Lithos 77 (2004) 841 – 857

www.elsevier.com/locate/lithos

An updated classification scheme for mantle-derived garnet,

$
for use by diamond explorers
Herman S. Grütter a,b,*, John J. Gurney c, Andrew H. Menzies c, Ferdi Winter a
a
De Beers GeoScience Centre, PO Box 82232, Southdale, Gauteng 2135, South Africa
b
Mineral Services Canada, 205-930 Harbourside Drive, North Vancouver, BC, Canada V7P 3S7
c
Mineral Services South Africa, PO Box 38668, Pinelands 7430, South Africa
Received 27 June 2003; accepted 14 December 2003
Available online 17 June 2004

Abstract

Mantle-derived garnets recovered in diamond exploration programs show compositional variations in Cr, Ca, Mg, Fe and Ti
that reflect the chemical, physical and lithological environments in which they occur, occasionally together with diamond. The
association of diamond with mantle garnet has progressed through a number of geochemical advances, most notably those of
Dawson and Stephens (1975) and Gurney (1984), which are integrated in this work with less well known petrological advances
made primarily in xenolith and experimental petrology. A simple, robust garnet classification scheme is formulated which
accommodates empirical garnet – diamond relationships for peridotitic (G10, G9, G12), megacrystic (G1), Ti-metasomatised
(G11), pyroxenitic (G4, G5) and eclogitic (G3) lithologies in eight distinct garnet classes. The calcium-saturation characteristics
of harzburgitic (G10), lherzolitic (G9) and wehrlitic (G12) garnets are described by a Ca-intercept projection that also shows
promise as a relative barometer for garnet lherzolite (Grütter and Winter, 1997). Thermobarometric aspects of garnet – diamond
associations are highlighted in the scheme through the use of the minor elements Mn and Na, though analysis by anything other
than an electron microprobe is not required for classification. A ‘‘D’’ suffix is added to the G10, G4, G5 or G3 categories to
indicate a strong compositional and pressure – temperature association with diamond. The scheme remains open to
improvement, particularly with regard to delineation of pyroxenitic (or websteritic) diamond associations and to advances in
Ca-in-garnet and Na-in-garnet thermobarometry.
D 2004 Elsevier B.V. All rights reserved.

Keywords: Diamond exploration; Pyrope; Ca-intercept; Peridotite; Megacryst; Eclogite; Pyroxenite; G10; G9; G1; G3

1. Introduction

Exploration for potentially diamond-bearing intru-

$
Supplementary data associated with this article can be found, sives worldwide usually involves a multi-disciplinary
in the online version, at doi: 10.1016/j.lithos.2004.04.012. approach. One powerful approach is to trace mantle-
* Corresponding author. Mineral Services Canada, 205-930 derived garnet and related indicator minerals to their
Harbourside Drive, North Vancouver, BC, Canada V7P 3S7. Fax:
+1-604-980-6751.
source. The pathfinder minerals are routinely analysed
E-mail address: [email protected] for major and minor elements by electron microprobe,
(H.S. Grütter). in part to identify and differentiate them from visually

0024-4937/$ - see front matter D 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.lithos.2004.04.012
842 H.S. Grütter et al. / Lithos 77 (2004) 841–857

similar minerals of non-mantle origin, but also to tigations and empirical observations on garnet con-
enable qualitative predictions to be made regarding centrates are added to arrive at a simple, robust
the diamond potential of their source rock (e.g. garnet classification scheme that is specifically tai-
Gurney et al., 1993). Garnet and chromite are the lored to the requirements of the diamond explora-
heavy minerals of choice in this application because tionist. The main features of the updated scheme are
they are very common amongst the mineral inclusions (i) reliance only on compositional data obtained by
found in diamond and they usually survive dispersion electron microprobe analysis, (ii) backward compat-
and alteration at the Earth’s surface substantially ibility with previous work, concepts and nomencla-
better than do mantle-derived olivine or pyroxene. ture, (iii) internal consistency with known diamond
Relative to common mantle-derived garnets, the peri- associations, and (iv) ease and transparency of im-
dotitic and eclogitic varieties found included in dia- plementation.
mond have reasonably distinct compositions (e.g.
Gurney, 1984) and simplified compositional screens
based on bivariate scatterplots are commonly used to 2. Data and methods
classify and prioritise mantle-derived garnets recov-
ered during exploration programmes (e.g. Lee, 1993; 2.1. Data sources
Fipke et al., 1995). This methodology developed from
the prior use of simple scatterplots to illustrate geo- The compositions of garnet and associated miner-
chemical relationships of eclogitic and peridotitic als that occur in some 4500 mantle-derived or gran-
garnet to diamond (Sobolev and Lavrent’ev, 1971; ulite-grade lower crustal xenoliths and microxenoliths
Gurney and Switzer, 1973) and to succinctly charac- and as inclusions in about 600 diamonds were com-
terise mantle lithologies in terms of garnet composi- piled from a large number of published sources and
tion (e.g. Sobolev et al., 1973a,b; Switzer, 1975; selected unpublished theses (see Appendix A of this
Schulze, 1995). work, Appendix 1 of Grütter and Moore, 2003 and
Multivariate statistical analysis involving five or Appendix 1 of Grütter and Quadling, 1999). Addi-
more compositional attributes, particularly cluster tional data for minerals in xenoliths are derived from
analysis and derivatives thereof, has also been used the compilations of Schulze (1995, 1996, 1997, 2003).
to relate garnet xenocryst compositions to their The compositions of garnet inclusions in diamonds
parental mantle lithology (e.g. Dawson and Stephens, are from essentially the same sources as listed by
1975; Danchin and Wyatt, 1979), to compare pop- Stachel et al. (2000, and references therein). Garnet
ulations of grains within and between different kim- compositions used in the multivariate statistical stud-
berlite intrusions (Jago and Mitchell, 1989), and to ies of Dawson and Stephens (1975), Danchin and
characterise the compositional attributes of entire Wyatt (1979) and Jago and Mitchell (1989) were
sections of mantle lithosphere (Griffin et al., 2002). kindly provided on request. Data for minerals occur-
Although not specifically concerned with the rela- ring in concentrate derived from kimberlites and
tionship of garnet to diamond, the diamond explora- related rocks are those held by the Kimberlite Re-
tion industry has, through time and by common use, search Group at the University of Cape Town and De
borrowed the cluster-based nomenclature of Dawson Beers Consolidated Mines. The compiled data set
and Stephens (1975) to describe simplified garnet encompasses a variety of analytical conditions, stand-
compositional categories. For peridotitic garnets the ardizations and matrix correction procedures for elec-
term ‘‘G10’’ is thus considered short-hand for ‘‘sub- tron microprobes, and in this regard matches the
calcic’’, ‘‘Ca-undersaturated’’ or ‘‘harzburgitic’’, commercial data typically being used in modern-day
while ‘‘G9’’ represents ‘‘Ca-saturated’’ or ‘‘lherzo- exploration applications. The available analyses were
litic’’. The work reported on here continues this checked to comply within F 3% of ideal stoichiom-
convenient practise whilst updating and formalising etry and were screened to eliminate rare non-mantle
many of the simple classification thresholds. Addi- (i.e. crustal) compositions. Majoritic garnet analyses
tional constraints, based primarily on garnet phase- (>3.1 Si cations per 12 oxygens) were identified for a
relations, the results of experimental petrology inves- limited number of garnets occurring as inclusions in
 H.S. Grütter et al. / Lithos 77 (2004) 841–857 843

diamond; they are not represented elsewhere in the ene– garnet barometer of Nickel and Green (1985) in
compiled data sets. combination with a clinopyroxene –solvus thermome-
ter (Nimis and Taylor, 2000) for lherzolites and pyrox-
2.2. Diamond-facies nomenclature, geotherms and enites, or the olivine –garnet thermometer (O’Neill and
pressure– temperature data Wood, 1979) for garnet harzburgites.

Modern-day single-grain thermometry techniques

are in principle capable of assigning individual peri- 3. Petrological constraints
dotitic garnet grains recovered during diamond explo-
ration to the stability field of graphite or diamond (e.g. 3.1. Garnet –chromite – carbon relations in peridotite
Griffin and Ryan, 1993; Grütter et al., 1999), but doing
so requires knowledge of the local geotherm. This is Compositional and phase-relations between Cr-py-
commonly an unknown in exploration programmes rope garnet, chromite and carbon in model peridotite
and we therefore make the simplifying assumption that systems are illustrated for typically cratonic geothermal
a typical cratonic geotherm pertains to the area being conditions in Fig. 1. The Cr/(Cr + Al) ratios of Ca-free
explored. The updated classification scheme is thus harzburgitic garnet, Ca-saturated lherzolitic garnet and
calibrated by design for geotherms intersecting the coexisting chromite increase with pressure and respec-
graphite/diamond transition at temperatures in the tively attain values of J, K and L at the graphite/
range 920 to 1000 jC, this being equivalent to the diamond transition. For the fixed geothermal condi-
38 to 40 mW/m2 model conductive geotherms of tions illustrated, this implies that any Cr-pyrope garnet
Pollack and Chapman (1977). Individual grains that with Cr/(Cr + Al) greater than line segment J– K must
have a strong compositional association with diamond derive from inside the diamond stability field. The line
and that also fall within the diamond stability field segment J – K has been constrained to the simple
under these conditions are considered to be of ‘‘dia-
mond-facies’’ and are labelled with the suffix ‘‘D’’ in
the classification scheme. The general validity of this
approach can be tested by comparing garnet composi-
tions in concentrates from cratonic sources that are
known to be diamondiferous or barren. The actual
geotherm could also be verified by employing single-
grain thermobarometers (Ryan et al., 1996; Nimis and
Taylor, 2000; Grütter and Moore, 2003).
Current xenolith thermobarometry techniques are
capable of discriminating pressure – temperature ( P –
T) conditions for naturally occurring mantle xenoliths
to within approximately 3.0 kbar (Pearson et al., 1994;
Table 2 of Taylor, 1998). Allowing for this error margin
on either side of the diamond stability curve ( P
[kbar] = 19.4 + 0.025*T [jC], Kennedy and Kennedy,
1976) permits assignment of mantle xenoliths to the
stability fields of graphite ( P < 16.4 + 0.025*T) or Fig. 1. Partly schematic pressure – composition section illustrating
diamond ( P >22.4 + 0.025*T) with relatively high con- the influence of Cr/(Cr + Al) ratio on spinel – garnet (Spl – Grt)
fidence. Such graphite- or diamond-stable assignments assemblage relations in peridotite (modified after Webb and Wood,
for mantle xenoliths are used below to validate tem- 1986). The graphite/diamond (Gph/Dia) transition occurs at fixed
pressure for a fixed geotherm and hence specifies garnet and
perature and pressure effects on garnet MnO content
coexisting chromite compositions J, K and L in peridotite
and Ca-intercept variations. The requisite pressures and assemblages. Line segment J – K represents a compositional tie line
temperatures were calculated for garnet-bearing xen- which corresponds to the graphite – diamond constraint illustrated in
oliths in our compiled data base using the orthopyrox- Fig. 3.
844 H.S. Grütter et al. / Lithos 77 (2004) 841–857

