Your Federal Quarterly Tax Payments are due April 15th Get Help Now >>

Geochemical evolution of topaz rhyolites by ezw15872


									                                 American Mineralogist, Volume 69, pages 223J36, 1984

                         evolution of topaz rhyolites from the ThomasRange
                               and Spor Mountain, Utah
   Eruc H. CnnrsrrausBNl. Jnuns V. BrruN2. MrcHRnr F. SspnroaN AND Doxnro M. Bunr
                             Department of Geology, Arizona State University
                                         Tempe, Arizona 85287


                Two sequences   ofrhyolite lava flows and associatedpyroclastic depositsare exposedin
             the Thomas Range(6 m.y.) and at Spor Mountain (21 m.y.) in west-centralUtah. Both
             contain topaz indicative of their F-enrichment (>0.2%) and aluminous nature. The
             rhyolites are part of the bimodal sequenceof basalt and rhyolite typical of the region.
             Moderate changes in major elements coupled with large variations in trace elements in
             vitrophyres from the Thomas Rangeare generallyconsistentwith fractionation of observed
             phenocrysts. Especially important roles for the trace minerals are suggested.
             disagreementbetween major and trace element models regardingthe degreeof crystalliza-
             tion as well as improbably high Dr" and Drh suggestthat some diffusive differentiation
             involving the migration of trace elements complexed with volatiles also occurred or that
             monazite, titanite or samarskitefractionation was important. Higher concentrationsof F in
             the evolving Spor Mountain rhyolite drove residual melts to less silicic compositions with
             higher Na and Al and promoted extended differentiation yielding rhyolites extremely
             enriched in Be, Rb, Cs, U, Th, and other lithophile elementsat moderate SiO2concentra-
             tions 04Vo\.

                     Introduction                          1979).The lavas and tufs were emplaced6 to 7 m.y. ago
  During the Late Cenozoic much of the western United      from at least 12 eruptive centers (Lindsey, 1979) and
                                                                       to form a partially dissected plateau with an
States experiencederuptions of bimodal suites of basalt coalesced
and rhyolite (Christiansenand Lipman, 1972)consequent      area in excess of 150 km2. Turley and Na-sh(1980)have
                                                                       their volume to be about 50 km'.
to the developmentof the Basin and Rangeprovince, the estimated
                                                              The topaz rhyolite of the Spor Mountain Formation has
SnakeRiver Plain, and the Rio Grande rift. These rocks
                                                               isotopic age of 21.3-10.2   m.y. (Lindsey, 1982).This
are especially characteristic of the Great Basin where an
                                                           rhyolite and its subjacentmineralizedtuff(the "beryllium
bimodal rhyolitic and mafic volcanism commencedabout
                                                                          in isolated and tilted fault blocks on the
22 m.y. ago (e.g., Best et al., 1980).Those rhyolites tuff') occur
which are rich in fluorine (up to 1.5 wI.%) and lithophile southwest side of Spor Mountain and as remnants of
                                                           domes and lava flows on the east side of the mountain.
elements(Be, Li, U, Rb, Mo, Sn, and others)appearto
                                                                      (1982) suggeststhat at least three vents were
be part of a distinct assemblage topaz-bearingrhyolite Lindsey
lavas that occurs across much of the western United        active during the emplacement of the rhyolite of Spor
                                                                                  episodesfor both the Spor Mountain
States and Mexico. The general characteristics of topaz Mountain. Eruptive
rhyolites have been discussed by Christiansen et al.       and Thomas Range sequencescommencedwith the em-
(1980,1983a,b).                                            placement of a series of pyroclastic breccia, flow, surge,
                  Perhapsthe best known are those from
                                                                                    were terminated by the effusion of
the area near Topaz Mountain in west-central Utah (Fig. and air-fall units and
 1).                                                       rhyolite lavas (Bikun, 1980).
   Two age groups of topaz rhyolite lava flows and their      This study documents the petrologic and geochemical
associated tuffs occur in this region, one forming the     relationshipsofthese lavas. Particular attention is paid to
                                                                                 in the evolution ofthese rocks and to
ThomasRangeand the other peripheral to Spor Mountain the role offluorine
(Fig. l). The rhyolites of the Thomas Range have been      the origin of their extreme enrichment in incompatible
                                                                                           that most of the variation is
mappedas the Topaz Mountain Rhyolite (Lindsey, 1982. trace elements.We conclude
                                                           the result of the fractionation of the observed pheno-
                                                            crysts; "thermogravitational diffusion" appears to have
  I Present address: Department of Geology, University of  played only a small role.
Iowa, Iowa City,lowa 52242.                                   The petrology of these rhyolites was initially studied as
  2 Presentaddress:Shell Oil Co., P.O. Box 2906, Houston,   part of an investigation of uranium deposits associated
Texas77001.                                                 with fluroine-rich volcanic rocks (Burt and Sheridan,
0003-004)vE4l0304-0223$02.                               223
224                                     CHRISTIANSEN ET AL.: TOPAZ RHYOLITES

  Fig. l. Generalizedgeologicmap of the southernThomas Range,Utah (after Lindsey, 1979,1982;        Staatzand Carr,1964). Sample
localities shown by dots (surface)and crosses(drill core). Samplenumbers correspond to those in Tables I and2. Samples6la, b,
and c were collected from a site north ofthe area shown and sample6-358and 18-ll3 are from a drill core obtainedfrom a sitejust to
the west of the map area (Morrison, 1980).

1981).The rhyolite of Spor Mountain is underlain by a             the tables. Other trace elementswere analyzedby instru-
cogenetic tuf which is mineralized with Be, F, and U              mental neutron activation at the University of Oregonand
(Lindsey, 1977,1982;     Bikun, 1980)
                                    and is associated  with       are considered to be precise to better than +5Vo except
sedimentaryuranium depositsand small fluorite deposits.           for Cr and Ev (+lUVo\.
Presumablythe lavas or their intrusive equivalents were
the source of the ores. In contrast, the Topaz Mountain                                    Mineralogy
Rhyolite is not associatedwith economic surfacemineral-              The mineralogy of the two sequencesof rhyolites are
ization, but is similar in many respectsto the older lavas.       grossly similar. The Topaz Mountain Rhyolite generally
There is also increasing evidence that such fluorine-rich         contains sanidine, qtrartz, sodic plagioclase biotite t Fe-
volcanic rocks are intimately associatedwith subvolcanic          rich hornblender-titanite. Biotite compositions show in-
porphyry molybdenumdeposits(Burt et al.,1982).                    creasing Fe/(Fe + Mg) with differentiation (Turley and
                                                                  Nash, 1980).Two-feldspar and Fe-Ti oxide temperatures
                   Analytical methods                             correlate well and rangefrom 630'to about 850"C(Turley
                                                                  and Nash, 1980;   Bikun, 1980).Augite is found in one lava
   The samplesanalyzedin this study were selectedfrom             flow (SM-61)with a high equilibration temperature.Spes-
a suite of more than 100samplescollected from outcrops            sartine-almandinegarnet occurs in some lavas, generally
in addition to drill-core samplesobtained by Bendix Field         as a product of vapor-phasecrystallization (Christiansen
Engineering Corporation (Morrison, 1980). Samples se-             et al., 1980).Titanomagnetite is present as micropheno-
lected for chemical analysis were chosen to representthe          crysts in most samples,but ilmenite is rare.
spectrum of fresh rock compositions on the basis of                  The rhyolites of the Spor Mountain Formation contain
petrographic examination.                                         sanidine, quartz, sodic plagioclase, and biotite. Biotites
   Wavelengthdispersive XRF was used to determine Si,             are extremely Fe-rich (Christiansen et al., 1980) and
Ti, Al, Fe, Mn, Ca, K and P. Analyseswere performed               suggestcrystallization near the QFM oxygen buffer. Two-
using glass discs (Norrish and Hutton, 1969). Energy              feldspartemperaturescluster around 680"to 690'at I bar.
dispersiveXRF with undiluted rock powders was used to             Magnetite is more abundant than ilmenite.
determine Rb, Sr, Y, Zr, Nb and Ga. Atomic absorption                Magmatic accessory phases in both rhyolites include
spectrometrywas used to analyzefor Na, Mg, Be, Li and             apatite, fluorite, zircon, and allanite. Topaz, sanidine,
Sn. Natural rock powders were used as standardsin each            quartz, Fe-Mn-Ti oxides, cassiterite, garnet, and beryl
case. Fluorine was analyzed using ion-selective elec-             occur along fractures, in cavities and within the ground-
trodes and a few sampleswere also analyzedfor F and CI            mass of the rhyolites. They are the result of crystalliza-
using an ion chromatographas describedby Evans et al.             tion from a vapor releasedfrom the lavas during cooling
(1981). Estimates of analytical precision are included in         and devitrification.
                                                      CHRISTIANSEN ET AL.: TOPAZ RHYOLITES                                                                             225

