Fig. 7. Cation variation diagrams for tourmalines from the Dernburg formation, Marmora terrane; data fields after Henry
and Guidotti 1985, 1, Li-rich felsic intrusives; 2, Li-poor felsic intrusives; 3, hydrothermally altered granites; 4 and 5,
Al-saturated and Al-undersaturated, respectively, metapelities and metapsammites, 6, Fe
3 +
-rich quartz-tourmaline rocks, calc-silicates, and metapelites; 7, low-Ca ultramafic rocks; 8,
metacarbonate rocks and metapyroxenites.
low d
11
B of + 4.0‰ Fig. 8. All in all, the d
11
B values for tourmaline grains from the Chameis
sub-terrane, though spanning over a relatively large interval, are, in general, very high and
provide strong evidence of a marine evaporitic origin, although a continental setting with highly
evolved
seawater-derived brines
cannot be
excluded.
5. Fluid inclusion chemistry
Eleven samples of quartz-chlorite-tourmaline veins, coarse-grained sparry, pink calcite veins
within dolomite, a quartz vein within a black chert
at the
base of
the evaporite-derived
dolomite, and stratiform tourmalinite all of which, based on microthermometric and textural
observations, appear to be dominated by only a single fluid inclusion generation were selected for
crush-leach analysis at the Department of Geolog- ical Sciences, UCT, to determine the relative
abundance of dissolved ionic species. The fluid inclusions are undersaturated to saturated two or
three-phase vapour, liquid, with or without halite daughter crystal aqueous and CO
2
-rich. They are 1 – 5 mm in size and display highly variable liquid,
vapour ratios which is most likely due to necking of these fluid inclusions after they have reached
the liquid – vapour curve on their retrograde P – T path. About 500 mg of cleaned mineral separate
were crushed under doubly distilled de-ionised high d
11
B values of + 10.7 to + 27.5‰ Fig. 8. One tourmaline sample from a tourmaline-rich
dolomite from a similar succession at Chameis Bay also yielded a high d
11
B value of + 20.1‰. In contrast, one tourmaline sample from a very
coarse-grained tourmaline aggregate from a mag- nesian calcpelite about 1 km to the east of the
sample locality near Bakers Bay has a relatively
Fig. 8. Boron isotopic composition of tourmalines from the Dernburg formation, compared with those of marine and non-marine evaporites and seawater after Jiang, 1998.
Table 3 Sample average concentrations in mgl of solutes in fluid inclusions from the Sholtzberg member
Na Sample type
a
NH
4
K Mg
Mn Ca
Cl Br
SO
4
1 29.18
0.00 7.54
4.87 0.00
29.02 72.58
0.13 1.90
0.00 7.06
4.40 0.00
28.25 26.95
1 63.23
0.21 2.19
2 3.00
0.00 0.92
1.39 0.00
8.61 4.91
0.03 2.63
0.21 2.02
0.31 0.54
16.30 27.90
3 28.45
0.10 17.49
0.27 2.43
0.92 0.85
3 63.96
21.83 39.08
0.15 113.47
0.08 3.71
5.17 0.00
3.34 68.79
4 3.95
0.00 81.26
0.06 2.85
4.43 0.00
30.74 4.03
0.00 113.54
4 2.40
0.09 2.44
3.79 0.00
2.15 23.18
4 4.28
0.00 315.29
2.26 5
0.14 0.69
1.93 0.00
2.38 5.45
0.00 2.00
0.60 5
0.10 0.45
1.87 0.00
2.71 4.20
0.00 2.28
0.07 0.43
1.58 0.00
0.00 1.77
5 3.20
0.00 0.56
a
1, Sparry calcite vein; 2, quartz vein in chert; 3, quartz–chlorite–tourmaline vein in meta-evaporite; 4, tourmaline nodule; 5, stratiform tourmalinite layer.
water and the resultant leachate was analysed using high performance ion chromatography
HPIC for Na
+
, K
+
, NH
4 +
, Ca
2 +
, Mg
2 +
, Mn
2 +
, Cl
−
, F
−
, NO
3 −
, SO
4 2 −
reflecting total dissolved sulphur and Br
−
. The choice of eluent Na
2
CO
3
and NaHCO
3
for the analysis of anion concentra- tions precluded the determination of carbonic spe-
cies in the leachates. Precision is estimated to be better than 9 2 for most ions except for K
+
, for which it is 9 4. The lower limits of detection are
on the order of 0.001 mgl for further details on the technique see Frimmel et al., 1999. Each
sample was analysed at least twice and the means of the concentrations obtained are reported Table
3. As the exact volume of fluid inclusions taken up into the leachate during crushing is not known, the
absolute concentrations are of little significance but molar element ratios Fig. 9 are more informa-
tive. No F
−
and NO
3 −
were detected and NH
4 +
was found only in small amounts. Similarly, none of the samples analysed contained measurable
Mn
2 +
, except for the quartz vein in the chert. Using HPIC, Br
−
was below the detection limit and was then analysed for using ICP-MS, which is
by an order of magnitude more sensitive for this element. The Br
−
contents were found to be very low and in the tourmalinite-hosted fluid inclusions
not detectable. The dominant cation in all samples is Ca
2 +
, followed by Na
+
, Mg
2 +
, and K
+
Fig. 9a. Among the anions, Cl
−
dominates except for some tourmaline-hosted inclusions in which SO
4 2 −
is present in higher concentrations Fig. 9b. Charge balance calculation suggests the presence
of a major additional anion, most likely CO
3 2 −
, which could not be analysed for directly.
In order to assess the extent to which water derived from the evaporation of seawater is in-
Fig. 9. Sample-average molar cation and anion distribution of leachates from fluid inclusions hosted by stratiform tourmalin-
ite and various vein types in the meta-evaporite sequence of the Dernburg formation. Analytical uncertainties correspond
to size of symbols.
Fig. 10. Na – Cl – Br systematics of fluid inclusion leachates from various vein types in the meta-evaporites sequence of the
Dernburg formation for legend see Fig. 9. Analytical uncer- tainties correspond to size of symbols.
6. Discussion