavelength dispersive X-ray fluorescence spec-trometry on fused glass discs using the method of Norrishand Chappell (1977). Na,O and LirO were determinedusing flame photometry, and Fe2t was determined by ti-tration against a standard solution of potassium dichro-mate (Peck, 1964). F was determined colorimetrically byautoanalyzer using the method of Fuge (1976). Measure-ment of volatile constituents (CO, and HrO) was doneusing the method of Riley (1958).

Trace elements (except for REE, Cr, Cs, Hl Sb, Sc, Ta)were measured in duplicate by X-ray fluorescence spec-trometry on pressed powder pellets using the method ofNorrish and Chappell (1977). REE, Cr, Cs, Hf, Sb, Sc,Ta were measured (also in duplicate) by instrumentalneutron activation analysis using the method of Chappelland Hergt (1989). Corrections were applied to Ta mea-surements to compensate for Ta contamination from thetungsten carbide mill.

For major elements, values quoted are the mean ofthree determinations (except for F, which was measuredtwice, and HrO and COr, which were measured once).Trace element values quoted are the mean of two mea-surements. If close agreement was not achieved betweenduplicates, the sample was reanalyzed.

All measurements (except LirO and F) were carried outat the Australian National University in Canberra. LirOcontent of the samples was measured in the Geology De-partment laboratories of James Cook University ofNorthQueensland, and F determinations were made at the Da-vies Laboratories of the Commonwealth Scientific andIndustrial Research Organisation (CSIRO) in Townsville.

Electron microprobe analyses of minerals (Table 2) werecarried out at James Cook University of North Queens-land using a Siemans Etec Autoscan scanning electronmicroscope fitted with a Link 290 S(LD spectrometer.Estimates of precision and relative accuracy for this unithave been prepared by Dunham and Wilkinson (1978).

M.c..ron AND TRACE ELEMENT coMposrrroNs

Pinnacles Granite and lithium fluorine granite

Analytical data for these rocks are presented in Table3. The lithium fluorine granite is slightly more aluminousthan the Pinnacles Granite, possibly as a consequence ofits higher F content. In terms of normative quartz, albite,and orthoclase (Table 3), both rocks plot close to theHrO-saturated ternary minimum at 100 MPa (1 kbar)(see Fig. 3).

Abundances of TiOr, FeO, FerO., MnO, MgO, S, PrOr,and CaO are very low relative to the low-Ca granite of

Potas-sium Sodaaplite aplite

Lithium Alkalifluorine feldspargranite Topazite granite

sio,Tio,Al2o3Fe,O"FeOMnOMgoCaONaroK.oProuSH,O*Hro-CO"FRestO = F

Total

74.54 68.60 76.19 76.14<0.01 0.04 0.04 <0.0114.89 17.65 13.26 13.23<0.01 0.06 0.15 0.07

0.09 0.60 0.58 0.78<0.01 0.01 0.01 0.01<0.03 0,08 <0.03 <0.03

0.26 0.26 0.43 0.314.29 4.06 4.87 3.914.51 6.64 3.06 4.550.01 0.01 0.01 <0.01

<0.02 <0.02 <0.02 <0.020.31 0.35 0.47 0.390.06 0.08 0.14 0.07o.24 0.43 0.27 0.500.60 1.o2 0.55 0.600.17 0.33 0.26 0.330.25 0.43 0.23 0.25