                               Rock chemistry                                           older rhyolites from the Spor Mountain Formation.
                                                                                           Whole-rock compositions are plotted in terms of nor-
 Major element composition                                                              mative quartz-albite-orthoclasein Figure 2 for compari-
    Chemical analyses of vitrophyres and fresh felsites                                 son with experimental work in the same system (with
 from both volcanic sequencesare presented in Tables I                                  H2O or with HzO-F). Normative femic constituents are
 and 2, along with ctpw normative corundum or diopside                                  generallyless than 2Voof the total, so when contoured in
 as calculated on a water- and fluorine-free basis (Fe2*/                               terms of An content, this systems is probably a reason-
 (Fe total) = O.52;Bikun, 1980).All of the sampleshave                                  able model for the magmas under consideration. Al-
 high Si, Fe/Mg, and alkalies and low Ti, Mg, Ca, and P,                                though the vitrophyres from the Spor Mountain Forma-
 features shared with all topaz rhyolites and typical of                                tion show considerable scatter, there is no overlap with
 rhyolites from bimodal (mafic-silicic) associationsfrom                                those from the Thomas Range. Both suites lie on the
 around the world (Ewart, 1979; Christiansen et al.,                                    feldspathic side of the hydrous minima. Only sampleSM-
  1983a).As expected, both suites are generally high in                                 35, which displays an anomalouslyhigh calcium content,
 fluorine (greater than O.2Vo).   Only two vitrophyres, one                             lies in the primary phase field of quartz on this diagram.
 from each sequenceof lavas, are peraluminous (corun-                                   With decreasingCa, Mg, Ti, and increasing F, the sam-
 dum normative). Devitrification of the vitrophyres may                                 ples from the Thomas Rangeplot nearer the fluid-saturat-
 result in compositions with normative corundum, for                                    ed minimum. We interpret this trend (also noted by
 examplecompare SM-61a,b, and c or samplesSM-34 and                                     Hildreth, 1977)as a reflection of lower temperaturesand
 35 which were collected from different "facies" of the                                 the approachto fluid-saturationofan initially unsaturated
 same lava flows. In both cases the intensely devitrified                               magmacrystallizing at about I to 1.5 kbar total pressure.
 samples are corundum normative. This is apparently                                     The displacementof the Spor Mountain rhyolites toward
 related to the loss of an alkali-bearingvapor-phaseduring                              the Ab apex is probably the result of their higher fluorine
 their subaerialcrystallization (cf. Lipman et al., 1969).                              content. Manning (1981) has shown that increasingF
    Vitrophyres from the two rhyolites can be distin-                                   concentration shifts the fluid-saturated minimum at I
 guishedby their major elementgeochemistry. In general,                                 kbar to increasingly albitic compositions (Fig. 2). A
 the rhyolites from the Thomas Range have greater con-                                  poorly constrained path defined by residual melts in a
 centrations of Si, Mg, and Ti, and less Al and F than the                              cooling granite-HF-HzO system shows this same trend

                                      Table l. Whole-rock chemical analysesof rhyolites from Spor Mountain, Utah

Lithology*             VFVFIFMV                                                                                VF                     FF
                                                                                       67         70          7t        6-358      18-113   26-Ll2        ****
Sample No,            31     34            35        37       41          63

                                                                       veight      per cenE
s10?                 7 3. 3    14.8       73,9       7 5. 3   74.5       73.7         65,4       72.2       73.7        74.O       74.4     73.5            1
Ti n-
"-t                   0.06      0.06       0.06       0.07     0.05       0.04         0.94       0.04       0.06        0.06       0.06     0.03           3
A1201                L3.2      1 3 .3     1 3 .t     12.7     1 4. 4     1 3. 8       r4 .6      1 3. 9     13.3        13.5       r3.4     r4,0            2
                      1.48      1.52        1. 4 3    I .01    0.81       1.17         6.35       t. 19      1.33        1.63       1.48     r. r.9         5
Mnfl                  0.07      0.04       0.06       0.08     0.02       0.05         0.09       0.06       0.05        0.07       0.06     0.06           5
                      0.05      0.07       0.08       0.07     0.08       0.21         0.36       o .06      0.07        0.23       0.11     0.14          10
Ca0                   0.61      0.60        L.27      0.70     0.14       0.45         2.87       0.58       0. 70       0.43       0,59     0.63           4
Naro                  4.O2      3.'16      4.33       3.85     2.63       3.83         3.r5       4.90       4.39        3.92       4,0r     3,94           2
Kr0                   4.99      4.81       3.65       4.86     5.95       4.9L         4,95       4.98       4.74        4.88       4.83     4.85           2
P;o-                  0.00      0.00       0.00       0.00     0.00       0.00         0.00       0.01       0.01        0.00       0.00     0.00          20
Total                97.78     98,96      97.88      98,64    98.58      9 8 .1 6     98.7r      9 7. 9 0   98.35       98.72      98.94    98.34
LOI**                 2.96      0.50       3.55       2.51     r.43       0.91         2,30       3.00        2.98       0.77       0.57        0,70       10

                                                                       parts per million
Rb                    910      1000       1060        605     1600       1400          576                   1000        860       1030     1450            4
ST                      510                  55                  5         20          3L5                     30         40         10       15           I0
Y                      92       135        135         46       13         56          145                    122         94        140       77            6
Zr                     99       r20        120        120       85         80          360                    110        105        r2O       87            6
Nb                     90       135        r20         73      r30        L20           45                    114        105        130      130            4
                       50        50        n.d.        40       70         65           30                     45         35         75       70           15

                                                                CIPI'i nornative     mlncrals   ***

c                     0.07      0.81                           3.40       1.40                                 -         r.01       o.47        1.11
di                                         o.64                                        1.80                   0. 6 6
*           V= vitrophyre,     F= felsite,      T= tuff,  s= spherulitic   vitrcphyre'
**          LOI= Loss on ignitlon     at 900'C for four hours-
***         calculated   from volat11e     - free analyses.
****        EstiEted    precision  based on replicate     analyses.    ReporEed as one standard        deviation    expressed   as Percentage    of   anount     rePorted.
n.d.=       not deternlned.
226                                                  CHRISTIANSEN ET AL.: TOPAZ RHYOLITES

                                    Table 2. Whole-rock chemical analysesof rhyolites from the Thomas Range, Utah

 Llthology                          TFVVFF                                                         svsF                               VF               s
 Sanple No.           42           43      58       59      61a     1-93           8-938         29-L24 29-206        61b     6Ic                     I-OJ

 slo?                7s.9          75.9    75.8    76.6    75.0    75.9           75.8            75.9       16.2    74.2    74.2    76.t     75.4    75.5
                      0,09          0.16    0.08    0.07    0.28    0.07           0.08            0.07       0.07    0.28    0.30    0.14     0.14    0.10
                     t2.9          t2.7    t2.9    tz.a    13.1    12.8           t2.6            l3.o       t2.7    13.1    13.8    12,4     r2,5    t2.7
;:1;1                 1.09          1.17    0.99    0.98    t.26    1.61           0.93            r.o0       0.99    L.25    1.35    1.04     1.00    LzO2
ur6 '                 0.08          0.06    0.05    0.06    0.04    0.08           0.07            0.07       0.07    0.06    0.06    0.06     0.05    0.05
Mgo                   0.09          0.13    0.17    0.06    0.19    0.13           0.13            0.r8       0.05    0.26    0.21    0.r2     0.30    0.14
caO                   o.74          0,74    0.99    o.67    0.82    0.39           o.69            o.72       0.72    1.05    0.85    0.76     t.97    0.46
Na?o                  4,11          3.4s    4.t2    3.98    3.48    3.76           3.92            3.65       3.74    3.43    3.52    3,35     3.58    4.11
KrO                   4.69          5.O2    4.82    4.62    5.5r    4.83           4.72            4.7L       4.80    5.45    5.55    5.40     5.15    5.02
                      0.00          0.00    0.00    0.00    0.o2    0.00           o.o0            o.o0       0.00    0.02    0.01    0.00     0.00    0.00
AOEAI                99.99     99.33       99.92   99.84   99.70   99,57          9A.94           99.30      99.14   00.12   99,85   99.37   100.09   99.11
LOI                   3,11      2.66        0.79    3.34    2.97    0,33           0.34            2.t6       2.84    2.80    0.17    2.83     t.66    0.16
                                                                          parts   per [Lllion
Rb                    585   374    580              533     185     6t4            440                 470    433             185     377      372     515
                       L219311583624                                                                    s      t<              79      25       59      11
                       43    47     61               93      35      33            30                   80     75              37      42       37      24
zt                    115   140   105               110     190     115            90                  105    100             205     130      110     130
Nb                     70    50     50               60      40      72            51                   4a     47              40      50       42      51
                       30    40     45               35      30      45            45                   40     30             J)       45       20      30
                                                                   CIP}I NorMtlve          I{lnerala
c                                            -      0.02            0.65                           0.57                -      o.49
di                    0.75         0.34     r.70     -      0.03     -            o.44              -         0.07    0.15           0.70      2.22    0.65