relationship Cr2O3 = 5.0 + 0.94*CaO (in wt.%) for low-

Ca Cr-pyrope garnet compositions by inspection of
experimental (Malinovsky and Doroshev, 1977) and
natural assemblage relations in chromite-saturated,
carbon-bearing garnet harzburgite xenoliths derived
from cratonic kimberlites (Pokhilenko et al., 1991,
1993; Grütter, 1994; Grütter and Sweeney, 2000;
Menzies, 2001, see Fig. 3). The relationship is desig-
nated as the graphite –diamond constraint (the GDC)
and we accordingly use Cr 2O 3 z 5.0 + 0.94 * CaO
(wt.%) in the updated classification scheme as a petro-
logically defined limit that aids in differentiating dia-
mond-facies low-Ca garnet compositions (‘‘D’’ suffix).
The relations shown in Fig. 1 illustrate that pyrope
garnets with Cr/(Cr + Al) less than line segment J– K
may also occur inside the diamond stability field, but
Fig. 2. Frequency histogram of MnO content in garnet for diamond-
only when they occur in chromite-free peridotite or graphite-facies peridotite xenoliths from cratonic kimberlites. The
assemblages. Thus about 46% of the 348 peridotitic diamond- or graphite-facies designation is based on conventional
Cr-pyrope inclusions in diamond in our database have thermobarometry, subject to an error of F 3 kbar (see Section 2).
Cr 2O 3 contents lower than the GDC (i.e. 1.0 < Garnet MnO contents overlap within the range 0.25 to 0.45 wt.%
Cr2O3 < 5.0 + 0.94*CaO, see also Fig. 5 of Gurney MnO, but we adopt a graphite/diamond division at 0.36 wt.% MnO
(arrow) for the purposes of the classification scheme. This choice
et al., 1993). An additional constraint is required to correctly classifies as diamond-facies some 268 of 336 peridotitic
correctly identify these Cr-pyrope garnets as being of garnets included in diamond.
diamond-facies, and we use their MnO content for this
purpose. Cr2O3), and it is possible that their MnO contents
have been artificially inflated by as much as 0.20
3.2. MnO in pyrope garnet wt.% during electron microprobe analysis as a result
of a Cr-Kh peak (at 5.947 keV) overlapping that of
Detailed empirical work on upper mantle xenoliths the Mn-Ka peak (at 5.985 keV). This analytical
has documented an inverse relationship between the artifact is particularly noticeable for high-Cr2O3 gar-
MnO content of peridotitic garnet and xenolith equil- net compositions reported as inclusions in Chinese
ibration temperature (Delaney et al., 1979; Fig. 13 of diamonds (Wang et al., 2000), but it appears to also
MacGregor, 1979; Smith et al., 1991). Our thermo- be present in a few other data sets investigated in the
barometric categorization of 751 garnet harzburgite course of this work. The peak overlap is normally
and lherzolite xenoliths derived from cratonic, dia- readily resolved during wavelength-dispersive analy-
mond-bearing kimberlites shows a value of MnO < sis of Mn at low concentrations by selection of a LiF
0.36 wt.% ( F 0.1 at 1r) to be characteristic of rather than a PET crystal.
peridotitic garnets that occur inside the diamond
stability field (Fig. 2). The threshold has a large error 3.3. Garnet Ca-intercept variations
( f 30%) and should be used with care, but since it
correctly classifies 268 of 336 peridotitic garnets The close association of diamond with Ca-under-
included in diamond (i.e. 80% of those analysed for saturated garnet compositions in preference to Ca-
MnO), it may be employed to good effect in the saturated compositions (Gurney and Switzer, 1973;
updated scheme to differentiate diamond-facies low- Gurney, 1984) suggests that the G10/G9 divide (Fig.
Ca garnet compositions (Fig. 2). 3) may be empirically correlated with pressure – tem-
It is instructive to note that 52 of the 68 incor- perature conditions falling just inside the diamond
rectly classified Cr-pyrope inclusions in diamonds stability field. Pressure values of 45 to 50 kbar are
have relatively high Cr2O3 contents (8 to 15 wt.% considered appropriate for cratonic geotherms by
 H.S. Grütter et al. / Lithos 77 (2004) 841–857 845

projection to 0 wt.% Cr2O3, adapted from that given in

Grütter and Winter (1997), is formulated as (in wt.%):

IF CaO V 3.375 + 0.25*Cr2O3

THEN CA_INT = 13.5*CaO/(Cr2O3 + 13.5)

ELSE CA_INT = CaO 0.25*Cr2O3.

Fig. 3 illustrates that the projection contains geo-

metric elements of the J-score developed by Gurney for
harzburgitic garnets (as quoted in Lee, 1993), and also
of the ‘‘reduced CaO values’’ calculated and used by
Hatton and Gurney (1987, their Fig. 237c) to represent
Ca-content relationships amongst eclogitic, websteritic
and lherzolitic garnets. The current formulation fol-
lows that of Grütter and Winter (1997) in that it sweeps
smoothly from a CA_INT value of 0.0 across the G10
Fig. 3. Conventional garnet Cr2O3 vs. CaO diagram showing the field to 3.375 at the G10/G9 divide and continuously
85% line of Gurney (1984) (commonly known as the G10/G9 extends to substantially higher values (Fig. 3).
divide) and the graphite – diamond constraint of Grütter and
Sweeney (2000). The downward-pointing arrows with CA_INT
Typical CA_INT values calculated for Cr-pyrope
values illustrate the geometric effect of the calcium-intercept garnets from cratonic, peri-cratonic and off-craton
projection discussed in this work (modified after Grütter and settings are shown in histogram format in Fig. 4A to
Winter, 1997). 4C. A general decrease of CA_INT values for lherzo-
litic garnet populations is observed for these three
settings, in agreement with the pressure effects noted
Gurney et al. (1993, p. 2430) and Gurney and Zweis- by previous workers and summarised above. A no-
tra (1995, p. 297). Prior observations on garnet con- tional divide at CA_INT f 4.3 partially separates
centrates from off-craton kimberlites illustrated that, at diamond-bearing kimberlite settings (central Slave,
fixed Cr2O3 content, the CaO content of lherzolitic central Kaapvaal and Somerset Island) from settings
garnet populations increases by f 12% relative to that are barren (Gibeon, Karoo, East Griqualand, The
cratonic conditions (Boyd and Gurney, 1982), the Thumb), but there is substantial overlap of CA_INT
implication being that the CaO variation may be an values. It follows that CA_INT values for individual
effect of decreased pressure. Garnet CaO variations of lherzolitic garnets cannot be assigned unambiguously
similar magnitude are well known from high-pressure to any one of the three cratonic settings considered,
experimental results in model and natural lherzolitic nor, by inference, to diamond-stable conditions (Fig.
or pyroxenitic assemblages, where they are found to 4A to C). Further work is necessary to improve the
be non-linear with respect to pressure, dependent on accuracy of the Ca-in-garnet thermobarometer and to
temperature and the Na2O content of coexisting cli- incorporate additional compositional factors not
nopyroxene, and most useful for relative thermobar- accounted for by the simple Ca-intercept projection
ometry at pressures inside the graphite stability field employed here (see also Brenker and Brey, 1997).
(Kushiro et al., 1967; Boyd, 1970; Akella, 1976; Brey However, the observed overlap and consequent lack
et al., 1986, 1990; Brenker and Brey, 1997). These of barometric constraint does not affect the utility of
circumstances indicate that the G10/G9 divide may be Ca-intercept values in classifying garnet composi-
utilized as an appropriately situated static reference tions. For instance, it is apparent from Fig. 4 that
from which to measure the Ca-saturation character- CA_INT values for lherzolitic garnets range upward
istics of harzburgitic as well as lherzolitic garnets. from 3.375 and show a natural upper limit at a value
Given the G10/G9 divide occurs at CaO = 3.375 + of f 5.4. We hence regard the latter value as the
0.25*Cr2O3 (in wt.%), a calcium-intercept (CA_INT) maximum for lherzolitic garnets derived from the
846 H.S. Grütter et al. / Lithos 77 (2004) 841–857

wide variety of settings that are likely to occur in

diamond exploration applications.
Bulk-tested kimberlites in the Kirkland Lake and
Somerset Island provinces are known to have very low
diamond contents (0 to 0.02 ct/ton, Kjarsgaard and
Levinson, 2002) while kimberlites in the Gibeon
province are not known to contain diamond. The
combination of this information with the results illus-
trated in Fig. 4 suggests that average CA_INT values
of less than 4.3 for large populations of lherzolitic
garnets may be useful to discriminate diamond-stable
from graphite-stable conditions in cratonic upper man-
tle lithospheres. This application of Ca-intercept ther-
mobarometry to populations of lherzolitic garnets
should be approached with care because population
variances are comparatively high (typically F 0.3 at
1r). Since prominent CA_INT modes for lherzolitic
garnet populations from individual kimberlites may
fall either side of the 4.3 threshold, particularly for
small data sets (Fig. 4D), it is considered advisable to
closely inspect the data set and to calculate average
CA_INT values only for data sets that contain 500 or
more lherzolitic garnets.