99.72 99.79 100.06 100.64Tlace elements

260 171 548220 55 10

1771 865 132250.0 27.5 5.538 15 4321 58 363.0 8.2 13.0

58 97 6182.0 71.0 46.51 5 1 3 57 101 1 5

< 1 < 1 132 20 2243.0 36.8 32.62 1 < 1

<0.1 <0.1 0.2512.0 12.2 24 54.2 6.4 4.52.6 2.8 3.39 4 6

89.52 76.48<0.01 0.03

7.36 12.040.01 0.240.03 0.83

<0.01 0.01<0.03 <0.03<0.01 0.51

0.o2 3.800.07 4.54

<0.01 <0.01<0.02 <o.02

0.39 0.390.05 0.151.43 0.662.26 0.460.03 0.260.95 0.19

100.22 100.21

12 165<5 <512 7542 5 8 . 0

14 835.4 20 5

24 1629.5 55.0

< 1 4<1 186

5 23 5 91 .2 31 .4

<' l 11 .6 <0 .10.4 15.43.0 10.60.3 2.35 2 0

17.0 45.548.0 9210.4 39.5210 1 1 .400.02 0.040.7 12.60.16 2.70

5.00.51 19.40.08 2.90

89.15 36.627.26 0.84o.42 27.120.14 32.120.00 0.00

LiBaQ h

5rPbThUZrNbTa

CuZnGaCrSbCsHfScSn

2725095983.0221 22.0

4912.O22

< l3

40.0'I

<0.17.23.60.9

1 0

48.4 50.023 21

LACeNdSmEuGdTbHoYbLU

K/RbFE

OrAbAn

Rare earth elements4.8 5.9 48.0 23.5

10.8 14.2 115 46.03.5 51 39.5 12.00.67 1.28 10.00 2.80o.27 0.22 0.10 0.03

1 .2 9 .7 2 .10.30 2.o5 0.650.4 4.5 1.2

1.34 3 45 21.8 8.50.22 0.53 3.35 't.34

Ratios39.0 31.1 29.1 28.6

0 8 1 9 8Normalive minerals

32.11 18.95 35.42 35.402s0 3.69 1.75 1.79

20.02 39.92 18.41 27.4536.26 34.34 41.17 33.05

0.00 0.00 0.38 0.00

Note: FE : [Fe3+/(Fer+ + Fe3*)] x 100. Maior elements expressed inpercent, trace elements in ppm. Dash: not measured. Normative minerals(expressed as percentages) were calculated on an F-free basis using FeO/Feros as measured.

Turekian and Wedepohl (1961) for which complete an-alytical data are published. Similarly the ratio Fe3*/(Fe2*+ Fer*) is very low (Table 3). F abundances, ranging from0.46 to 0.600/o, are relatively high, with F resident in dark

Page 6: Topaz-bearing rocks from Mount Gibson, North Queensland ...

308 JOHNSTON AND CHAPPELL: TOPAZ-BEARING ROCKS

200

100

503020

.3 roEc r (o( ) ! t

\ 2.l<

81E

0.5

t"C" nio .leujoru I Ho |

"'ot,

Fig. 3. Normative compositions of samples 631, 632, 633,637,638, and 640 plotted in the system Q-Ab-Or. Ternary min-ima for HrO-saturated melts at 100 MPa (1 kbar) containing 00/oand lolo F also shown (from Manning, l98la).

micas and hydrothermal fluorite. Other volatiles accountfor sl0lo of the totals (see Table 3).

Trace element abundances show strong enrichments ofthe lithophile elements, viz.,Li, Rb, Cs, Th, U, Nb, (Y),Ga, and Hf (Table 3), whereas Ba and Sr are stronglydepleted. REE data are presented in Table 3 and chon-drite normalized patterns in Figure 4.

Aplites

Aplite samples 63 I and 632 are significantly enrichedin AlrO, (Table 3) relative to the other samples. TiOr,FerOr, FeO, MnO, CaO, MgO, and PrO, levels are allvery low, as is the Fe3*/(Fe2* + Fe3*) ratio (cf. Exley andStone, 1964; Hall, l97l). F contents ofthe aplites rangefrom 0.55 to 1.02o/o and are similar to those of lithiumfluorine granites in other provinces (Bailey, 1977).

These rocks can be satisfactorily represented in termsof normative quartz, albite, and orthoclase (Fig. 3). Allsamples are displaced from the ternary minimum com-position (see Fig. 3), with sample 632beingmore potassicand sample 633 being more sodic than the minimum.

Trace element chemistry of the aplites is also presentedin Table 3. Selected REE patterns are shown in Figure 4and Rb, Sr, and Ba trends in Figure 5. Although the database is limited, it appears that Ba, Sr, and Eu are enrichedin the aplites relative to the Pinnacles Granite, whereasREE, Th, U, Y, Zr, Clu, Zn, Ht Sn are significantly de-pleted. Other elements (Rb, Ga, Nb, Ta, +Li) are en-riched.

Aplite sample 631 has lower Pb, Th, U, Zr, Nb, Ta, Y,Cu, Zn, Cs, Li, Hf, and Sc than potassium aplite sample632, which in turn has generally lower trace elementabundances than soda aplite sample 633. The latter isgrouped with the aplites because it shows similar (but

0.3o.2

Fig. 4. REE patterns normalized against chondritic values ofHanson (1978). General concave-upward pattern results fromfractionation of monazite, limiting enrichment of LREE, whileHREE are largely unbuffered.

weak) Ba and Sr enrichment and because it does not havea ternary minimum composition. In other respects, itscomposition is more like the Pinnacles Granite, especial-ly its REE abundances, which are virtually identical tothose of the Pinnacles Granite.

Topazite

Major element composition of sample 638 (topazite) isunusual in that it includes principally SiOr, AlrOr, and F.HrO and CO, are the other major constituents. All traceelements are strongly depleted.

DrscussroN

Pinnacles Granite and lithium fluorine granite

The major element compositions ofboth rocks are con-sistent with protracted fractional crystallization. For ex-ample, highly fractionated rocks should, in a closed sys-tem, approach ternary minimum compositions (Tuttle andBowen, 1958). The position of such minima will dependon the type and amount of the various volatile constitu-ents in the respective melts (Manning, 198 la; Burnhamand Nekvasil, 1986). Samples 637 and 640 display suchbehavior (Fig. 3) and probably represent ternary minimaat varying F levels. Taken at face value, their positionsin Figure 3 suggest that these melts contained less thanlolo F. This is unlikely to have been the case for sample637 (containing topaz) but may be accurate for sample640. Note also that the experimental results, derived un-der HrO-saturated conditions, may not be directly appli-cable to natural systems (Manning, 198 la).