     See footnotes    to   Table    1

(Kovalenko, 1977b).These considerations suggest that                                       rocks from central Asia, Kovalenko and Kovalenko,
1.5 kbar may be a maximum estimate for the crystalliza-                                     t976t.
tion pressure.                                                                                Topaz Mountain Rhyolite. Rhyolites from the Thomas
                                                                                           Range display considerable variation in their trace ele-
Trace element composition                                                                  ment composition even though they are all fairly enriched
    The trace element compositions of these rhyolites                                      in lithophile (or "fluorophile") trace elements. Signifi-
(Tables 1,2, and 3) are especiallyinformative; notable are                                 cantly, vitrophyres with low equilibration temperatures,
the substantial enrichmentof Rb, Cs, Nb, Y, U, Th, Ta,                                     like SM-29-206(630'C), are more enriched in lithophile
and HREE and the depletion of Eu, Sr and Co. These                                         elementsthan those with high-equilibrationtemperatures.
featuresdistinguishtopaz rhyolites from most other rhyo-                                   With increasing evolution (as indexed by uranium con-
litic magma types. The highly fractionated character of                                    centration)F, Be, Sn, Li, Rb, Cs, Na, Th, Ta, Nb, Y and
these rhyolites is demonstratedby their similarily to Li-                                  the HREE increasewhereasK, Sr, Zr, Hf, Sc, Mg, Ca,
pegmatitesand their associatedgranites (e.g., Goad and                                     Ti, P, and the LREE plus Eu are depleted.REE patterns
Cerny, 1981), and to ongonites (topaz-bearing volcanic                                     thus become flatter flowqr Lallu) and Eu anomalies
                                                                                           deeper (lower Eu/Eu*) with increasing evolution in the
                                                                                           Topaz Mountain Rhyolite (Fie. 3). Variation diagramsfor
                                                                                           some of these elements are shown in Figure 4. Enrich-
                                                                                           ment factors for two rhyolites are shown in Figure 5.
                                                                                              The correlation of these changes with the eruptive
                                                                                           history is the subject of work in progress, but the coher-
                                                                                           ent variation of many trace and major elementsacrossthe
                                                                                           suite (plus their mineralogic and temporal similarity)
                                                                                           suggests  that the rhyolites are part of a comagmaticgroup
                                                                                           of rocks.
                                                                                              Spor Mountain Formation The samplesanalyzedfrom
   Fig. 2. Normative composition of rhyolites in terms of quartz                           the Spor Mountain Formation are more enriched in U,
(Q), albite (Ab) and orthoclase (Or) comparedto experimentally
                                                                                           Th, Rb, Cs, Be, Li, Sn, Ta, Ga, Nb, Y and REE
determined ternary minima in the hydrous system (contours
                                                                                           (especially HREE) than the younger rhyolites from the
from Anderson and Cullers, 1978) and in a hydrous fluorine-
bearing system (Manning, 1981). Numbers on solid contours                                  Thomas Range. The rhyolites of the Spor Mountain
indicate PH2O and beside crosses they indicate werght Vo F in                              Formation have extremely low K/Rb (30 to 44), high
water-saturated system at I kbar. The dashed line is from                                  Rb/Sr (ca. 150),and low Mg/Li (2 to 5). Semiquantitative
Kovalenko (1977) and,represents the composition of residual                                analysesreported by Lindsey (1982) are compatible with
melts formed by fractional crystallization in a hydrous granite                            these results. Thus the composition suggeststhat the
system with I to 2 wt.VoHF at I kbar and 800 to 550 "C.                                    Spor Mountain magmas were highly ditrerentiated. The
                                                 CHRISTIANSEN ET AL.: TOPAZ RHYOLITES                                                                                     227

           Table 3. Trace element composition of topaz rhyolites from Spor Mountain and the Thomas Range, Utah

                                            THOI'IASRANGE                                                                        sPoR !m{.
                  SM-6la         sM-62a         sM-29-206            s|t-42      sH-41              Sli{-3I           sM-35           SU-70           SM-71

          LA         20             40              60                                                 -                 EO                r00            60   l0
          Cs                         7.4                ot            L4.L            L4,2            55                 58                                     2
          Be             2           6                  7                                                                52                a2             65   r0

          Cr         n.d             n,d                2.6            n.d            L4               3.6                1.2                                  l0
          Co         0,4             n.d                n.d            n.d            n.d              0.4                0.4                                   5
          Sc         2.8             1.8                L.7            z.z            z.q              2.6                2.7                                  J

          Sn         2                                                                                                                       30                15
                                                                                                       6.2                7.L                                  '5
          Hf         ?a              5.5                4.9            5.7            6.0
                     2.3                                t.4            6.4            6.0             26                 25                                    2

          Th        25             60              47                  56         65                 64                 69                45            67     3
          U                        l5              I9                  25         27                 t6                 38                26            31     4
                  r900           1600            3200                5400       6400               8200              11000             12500          8000

          cl                                                                                           -              1680                 tO60
          La        73                              l8                25              27              60                59
          Ce       t32             t29              53.1              64              62             r44               117                                     3

          Nd        63              35              31                31              35              52                 5l                                    3
          Sm         8.8                                               4.9             5.0            13.5               l7 .8
          Eu         1.2             0. 2 1             0 .065         0,059           0.05r           0 .l 5             0.o44                                 7

          ID         r.3             t.2                L9             -              0.79             3.3                3,4
          Yb         4,0             9.8                9,6            7. 9           8.8             15.6               t5.7                                  3
          Lu         0.53            0.90               1.7            1.6            1.5              2.5                2.6                                  3

          LalL5     14.0                                I.1             1.6            1.8             2.4                2.3

          LalCeN      l 4                               0.88            r,05           l. 12            1.07                  l. 11

          Eu/Eu*     0.35            0.06               o.o2           0.02           0.025            o.o24              0.006

          Note:     All  analyses   in   ppm;  ratios     relative  to chondrites.
          **    Eatlmted    precislon,     reported      as one standard  deviation        as percentage        of   arcult      presmt.
          n.d.    = not detected.

                                                                                                                                                  ln Ta             ++

                                                  Thomat Fan9o                                                          ln U                                        InU

                                                                                             hF                           +                       ln Rb

                                                                                                  m : 0.64


   Fig. 3. Rare-earthelement composition of the rhyolites from
the Thomas Rangeand Spor Mountain. Note the enrichment of                                Fig. 4. Logarithmic variation diagrams for topaz rhyolites
HREE and the depletion of LREE with increasingdepth of the                            from the Thomas Range O and Spor Mountain (+). The lines
Eu anomalyin samplesfrom the ThomasRange.The scaleon the                              representleast-squares to the data from the Thomas Range.
right appliesto samplesfrom Spor Mountain and the scaleon the                         The bulk partition coefrcient (-D for each element is derived
left to samples from the Thomas Range. Concentrations are                             from the equation D = t - m, where m is the slope of the line
normalizedto 0.83 times Leedey chondrite (Masudaet al.,l97l).                         (seetext).
                                      CHRISTIANSEN ET AL.: TOPAZ RHYOLITES

                                                               source). The simplest discriminant between partial melt-
                                                               ing and fractional crystallization is the behavior of com-
                                                               patible trace elements(elementswith bulk partition coef-
                                                               ficients D greater than l). In feldspar-rich rocks, Eu, Ba,
                                                               and Sr all have D greater than 1. Large ranges in the
                                                               concentrationsof these elementscannot be produced by
                                                               partial melting processesas the range is limited by l/D
                                                               (Shaw, 1970;Hanson, l97E).
                                                                  Figure 6 shows the behavior of Eu relative to U (with a
                                                               very low D). In quantitative formulations, the fraction of
                                                               liquid produced by partial melting or the residual liquid
                                                               from crystallization (both symbolized by fl are inversely
                                                               related to uranium concentration. if uranium behavesas
                                                               an incompatible element. As noted below, D" is slightly
  Fig. 5. Enrichmentfactors for 30 elementsderived by          greater than 0; nonethelessthe trend of the data is clearly
comparing samples
           two         from the Thomas Range (heavybars).      inconsistent with batch partial melting, but indicates a
Elemental ratiosderived comparing
                       by          earlyandlatepartsof the     DEUof about 3 for some type of fractionation process.
eruptionof the Bishop Tutr (Hildreth, 1979)are shown for          There appears to be little evidence of partial melting
comparison bars).                                              processesremaining in these highly evolved rhyolites. A
                                                               determinationof the partial melting history would require
anomalousconcentrationsof lithophile elementsare con-          a detailed examination of a number of less silicic consan-
sistent with the suggestionthat the Be-U-F ores in the         guineousrocks (like SM-6la) or perhaps an examination
underlying beryllium tufi were formed as these elements        of their intrusive equivalents.
were releasedand mobilized by devitrification and leach-          Fractional crystallization. Many of the elementalvaria-
ing of the lavas (Bikun, 1980;Burt and Sheridan,l98l).         tions such as increasingNa, U, Th, and Rb, and decreas-
Mineralization is absentfrom the less evolved rhyolites of     ing Ca, Mg, Ti, and Eu with increasing SiO2 content of
the Thomas Range.                                              the rhyolites of the Thomas Range are qualitatively
  The REE patterns (Fie. 3) also provide evidence of           consistentwith fractional crystallization of observedphe-
extreme differentiation in terms of the large negative Eu      nocrysts. The major and trace element data combine to
anomalies (Eu/Eu* : 0.03 to 0.01). In addition the             suggestthat samples SM-42, -62, and -29-2M could be
patterns are nearly flat (low Lall-up : 2.5). Similar          derived from SM-61 by crystal separation. This process
patterns are recognized in other highly evolved rhyolitic      was modeledusing a program written by J. Holloway and
magmas   (Hildreth, 1979;Keith, 1980;Bacon et al., l98l;
Ewart et al.,1977; and others), and are typical of many
topaz rhyolites (Christiansen al., 1983a).