4. Garnet classifications

4.1. Harzburgitic (G10)

Gurney (1984) correlated 85% of peridotitic garnet

inclusions in diamonds from global sources with the
Ca-poor, Cr-rich harzburgitic pyrope cluster 10 of
Dawson and Stephens (1975) and in doing so ‘‘brand-

Fig. 4. (A – D) Histograms of CA_INT values for peridotitic garnets

with Cr2O3>1.0 wt.% from cratonic settings (A), peri-cratonic
settings (B), off-craton settings (C), and from individual kimberlites
(D). See text for discussion of thresholds indicated at CA_INT
values of 3.375, 4.3 and 5.4. Central Slave data from Armstrong
(2001), central Kaapvaal from the UCT KRG database, Somerset
Island from Jago and Mitchell (1989), Gibeon province, Finsch and
Zeekoegat from the De Beers database. Off-craton xenolith data are
from The Thumb (Ehrenberg, 1978), Karoo kimberlites (Robey,
1981; Nowicki, 1990), East Griqualand kimberlites (Boyd and
Nixon, 1979) and alkali basalt vents at Vitim, Eastern Siberia (Ionov
et al., 1993; Glaser et al., 1999). The Zeekoegat kimberlite occurs
off-craton in the East Griqualand province, southeast of the
Kaapvaal craton. Garnet compositions for most of the southern
African localities mentioned here were originally described in Boyd
and Gurney (1982), which also includes a locality map.
 H.S. Grütter et al. / Lithos 77 (2004) 841–857 847

ed’’ the G10 garnet standard with which the diamond occur in xenolith and microxenolith fragments
potential of exploration projects is often judged (Fig. where their green or grey – green colour is often
3). The association made with diamond in this case is distinctive. Wehrlitic garnets classify predominantly
primarily geochemical and statistical in nature. The in cluster groups 7 and 12 of Dawson and Stephens
presence of graphite in garnet harzburgite xenoliths (1975), with minor occurrences in their cluster
(Nixon et al., 1987; Viljoen et al., 1994) implies an group 11. Here we follow the data compilations of
association of both polymorphs of carbon with low- Sobolev et al. (1973a,b) and Schulze (1993, 2003)
Ca G10 garnet compositions. The Cr-saturation char- to establish compositional limits for an updated and
acteristics of G10 garnets (Figs. 1 and 3) and/or their simplified wehrlitic garnet category labelled G12, as
MnO content (Fig. 2) may be used to specifically follows:
highlight diamond-facies G10 garnets. G10 garnets in
our classification scheme are thus compositionally Cr2O3 [wt.%]: z 1.0 to < 20.0
characterised by: CA_INT [wt.%]: >5.4
CaO [wt.%]: < 28.0
Cr2O3 [wt.%]: z 1.0 to < 22.0 MgO [wt.%]: >5.0
CA_INT [wt.%]: 0 to < 3.375
MGNUM: z 0.75 to < 0.95 It is noted that crustal uvarovitic garnets commonly
have extremely high CaO contents, generally well
where MGNUM=(MgO/40.3)/(MgO/40.3 + FeOt/ over 28 wt.% CaO (Schulze, 1993).
71.85) [oxides in wt.%]. G10D diamond-facies gar-
nets additionally have (in wt.%): 4.4. Low-Cr megacrysts (G1)

Cr2O3 z 5.0 + 0.94*CaO, or Garnets belonging to the low-Cr suite of mega-

Cr2O3 < 5.0 + 0.94*CaO and MnO < 0.36. cryst minerals may occur in high relative abundance
in certain kimberlites (e.g. Monastery), but Schulze
4.2. Lherzolitic (G9) (1987) notes their occurrence also in other mantle-
derived magma types like alnoites (e.g. Solomon
Cr-pyrope garnets derived from lherzolites are Islands) and in certain alkali basalts (e.g. Vitim,
by far the most abundant garnet type recovered in see Litasov, 2000). Megacrystic garnets are typically
diamond exploration applications. Their statistical coarse-grained (2 –10 cm) and significantly fractured,
association with diamond is weak (15% of peridotitic and due to their relatively Fe-rich and Ti-rich nature
inclusions in diamond, Gurney, 1984), particularly generally endure chemical weathering better than
given their high relative abundance as xenocrysts in several other mantle-derived garnet types. Disaggre-
diamondiferous kimberlites (e.g. Gurney and Swit- gation of megacrysts in the secondary environment
zer, 1973, also Fig. 4). Dawson and Stephens can hence release disproportionately abundant path-
(1975) classified lherzolitic garnets in cluster 9, finder minerals derived from kimberlite or kimber-
and in our scheme G9 garnets have the following lite-like intrusives. Their isolation as a specific
compositions: compositional group is therefore important from an
exploration perspective, even though megacrystic
Cr2O3 [wt.%]: z 1.0 to < 20.0 garnets have no established association with dia-
CA_INT [wt.%]: z 3.375 to < 5.4 mond. Garnet megacrysts correspond very closely
MGNUM: z 0.70 to < 0.90 to cluster group 1 of Dawson and Stephens (1975)
and have Cr, Ca and Mg-number characteristics
4.3. Wehrlitic (G12) overlapping those of garnets in pyroxenite and web-
sterite xenoliths. However, the latter are generally
Mantle-derived garnets with high CaO and Cr2O3 less titaniferous at any given Mg-number (Fig. 5).
content are only very rarely described as inclusions in Based predominantly on the compilations of mega-
diamonds (Sobolev et al., 1970), but are known to cryst data in Jakob (1977), Bell and Rossman (1992)
848 H.S. Grütter et al. / Lithos 77 (2004) 841–857

Fig. 5. (A, B) Diagrammatic summary of Cr, Ca, Ti, Mg and Fe compositional characteristics of garnets with moderate to low Cr2O3 content
derived predominantly from kimberlite sources, but also including occurrences recorded in related rocks and alkali basalts. At any given Mg-
number, megacrystic garnets (MEG, n = 515) generally have higher Ti content than garnets in eclogite xenoliths (ECL, n = 1113) and pyroxenite
xenoliths (PRX, n = 391). The pyroxenite category includes 139 data for garnet websterite xenoliths.

and Schulze (1997, 2003), we characterise megacryst depending slightly on the classification method used.
compositions as follows (see Fig. 5): We have adopted their group 11 as representative of
high-TiO2 peridotitic garnet compositions and de-
Cr2O3 [wt.%]: 0 to < 4.0 scribe them as follows.
CA_INT [wt.%]: z 3.375 to < 6.0
MGNUM: z 0.65 to < 0.85 Unlike G1 (i.e. G1 grains should be identified first
TiO2 [wt.%]: z 2.13 2.1*MGNUM and excluded as possible G11’s):
TiO2 [wt.%]: < 4.0 Cr2O3 [wt.%]: z 1.0 to < 20.0
CA_INT [wt.%]: z 3.0
Many investigators have shown that Ca – Fe – Ti CaO [wt.%]: < 28.0
melt-metasomatism drives garnet compositions in MGNUM: z 0.65 to < 0.90
high-temperature mantle peridotites toward those of TiO2 [wt.%]: z 2.13 2.1*MGNUM
megacrystic garnets (e.g. Burgess and Harte, 1999 and TiO2 [wt.%]: < 4.
references therein). A compositional overlap between
megacrystic garnets and certain lower-Cr2O3, high- 4.6. Pyroxenitic, websteritic and eclogitic (‘‘G4’’ and
TiO2 peridotitic garnets may thus occur, but we make ‘‘G5’’)
no discrimination in our classification scheme because
neither diamond nor graphite is associated with melt- Dawson and Stephens (1975) did not classify
metasomatic assemblages in peridotites. McCammon garnets that occur in pyroxenite (and websterite)
et al. (2001) have indicated that the absence of carbon mantle xenoliths into a specific group, but included
may be a consequence of progressive oxidation during them within their G9 (dominantly lherzolitic) and G3
melt-metasomatism. (dominantly eclogitic) categories. Pyroxenitic garnets
were also left undifferentiated in a recent garnet
4.5. High-TiO2 peridotitic (G11) classification scheme (Schulze, 2003). These moder-
ate- to low-Cr garnets are important to diamond
Classification runs conducted by Dawson and explorers due to a distinct association with diamond
Stephens (1975, pp. 601 – 602) on high-TiO2 garnets (e.g. Gurney et al., 1984; Aulbach et al., 2002), and as
derived from ‘‘sheared’’ peridotites returned instances possible indicators of lithosphere destruction (see
in their cluster groups 1, 2, 9, 10 and mostly 11, Pokhilenko et al., 1999). Pyroxenitic/websteritic gar-
 H.S. Grütter et al. / Lithos 77 (2004) 841–857 849