Trace element abundances (e.g., elevated Li and Rb

oo, (uaoo")--

Weight %

Page 7: Topaz-bearing rocks from Mount Gibson, North Queensland ...

JOHNSTON AND CHAPPELL: TOP AZ-BEARING ROCKS 309

and low Ba and Sr) are also consistent with protractedfractional crystallization (Neuman et al., 1954). Also, theREE patterns (concave upward) appear to be fraction-ation generated. LREEs in the Pinnacles Granite (sample640) were buffered at around 100-200 times chondriticvalues (see Fig. 4) by fractionation of monazite as shownby kinking of the patterns between Nd and Sm, as mod-eled by Yurimoto et al. (1990). HREEs, whose solubilitiesin the melt were possibly enhanced by high F content(Collins etaI.,1982 London, 1987), remained unbufferedduring fractionation, giving rise to the observed concave-upward patterns. Deep Eu anomalies are consistent withprotracted fractionation of plagioclase feldspars (Hanson,I 980).

Systematically higher abundances of the rare alkalis inthe lithium fluorine granite (sample 637) and a lower K/Rbratio are consistent with it being more evolved than thePinnacles Granite (sample 640). However, the depletionof Pb, Zn, Cu, Th, U, Zr, Nb, Y, Hl Sn, and all REErelative to sample 640 is not consistent with this hypoth-esis. This, we believe, reflects the divergent crystallizationhistory of the two rocks. The chemistry of sample 640 isconsistent with fractional crystallization of a silicate melt,with a modest component of postsolidification hydro-thermal alteration that has slightly dispersed trace ele-ment chemistry (Johnston and Black, 1986). In contrast,sample 637 has also crystallized from a silicate melt but,we speculate, evolved hydrous fluid toward the end ofcrystallization. Thus, dominantly magmatic trace ele-ment abundances have been overprinted by melt/fluidand crystal/fluid equilibria. Hence the lower abundancesof REE and Nb would be consistent with partitioning ofsuch elements from the remaining melt to the hydrousfluid (Kilinc and Burnham, 1972 Holland, 1972;Picha-vant and Manning, 1984; Webster et al., I 989), which weinfer to have been very chloride rich (Witt, 1988). Ad-ditionally, dissolution of early formed monazite by theexsolved hydrous fluids would also reduce REE, Th, andPrO, values in sample 637. This fluid would ultimatelybe lost into the aureole (cf. Mitropoulos, l98l). Abun-dant monazite in hornfelsed contacts with host rocks isclear evidence that this occurred. The location of lostHREE is less certain, but they may have partitioned intoabundant hydrothermal fluorite. Similarly, partitioningof Zr, Hf, and U into the exsolving hydrous fluid andpossible dissolution of primary zircon by that fluid aresuggested by the common occurrence of zircon as adaughter mineral in fluid inclusions of related granitesand of mineralization (Witt, 1987, 1988). Also,losses ofPb and Zn are consistent with results of Urabe (1985,1987), who found that at low pressures (160 MPa, 1.6kbar) Pb and Zn partitioned in favor of a Cl-rich fluid.

Some Rb and Li would also have been partitioned fromthe remaining melt into the hydrous fluid (Webster et al.,1989), but the Rb already frozen in potassium feldsparswould account for the observed abundances since fluid/crystal partition coefficients for Rb (and Cs) into feldsparsare close to unity (Carron and Lagache, 1980).

. 6 3 1. 633

o Topaz-bear ing rockso Nett le Sui te

o o Go Sam Sui te

froo nt r - t r

o 20 40 60 80 100sr (ppm)

Fig. 5. (a) Plot ofBa vs. Rb for granites and topaz-bearingditrerentiates of the Coolgarra Batholith. Additional data for theNettle and Go Sam suites from Johnston (1984). Samples 631,632, and 633 arc significantly displaced from the fractionationtrends displayed by the batholith as a whole, which suggeststhese abundances are not magmatic. (b) Plot of Sr vs. Rb forgranites and topaz-bearing diferentiates ofthe Coolgarra Batho-lith. Sr displays similar trends to Ba, but its enrichment is sig-nificantly less. Sample 638 (topazite) with only 12 ppn Rb doesnot appear on these plots.

Aplites

We suggest that samples 631 and 632 did not crystal-lize from a silicate melt but more likely from a dense,alkali- and aluminosilicate-rich hydrous fluid (cf. Lon-don, 1986a). The source of this fluid may have been thePinnacles Granite or other small pluton ofthe same suite.The reasons for this view are as follows:

l. The aplites show a marked departure from the lociof HrO-saturated ternary minima at various pressures andF contents. Nonminimum compositions are readily ex-

oo)

olo

x od o

TE

50 100 150 200Ba (ppm)

250 300

oorJ)

x o

E

ooN

. 6 3 7

A

' 631. 633

. Topaz-bearing rockso Nett le Suite

f ; 44 fu , " "_ _o onosam su i te

A O A

o - u - E O O

Page 8: Topaz-bearing rocks from Mount Gibson, North Queensland ...