                  Petrologic eyolution
   The origin of the elemental variations and substantial
enrichmentsof incompatible trace elements(those which              Eu    o.l
have bulk partition coemcients D less than l) in topaz
rhyolites could result from: (l) small and variable degrees
of partial melting, (2) crystal fractionation, or (3) liquid
fractionation (e.9., thermogravitational difusion, liquid
immiscibility, or vapor-phase transport). In this section
evidence is presented supporting the suggestionthat the                                      0.4       0.6    0.6
chemical evolution of the Topaz Mountain Rhyolite (and
by analogy the rhyolite of Spor Mountain as well) was                                              U
governed by crystal fractionation possibly acting in con-
                                                                 Fig. 6. Behavior of compatible (Eu) versus incompatible (U)
cert with some form of liquid difierentiation.                 elements during batch partial melting and fractional crys-
                                                               tallization. Curves were calculatedfrom
Topaz Mountain Rhyolit e
   Partial melting. Variable degreesof partial melting of a      clc
                                                                = = =----;          - Gatch partiat melting) and | = "flD-'t
uniform source could possibly produce the variation             C6 D(l-D+f                                       Lo
observed in the incompatible trace elements without
substantially altering the major element chemistry of the      (fractional crystallization). The Eu-U variation does not appear
liquid (as long as no phase was eliminated from the            to be related to variable degreesof partial melting.
                                                       CflruSTTANSEN AL': TOPAZRHYOLITES
                                                                    ET                                                                             229

                                                                                          ating mineral assemblage    (becauseof constraintsimposed
M. Hillier that conforms in most respects with sugges-
tions of Reid et al (1973).Weighting is basedon llwt.Vo of                                by the program), mass balance relations suggest that
each component, normalized so that SiO2has a weight of                                    removal of 0.0006 to 0.0012 wt. fraction apatite could
l. The major element compositions of the rocks (except                                    account for the excess P2O5in the calculated composi-
PzOs)and phenocrysts were used as input for the pro-                                      tions. Although these models are in no way unique, they
gram. Feldspar (Ab/An/Or), magnetite, ilmenite and bio-                                   do suggestthat substantialfractional crystallization could
tite (Fe/Mg) compositionswere varied within limits deter-                                 have occurred without drastically changing the major
mined by microprobe analyses of phenocrysts (Bikun,                                       element(Si, Al, K, Na) composition of the derivatives-a
 1980;Turley and Nash, 1980).Table 4 shows the results                                    qualitative argument usually proffered to discount frac-
of this attempt and demonstratesthat fractional crystalli-                                tional crystallization of granitic magmasto produce sub-
zation of quartz, sanidine, plagioclase,biotite, and Fe-Ti                                stantial inrichments of incompatible elements. Although
oxides (magnetite and ilmenite in a 50:50 mixture) can                                    Hildreth (1979) has made a strong case for the lack of
 account for most of the chemical variation. Mn and Mg                                    crystal settling in the magmachamberof the Bishop Tuff,
 show the largest Out statistically insignificant) relative                               there is considerable evidence (field, theoretical, and
 erors. The SiOzcontent of the "cumulates" rangesfrom                                     experimental) that many granitic intrusions solidify in-
 72 to 74Vo.These models suggestthat from 40 to 60Vo                                      wards by crystallization on the walls and especially the
 crystallization of a magma similar to SM-61 could pro-                                   floor-without significant crystal settling (Bateman and
 duce residual liquids similar to the other three samples.                                 Chappell, 1979 Batemanand Nokleberg, 1978 McCarthy
                                                                                                            ;                                ;
 The relative proportions of the phenocryst phasesare in                                   and Fripp, 1980;Groves and McCarthy,l97E; Pitcher'
 accord with those generally observed in the vitrophyres                                   1979;McBirney, 1980).Theoretically, the magma cham-
 with sanidine > quaxtz > oligoclase > biotite > Fe-Ti                                     ber that gave rise to these topaz rhyolites could have
 oxides. Although apatite was not included in the fraction-                                crystallized substantially between eruptions of the lavas'
                                                                                           In fact, the elevated concentrations of fluorine would
                                                                                           promote extensivecrystal fractionation. Wyllie (1979)has
Table 4. Least-squares   approximations the efrect of
                                         of                                                suggested   that the addition of fluorine to a granitic magma
          crystallization rhyolites
 fractional             of                       Range
                                   from the Thomas
                                                                                           results in a substantial reduction of the solidus tempera-
               obasrved             celculated          fractlonatlng                      ture but produces a smaller reduction of the liquidus
                                                                                           temperature.This would extend the duration offractional
             dcrlvetive      (Z)   derlv6!1ve           P!C9!9 !4j       !)-

     r. gu:!b-ge.-9E!3e                                                                    crystallization of a cooling pluton over a wider tempera-
     s1o2         75't                76,1                ssnldltre            16.55       ture (and time) interval. The accumulation of fluorine in
     T102         0, 14                0. 13              qu6E3r               10,00
                                                                                           the residual melt would further enhance this efect. In
                                       12.4               p16gloc16e.
                                                                                           addition, fluorine may reduce the viscosity of silicate
     A1203        12.4                                                          8.55

     FG203         1.04                 1.04              blottte               I .21

     llno         0,06                  0,05              Fc-Tl     dtdc.       O,47
                                                                                           melts and increase diffusion rates (Bailey, 1977)which
     !rg0         O,l2                  0. ll                                              would aid crystallization and fractionation.
     c.0           0.76                 0.76           Redldul      llqutd     63. l9          A more rigorous test (at least one with fewer assump-
     N620          1,35                 3,35                                                tions) of the crystal fractionation model can be made by
     rzo           5,40                 5,40                                                using the trace element compositions and the analytical
          Tot.l 99,37                 99.35
                                                                                            method of Allegre et aL.(1977). Using the conventional
     t' !!:9I4$!!:3.!L?99
     st02      76.2
                                                                                            Rayleigh fractionation law and trace element variation
     1102           0.07                0,07               .enldln.            22.29        diagramsthat use elements with very low bulk partition
     Ar203         t2.7                                    quert,              I 1.95       coefficients as "monitors" of fractionation, U and Cs in
     8.203          0.99                0.99               plagloclere          6,06
                                                                                            this case, the bulk partition coefficient for a fractionating
     r,bo           0,07                0,05              blotlt.               1.28
                                                                                            system can be derived without further assumptions as
     Xgo            0,05                0.07               F.-It    oald..      0.59

     cso            0.72                0.84
                                                                                            outlined below. This bulk partition coefficient can then be
     le20           3,74                3,76           Re.tdssl     llquld     57,73
                                                                                            used to determine / (the degree of fractional crystalliza-
     tzO            4,80                4,81                                                tion) and to place some limits on the composition of the
          tot.l    99.34               99.50                                                fractionating mineral assemblage.
     c, su-6la ro su-42                                                                        During perfect fractional crystallization the Rayleigh
     8102    75,9                      75,9
                                                                                            fractionation law (Neumann et al., 1954)can be used to
     Ito2     0.09                      o,o7
     A1203         t2,9                12,9                 .&ldlne             29.53
                                                                                             describethe trace element concentration of a diferentiat-
     L203           1.09                L09                 qurtz               16.69        ed liquid, Cl, relative to the concentration in the parent
     rbo            0.08                0,06                pl.glocl..e          9.2L        melt, C6:
     xgo            0.09                0,05                blotlte               I .73
     cro            O,74                0.82                Pe-!l     oxldc.     0.5E                             c":    6o/D-tr                     (l)
     }|r20          4. I I
     32             4 '69               4.70           Rrildul      llquld      39 ,95
                                                                                           where D is the bulk partition coefrcient. ff D << I
          Totll    99.69               99.74
                                                                                                                                  denoted by *)'
                                                                                           (incompatible in the mineral assemblage,
 230                                                        CHRISTIANSEN ET AL.: TOPAZ RHYOLITES