nets are easily differentiated from megacrysts by Mg-

number and TiO2 content (Fig. 5), but, like previous
investigators, we also found significant compositional
overlaps to occur with low-Cr peridotitic and eclogitic
garnets. A suitable compromise was found by allow-
ing two categories of pyroxenitic garnets.
Pyroxenitic garnets similar to, but richer in Fe than
moderate- to low-Cr G9 garnets are designated ‘‘G5’’
in the updated scheme, although the original term
Fig. 6. (A, B) Mg – Ca – (Fe + Mn) diagrams illustrating the
referred to an Fe-rich eclogitic category. The updated compositional range of garnet in (A) carbon-free eclogite xenoliths
G5 garnet category is defined by: from kimberlites (n = 687), and (B) carbonaceous eclogite xenoliths
from kimberlites (n = 100). Compositions of worldwide eclogitic
TiO2 [wt.%]: < 2.13 2.1*MGNUM garnets outlined in stipple (after Haggerty, 1995). Diagram is from
Cr2O3 [wt.%]: z 1 to < 4.0 Grütter and Quadling (1999).
CA_INT [wt.%]: z 3.375 to < 5.4
MGNUM: z 0.3 to < 0.7. CaO content, as in the garnet classification scheme of
Schulze (2003), provides little advantage for the
Pyroxenitic garnets lower in Cr than G9 garnets, but diamond explorer and we accordingly choose to
with compositions overlapping low-Ca eclogitic gar- define our eclogitic garnet category across a range
nets are designated as group ‘‘G4’’. It is recognized that of compositions, as outlined below. It is noted that
this group contains eclogitic, pyroxenitic and webster- these compositional limits also encompass the ranges
itic garnets and that the adopted nomenclature departs observed for garnets in alkremite and certain lower
somewhat from the titaniferous ferroan eclogitic garnet crustal garnet granulite xenoliths. Our eclogitic G3
category originally envisioned by Dawson and Ste- category has:
phens (1975). The updated G4 category is defined by:
Cr2O3 [wt.%]: 0 to < 1.0
TiO2 [wt.%]: < 2.13 2.1*MGNUM CaO [wt.%]: z 6 to < 32.0
Cr2O3 [wt.%]: < 1.0 MGNUM: z 0.17 to < 0.86
CaO [wt.%]: z 2.0 to < 6.0 TiO2 [wt.%]: < 2.13 2.1*MGNUM
MGNUM: z 0.3 to < 0.90. TiO2 [wt.%]: < 2.0

4.7. Eclogitic (G3) 4.8. Na2O in G3, G4 and G5 garnets

Small diamond-bearing eclogite xenoliths are Eclogitic garnet inclusions in diamond are known
known to have in-situ grades equivalent to 650 to to commonly have Na2O > 0.07 wt.% (Sobolev and
20,000 ct/ton (see Helmstaedt, 1993), and for this Lavrent’ev, 1971; McCandless and Gurney, 1989),
reason alone eclogitic garnets represent extremely though this threshold provides incomplete discrimi-
important pathfinder minerals for diamond explorers. nation from garnet compositions in graphite-bearing
Eclogitic garnets are aluminous and show large var- eclogite xenoliths (Grütter and Quadling, 1999).
iations in FeO, MgO and CaO, to the extent that Further investigation of the phase-relations of car-
Dawson and Stephens (1975) required five separate bon, garnet and sodic pyroxene may yield a basis
cluster groups to describe their compositional varia- for accurately constraining diamond-facies eclogitic,
tion (their groups 3, 4, 5, 6 and 8). A compilation of websteritic and pyroxenitic garnet compositions,
garnet compositions in carbon-free and carbonaceous thereby permitting the suffix ‘‘D’’ to be added to
eclogites shows that carbon is not preferentially asso- either of these garnet categories with high confi-
ciated with eclogitic garnets of particular Fe –Mg –Ca dence. In the interim the Na2O > 0.07 wt.% thresh-
compositions (Fig. 6). This implies that subdivision of old noted in Gurney (1984) and documented further
eclogitic garnets on the basis of variable FeO, MgO or in Gurney et al. (1993, their Fig. 8) could be
850 H.S. Grütter et al. / Lithos 77 (2004) 841–857

applied for this purpose, leading to garnet categories

G3D, G4D and G5D within the framework of the
current classification scheme. The G3D and G4D
categories would constitute a replacement for the term
‘‘Group 1 eclogite’’ which originally referred to a
coarse-grained eclogite texture (MacGregor and Cart-
er, 1970), but now also has compositional connotations
(McCandless and Gurney, 1989).

4.9. Unclassified (G0)

Kimberlites and related mantle-derived magmas

occasionally contain xenocrystic garnets derived from
uncommon, unusual or ‘‘polymict’’ mantle lithologies.
The scheme proposed here makes no specific provi-
sion to classify such grains, instead leaving them to Fig. 7. G-number nomenclature of the classification scheme as
collect in an unclassified category labelled G0 by viewed in a conventional Cr2O3 vs. CaO diagram. Megacryst group
default. Manual inspection of this group may reveal G1 (stippled parallelogram) does not actually overlap groups G3,
their affinity to the groups defined above. G4, G5, G9 or G12 since it occurs at higher TiO2 content (see Fig.
5B). Pyroxenitic categories G5 and G4 are indicated by fill pattern.
Group G5 garnets are separated from G9 garnets by a Mg-number
< 0.7 threshold. Note unclassified category G0 at low CaO and
5. Summary and implementation Cr2O3 content.

The current classification scheme is formulated to overlapping with lherzolitic (G9) garnets at Mg-num-
be as simple as possible, whilst also trying to address bers >0.7 and by defining a very low-Cr ‘‘pyroxenitic’’
the multivariate nature of the classification problem (G4) group that also includes all low-CaO eclogitic
and the diversity of chemical, physical and lithological garnet compositions. A further group of common
environments in which mantle garnets and diamonds eclogitic garnets (G3) extends across a range of Mg –
occur. The compositional fields for garnet categories Fe compositions to much higher CaO content (Fig. 7).
outlined in this work are illustrated in terms of Cr2O3 An unclassified category (G0) completes the scheme.
and CaO contents in Fig. 7. Compositional overlaps It is recommended that implementation of the
have been resolved in order to keep the scheme robust, scheme sequentially tests an unknown grain for com-
implying that certain simplifying choices have been positional compliance in the order G1 – G11 –G10 –
made which reflect the needs of diamond exploration- G9 – G12 – G5 – G4 – G3 – G0, slightly different from
ists, rather than those of mantle researchers (for which the order in which they are defined above. A strong
see Schulze, 2003). Thus harzburgitic (G10), lherzo- geochemical and petrologic association with diamond
litic (G9) and wehrlitic (G12) garnet compositions are is indicated by adding a ‘‘D’’ suffix, currently only
separated in the scheme by recognizing natural bounds applicable to G10, G5, G4 and G3 garnet compositions.
in Ca-intercept values (e.g. Figs. 4 and 7), the latter The scheme has been applied to the garnet com-
being a continuous geometric function anchored to the positions used in the multivariate studies of Dawson
well-known G10/G9 divide of Gurney (1984, see Fig. and Stephens (1975), Danchin and Wyatt (1979) and
3). Megacrystic (G1) and high-TiO2 peridotitic (G11) to the databases compiled for this investigation, with
garnets occur on the Ti-rich and Mg-rich side of all results summarised in Tables 1– 4. Known petroge-
other garnet compositions (Fig. 5), and their classifi- netic and lithological associations are given in the left-
cation presents the only computational directive of the hand column of each table and garnet classifications
scheme: G1 and G11 categories have to be assigned according to the current scheme are aligned in similar
prior to any other. At low Cr2O3 content a compromise petrogenetic associations along the upper row. Mod-
is made by excluding pyroxenitic (G5) garnets from erate to high degrees of correlation are evident as
 H.S. Grütter et al. / Lithos 77 (2004) 841–857 851