3 1 0 JOHNSTON AND CHAPPELL: TOP AZ-BEARING ROCKS

plained by transfer of alkalis and silica through the hy-drous fluid to crystallize as either microcline or albiteelsewhere in the system (Burnham and Nekvasil, 1986;Stern et al., 1986). The observed pegmatitic selvages mayhave been produced in this way.

2. It is surprising that samples 631,632, and 633, which,on the basis of their K,/Rb ratios, are more fractionatedthan the Pinnacles Granite, have much higher Ba and Srabundances than the Pinnacles Granite (Figs. 5a and 5b).High Ba, Sr, and Eu abundances (relative to the PinnaclesGranite) cannot be generated in the aplites by melt/crys-tal equilibria involving the fractionation of feldspars ifthe partition coefficients of Arth (197 6) are applicable tothis system.

3. High Ba, Sr, and Eu values could be generated byfluid/crystal equilibria involving feldspars if partition co-efficients ofCarron and Lagache (1980) are applicable tothis system, even though Ba, Sr, and Eu levels in the fluidmay be very low (Nabelek, 1986).

4. The Nb/Ta and Th/U ratios in the aplites are quitedifferent from those observed in the Pinnacles Granite.

5. Depleted REE abundances set samples 631 and 632apart from sample 640 (the Pinnacles Granite) and aplitesample 633. Although REE would partition strongly froma silicate melt into an exsolving chloride-rich fluid (Web-ster et al., 1989), partition coefficients for crystals andfluid calculated for the REE by Nabelek (1986) from thedata ofFlynne and Burnham (1978) and Hanson (1980)show that the REE would not have partitioned from thefluid into coexisting feldspars. If the REE were not pre-cipitated in monazite or zircon, they would ultimately belost into the country rocks (cf. Mitropoulos, 1981).

6. The compositions of the micas in samples 631 and632 are quite different from those in the lithium fluorinegranite (Table 2), with significantly lower t61Al and muchhigher Mg'z* than sample 637 for the same Fe,o, abun-dances. Similarly, the LirO content of zinnwaldite in ap-lite sample 631 is only 0.600/0, whereas in the same min-eral from the lithium fluorine granite (sample 637) it is4.34o/o (Table 2). Micas of the Pinnacles Granite are sim-ilar to those of the lithium fluorine granite but less evolved(Johnston, I984).

7. No evidence for hydrothermal alteration is apparentin samples 631 and 632. Hence subsolidus addition ofBa and Sr and removal of other elements are unlikely.

8. No clear evidence is available to suggest that any ofthe aplites is a cumulate, nor can it be shown that theyare unrelated to the granitic hosts with which they areassociated (Johnston, I 984).

Sample 633 is somewhat different from samples 631and 632 in that its trace element abundances are almostall higher than those of the two latter rocks; this is es-pecially so for the REE. A possible explanation is that,rather like sample 637, sample 633 crystallized predom-inantly from a silicate melt that evolved a fluid at quitea late stage. However, the system may have remainedrelatively closed, and thus high abundances oflithophileelements (particularly the REE) were preserved.

Topazite

Sample 638 (topazite), plotting at the Q apex of the Q-Ab-Or triangle (Fig. a), clearly cannot have crystallizeddirectly from a silicate melt. This and other topazites ofthe Coolgarra Batholith have many characteristics of peg-matites. They appear to be cavity-fiIl material with ear-liest formed minerals comprising giant quafiz and topazcrystals (up to 600 mm). The large crystals probably crys-tallized from a low-viscosity, supercritical aqueous fluid(Jahns and Burnham, 1969; Stern et a1., 1986) with highF content (Glyuk and Anfilogov, 1973). Fine-grained,banded, porous (9olo vugs) aggregates ofquartz and topaz(with a trace of albite) have crystallized around and be-tween the giant crystals. These may have crystallized froma dense halogen- and alkali-rich aqueous fluid, possiblyin contact with a fluid-saturated silicate melt (London,1986a, 1986b). Quartz andtopaz (+ albite, dependingonF content of the system) would precipitate from this fluidand growth of these minerals would continue as Al, Si,and F were transferred through the supercritical fluid fromthe remaining melt to the sites of deposition. High F con-tent would inhibit the crystallization of potassium feld-spar and micas (Kovalenko and Kovalenko, 1976). Thustopazites are essentially hydrothermal (Manning, 198 1b;Birch, 1984), being formed from an aqueous phase thataccompanies a silicate melt. However, F content musthave been high-possibly up to 3o/o-if Glyuk and Anfi-logov's (1973) experimental work is applicable to naturalsystems.