then:                                                                                    20% in the range of f-values considered here (Allegre et
                                                                                         al., 1977), they do not substantially affect the conclu-
                                 cl:                                              (2)    sions. In addition, the presumably low Dc' and the
                                        fct                                             correlation of Cs and U tend to support the assumptionof
Thus, as noted earlier the concentration of a hygromag-                                  small D's for both elements.
matophile element serves as an index of the degree of                                      Using the D's of Table 5 and the element concentra-
differentiation. Replacing equation (2) in equation (l)                                 tions of Tables I through 3, / (the fraction of liquid
eliminates / and expressesthe Rayleigh law in terms of                                  remaining in a crystallizing system) can be calculated
the relative concentration of two elements:                                             from equation (1). The results of these calculations,
                                                                                        which use SM-6la as a parent, are shown in Table 6 and
                                                                                        are compared with f's derived from the major-element
                                   :       -
                                 cr- coSPrD tr                                    (3)   mixing model. The calculated f's for each sample have a
                                                                                        total rangeof about 0. l, except when using the data for F,
This can be expressedin logarithmic form as:                                            La, and Th (in one instance). Although both the major
                                                                                        and trace element models suggestthat the rocks become
                    rnc,=rnc6+(r-D)-(#)                                           (4)   more evolved in the sequenceSM-61a, -62, -29-206,       -42,
                                                                                        there are large discrepanciesbetween the J-values pre-
 This is a linear equation in ln C1 - ln Cf, if D remains                               dicted by the two diferent types of calculations. Even
 constant throughout the process of fractionation. The                                  including the possible l0 to 20Vo uncertainty produced by
 slope of the line is given by (l - D). Theoretically, this                             assumingthe D for uranium tiquals 0, the two solutions
 allows the determination of D for any element. Examples                                appearexclusive. In fact to satisfy the / predicted by the
of diagrams like these using samples from the Thomas                                    major element model the D for U would have to be lower
Range are shown in Figure 4. A linear regression tech-                                  than 0 (an impossible situation), not higher as it may be in
nique was used to determine the correlation and the slope                               reality.
of each variation line; the results are shown in Table 5.                                  A seriesof equations may be defined from the derived
The elementsMg, Ca, K, Na, Sc, and Hf have correlation                                  D's and the observed phenocryst assemblage. These
coefficients less than 0.85. Although such correlations                                 equationstake the form
may be statistically significant, they were not deemed                                                 -D
                                                                                                          = Olr Xu * Dtop * ...
   The most critical factor that may affect the applicability                           where D|7p = the partition coefficient of element i be-
of thesederived D's is the assumptionthat DU : 0. Using                                 tween phasea and liquid PoX" : weight fraction of phase
the averagemodal mineralogy and mineral partition coef-                                 a. A set of these equations for various elements may be
ficients from Hildreth (1977)andCrecraft et al. (l9gl), the                             solved simultaneouslyto place limits on the composition
probable DU in the lavas ranges from 0.06 to 0.16                                       of the fractionating mineral assemblage.We have at-
(including up to 0.5Voallanite or 0.04Vo                                                tempted to do this using the mineralfliquid partition
                                         zircon). Although
thesedeviations may result in relative errors of about 0 to                             coefficients for high silica rhyolites of (Hildreth, 1977;
                                                                                        Crecraft et al., l98l; Mahood and Hildreth, 1983).Solu-
                                                                                        tions ofthese equations are generally consistent with the
Table 5. Derived bulk distributioncoefficients for rhyolites
                                                                                        relative proportions of the major fractionating phasesas
                   of the ThomasRange
                                                                                        Table6. Comparisonof J-values (residual liquid) derived
                                                                                           from traceandmajorelement
                                                                                                                   models differentiation
            Co                                                       0.995
            8e                                         0.966
            Lt                             0.24                      0.954                                     f-value   telarlve   to 5l{-61
            II                             o.27                      o.969
                                                                                         EIeent            D              st{-62         sH-29-105        slr-42
            Ta                             0.38                      0.991
            Rb                             0.36                      0.950                   u             0               0.32                 0.26       0.20
            F                              0.35                      0.843                   C6          -.0.05            0.34                 0.26       0. r8
            Th                             0.50                      0.998                   Be           0.11             0.29                 o.24
                                                                                             Lt           o.24             0.40                 0.24
                             0.63                      0.868                   Lu           0.27             0.48                 0.20      0.22
            Nb                             0.70                      0.859                   la           0.38             0.34                 0.25      0.20
            Y                              0.8s                      o.964                   Rb           0.36             0.33                 o.27      0,l8
            K                              r. ll                     0.830                   F            0.36             1.42*                0.48*     0. 16*
                                                                                            Th            0.50             0.t7*                o.2a      0.20
            Zt                                                       0.938                  IID           0.63             o.34                 o.22      0.l5
            Ir                             r.45                      0.920                                1.45             0, l0*               0.04*     0.ll*
            TI                             1.84                      o.925                  T1            L84              0.44                 0.r9
            Sr                             2.23                      0.994                                2.94             0.4r.                0.2r      0.21
        EI                                2.94                       0.960
                                                                                         llean f**                         0.37 t 0.06      0.24 ! 0.03   o.2o ! o.o3
                                                                                         llaJor elernt    nodel   f        0.63             0.s8          o.4o
    Note!          calculgtLon    of 5 *plaloed    ln textS 12 ls a cotrelstLoo          * Not lncluded in coqutlnt
             obtaired                                                                                nean.
                                            durlng llnear  regreasloD of data.           ** Meetr t I standard devlation.
                                     CHRISTIANSEN ET AL.: TOPAZ RHYOLITES                                              231

determinedfrom the major-elementmodeling (X sanidine                Liquid fractionation The major processes of liquid
-0.46, plagioclase -0.20, X quartz -0.30, X biotite             differentiation that do not depend solely on crystal-liquid
-0.03, and X Fe-Ti oxides -0.01). This mineral assem-           equilibria are (1) liquid immiscibility, (2) separationof a
blagegivesacceptable for Be, Cs, Rb, Sr, Mn, Li, Ti,
                         D's                                    vapor or fluid, and (3) diffusion (thermogravitational
Zr, El, and K. The La variation limits the maximum              diffusion, double-difusive crystallization, the Soret ef-
proportion of allanite (which efectively includes any           fect).
monazite as well) to 0.0004 (X all : D/Dkf.nit"). The               The evolution of immiscible silicate and fluoride liquids
maximum proportion of fractionating zircon is limited to        is possible in fluorine-rich magmas (Koster van Groos
about 0.0004by the Lu variation. The zirconium deple-           and Wyllie, 1968;Kogarko and Ryabchikov, 1969;Hards
tion of about 90 ppm implies a similar proportion               and Freestone.1978;Kovalenko, 1977b)         and could theo-
(0.00043)of zircon (assuming f -0.25 for the change).           retically play an important role in the trace element
This amount would result in early cumulates with Zr             evolution of magmas. However, fluorine concentrations
concentrationsof about 210 ppm and thus seemsreason-            in hydrous alumino-silicatemelts must be extremely high
able. The P depletion suggests     that apatite made up about   before immiscibility occurs (6 wt.Vo for granite-NaF-
0.00065weight fraction of the cumulate (implying about          HzO, Hards and Freestone, 1978;3 to 4 wt.Vofor granite-
0.03Vo   P2O5 the cumulate).
               in                                               HF-H2O, Kovalenko, 197'7b; to 4 wt.Vofor ongonite-
    Bulk distribution coefficients for Y, Nb, Ta, and Th        HF-H2O, Kovalenko and Kovalenko, 1976). Such high
calculated using this assemblageof minerals are signifi-        levels of fluorine concentration are not reached in the
cantly smaller than the D's derived from the trace ele-         Topaz Mountain Rhyolite (or the more fluorine-enriched
ment variation. This observation may simply imply that           rhyolites from the Spor Mountain Formation) and it
the individual mineral D's employed in the calculations          appearsthat silicate liquid immiscibility played a negligi-
 are unrealistically low. It is not easyto explain Dr" of Drh    ble role in their evolution.
 in this fashion as they imply X zircon : 0.005 (or DIf"o'           The separation of a fluorine-rich fluid and its conse-
 = 860) and X allanite : 0.001 (or Dllu.;," : ll00). These       quent rise to the apical portion of a magmachambercould
 values are 3 to 5 times larger than those implied by the        have had an effect on the evolution of topaz rhyolite
 REE variation noted above. Extremely small quantities           magmas.It is difficult to assess   the saturation concentra-
 of fractionating monazite (Mittlefehldt and Miller, 1983),      tions of various F-speciesin igneousmelts in the absence
 titanite (identified by Turley and Nash, 1980, in some          of direct experimental studies in HzO-free systems.
 Thomas Range rhyolites), xenotime or even samarskite            Based on the correlation of SiO2 and F in ongonitic dike
 (reported from plutonic equivalents of topaz rhyolites in       rocks, Kovalenko (1973) suggestedthat fluid-saturation
 the SheeprockMountainsof Utah, w. R. Griffitts, 1981,           occurs at 0.5 to 3 wt.VoF, increasingwith decreasingSiOz
 written communication) could account for the high D's           over the range of 75 to 69 wt.Vo.However' in view of the
  calculatedfor Th, Y, Nb and Ta. Thus although we have          preference of F for melts over fluids (Hards, 1976)it
  not yet identified monazite or some of the other phases         seemsmore likely that H2O would induce the saturation
  cited here, they have been reported from similar highly         of a fluid which would contain some small amount of F. It
  evolved magmas.The small quantities required would be           thus seemsappropriateto examinethe H2G-F-granite (or
  extremely difrcult to detect by petrographicexamination.        albite) system. Wyllie (1979)reports that F increasesthe
  Alternatively, it is possible that someof the trace element     solubility of H2O in albite melts and Manning (1981)has
  variation is not produced by crystal/liquid fractionation-      suggested impreciselower limit for fluid saturationof 4
  some liquid-state process may be operating. Thus the            wt.VoH2Oin H2O-F-granite system. The late crystalliza-
  derived D's may be a measure of the "mobility" of               tion of biotite relative to quartz and feldspars and the
  different elements in the magma. Their variation could          experiments of Maaloe and Wyllie (1975) and Clemens
  reflect their difusion rates across gradients of P, ?, and      and Wall (1981) in F-free systems suggestthat H2O
  composition, their partitioning into an accumulatingfluid,      concentration in these granitic melts may have been less
  or their ability to form complexes with other migrating         than this. The position of the samplesanalyzedhere in the
  elements.                                                       Q-Ab-Or system is also consistent with the suggestion
     The lack of agreementbetweenthe f's derived from the         that the magma was not fluid-saturated. The emplace-
  major elementand trace elementmodels and the improba-           ment of the rhyolites as lava flows instead of as pyroclas-
  bly high D's derived for some elements may indicate             tic deposits is also suggestiveof a water-poor condition
  some sort of differentiation process in the liquid state.        (Eichelberger and Westrich, 1981)and the absenceof a
  However, none of these observationssuggestthat crystal          free-fluid prior to venting. Hildreth (1981)has reviewed
  fractionation did not operate. Indeed the overall major          other evidence indicating that some magma chambers
  and trace element systematics imply that any liquid              which give rise to large ash flows were not fluid-saturat-
   diferentiation was probably minor and may have only             ed. Nonetheless,Holloway (1976)has shown that even
   afected some trace elements. Keith (1982)and Nash and           small quantities of CO2 in granitic magmaswill produce
   Crecraft (1981)came to similar conclusionsregardingthe          fluid-saturation.In the light of our ignoranceof magmatic
   diferentiation of other rhvolitic rocks.                        COz concentrations it is worth pointing out that the
 232                                  CHRISTIANSEN ET AL.: TOPAZ RHYOLITES