Table 1 Table 3
Dawson and Stephens (1975) database Xenolith database compiled for this investigation
N = 398 G10 G9 G11 G1 G5 G4 G3 G12 G0 N = 4532 G10 G9 G11 G1 G5 G4 G3 G12 G0
n= 83 77 20 45 2 37 86 26 22 n= 350 1495 380 601 20 698 872 83 33
DINCL 53 64 – 4 – – 9 17 – 6 HZB 284 97 – 3 – – – – – 0
HZB 2 – 100 – – – – – – – LHZ 1378 2 79 13 3 – 2 – 2 0
LHZ 34 – 71 15 6 – 3 3 3 – LHZ_DEF 316 – 33 42 25 – – – 1 –
PRX 13 – 39 – 15 – 39 8 – – PER 240 14 61 16 3 – 5 – 2 0
ECL 99 – 1 – – – 20 73 – 6 MEG 515 – 0 3 88 0 6 1 0 1
WEH 2 – – – – – – – 100 – PRX 252 0 24 – 2 7 55 8 2 1
KCONC 180 26 19 7 22 1 3 2 13 7 WEB 139 – 31 1 2 1 62 – 2 1
UNK 15 13 67 – 7 7 – – – 7 ECL 1113 – 1 – 1 – 32 64 1 1
Classification of garnet compositions by the current scheme. ALK 37 – 8 – – – 8 60 3 22
Percentages listed are rounded and calculated as a proportion of WEH 37 – 8 14 – – – – 78 –
the known garnet category given in the left-hand column. Category GRAN 153 – – – – – 26 72 – 2
abbreviations are ALK = Alkremite; DINCL = Inclusion in diamond; MICXEN 68 32 57 2 4 – – – 4 –
DIXEN = Diamondiferous xenolith; ECL = Eclogite; GRAN = Lower Layout and abbreviations as in Table 1.
crustal granulite; HZB = Harzburgite; KCONC = Concentrate from
kimberlite; LHZ = Lherzolite; LHZ_DEF = Deformed lherzolite;
MEG = Megacryst; MICXEN = Microxenolith; OTH = Other; and wehrlitic garnets, and similar correlations are also
PER = Unspecified peridotite; PRX = Pyroxenite; UNK = Unknown; evident in Tables 1, 2 and 4. The re-defined and newly
WEB = Websterite; WEH = Wehrlite. The garnet compositions of introduced low-Cr ‘‘pyroxenitic/websteritic’’ G4 cat-
Dawson and Stephens (1975) and cross-tabulated raw counts for
Tables 1 – 4 are available as digital supplementary data in the online
egory shows acceptably low overlap with the low-Cr
version. G3 ‘‘eclogitic’’ category, but separation of G5 Cr-
bearing pyroxenitic and websteritic garnets from G9
lherzolitic garnets remains a challenge. Garnets in
high-valued vectors running from top left to bottom lower crustal granulite and in alkremite xenoliths
right in each of Tables 1 – 4, implying that upper classify predominantly as eclogitic, as expected. The
mantle garnet compositions are usefully separated current scheme has a low overall incidence of unclas-
and categorized by the current classification scheme. sified (G0) garnets, a noteworthy feature given the
Table 3 shows good correlations to exist for harzbur- significantly expanded database compiled for this
gitic, lherzolitic, unspecified peridotitic, megacrystic investigation (Tables 3 and 4).
Updated statistics for garnets included in diamond
are given in Table 4. The data set is dominated by
Table 2 peridotitic inclusions and the ratio G10/(G9 + G10) is
Danchin and Wyatt (1979) database
82%, still essentially the same as that calculated by
N = 1777 G10 G9 G11 G1 G5 G4 G3 G12 G0 Gurney (1984), even though the currently applied
n= 257 455 137 237 8 289 281 86 27 definition for peridotite has a lower Cr2O3 threshold
DINCL 191 57 4 3 2 – 9 24 – 2
DIXEN 30 80 7 13 – – – – – –
HZB 57 39 30 32 – – – – – –
Table 4
LHZ 168 2 81 13 1 1 1 – 1 –
Inclusions in diamonds compiled for this investigation
LHZ_DEF 101 2 33 35 31 – – – – –
PER 27 15 67 4 11 – – – 4 – N = 637 G10 G9 G11 G1 G5 G4 G3 G12 G0
MEG 39 – 3 – 67 – 18 – 10 3 n= 271 60 16 12 1 63 205 1 8
PRX 80 18 30 14 13 – 21 4 1 –
WEB 33 – 46 – 3 3 46 – – 3 D suffix n = 494 n = 255 – – – n = 1 n = 50 n = 188 – –
ECL 332 – 3 0 3 – 33 59 0 1 PER 348 78 16 5 1 – – – – 1
ALK 18 – – – – – 22 72 – 6 WEB 13 – 31 – 15 8 15 – 8 23
WEH 3 – – 33 – – – – 67 – ECL 273 – – – 3 – 22 75 – –
KCONC 698 11 27 5 22 1 17 3 11 2 OTH 3 – – – – – 33 – – 67
Layout and abbreviations as in Table 1. Layout and abbreviations as in Table 1.
852 H.S. Grütter et al. / Lithos 77 (2004) 841–857

(1.0 instead of 2.0 wt.% Cr2O3). The familiar eclogitic Jago, Dan Schulze, the Kimberlite Research Group at
(G3) association with diamond is clearly evident and it the University of Cape Town and the authors listed in
is noted that the newly defined G4 category contains the three cited data appendices (see Section 2.1. Data
more diamond-inclusions than the G9 category, even Sources). This work would have been impossible
though the latter garnets are much more abundant in without their collective effort through many years.
the upper mantle (Table 3). Following the methodol- Journal reviews by Bruce Jago and Gerhard Brey
ogy and thresholds presented in the updated scheme, a improved the overall clarity of the manuscript.
total of 494 of the 637 garnets in our diamond-
inclusion data set are assigned the diamond-facies
‘‘D’’ suffix (i.e. 78%). In particular, the ‘‘D’’ suffix is Appendix A . Data sources
found to be applicable in 255 of 271 G10 (94%), 1 of 1
G5, 50 of 63 G4 (79%) and 188 of 205 G3 (92%) Benoit and Mercier (1986)
compositions (Table 4). These statistics should inspire Bloomer and Nixon (1973)
confidence in the use of the G10D, G5D, G4D and Boyd et al. (2004)
G3D categories by diamond explorers. Boyd and Danchin (1980)
Boyd and Nixon (1978)
Boyd and Nixon (1979)
6. Conclusion Boyd et al. (1993)
Boyd et al. (1997)
The classification scheme outlined above utilises a Burgess and Harte (1999)
few relatively simple criteria to categorise the compo- Carswell et al. (1979)
sitions of garnet grains that may be associated with Cox et al. (1973)
diamond-bearing intrusives. The scheme is reliant Danchin and Boyd (1976)
only on the major and minor element compositional Daniels et al. (1995)
data that industry-standard electron microprobe anal- Dawson et al. (1978)
yses can provide and has superior accuracy compared Dawson et al. (1980)
to historical or contemporary classification schemes Delaney et al. (1979)
because it specifically incorporates both geochemical Delaney et al. (1980)
and petrological constraints that appear to determine Eggler et al. (1987)
the occurrence of peridotitic, eclogitic and ‘‘webster- Ehrenberg (1978)
itic’’ diamonds in the lithospheric upper mantle. Ehrenberg (1982)
Improvements in Ca-in-garnet and Na-in-garnet ther- Exley et al. (1982)
mobarometry are required to further improve the Field and Haggerty (1994)
accuracy of the current scheme. Field et al. (1989)
Franz et al. (1996)
Franz et al. (1997)
Acknowledgements Griffin et al. (1989)
Griffin et al. (1993)
The authors acknowledge the support and encour- Hall (1991)
agement of Mineral Services and the De Beers Group Hervig et al. (1986)
of Companies, particularly during the latter stages of Ionov et al. (1993)
this project. HSG publishes with permission of De Kopylova et al. (2000)
Beers Consolidated Mines and acknowledges influ- Logvinova and Sobolev (1995)
ential discussions with Gerhard Brey, Dave Apter, Luth et al. (1990)
Bruce Wyatt and Peter Williamson. We gratefully MacGregor (1979)
acknowledge the xenolith and diamond-inclusion McCallum and Eggler (1976)
mineral composition data published or compiled and McGee and Hearne (1989)
made available by Barry Dawson, Bruce Wyatt, Bruce Menzies (2001)
 H.S. Grütter et al. / Lithos 77 (2004) 841–857 853

Mitchell (1984) Bell, D.R., Rossman, G.R., 1992. The distribution of hydroxyl in
Mofokeng (1998) garnets from the subcontinental mantle of southern Africa. Con-
trib. Mineral. Petrol. 111, 161 – 178.
Nixon (1987) Benoit, V., Mercier, J.-C.C., 1986. Les enclaves ultramafiques du
Nixon and Boyd (1973) volcanisme alcalin Tertiaire du centre du plateau du Colo-
Nowicki (1990) rado: implications tectoniques. Bull. Soc. Geol. Fr. 6,
Pearson et al. (1990) 1015 – 1023.
Pearson et al. (1994) Bloomer, A.G., Nixon, P.H., 1973. The geology of Letseng-la-terae
kimberlite pipes. In: Nixon, P.H. (Ed.), Lesotho Kimberlites.
Pearson et al. (1999)
Lesotho National Development, Maseru, pp. 20 – 36.
Pokhilenko et al. (1977) Boyd, F.R., 1970. Garnet peridotites and the system CaSiO3 –
Pokhilenko et al. (1991) MgSiO3 – Al2O3. Spec. Pap.-Miner. Soc. Am. 3, 63 – 75.
Pokhilenko et al. (1993) Boyd, F.R., Pearson, D.G., Hoal, K.O., Hoal, B.G., Nixon, P.H.,
Reid et al. (1975) Kingston, M.J., Mertman, S.A., 2004. Garnet Iherzolites from
Robey (1981) Louwrensia, Namibia: Bulk compositions and P/T relations.
Lithos 77, 573 – 592 (this volume).
Rudnick et al. (1994) Boyd, F.R., Danchin, R.V., 1980. Lherzolites, eclogites and mega-
Schulze (1995) crysts from some kimberlites of Angola. Am. J. Sci. 280-A,
Schulze (1996) 528 – 549.
Schulze et al. (1997) Boyd, F.R., Gurney, J.J., 1982. Low-calcium garnets: keys to craton
Schulze et al. (2000) structure and diamond crystallization. Year B.-Carnegie Inst.
Wash. 81, 261 – 267.
Shee (1994)
Boyd, F.R., Nixon, P.H., 1978. Ultramafic nodules from the Kim-
Shee et al. (1989) berley pipes, South Africa. Geochim. Cosmochim. Acta 42,
Simon et al. (in press) 1367 – 1382.
Skinner (1989) Boyd, F.R., Nixon, P.H., 1979. Garnet lherzolite xenoliths from the
Smith (1999) kimberlites of East Griqualand, South Africa. Year B.-Carnegie
Smith and Boyd (1992) Inst. Wash. 78, 488 – 492.
Boyd, F.R., Pearson, D.G., Nixon, P.H., Mertzman, S.A., 1993.
Smith and Wilson (1985) Low-calcium garnet harzburgites from Southern Africa: their
Smith et al. (1991) relations to craton structure and diamond crystallization. Con-
Sobolev et al. (1973a,b) trib. Mineral. Petrol. 113, 352 – 366.
Sobolev et al. (1984) Boyd, F.R., Pokhilenko, N.P., Pearson, D.G., Mertzman, S.A.,
Sobolev et al. (1997) Sobolev, N.V., Finger, L.W., 1997. Composition of the Siberian
cratonic mantle: evidence from Udachnaya peridotite xenoliths.
Stachel et al. (2000)
Contrib. Mineral. Petrol. 128, 228 – 246.
Stiefenhofer et al. (1999) Brenker, F.E., Brey, G.P., 1997. Reconstruction of the exhumation
van der Westhuizen (1992) path of the Alpe Arami garnet – peridotite body from depths
Viljoen et al. (1994) exceeding 160 km. J. Metamorph. Geol. 15, 581 – 592.
Brey, G.P., Nickel, K.G., Kogarko, L., 1986. Garnet – pyroxene
equilibria in the system CaO – MgO – Al2O3 – SiO2 (CMAS):
prospects for simplified (‘‘T-independent’’) lherzolite barome-
References try and an eclogite-barometer. Contrib. Mineral. Petrol. 92,
448 – 455.
Akella, J., 1976. Garnet pyroxene equilibria in the system CaSiO3 – Brey, G., Köhler, T., Nickel, K.G., 1990. Geothermobarometry in
MgSiO3 – Al2O3 and in a natural mineral mixture. Am. Mineral. four-phase lherzolites: I. Experimental results from 10 to 60 kb.
61, 589 – 598. J. Petrol. 31, 1313 – 1352.
Armstrong, J.P., 2001. Kimberlite Indicator Mineral Chemistry Da- Burgess, S.R., Harte, B., 1999. Tracing lithosphere evolution
tabase (KIMC): A preliminary digital compilation of Kimberlite through the analysis of heterogeneous G9/G10 garnets in peri-
Indicator Mineral Chemistry extracted from publically available dotite xenoliths: I. Major Element Chemistry. In: Gurney, J.J.,
assessment filings; Slave Craton and environs, Northwest Ter- Gurney, J.L., Pascoe, M.D., Richardson, S.H. (Eds.), J.B. Daw-
ritories and Nunavut, Canada. DIAND NWT Geology Division, son Volume Proc. 7th Int. Kimb. Conf. Red Roof Design, Cape
DIAND EGS Open Report 2001-02 (CD-ROM). Town, pp. 66 – 80.
Aulbach, S., Stachel, T., Viljoen, K.S., Brey, G.P., Harris, J.W., Carswell, D.A., Clarke, D.B., Mitchell, R.H., 1979. The petrology
2002. Eclogitic and websteritic diamond sources beneath the and geochemistry of ultramafic nodules from Pipe 200, North-
Limpopo Belt—is slab-melting the link? Contrib. Mineral. Pet- ern Lesotho. In: Boyd, F.R., Meyer, H.O.A. (Eds.), The Mantle
rol. 143, 56 – 70. Sample: Inclusions in Kimberlites and Other Volcanics. Proc.
854 H.S. Grütter et al. / Lithos 77 (2004) 841–857