The mineralogy of the topazite is such that there arevirtually no hosts for any trace elements. Relatively highabundances ofLREE, Zr, and Nb suggest the presence ofmonazite, zircon, and columbite, respectively.

ColrpanrsoN wITH orHER PRovTNCES

Ongonites

The Mount Gibson aplites (samples 631 and 632) ap-pear closely similar to ongonites (topaz-bearing quafizkeratophyres) first described from Mongolia by Kovalen-ko et al. (1971) and more recently by Kortemeier andBurt (1988) from central Arizona, U.S.A. Ongonites, thesubvolcanic analogues of lithium fluorine granites, arethought to have crystallized from highly fractionated meltscontaining up to 3.50/o F (Bailey, 1977).

The Mount Gibson aplites generally have lower F con-tents (0.55-l .02o/o) than the Mongolian ongonites (0.8-3.20lo) (Kovalenko, 1973), although their ranges overlap.For this reason (Manning, 1981a), the Mount Gibson ap-lites contain more SiO, and KrO and less AlrO, and NarOthan the "type" ongonites. Both suites of rocks, beinghighly fractionated, have very low abundances of TiOr,FeO, FerOr, MgO, MnO, CaO, and PrO'. They are bothperaluminous and highly reduced. Compared to the on-gonitic Dysart dike of Kortemeier and Burt (1988), theMount Gibson rocks are lower in SiO, and AlrO. andmuch lower in F. Fe,", is similar, but the Dysart dike ismuch more oxidized than the Mount Gibson rocks. NarO

Page 9: Topaz-bearing rocks from Mount Gibson, North Queensland ...

JOHNSTON AND CHAPPELL: TOP AZ-BEARING ROCKS 3 l l

values are similar in both suites, but sample 632 containsan order of magnitude more KrO than the Dysart ongon-rte.

With some exceptions, trace element abundances aresimilar in both the Mount Gibson and Mongolian rocks.Comparisons with data compiled by Christiansen et al.(1983) indicate that sample 631 has abundances of Li,Zr, Nb, Zn, Cs, Hf, Sn that fall below the range of theMongolian ongonites, whereas sample 632 displays lowerCs, Hl Sn, and Zr but higher Li than the Mongolianongonites. Comparison with the Dysart dike of Korte-meier and Burt (1988) indicates that the Mount Gibsonrocks have lower Li, Ba, Sr, Nb, Hl and Sn and higherRb, Th, Zr, Ta, and Y, whereas U is similar in bothsuites. The behavior of REE in Mongolian ongonites andthe Mount Gibson aplites appears to be closely similar,but no data are available for the Dvsart dike.

Topazites

Topazites are closely related to ongonites (Kortemeierand Burt. 1988). These authors noted the transition fromtopazite to ongonite in the Breadpan and Dysart dikesand suggested that such changes may have reflectedchanges in HF concentrations and timing and degree ofloss of volatiles to the country rock. This transition hasnot been observed at Mount Gibson. However, topaziteoutcrops in close proximity to samples 631 and 632. TheMount Gibson Iopazite has significantly higher SiO, (andpossibly F) and lower AlrO. than any of the Breadpansamples. The trace element levels in sample 638 are sig-nificantly lower than the Breadpan samples (except Baand Y). This suggests a more complete loss of virtuallyall elements into a vapor phase than occurred duringcrystallization of the Breadpan samples.

There are several other occurrences of these rock typesrecorded in the literature (e.g., Eadington and Nashar,1978; Manning, l98la; Birch, 1984; Kleeman, 1985), andthey are generally similar to the Mount Gibson rocks,although they vary in detail.

Topaz rhyolites

Soda aplite (sample 633), lithium fluorine granite (sam-ple 637), and the Pinnacles Granite (sample 640) showsimilarities in chemistry with the topaz rhyolites of thewestern U.S.A. (Burt et al., 19821, Christiansen et al.,1983). Thus, although ongonites andtopaz rhyolites haveoverlapping major and trace element abundance rangesand appear to be regarded by Christiansen et al. (1983)as identical, we treat them separately here to emphasizeour view that ongonites are generally more evolved thantopaz rhyolites.

The major and trace element compositions of samples633. 637. and 640 are similar but not identical to the"average topaz rhyolite" ofChristiansen et al. (1983). Forexample, SiOr, AlrO3, and F abundances for these sam-ples often lie above the range of the average, whereasmost other major elements lie below it. Also, trace ele-ment values mostly lie within the range of the "average

topaz rhyolite," but some (e.g., Li, Rb) are higher andothers (e.g., Sn) are lower. REE patterns for topaz rhyo-lites, e.g., from Spor Mountain and Thomas Range, Utah,U.S.A. (Christiansen et al., 1984), are similar to that ofsample 640 (see Fig. a).