   accumulationof a fluid near the roof of a magmachamber          1975; Fryer and Edgar, 1977;Tayloret al., 1981;    contrast
  would result in net dilution of the magmatic concentra-         with Muecke and Clarke, 1981)and the depletednature of
  tions of all trace elements except in the case of D_7q          someresidual granulites (Collerson and Fryer, 1978).For
  (partition coefficient between melt and fluid) less than l.     the REE, thE stability of the fluoride complexesincreases
  Kovalenko (1977a)has shown experimentally that Li and           from La to Lu and at lower temperatures(Moller et al.,
  Rb have fairly high Dnyqin an ongonite-HzO-HF system            1980).Mineyev et al. (1966)experimentally demonstrated
  at I kbar. With temperature decreasing from 800. to             the increasedmobility of Y (as representativeof HREE)
  600"C, DHfl decreasesfrom 50 to 5 and Dffi decreases            relative to LREE (La and Ce) in acidic solutionswith high
  from 250to 50. Thus it seemsimpossiblefor Dnyn attain
                                                    to            fluorine and sodium activities acrossa temperaturegradi-
  values of less than I before a melt crystallized complete-      ent. Thus, the pronounced HREE enrichment (consid-
  ly. Webster and Holloway (1980), Flynn and Burnham              ered to be a fingerprint ofliquid fractionation in rhyolitic
  (1978) and Wendlandt and Harrison (1979) have shown             magmas) during the evolution of the Topaz Mountain
  that high Dnysare common for REE in H2GCI, H2O-F,               Rhyolite may have been the result of the co-migration of
  and CO2-rich fluids at crustal pressures. On the other          F with HREE (and presumablywith Li, Be, U, Rb, Sn,
  hand, Candellaand Holland (1981)and Manning (l98lb)            etc.) to the apical portions of a magma chamber. From
  have shown that Cu, Mo, and Sn prefer coexisting                studies of rhyolitic ash-flow tufs, Mahood (1981) has
  hydrous fluids 1-rp, Cl) over silicate melts. Thus a            suggestedthat the dominant control on ash-flow trace-
  transient vapor-phasepreceding or coincident with erup-        element enrichment patterns may be their roofward mi-
  tion may have altered the concentrations of these ele-         gration as volatile complexes. The most important vola-
  ments but probably did little to efect the concentrations      tiles in highly evolved granitic magmasare probably H2O,
 of other elementsincluding the REE.                             HF, HCl, B, and CO2(?).    The suiteof elements   associated
     If element enrichment/depletionpatterns can be con-         with these volatiles should difer dependingon the nature
  sidered indicative (and there is no evidence that it is so     of the complexes or ion-pairs formed by each volatile
  simple), the apparent similarity of the variations in the      (Hildreth, l98l).
 rhyolites from the Thomas Range to those of the Bishop              In anotherpaper (Christiansen al., 1983b), sug-
                                                                                                      et              we
 Tuff (Hildreth, 1979) could be taken as evidence that           gest that the trace element evolution of rhyolitic magmas
 "convection-aided thermogravitational difusion" pro-            is sensitive to FiCl in the melt. With ditrerentiation, F-
 duced some of the characteristics of the topaz rhyolites        dominatedsystems(F/Cl greater than 3) display increases
 (Fig. 5). Samples SM-41 or SM-29-206could represent             in Al, Na, Li, Rb, Cs, Ta, Th, and Be, while Cl-
 eruptions from the upper part of a magmachamber zoned           dominatedsystemsare usually peralkaline and are associ-
 at depth to a compositionsimilar to SM-61a.However,             ated with greater enrichments of LREE, Na, Fe, Ti, Mn,
 the uncertainty in the temporal relationship among the          Zn, Nb, and Zr. Although magmaswith variable propor-
 sampledlava flows as well as the rapid reestablishmentof        tions of H2O, F, Cl and CO2 certainly exist, topaz
 chemical zonation in some magma chambers following             rhyolites appear to represent the fluorine-dominatedend
 their eruption (Smith, 1979), precludes postulating that       member. However, it is important to note that the vola-
 the lavas were erupted from a unitary system like the          tiles and alkalies also affect the stability of various
 Bishop Tuff magma chamber. The samples may simply              mineral phases important to the chemical evolution of
 record temporal variations in a magmachamberlike those         these rhyolites (allanite, zircon, biotite, titanite, apatite,
 describedby Mahood (1981)and Smith (1979).Whether              pyroxene, and hornblende).Thus, it is not an easy task to
the processwhich gave rise to these features (Fig. 5) was       distinguish liquid-state processesfrom those that result
thermogravitationaldiffusion (Shaw et al., 197 ; Hildreth,
                                                 6              from fractional crystallization of a distinctive mineral
 1979)or crystallization in a double-difusive system with       assemblagewhich was stabilized by different volatile-
the rise of evolved less-denseliquid from the walls of a        element concentrations or ratios.
crystallizing magma chamber (Chen and Turner, l9g0;                 We suggestthat the extreme compositions observed in
McBirney, 1980;Nash and Crecraft, lggl; Christiansen,           lavas from the Topaz Mountain Rhyolite are the result of
 1983),or some other process is difficult to evaluate until     both extensive fractional crystallization and at least to
more quantitative formulations of these hypotheses can          some extent the result of the diffusive transfer of trace
be made.                                                        elements.Although some of the magmasmay have been
    In any case, fluorine probably played an important role     fluid-saturated shortly before eruption, the role of a
in the diferentiation process becauseof its tendency to         separatefluid phase in their chemical evolution appears
form stablecomplexeswith a number of elements,includ-           to have been minor.
ing the REE (Smirnov, 1973; Mineyev et al., 1966;
Bandurkin, 1961).Halogen complexing (and subsequent             Spor Mountain Formation
transport in a vapor-phase, or within a melt, down                The limited number of trace element analyses for
thermal, chemical, or density gradients)has been invoked        samplesof the Spor Mountain Rhyolite preclude detailed
to explain the incompatible trace-elementenrichment of          petrochemical modeling of its differentiation. However,
evolved magmasand pegmatites(Bailey and Macdonald,              comparisonswith the younger Topaz Mountain Rhyolite
                                      CHRISTIANSEN ET AL.: TOPAZ RHYOLITES                                             233

and with Asian "ongonites" suggestsome ideas about its           the association of numerous tuff-lined breccia pipes
evolution.                                                       (Lindsey, 1982) with the Spor Mountain magmatism
    The composition of the Spor Mountain Rhyolites im-           suggeststhat the magma became saturated during erup-
plies that they evolved from magmas slightly diferent            tion as a result of a rapid pressuredrop. Enhancedvapor-
from those erupted in the Thomas Range.They are richer           phasetransport of Sn and Mo and perhapsBe, Li, Cs, U,
in Fe, Cr, Co, Ba, LREE, and depleted in Si relative to          and other elements may have occurred under these
the evolved Thomas Rangelavas, trends which are incon-           conditions adding to the effect of other differentiation
sistent with the continued differentiation of the Topaz          processes.This mechanismcould lead to the rapid forma-
Mountain Rhyolite. In contrast, they are extremely en-           tion of a lithophile-elementrich cap to the magmacham-
riched in U, Rb, Th, Cs, Li, Be, Sn, Ta, and the HREE.           ber. The subsequenteruption of the volatile-rich cap may
The trace element compositions do not lie on the trends          be representedby the emplacementof the Be (Li-F-U)-
defined by the Topaz Mountain Rhyolite, implying                  mineralized tuff that underlies the Spor Mountain rhyo-
that D's were different. that initial concentrations varied       lite.
or both. However, the overall chemical and mineralogic
 similarity of the two sequences(REE patterns, enrich-                                  SummarY
 ment and depletion of the sameelementsrelative to other             The fluorine-rich rhyolites of the Thomas Range and
 rhyolite types, Fe-enriched mafic minerals) suggeststhat        Spor Mountain, Utah, are enriched in a distinctive group
 the Spor Mountain magma and the rhyolites of the                of lithophileelements     (U, Th, Rb, Li, Be, Cs, Ta and Sn)
 Thomas Range may have evolved from grossly similar              typical of other    topaz rhyolites from the western United
 parents. The most important features of the older rhyo-         States. The mineralogy and geochemistry of the lavas
 lites that require explanation are (1) the low Si and high AI   suggests    that they are similar to "ongonites", aluminous
 contents, (2) Na/K greater than 1, and (3) the extreme          bimodal rhyolites, and anorogenic granites.
 enrichment of "fl uorophile" elements.                              The topaz rhyolites from the Thomas Range are inter-
     The high Na/K (relative to other topaz rhyolites) of the    preted as being derived by differentiation from a less
 Spor Mountain Rhyolites can be explained in terms of the        silicic but still rhyolitic magma.Calculations suggestthat
 effect of fluorine on residual melt compositions in the         sanidine, quartz, sodic plagioclase,Fe-Ti oxides' biotite'
 granite system(Manning 1981;Kovalenko 1977b).Both               allanite, zircon and apatite were the dominant fractionat-
  sets ofexperiments show a regular increasein the size of        ing phases over the course of differentiation sampled'
  the quartz field in the Q-Ab-Or system and increasingly         However, failure of fractional crystallization models to
  albitic residual melts with increasedfluorine content (Fig.     explain the high bulk partition coefficients of Ta and Th
  2). Thus as the activity of the fluorine increases in a         suggests that liquid fractionation, aided by the high
  residual melt, either by differentiation or by the isochoric    fluorine content of the magmasmay have occurred or that
  crystallization of a fluid-saturated melt (Kovalenko,           some unidentified phase (monazite' titanite, samarskite,
   1973),quartzfractionation becomesmore important lead-          xenotime) was removed from the evolving melt'
  ing to a reversal in the normal trend of SiO2increasewith           The Spor Mountain rhyolite contains higher concentra-
  chemicalevolution. As a result aluminum increasesin the         tions of fluorine and lithophile elements than the Topaz
  residual melt and continued fractionation of sanidine            Mountain Rhyolite. It may have evolved from a magma
  leads to higher Na/K in the liquid. This move to more            generally similar to the Topaz Mountain Rhyolite, but its
   sodic liquid compositions as a result of potassium feld-        differentiation was more protracted. Expansion of the
   spar fractionation is petrographically manifest by the          quartz stability field by elevated fluorine concentrations
   common mantling of plagioclase by sanidine (anti-rapi-          and the resultant separationof quartz and sanidine may
   kivi) in the lavas from Spor Mountain.                          have led to its lower Si and higher Na and Al contents'
      Each of these trends (decreasingSiO2 and increasing          Extended fractional crystallization, possibly aided by
   Na and Al with higher F) is observedin the extremely F-         liquid fractionation, led to extreme concentrationsof Be,
   rich topaz rhyolites like those at Honeycomb Hills, Utah        Li, U, Rb, Cs and other fluorophile elements' The
   (Christiansenet al., 1980)and are well developed in F-          magmaticconcentrations of these elements set the stage
   rich ongonites from central Asia (Kovalenko and Kova-           for the development of the Be-mineralized tuff which
   lenko, 1976).These trends are also mimicked during the          underlies the lava.
   crystallization of rare-metal Li-F granites, strengthening
   the suggestion that topaz rhyolites may be associated                             Acknowledgments
   with mineralized plutons at depth.                                           thankA. Yates,B' Murphy,C. Liu' K. Evans,R'
      The Spor Mountain Rhyolite appearsto represent the          Satkin, D. Lambert, Z. Peterman  and G. Golesfor help in
   product of protracted fractional crystallization under the     performing analytical
                                                                             the         work presented here;D. Lindseyand
   influence of high fluorine activity in the melt. Liquid        R. Colefor introducing to the geology the SporMountain
                                                                                        us              of
   fractionation is suggested,but not demonstrated,by the         region;J. Hollowayfor the use of his computer programs;S'
   flat REE patterns (Fig. 3). Although the magmamay not          Selkirk drafting figures B. P. Correa, Lesher, D'
                                                                         for        the      and             C'       J'
   have beenfluid-saturatedduring the most of its evolution,      Keith.R. T. Wilson, Hildreth, P. Nash'J. R. Holloway'
                                                                                      W.          W.
 234                                    CHRISTIANSEN ET AL.: TOPAZ RHYOLITES