2nd Int. Kimb. Conf., vol. 2. Amer. Geophys. Union, Washing- tion techniques emphasising indicator mineral geochemistry
ton, DC, pp. 127 – 144. and Canadian examples. Bull.-Geol. Surv. Can., vol. 423.
Cox, K.G., Gurney, J.J., Harte, B., 1973. Xenoliths from the Mat- 86 pp.
suko pipe. In: Nixon, P.H. (Ed.), Lesotho Kimberlites. Lesotho Franz, L., Brey, G.P., Okrusch, M., 1996. Steady state geotherm,
Nat. Develop, Maseru, pp. 76 – 100. thermal disturbances and tectonic development of the lower
Danchin, R.V., Boyd, F.R., 1976. Ultramafic nodules from the Pre- lithosphere underneath the Gibeon kimberlite province, Nami-
mier kimberlite pipe, South Africa. Year B.-Carnegie Inst. Wash. bia. Contrib. Mineral. Petrol. 126, 181 – 198.
75, 531 – 538. Franz, L., Brey, G.P., Okrusch, M., 1997. Metasomatic reequilibra-
Danchin, R.V., Wyatt, B.A., 1979. Statistical cluster analysis of tion of mantle xenoliths from the Gibeon kimberlite province
garnets from kimberlites and their xenoliths. Ext. Abstr. 2nd (Namibia). Russ. Geol. Geophys. 38, 261 – 276.
Kimb. Symp., Cambridge, pp. 22 – 27. Glaser, S.M., Foley, S.F., Gunther, D., 1999. Trace element com-
Daniels, L.R.M., Richardson, S.H., Menzies, A.H., de Bruin, D., positions of minerals in garnet and spinel peridotite xenoliths
Gurney, J.J., 1995. Diamondiferous garnet macrocrysts in the from the Vitim volcanic field, Transbaikalia, eastern Siberia.
Newlands kimberlite, South Africa—Rosetta stones from the Lithos 48, 263 – 285.
Kaapvaal craton keel. Extended Abs. 6th Ink. Kimb. Conf., Griffin, W.L., Ryan, C.G., 1993. Trace elements in garnets and chro-
Novosibirsk, pp. 121 – 123. mites: evaluation of diamond exploration targets. In: Diamonds:
Dawson, J.B., Stephens, W.E., 1975. Statistical classification of Exploration, Sampling and Evaluation. Short Course Proceed-
garnets from kimberlite and associated xenoliths. J. Geol. 83, ings, Prospectors and Developers Assoc. Canada, Toronto,
589 – 607. pp. 185 – 212.
Dawson, J.B., Smith, J.V., Delaney, J.S., 1978. Multiple spinel – Griffin, W.L., Cousens, D.R., Ryan, C.G., Sie, S.H., Suter, G.F.,
garnet peridotite transitions in upper mantle: evidence from a 1989. Ni in chrome pyrope garnets: a new geothermometer.
harzburgite xenolith. Nature 273, 741 – 743. Contrib. Mineral. Petrol. 103, 199 – 202.
Dawson, J.B., Smith, J.V., Hervig, R.L., 1980. Heterogeneity in Griffin, W.L., Sobolev, N.V., Ryan, C.G., Pokhilenko, N.P.,
upper-mantle lherzolites and harzburgites. Philos. Trans. R. Win, T.T., Yefimova, E.S., 1993. Trace elements in garnets and
Soc. Lond., A 297, 323 – 331. chromites: diamond formation in the Siberian lithosphere. Lithos
Delaney, J.S., Smith, J.V., Dawson, J.B., Nixon, P.H., 1979. Man- 29, 235 – 256.
ganese thermometer for mantle peridotites. Contrib. Mineral. Griffin, W.L., Fisher, N.I., Friedman, J.H., O’Reilly, S.Y., Ryan,
Petrol. 71, 157 – 169. C.G., 2002. Cr-pyrope garnets in the lithospheric mantle: 2.
Delaney, J.S., Smith, J.V., Carswell, D.A., Dawson, J.B., 1980. Compositional populations and their distribution in time and
Chemistry of micas from kimberlites and xenoliths: II. Primary- space. Geochem. Geophys. Geosyst. 3, 1073, doi:10.1029/
and secondary-textured micas from peridotite xenoliths. Geo- 2002GC000298.
chim. Cosmochim. Acta 44, 857 – 872. Grütter, H.S., 1994. Spinel – garnet – carbon phase relations in
Eggler, D.H., McCallum, M.E., Kirkley, M.B., 1987. Kimberlite- coarse Kaapvaal-type peridotites and implications for garnet –
transported nodules from Colorado – Wyoming: a record of en- orthopyroxene equilibration. Ext. Abs. Int. Symp. Phys. Chem.
richment of shallow portions of an infertile lithosphere. Spec. Upper Mantle, Sao Paulo, pp. 5 – 7.
Pap.-Geol. Soc. Am. 215, 77 – 89. Grütter, H.S., Moore, R., 2003. Pyroxene geotherms revisited—an
Ehrenberg, S.N., 1978. Petrology of potassic volcanic rocks and empirical approach based on Canadian xenoliths. Ext. Abstr. 8th
ultramafic xenoliths from the Navajo volcanic field, New Int. Kimb. Conf., Victoria, 272 (CD-ROM).
Mexico and Arizona, Unpubl. PhD thesis. Univ. California, Grütter, H.S., Quadling, K.E., 1999. Can sodium in garnet be used
Los Angeles. 259 pp. to monitor eclogitic diamond potential? In: Gurney, J.J., Gurney,
Ehrenberg, S.N., 1982. Petrogenesis of garnet lherzolite and meg- M.D., Pascoe, M.D., Richardson, S.H. (Eds.), J.B. Dawson Vol-
acrystalline nodules from The Thumb, Navajo volcanic field. ume Proc. 7th Int. Kimb. Conf. Red Roof Design, Cape Town,
J. Petrol. 23, 507 – 547. pp. 314 – 320.
Exley, R.A., Smith, J.V., Hervig, R.L., 1982. Cr-rich spinel and Grütter, H.S., Sweeney, R.J., 2000. Tests and constraints on single-
garnet in two peridotite xenoliths from the Frank Smith mine grain Cr-pyrope barometer models: some initial results. Ext.
South Africa: significance of Al and Cr distribution between Abstr. GAC/MAC Annual Joint Meeting, Calgary (CD-ROM,
spinel and garnet. Min. Mag. 45, 129 – 134. GeoCanada 2000).
Field, S.W., Haggerty, S.E., 1994. Symplectites in upper mantle Grütter, H.S., Winter, F., 1997. Pressure, temperature and compo-
peridotites: development and implications for the growth of sition effects on the position of the garnet lherzolite trend.
subsolidus garnet, pyroxene and spinel. Contrib. Mineral. Petrol. Anglo American Research Laboratories Report KR97/0504.
118, 138 – 156. 35 pp.
Field, S.W., Haggerty, S.E., Erlank, A.J., 1989. Subcontinental Grütter, H.S., Apter, D.B., Kong, J., 1999. Crust – mantle coupling:
metasomatism in the region of Jagersfontein, South Africa. evidence from mantle-derived xenocrystic garnets. In: Gurney,
Kimberlites and related rocks: Their mantle/crust setting, dia- J.J., Gurney, J.L., Pascoe, M.D., Richardson, S.H. (Eds.), J.B.
monds and diamond exploration. Proc. 4th Int. Kimb. Conf. Vol. Dawson Volume Proc. 7th Int. Kimb. Conf. Red Roof Design,
2. Spec. Publ.-Geol. Soc. Aust., vol. 14, pp. 771 – 783. Cape Town, pp. 307 – 313.
Fipke, C.E., Gurney, J.J., Moore, R.O., 1995. Diamond explora- Gurney, J.J., 1984. A correlation between garnets and diamonds. In:
 H.S. Grütter et al. / Lithos 77 (2004) 841–857 855