Other examples of topaz rhyolites have been more re-cently reported by Rubin et al. (1987) and Congdon andNash ( 1 988). Once again, since they are enriched in manyincompatible elements, these rhyolites show similaritieswith the Mount Gibson rocks. However, the observedtrace element abundances are the end result of severalprocesses including partial melting of a variety of sourcerocks, fractionation, fluid loss, and postemplacement al-teration. These in turn are controlled by other factorsincluding temperature, pressure, partition coefrcients, andabundance of volatiles such as HrO, COr, Cl, and F. It istherefore not surprising that the ranges oftrace elementsare quite wide.

AcxNowr.nncMENTS

We thank A.J R. White, D. Wyborn, P.J. Stephenson, and C. Cuff for

their constructive comments on the manuscript. Thoughtful reviews by

E.H. Christiansen, S. Ludington, and C.R. Bacon have significantly im-

proved the text. C.J acknowledges financial support of the Common-

wealth ofAustralia in the form ofa Postgraduate Award and a National

Research Fellowship.

RnrnnBNrcns crrED

Amold, G.O, and Fawckner, J.F. (1980) The Broken River and Hodg-kinson Province. In R.A. Henderson and P.J. Stephenson, Eds., Thegeology and geophysics ofnortheastern Australia, p' 175-189' Geolog-ical Society of Australia (Queensland Division), Brisbane.

Arth, J.G. (1976) Behaviour of trace elements during magmatic process-

es-a summary of theoretical models and their applications. Journal ofResearch ofthe U.S. Geological Survey, 4,41-47.

Bailey, J.C. (1977) Fluorine in granitic rocks and melts: A review. Chem-ical Geology, 19, l-42.

Birch, W.D. (1984) Quartz{opazJoellingite rocks near Eldorado, Victo-ria. Australian Journal of Earth Sciences, 3l, 269-278.

Black, L.P., and McCulloch, M.T. (1990) Isotopic evidence of the depen-dence of recurrent felsic magmatism on new crust formation: An ex-

ample from the Georgetown region of northeastern Australia. Geo-chimica et Cosmochimica Acr.a,54, 183-196.

Blake, D. (1972) Regional and economic geology ofthe Herberton/MountGarnet area-Herberton Tinfield, North Queensland. Bweau of Min-

eral Resources Geology and Geophysics, Australia, Bulletin 124.

Burnham, W.C., and Nekvasil, H. (1986) Equilibrium properties of gan-

ite pegmatite magmas. American Mineralogist, 7 1, 239-263.Burnham. W.C., and Ohmoto, H. (1980) Iste slage processes of felsic

magmatism. Mining Geology, 8, l-l l.Burt, D.M., Sheridan, M.F., Bikun, J.V., and Christiansen, E.H. (1982)

Topaz rhyolites-distribution, origin, and significance for exploration.Economic Geology, 77, l8l8-1836.

Carron, J.P., and lagache, M. (1980) Etude exp6rimentale du fractionne-ment des 6l6ments Rb, Cs, Sr et Ba entre feldspaths alcalins, solutionshydrothermales et liquides silicat6s dans le systeme Q-Ab-Or-HrO d 2

kbar entre 700 et 800 .c Bullerin de Min6ralogie, 103, 571-578.Chappell, B.W., and Hergt, J.M (1989) The use of known iron content

as a flux monitor in neutron activation analysis' Chemical Geology,78 , l 5 l - 157 .

Christiansen, E.H., Burt, D.M., Sheridan, M.F., and Wilson, R'T. (1983)

The petrogenesis of topaz rhyolites from the western United States.

Contributions to Mineralogy and Petrology, 83, l6-30.Christiansen, E.H., Bikun, J.V., Sheridan, M.F., and Burt, D.M. (1984)

Page 10: Topaz-bearing rocks from Mount Gibson, North Queensland ...

312 JOHNSTON AND CHAPPELL: TOP AZ-BEARING ROCKS

Geochemical evolution oftopaz rhyolites from the Thomas Range andSpor Mountain, Utah. American Mineralogist, 69,223-236.

Collins, W.T, Beams, S.D., White, A.J.R., and Chappell, B.W. (1982)Nature and origin ofA-t1pe granites with particular reference to south-eastern Australia. Contributions to Mineralogy and Petrology, 80, I 89-200.

Congdon, R.D., and Nash, W.P. (1988) High-fluorine rhyolite: An erup-tive pegrnatite magma at the Honeycomb Hills, Utah. Geology, 16,l 0 l 8 -1021 .

De Keyser, F. (1963) The Palmewille Fault-a'fundamental' structure innorth Queensland. Journal ofthe Geological Society ofAustralia, 10,273-278.

De Keyser, F., and Lucas, K.G. (1968) Geology of the Hodgkinson andIaura Basins, North Queensland. Bureau of Mineral Resources, Ge-ology and Geophysics, Bulletin 84, 254 p.

Dunham, A.C., and Wilkinson, F.C.P. (1978) Accuracy and detectionlimits of energy dispersive electron microprobes. Analysis of silicates.X-ray Spectrometry, 7, 50-56

Eadington, P.J., and Nashar, B. (1978) Evidence for the magmatic originof quartz-topaz rocks from the New England Batholith, Australia. Con-tributions to Mineralogy and Petrology, 67, 433-438.