and R. T. Gregory for valuable discussionsand reviews of the        Christiansen, H. (1983)
                                                                                   E.            The Bishop Tuffrevisited: Composi-
manuscript.                                                            tional zonation by double diffusive fractional crystallization
  This work was partially supportedby U.S. DOE Subcontract             (DDFC). Geological Society of America Abstracts with Pro-
#79-270-E from Bendix Field Engineering Corporation and                grams,15,390.
funds from Arizona State University. The senior author also         Christiansen,E. H., Bikun, J. V., and Burt, D. M. (1980)
acknowledgessupport from the National ScienceFoundation in             Petrologyand geochemistryoftopaz rhyolites, westernUnited
the form of a Graduate Fellowship and from the Geological              States. U.S. Department of Energy Open-File Report GJBX-
Society of America in the form of a research srant.                    22s(80),37-122.
                                                                    Christiansen, H., Burt, D. M., and Sheridan,M. F., (1983a)
                          References                                   Geology of topaz rhyolites from the western United States.
                                                                       Geological Society of America Special Paper, in press.
Allegre, C. H., Treuil, M., Minster, J. F., Minster, B., and       Christiansen, H., Burt, D. M., Sheridan,M. F., and Wilson,
   Albarede, F . (1977)Systematicuse of trace elementin igneous        R. T. (1983b).Petrogenesisoftopaz rholites. Contributions to
   process,Part I: Fractional crystallization processes volcan-
                                                       in              Mineralogy and Petrology, 83, 16-30.
   ic suites. Contributions to Mineralogy and petrology, 60, 57-   Christiansen,R. L. and Limpan, P. W. (1972)Cenozoic volca-
   75.                                                                 nism and plate tectonic evolution ofthe western United States,
Anderson, J. L. and Cullers, R. L. (1978)Geochemistryand               IL Late Cenozoic. Philosophical Transactions of the Royal
   evolution of the Wolf River batholith. a late precambrian           Societyof London, ser.A,271,249-284.
   rapakivi massif in north Wisconsin, U.S.A. precambrian Re-      Clemens, D. and Wall, V. J. (1981)
                                                                               J.                           Origin and crystallization
   search.7. 287-324.                                                 of some peraluminous (S-type) granitic magmas. Canadian
Bacon, C. R., Macdonald,R., Smith, R. L., and Baedecker,      p.      Mineralogist,19, I ll-131.
   A. (1981)Pleistocenehigh-silicarhyolites of the Coso volcanic   Collerson,K. D. and Fryer, B. J. (1978)    The role of fluids in the
   field, Inyo County, California. Journal of Geophysical Re-         formation and subsequentdevelopment of early continental
   search.86. 10223-10241.                                            crust. Contributions to Mineralogy and Petrology, 67, l5l-
Bailey, J. C. (1977) Fluorine in granitic rocks and melts: A           167.
   review. Chemical Geology, 19, 2-42.                             Crecraft, H. R., Nash, W. P., and Evans, S. H. (1981)Late
Bailey, D. K. and Macdonald, R. (1975)Fluorine and chlorine in        Cenozoicvolcanismat Twin Peaks,Utah. Geologyand Petrol-
   peralkaline liquids and the need for magma generation in an        ogy. Journalof Geophysical    Research,  86, 10303-10320.
   open system. Mineralogical Magazine, 40, 405-414.               Eichelberger, L. and Westrich, H. R. (1981)Magmaticvola-
Bandurkin, G. A. (1961)On the behavior of rare earth elemenrs         tiles in explosive rhyolitic eruptions. Geophysical Research
  in fluorbearingenvironments.Geochemistry 1962,2, 159-167     .      Letters. 8. 757-760.
Bateman, P. C. and Chappell, B. W. (t979) Crystallization,         Evans,K. L., Tarter,J. G., andMoore, C. B. (1981).     Pyrohydro-
  fractionation, and solidification of the Tuolumne Intrusive         lytic-ion chromatographicdetermination of fluorine, chlorine,
   Series,Yosemite National Park, Califomia. GeologicalSocie-         and sulfur in geological samples. Analytical Chemistry, 53,
  ty of America Bulletin, 90, pt. l, 465482.                          925-928.
Bateman,P. C. and Nokleberg, W. J. (1978)Solidification of the     Ewart, A. (1979)A review of the mineralogy and chemistry of
  Mount Givens Granodiorite, Sierra Nevada, California. Jour-        Tertiary-Recent dacitic, latitic, rhyolitic, and related salic
  nal of Geology,86, 563-579.                                        volcanic rocks. In F. Barker, Ed., Trondhjemites,dacites,and
Best, M. G., McKee, E. H., and Damon, P. E. (1980)Space-             relatedrocks, p. l3-121. Amsterdam,Elsevier.
  time-+omposition patterns of late Cenozoic mafic volcanism,      Ewart, A., Oversby,V. M., add Mateen,A. (1977)      Petrology   and
  southwesternUtah and adjoining areas. American Journal of          isotope geochemistryofTertiary lavas from the northern flank
  Science,   280. 1035-1050.                                         of the Tweed Volcano, southeasternQueensland.Journal of
Bikun, J. V. (1980)Fluorine and lithophile element mineraliza-       Petrology,   18,73-113.
  tion at Spor Mountain, Utah. U.S. Department of Energy           Flynn, R. T. and Burnham, C. W. (1978) An experimental
  Open-File Report GJBX-225(80),167-377      .                       determination of rare-earth partition coefficients between a
Bodkin, J. B. (1977) Determination of fluorine in silicates by       chloride-containing vapor phase and silicate melts. Geochi-
  specific ion electrode following fusion with lithium metabor-      mica et Cosmochimica      Acta, 42,685-701.
  ate. Analyst, 102, 409-413.                                      Fryer, B. J. and Edgar, A. D. (1977)Significanceof rare earth
Burt, D. M. and Sheridan, M. F. (1981) A model for the               distributions in coexisting minerals of peralkaline undersatu-
  formation of uranium/lithophile element deposits in fluorine-      rated rocks. Contributions to Mineralogy and Petrology, 61,
  enriched volcanic rocks. American Association of petroleum         35-48.
  Geologists,  Studiesin Geology 13,99-109.                        Goad, B. E. and Cerny, P. (1981) Peraluminouspegmatitic
Burt, D. M., Bikun, J. V., and Christiansen, H. (1982)
                                               E.         Topaz      granites and their pegmatite aureoles in the Winnipeg River
  rhyolites-Distribution, origin, and significancefor explora-       district, southeasternManitoba. CanandianMineralogist, 19,
  tion. EconomicGeology,77, l818-1836.                               177-194.
Candela,P. A. and Holland, H. D. (1981)The efecr of fluorine       Groves,D. I. and McCarthy, T. A. (1978)     Fractionalcrystalliza-
  on partitioning of molybdenum between a magma and a                tion and the origin of tin deposits in granitoids. Mineralium
  hydrothermal fluid. Geological Society of America Abstracts        Deposita, 13, ll-26.
  with Programs,13,422.                                            Hanson, G. N. (1978)The application of trace elements ro the
Chen, C. F. and Turner, J. S. (1980)Crystallization in a double-     petrogenesisof igneous rocks of granitic composition. Earth
  diffusive system. Journal of GeophysicalResearch, 85, 2573-        and Planetary Science Letters, 38,2643.
  2593.                                                            Hards, N. J. (1976)Distribution of elements between the fluid
                                        CHRISTIANSEN ET AL.: TOPAZ RHYOLITES                                                       2t5