Glover, J.E., Harris, P.G. (Eds.), Kimberlite occurrence and ori- garnet – pyroxene equilibrium in the system MgSiO3 – CaSiO3 –
gins: a Basis for Conceptual Models in Exploration. Geology Al2O3. Earth Planet. Sci. Lett. 2, 460 – 464.
Department and University Extension, University of Western Lee, J.E., 1993. Indicator mineral techniques in a diamond explora-
Australia, Publication 8 143 – 166. tion program at Kokong, Botswana. In: Diamonds: Exploration,
Gurney, J.J., Switzer, G.S., 1973. The discovery of garnets closely Sampling and Evaluation. Short Course Proceedings, Prospec-
related to diamonds in the Finsch pipe, South Africa. Contrib. tors and Developers Assoc. Canada, Toronto, pp. 213 – 235.
Mineral. Petrol. 39, 103 – 116. Litasov, K.D., 2000. Petrology of a megacryst assemblage in Plio-
Gurney, J.J., Zweistra, P., 1995. The interpretation of the major cene – Pleistocene basanites from the Vitim Plateau (Transbaika-
element compositions of mantle minerals in diamond explora- lie). Volc. Seis. 22, 35 – 53.
tion. J. Geochem. Explor. 53, 293 – 309. Logvinova, A.M., Sobolev, N.V., 1995. Morphology and composi-
Gurney, J.J., Harris, J.W., Rickard, R.S., 1984. Silicate and oxide tion of mineral inclusions in chromite macrocrysts from kim-
inclusions in diamonds from the Orapa mine, Botswana. In: berlites and lamproites. Ext. Abstr. 6th Int. Kimb. Conf.,
Kornprobst, J. (Ed.), Kimberlites II: The mantle and crust – man- Navosibirsk, pp. 331 – 332.
tle relationships. Proc. 3rd Int. Kimb. Conf., vol. 2. Elsevier, Luth, R.W., Virgo, D., Boyd, F.R., Wood, B.J., 1990. Ferric iron in
Amsterdam, pp. 3 – 9. mantle-derived garnets: implications for thermobarometry and
Gurney, J.J., Helmstaedt, H., Moore, R.O., 1993. A review of the for the oxidation state of the mantle. Contrib. Mineral. Petrol.
use and application of mantle mineral geochemistry in diamond 104, 56 – 72.
exploration. Pure Appl. Chem. 65, 2423 – 2442. MacGregor, I.D., 1979. Mafic and ultramafic xenoliths from the
Haggerty, S.E., 1995. Upper mantle mineralogy. J. Geodyn. 20, Kao kimberlite pipe. In: Boyd, F.R., Meyer, H.O.A. (Eds.),
331 – 364. The Mantle Sample: inclusions in Kimberlites and Other Vol-
Hall, D.C., 1991. A petrological investigation of the Cross kimber- canics. Proc. 2nd Int. Kimb. Conf. Am. Geophys. Union, Wash-
lite occurrence, Southeastern British Columbia, Canada. ington, DC, pp. 156 – 172.
Unpubl. PhD thesis. Queen’s University, Kingston, Ontario, MacGregor, I.D., Carter, J.L., 1970. The chemistry of clinopyrox-
Canada, 533 pp. enes and garnets of eclogite and peridotite xenoliths from the
Hatton, C.J., Gurney, J.J., 1987. Roberts Victor eclogites and their Roberts Victor mine, South Africa. Phys. Earth Planet. Inter. 3,
relation to the mantle. In: Nixon, P.H. (Ed.), Mantle Xenoliths. 391 – 397.
Wiley, London, pp. 453 – 463. Malinovsky, I.Y., Doroshev, A.M., 1977. Evaluation of P – T con-
Helmstaedt, H.H., 1993. Natural diamond occurrences and tectonic ditions of diamond formation with reference to chrome-bearing
setting of ‘‘primary’’ diamond deposits. In: Diamonds: Explo- garnet stability. Ext. Abs. Second Internat. Kimb. Conf., Octo-
ration, Sampling and Evaluation. Short Course Proceedings, ber 3 – 7, Santa Fe, New Mexico. Unpaginated.
Prospectors and Developers Assoc. Canada, Toronto, pp. 3 – 72. McCallum, M.E., Eggler, D.H., 1976. Diamonds in an upper mantle
Hervig, R.L., Smith, J.V., Dawson, J.B., 1986. Lherzolite xenoliths peridotite nodule from kimberlite in Southern Wyoming. Sci-
in kimberlites and basalts: petrogenetic and crystallographic sig- ence 192, 253 – 256.
nificance of some minor and trace elements in olivine, pyrox- McCammon, C.A., Griffin, W.L., Shee, S.R., O’Neill, H.S.C., 2001.
enes, garnet and spinel. Trans. R. Soc. Edinb. Earth Sci. 77, Oxidation during metasomatism in ultramafic xenoliths from the
181 – 201. Wesselton kimberlite, South Africa: implications for the survival
Ionov, D.A., Ashchepkov, I.V., Stosch, H.-G., Witt-Eickschen, G., of diamond. Contrib. Mineral. Petrol. 141, 287 – 296.
Seck, H.A., 1993. Garnet peridotite xenoliths from the Vitim McCandless, T.E., Gurney, J.J., 1989. Sodium in garnet and po-
volcanic field, Baikal region: the nature of the garnet – spinel tassium in clinopyroxene: criteria for classifying mantle eclo-
peridotite transition zone in the continental mantle. J. Petrol. gites. In: Ross, J. (Ed.), Kimberlites and Related Rocks.
34, 1141 – 1175. Geological Society of Australia Special Publication, vol. 14/2,
Jago, B.C., Mitchell, R.H., 1989. A new garnet classification tech- pp. 827 – 832.
nique: divisive cluster analysis applied to garnet populations from McGee, E.S., Hearne, B.C., 1989. Primary and secondary mineral-
Somerset Island kimberlites. In: Ross, J. (Ed.), Kimberlites and ogy of carbonated peridotites from the Macdougal Springs diat-
Related Rocks, Vol. 1: Their Composition, Origin and Emplace- reme. Spec. Publ. - Geol. Soc. Aust. 14, 725 – 734.
ment. Spec. Pub.-Geol. Soc. Aust., vol. 14, pp. 297 – 310. Menzies, A.H., 2001. A detailed investigation into diamond-bearing
Jakob, W.R.O., 1977. Geochemical aspects of the megacryst suite xenoliths from Newlands kimberlite, South Africa. Unpubl. PhD
from the Monastery kimberlite pipe. Unpubl. Report, Atomic thesis, Univ. of Cape Town.
Energy Board, Pelindaba, 81 pp. Mitchell, R.H., 1984. Garnet lherzolites from the Hanaus-I and
Kennedy, C.S., Kennedy, G.C., 1976. The equilibrium boundary Louwrensia kimberlites of Namibia. Contrib. Mineral. Petrol.
between graphite and diamond. J. Geophys. Res. 81, 2467 – 2470. 86, 178 – 188.
Kjarsgaard, B.A., Levinson, A.A., 2002. Diamonds in Canada. Mofokeng, S.W., 1998. A comparison of the nickel and the con-
Gems. Gemol. 38, 208 – 238. ventional geothermometers with respect to the Jagersfontein and
Kopylova, M.G., Russell, J.K., Stanley, C., Cookenboo, H., 2000. the Matsoku kimberlite peridotite xenoliths. Unpubl. MSc the-
Garnet from Cr- and Ca-saturated mantle: implications for dia- sis, Univ. Cape Town, 124 pp.
mond exploration. J. Geochem. Explor. 68, 183 – 199. Nickel, K.G., Green, D.H., 1985. Empirical geothermobarometry
Kushiro, I., Syono, Y., Akimoto, S., 1967. Effect of pressure on for garnet peridotites and implications for the nature of the
856 H.S. Grütter et al. / Lithos 77 (2004) 841–857