Exley, C.S., and Stone, M. (l 964) The granitic rocks ofsouthwest England.In K.F.G Hoskings and G.S. Shrimpton, Eds., Present views of someaspects ofthe geology ofCornwall and Devon, p. I 3 l- I 84. Royal Geo-logical Society, Cornwall.

Flynne, R T , and Bumham, W.C. (1978) An experimental determinationofrare earth partition coemcients between a chloride containing vapourphase and silicate melts. Geochimica et Cosmochimica Acta.42.685-701 .

Fuge, R. (1976) The automated colourimetric determination of fluorineand chlorine in geological samples. Chemical Geology, 17, 37 -43.

Glyuk, D.S., and Anfilogov, V.N. (1973) Phase equilibria in the systemgxanite-Hro-HF at a pressure of 1000 Kg/cmr. Geochemistry Inter-national, lO,32l-325 (rranslated from Geokhimiya, no. 3, 434-438,r973).

Hall, A. (1971) Greisenization in the granite of Cligea Head, Cornwall.Proceedings of the Geological Association, 82, 209-230.

Hanson, G.N. (l 978) The application oftrace elements to the petrogenesisofigneous rocks ofgranitic composition. Earth and Planetary Sciencektters. 38. 26-42.

-(1980) Rare earth elements in petrogenetic studies of igneous sys-tems. Annual Reviews of Earth and Planetary Sciences, 8, 37 l-406.

Holland, H.D. (1972) Granites, solurions and base metal deposits. Eco-nomic Geology, 67, 281-301.

Jahns, R.H., and Bumham, C.W. (1969) Experimental studies of peg-matite genesis: l. A model for the derivation and crystallization ofgranitic pegmatites. Economic Geology, 64, 843-864.

Johnston, C. (1984) Granitoids ofthe Coolgarra Batholirh, Mount Garnet,Nonh Queensland,,454 p. Unpublished Ph.D. thesis, James Cook Uni-versity of North Queensland, Townsville, Queensland, Australia.

Johnston, C., and Black, L.P. (1986) Rb-Sr systematics ofthe CoolgarraBatholith, North Queensland. Australian Journal ofEarth Sciences, 33,309-324.

Kilinc, LA, and Burnham, C.W. (1972) Partitioning of chloride betweena silicate melt and coexisting aqueous phase from 2 to 8 kilobars. Eco-nomic Geology, 67, 231-235.

Kleeman, J.D. (1985) Origin of disseminated Wolframite-bearing quartz-topaz rock at Torrington, New South Wales, Australia. High heat pro-duction (HHP) granites, hydrothermal circulation and ore genesis, p.197-2Ol.Institution of Mining and Metallurgy, London.

Kortemeier, W T., and Burt, D.M. (1988) Ongonite and topazite dikes inthe Flying W ranch area, Tonto basin, Arizona. American Mineralogist,73 ,507 -s23 .

Kovalenko, V.L (l 973) Distribution of fluorine in a topaz-bearing quarrzkeratophlre dike (ongonite) and solubility offluorine in granitic melts.Geochemical Institute. 10. 4l-49.

Kovalenko, V.I., and Kovalenko, N.I. (1976) Ongonites (topaz-bearingquartz keratophyre)-Subvolcanic analogue of rare metal Li-F ganites.Nauka Press, Moscow (in Russian).

Kovalenko, V.I., Kuz'min, M.I., Antipin, V.S. and Petrov, L.L. (1971)Topaz-bearing quartz keratophyre (ongonite), a new variety of subvol-

canic igneous vein rock. Doklady of the Academy of Sciences of theUSSR, Earth Science Sections, 199, 132-135 (translated from DokladyAkademii Nauk SSSR, 199, no. 2, 430-433, 197 l).

London, D. (1986a) Magmatic hydrothermal transition in the Tanco rare-element pegmatite: Evidence from liquid inclusions and phase-equilib-rium experiments. American Mineralogist, 7 1, 37 6-39 5.

-(1986b) Formation of tourmaline-rich gem pockets in miaroliticpegmatites. American Mineralogist, 7 l, 396-405.

-(1987) Internal differentiation of rare-element pegmatites: Efectsofboron, phosphorous and fluorine. Geochimica et Cosmochirnica Acta,51,403-420.

Manning, D.A.C. (l98la) The effect of fluorine on liquidus phase rela-tionship in the system Qz-Ab-Or with excess water at I kb. Contri-butions to Mineralogy and Petrology, 76,205-215.

-(l98lb) The application of experimental studies in determiningthe origin of topaz-quartz-tourmaline rocks and tourrnaline-quartz rocks.Annual Conference ofthe Ussher Society, 1981,121-127.

Mitropoulos, P. (1981) REE patterns of the metasedimentry rocks of theIand's End gmnite aureole (southwest England). Chemical Geology,35,26s-280.