  phase and silicate melt phase of granites and nepheline syen-      Mahood, G. and Hildreth, W. (1983)Large partition coefficients
  ites.N.E.R.C. ReportD6, Progress Experimental
                                         in             Petrology      for trace elements in high-silica rhyolites. Geochimica et
  3, E8-90.                                                            Cosmochimica Acta, 47, ll-30.
Hards, N. J. and Freestone, C. (197E)
                                I.          Liquid immiscibilityin   Manning, D. A. C. (l98la) The efect of fludrine on liquidus
  fluor-silicatesystems.N.E.R.C. Report, Progressin Exper-             phaserelationshipsin the systemQz-Ab-Or with excesswater
  mental Petrology 4, ll-13.                                           a lkb: Contributions to Mineralogy and Petrology,76,206-215.
Hildreth, E. W. (1977)The magmachamber of the Bishop Tuff:           Manning, D. A. C. (l98lb) The use of radioactive tracers in
  Gradients in temperature, pressure and composition. Ph.D.            detennir-lgthe behavior of tin in silicate-vapour systems.
  thesis. University of California, Berkeley.                          N.E.R.C. Report D-lE Progressin Experimental Petrology 5'
Hitdreth, W. (1979)The Bishop Tuf. Evidence for the origin of           l5-r6.
  compositionalzonation in silicic magmachambers.Geological          Masuda,A., Nakamura,N., and Tanaka,T' (1973)Fine struc-
   Society of America Special Paper lEO, 43-75.                        tures of mutually normalizedrare-earthpatternsof chondrites.
Hildreth, W. (1981)Gradients in silicic magmachambers.Impli-           Geochimica Cosmochimica
                                                                                     et                Acta, 37,239-248.
   cations for lithospheric magmatism.Journal GeophysicalRe-         McBirney, A. B. (1980) Mixing and unmixing of magmas.
   search.86. 10153-10192.                                             Journal of Volcanology and GeothermalResearch'7, 357 l.     :37
Holloway, J. R. (1976) Fluids in the evolution of granitic           Mccarthy, T. S. and Fripp, R. E. P' (1980)The crystallization
   magmas. Consequencesof finite COz solubility. Geological            history of a granitic magma, as revealed by trace element
   Societyof American Bulletin, 87, 15l3-1518.                         abundances.    Journalof Geology,88, 2ll-224,
Keith, J. D. (1980)Miocene porphyry intrusions, volcanism, and       Mineyev, D. A. (1963)Geochemical differentiation of the rare
   mineralization, southwesternUtah and easternNevada. M.S'            earths.Geochemistry, 1129-1149.
   thesis.University of Wisconsin,Madison.                           Mineyev. D. A., Dikov. Y. P., Sobolev,B. P., and Borutskaya,
Keith, J. D. (1982) Magmatic evolution of the Pine Grove                V. L. (l%6) Differentiation of rare-earth elements under
   porphyry molybdenum system, southwesternUtah. Ph.D'                  supercritical conditions. Geochemistry International, 3' 357-
   Thesis. University of Wisconsin, Madison.                            359.
Kogarko, L. N. and Ryabchikov, I. D. (1969)Diferentiation of         Mittlefehldt. D. W. and Miller, C. F. (1983)Geochemistrvof the
   alkali magmasrich in volatiles. GeochemistryInternational 6,         SweetwaterWash pluton, California: Implications for "anom-
    I l3-l 123.                                                         alous" trace element behavior during differentiation offelsic
Koster van Groos, A. F. K., and Wyllie, P. J. (1968)Melting             magmas.Geochimica et Cosmochimica Acta, 47, lO9-124.
   relationships in the system NaAlSirOrNaF-HzO to 4 kb              Moller, P., Morteani, G., and Schley, F. (1980)Discussionof
   pressure.Journal of Geology,          5p-:70.                        REE distribution patterns of carbonatities and alkalic rocks.
Kovalenko, N. I. (1977a)    Measurementof the Li and Rb distribu-       Lithos. 13. l7l-179.
   tion in the ongonite-H2O-HF system. Geochemistry Interna-         Morrison, B. C. (1980)A summary geologic report on the Spor
   tional, 14, 133-140.                                                 Mountain drilling project in Juab County, Utah. U.S. Depart-
Kovalenko, N. I. (1977b) The reactions between granite and              ment of Energy Open-File Report GJB-19(80).
    aqueoushydrofluoric acid in relation to the origin offluorine'   Muecke.G. K. and Clarke, D. B. (19E1)       Geochemical   evolution
    bearing granites. Geochemistry International, 14, 108-118.          of the South Mountain batholith, Nova Scotia. Rare-earth-
 Kovalenko, V. I. (1973) Distribution of fluorine in a topaz-           element evidence. CananadianMineralogist, 19, ll3-145.
    bearing quartz-keratophyre (ongonite) and solubility of fluo-    Nash, W. P. and Crecraft, H. R. (1981)Evolution of silicic
    rine in granitic melts. Geochemistry international, 10, 4l-49.       magmasin the upper crust. An experimentalanalog.EOS' 62'
 Kovalenko, V. I. and Kovalenko, N. I. (1976)Ongonites(topaz-            1075.
    bearing quartz keratophyre)-subvolcanic analogue of rare-        Neumann, H., Mead, J., and Vitaliano, C. J. (1954) Trace
    metal Li-F granites (in Russian). Nauka Press, Moscow.               elementvariations durirtg fractional crystallization as calculat-
 Lindsey, D. A. (1977)Epithermal beryllium deposits in water-            ed from the distribution law. Geochimica et Cosmochimica
    laid tuff, western Utah. Economic Geology, 72,219-232.               Acta. 6. 90-99.
 Lindsey, D. A. (1982)Tertiary volcanic rocks and uranium in the     Norrish, K. and Hutton, J. T. (1969)An accurateX-ray spectro-
    Thomas Range and northern Drum Mountains, Juab County,               graphic method for the analysis of a wide range of geologic
     Utah. U.S. GeologicalSurvey Professional     Paper 1221.            samples.              bt
                                                                                   Geochimica Cosmochimica        Acta, 33,431453.
 Lindsey, D. A. (1979)Geologic map and cross-sectionof Ter-          Pitcher, W. S. (1979) The nature, ascent and emplacementof
     tiary rocks in the Thomas Range and northern Drum Moun-             granitic magmas.Journal of the GeologicalSociety of London,
     tains, Juab County, Utah. U.S. Geological Survey Miscella-          136,627-662.
     neousInvestigation   Map I-1176,scale l:62,500.                  Reid, M. I.,Gancarz, A. J., and Albee, A. L' (1973)    Constrained
 Lipman, P. W., Christiansen, L., and Van Alstine, R. E.
                                    R.                                   least-squares analysis of petrologic problems with an applica-
     (1969) Retention of alkalis by calc-alkalic rhyolites during        tion to lunar sample 12040.Earth Planetary Science Letters,
     crystallizationand hydration. American Mineralogist, 54,286-        17,433445.
     291.                                                             Shaw, D. M. (1970)Trace elementfractionation during anatexis.
 Maaloe, S. and Wyllie, P. J. (1975)Water content of a granitic          Geochimica Cosmochimica
                                                                                      et                 Acta, 37,237-243'
     magma deduced from the sequenceof crystallization deter-         Shaw,H. R., SmithR. L., and Hildreth, W. (1976)      Thermogravi-
     mined experimentally with water-undersaturatedconditions.           tational mechanismsfor chemical variations in zoned magma
     Contributions to Mineralogy and Petrology, 52, 175-191.             chambers (abstr.) Geological Society of America Abstracts
 Mahood, G. A. (1981) Chemical evolution of a Pleistocene                                8,
                                                                         with Programs, 1102.
     rhyolitic center. Sierra La Primavera,Jalisco,Mexico. Contri-    Smirnov, M. V. (1973)Electrode Potentialsin Molten Chlorides.
     butions to Mineralogy and Petrology, 77, 129-149.                   Nauka Press.Moscow.
236                                      CHRISTIANSEN ET AL.: TOPAZ RHYOLITES

Smith, R. L. (1979)Ash-flow magmatism.Geological Society of        Webster, E. A. and Holloway, J. R. (19E0)The partitioning of
  America Special Paper l8O,5-27.                                   REE's, Sc, Rb and Cs betweena silicic melt and a Cl fluid.
Staatz, M. H. and Carr, W. J. (lt}64) Geology and mineral           EOS,61, 152.1
  depositsof the Thomas and Dugway Ranges,Juab and Tooele          Wendlandt, R. F. and Harrison, W. J. (1979)Rare-earthparti-
  Counties, Utah. U.S. Geological Survey Professional paper         tioning between immiscible carbonateand silicate liquids and
  415.                                                              CO2vapor. Resultsand implications for the formation of light-
Taylor, R. P., Strong, D. F., and Fryer, B. J. (t981) Volatile      rare-earth-enrichedrocks. Contributions to Mineralogy and
  control of constrastingtrace element distributions in peralka-    Petrology,69,4@419.
  line granitic and volcanic rocks. Contributions to Mineralogy    Wylfie, P. J, (1979)Magmasand volatile components.American
  and Petrology, 77, 267-27  l.                                     Mineralogist, @, 4f9-500.
Turley, C. H. and Nash, W. P. (1980)Petrology of late Tertiary
  and Quaternary volcanism in western Juab and Millard Coun-
  ties, Utah. Utah Geological Mineralogical Survey Special                     Manuscript received, March 3, 1983;
  study 52, l-33.                                                          acceptedfor publication, September27, 1983.

To top