Nabelek, P.I. (1986) Trace element modelling ofthe petrogenesis ofgran-ophyres and aplites in the Notch Peak granitic stock, Utah. AmericanMineralogist, 7 l, 460-47 1.

Neuman, H., Mead, J., and Vitaliano, C.J. (1954) Trace element variationduring fractional crystallisation as calculated from the distribution law.Geochimica et Cosmochimica Acta. 6. 90-99.

Norrish, K, and Chappell, B.W. (1977) X-ray fluorescence spectrometry.In J. Zussman, Ed., Physical methods in determinative mineralogy, p.2Ol-272. Academic Press, New York.

Oversby, B S., Black, L.P., and Sheraton, J.W (1980) Late Palaeozoiccontinental volcanism in northeastern Australia. In R.A. Hendersonand P.J Stephenson, Eds. , The geology and geophysics of northeasternAustralia, p. 247 -268. Geological Society of Australia (Queensland Di-vision), Brisbane.

Peck, L.C. (1964) Systematic analysis ofsilicates. U S. Geological SurveyBulletin 1170.

Pichavant, M., and Manning, D.A.C. (1984) Petrogenesis of tourmalineganites and topaz granites: The contributions of expenmental data.Physics ofthe Earth and Planetary Interiors, 35, 3l-50.

Richards, D.N.G. (l 980) Palaeozoic granitoids of northeastem AustraliaIn R.A. Henderson and P J. Stephenson, Eds , The geology and geo-physics of northeastern Australia, p. 229-246. Geological Society ofAustralia (Queensland Division), Brisbane.

Riley, J.P. (1958) Simultaneous determination of water and carbon di-oxide in rocks and minerals. Analyst, 83,42-49.

Rubin, J.N., Price, J.G., Henry, C.D., and Koppenaal, D.W. (1987) Cry-olite-bearing and rare metal-enriched rhyolite, Sierra Blanca Peaks,Hudspeth County, Texas. American Mineralogist, 72, ll22-ll3}.

Smith, J.Y. (1974) Feldspar minerals, vol. 2. Springer-Verlag, Berlin.Stern, L.A., Brown, G.E., Jr., Bird, D.K., Jahns, R.H., Foord, E.E., Shig-

ley, J.E., and Spaulding, L.B., Jr. (1986) Mineralogical and geochemicalevolution of the Little Three pegmatite-aplite layered intrusive, Ra-mona, California. American Mineralogist, 7 1, 406-427.

Turekian, K.K., and Wedepohl, K.H. (1961) Distribution of the elementsin some major units of the Earth's crust. Geological Society of AmericaBulletin 72, 175-192.

Tuttle, O.F., and Bowen, N.L. (1958) Origin of granite in the light ofexperimental studies in the system NaAlSi,Or-KAlSi3Os-SiOr-HrO.Geological Society of America Memoir 74, 153 p.

Urabe, T. (1985) Aluminous granite as a source magma of hydrothermalore deposits: An experimental study. Economic Geology, 80, I 48- I 57.

- ( I 987) The effect of pressure on the partitioning ratios of lead andzinc between vapour and rhyolite melts. Economic Geology,82, 1049-1052.

van de Pijpekamp, B. (1982) Petrological criteria for establishing the tinpotential of granitoid complexes In A.M. Evans, Ed., Metallizationassociated with acid magmatism, p. 279-290. Wiley Interscience, NewYork.

Webster, J.D., Holloway, J R, and Hervig, R.L. (1987) Phase equilibriaofa Be, U and F-enriched vitrophyre from Spor Mountain, Utah. Geo-chimica et Cosmochimica Acta. 51. 389-402.

Page 11: Topaz-bearing rocks from Mount Gibson, North Queensland ...

Webster, J.D., Holloway, J.R., and Hervig, R.L. (1989) Partitioning oflithophiletrace elehents between HrO and HrO + CO, fluids and topazftyolite melt. Economic Geology, 84, 116-134.

Witt, W.K. (1987) Fracture-controlled feldspathic alteration in granitesassociated with tin mineralization in the Irvinebank-Emuford area,northeast Queensland. Australian Joumal ofEarth Sciences, 34,447-462.

-(1988) Evolution of high temperature hydrothermal fluids associ-atod with geisenization and feldspathiJ alteration of a tin-mineralizedganite, northeast Queensland. Economic Geology, 83, 310-334.

3 1 3

Yurimoto, H., Duke, E.F., Papike, J.J., and Shearer, C.K. (1990) Arediscontinuous chondrite-normalized REE paiterns in pegmatitic gran-ite systems the results of monazite fractionation? Geochimica et Cos-mochimica Acta, 54, 2141-21 45.

Mercuscrrpr R.EcErvED Mmcrr 16, 1990MaNuscnrvr AccEprED Oqrone* 25, 1991

JOHNSTON AND CHAPPELL: T OP AZ.BEARING ROCKS