lithosphere, kimberlites and diamonds. Earth Planet. Sci. Lett. Proc. 7th Int. Kimb. Conf. Red Roof Design, Cape Town, pp.
73, 158 – 170. 689 – 698.
Nimis, P., Taylor, W.R., 2000. Single clinopyroxene thermobarom- Pollack, H.N., Chapman, D.S., 1977. On the regional variation of
etry for garnet peridotites: Part 1. Calibration and testing of a Cr- heat flow, geotherms, and lithospheric thickness. Tectonophy-
in-Cpx barometer and an enstatite-in-Cpx thermometer. Contrib. sics 38, 279 – 296.
Mineral. Petrol. 139, 541 – 554. Reid, A.M., Donaldson, C.H., Brown, R.W., Ridley, W.I., Dawson,
Nixon, P.H., 1987. Kimberlitic xenoliths and their cratonic setting. J.B., 1975. Mineral chemistry of peridotite xenoliths from
In: Nixon, P.H. (Ed.), Mantle Xenoliths. Wiley, London, pp. Lashaine volcano, Tanzania. Phys. Chem. Earth 9, 525 – 543.
215 – 239. Robey, J.V.R., 1981. Kimberlites of the Central Cape province,
Nixon, P.H., Boyd, F.R., 1973. Petrogenesis of the granular and R.S.A. Unpubl. PhD thesis. University of Cape Town, 261 pp.
sheared ultrabasic nodules suite in kimberlites. In: Nixon, P.H. Rudnick, R.L., McDonough, W.F., Orpin, A., 1994. Northern Tan-
(Ed.), Lesotho Kimberlites. Lesotho National Development, zanian peridotite xenoliths: a comparison with Kaapvaal peri-
Maseru, pp. 48 – 56. dotites and inferences on metasomatic interactions. In: Meyer,
Nixon, P.H., van Calsteren, P.W.C., Boyd, F.R., Hawkesworth, C.J., H.O.A., Leonardos, O.H. (Eds.), Kimberlites, Related Rocks
1987. Harzburgites with garnets of diamond facies from south- and Mantle Xenoliths, Proc. 5th Int. Kimb. Conf.,Spec. Publ.-
ern African kimberlites. In: Nixon, P.H. (Ed.), Mantle Xenoliths. CPRM, vol. 1A/94, pp. 336 – 353.
Wiley, London, pp. 523 – 533. Ryan, C.G., Griffin, W.L., Pearson, N.J., 1996. Garnet geotherms:
Nowicki, T.E., 1990. Mantle xenoliths from the Abrahamskraal pressure – temperature data from Cr-pyrope garnet xenocrysts in
kimberlite: a craton-margin geotherm. Unpubl. MSc Thesis, volcanic rocks. J. Geophys. Res. 101, 5611 – 5625.
Rhodes Univ., Grahamstown, 113 pp. Schulze, D.J., 1987. Megacrysts from alkaline volcanic rocks.
O’Neill, H.S.C., Wood, B.J., 1979. An experimental study of In: Nixon, P.H. (Ed.), Mantle Xenoliths. Wiley, London,
Fe – Mg partitioning between garnet and olivine, and its cal- pp. 433 – 451.
ibration as a geothermometer. Contrib. Mineral. Petrol. 70, Schulze, D.J., 1993. Green garnets from South African kimberlites
59 – 70. and their relationship to wehrlites and crustal uvarovites. Spec.
Pearson, D.G., Boyd, F.R., Nixon, P.H., 1990. Graphite-bearing Publ. - Geol. Soc. Aust. 14, 820 – 826.
mantle xenoliths from the Kaapvaal craton: implications for Schulze, D.J., 1995. Low-Ca garnet harzburgites from Kimberley,
graphite and diamond genesis. Ann. Rep. Dir. Geophys. Lab. South Africa: abundance and bearing on the structure and evo-
Carn. Inst. Wash., pp. 11 – 19. lution of the lithosphere. J. Geophys. Res. 100, 12513 – 12526.
Pearson, D.G., Boyd, F.R., Haggerty, S.E., Pasteris, J.D., Field, Schulze, D.J., 1996. Chromite macrocrysts from southern African
S.W., Nixon, P.H., Pokhilenko, N.P., 1994. The characterisation kimberlites: mantle xenolith sources and post-diamond re-equil-
and origin of graphite in cratonic lithospheric mantle: a petro- ibration. In: Kogbe, C.A. (Ed.), Diamond Deposits in Africa.
logical carbon isotope and Raman spectroscopic study. Contrib. Spec. Issue-Africa Geosci. Rev., vol 3, pp. 203 – 216.
Mineral. Petrol. 115, 449 – 466. Schulze, D.J., 1997. The significance of eclogite and Cr-poor meg-
Pearson, N.J., Griffin, W.L., Doyle, B.J., O’Reilly, S.Y., Van Ach- acryst garnets in diamond exploration. Explor. Min. Geol. 6,
terbergh, E., Kivi, K., 1999. Xenoliths from kimberlite pipes of 349 – 366.
the Lac de Gras area, Slave Craton, Canada. In: Gurney, J.J., Schulze, D.J., 2003. A classification scheme for mantle-derived
Gurney, J.L., Pascoe, M.D., Richardson, S.H. (Eds.), J.B. Daw- garnet in kimberlite: a tool for investigating the mantle and
son Volume Proc. 7th Int. Kimb. Conf. Red Roof Design, Cape exploring for diamonds. Lithos 71, 195 – 213.
Town, pp. 644 – 658. Schulze, D.J., Valley, J.W., Viljoen, K.S., Stiefenhofer, J., Spicuzza,
Pokhilenko, N.P., Sobolev, N.V., Lavrent’ev, Y.-G., 1977. Xeno- M., 1997. Carbon isotope composition of graphite in mantle
liths of diamondiferous ultramafic rocks from Yakutian kimber- eclogites. J. Geol. 105, 379 – 386.
lites. Extended Abstracts IInd Int. Kimberlite Conf., Santa Fe. Schulze, D.J., Valley, J.W., Spicuzza, K.J., 2000. Coesite eclogites
Unpaginated. from the Roberts Victor kimberlite, South Africa. Lithos 54,
Pokhilenko, N.P., Pearson, D.G., Boyd, F.R., Sobolev, N.V., 1991. 23 – 32.
Megacrystalline dunites and peridotites: hosts for Siberian dia- Shee, S.R., 1994. Unpublished Compositional Data for Grain
monds. Ann. Rep. Dir. Geophys. Lab. Carn. Inst. Wash., Mounts of Peridotite Xenoliths from Wesselton Mine.
pp. 11 – 18. Shee, S.R., Bristow, J.W., Bell, D.R., Smith, C.B., Allsopp, H.L.,
Pokhilenko, N.P, Sobolev, N.V., Boyd, F.R., Pearson, D.G., Shi- Shee, P.B., 1989. The petrology of kimberlites, related rocks and
muzu, N., 1993. Megacrystalline pyrope peridotites in the lith- associated mantle xenoliths from the Kuruman province, South
osphere of the Siberian platform: mineralogy, geochemical Africa. In: Ross, N. (Ed.), Kimberlites and Related Rocks. Geol.
peculiarities and the problem of their origin. Russ. J. Geol. Geo- Soc. Aust. Spec. Publ., pp. 60 – 82.
phys. 34, 56 – 67. Simon, N.S.C., Irvine, G.J., Davies, G.R., Pearson, D.G., Carlson,
Pokhilenko, N.P., Sobolev, N.V., Kuligin, S.S., Shimuzu, N., 1999. R.W., 2003. The origin of garnet and clinopyroxene in ‘‘deplet-
Peculiarities of distribution of pyroxenite paragenesis garnets in ed’’ Kaapvaal peridotite. Lithos 71, 289 – 322.
Yakutian kimberlites and some aspects of the evolution of the Skinner, C.P., 1989. The petrology of peridotite xenoliths from the
Siberian craton lithospheric mantle. In: Gurney, J.J., Gurney, Finsch kimberlite, South Africa. S. Afr. J. Geol. 92, 197 – 206.
J.L., Pascoe, M.D., Richardson, S.H. (Eds.), P.H. Nixon Volume Smith, D., 1999. Temperatures and pressures of mineral equilibra-
 H.S. Grütter et al. / Lithos 77 (2004) 841–857 857

tion in peridotite xenoliths: Review, discussion, and implica- Stachel, T., Brey, G.P., Harris, J.W., 2000. Kankan diamonds (Guin-
tions. In: Fei, Y., Bertka, C.M., Mysen, B.O. (Eds.), Mantle ea) I: from the lithosphere down to the transition zone. Contrib.
Petrology: Field Observations and high pressure experimenta- Mineral. Petrol. 140, 1 – 15.
tion: a tribute to Francis R. (Joe) Boyd. Spec. Publ.-Geochem. Stiefenhofer, J., Viljoen, K.S., Tainton, K.M., Dobbe, R., Hannweg,
Soc., vol. 6, pp. 171 – 188. G.W., 1999. The petrology of a mantle xenolith suite from Ven-
Smith, D., Boyd, F.R., 1992. Compositional zonation in garnets in etia, South Africa. In: Gurney, J.J., Gurney, J.L., Pascoe, M.D.,
peridotite xenoliths. Contrib. Mineral. Petrol. 112, 134 – 147. Richardson, S.H. (Eds.), P.H. Nixon Volume Proc. 7th Int. Kimb.
Smith, D., Wilson, C.R., 1985. Garnet – olivine equilibration during Conf. Red Roof Design, Cape Town, pp. 836 – 845.
cooling in the mantle. Am. Mineral. 70, 30 – 39. Switzer, G.S., 1975. Composition of garnet xenocrysts from three
Smith, D., Griffin, W.L., Ryan, C.G., Sie, S.H., 1991. Trace-ele- kimberlite pipes in Arizona and New Mexico. Smithson. Con-
ment zonation of garnets in The Thumb: heating and melt-infil- trib. Earth Sci. 9, 1 – 21.
tration below the Colorado plateau. Contrib. Mineral. Petrol. Taylor, W.R., 1998. An experimental test of some geothermometer
107, 60 – 79. and geobarometer formulations for upper mantle peridotites with
Sobolev, N.V., Lavrent’ev, Y.G., 1971. Isomorphic sodium admix- application to the thermobarometry of fertile lherzolite and gar-
ture in garnets formed at high pressures. Contrib. Mineral. Pet- net websterite. Neues Jahrb. Mineral. Abh. 172, 381 – 408.
rol. 31, 1 – 12. van der Westhuizen, A., 1992. The Bellsbank kimberlites, with
Sobolev, N.V., Bartoshinsky, Z.V., Yefimova, E.S., Lavrentyev, special reference to a suite of purple garnet megacrysts from
Y.G., Pospelova, L.N., 1970. Olivine – garnet – chrome diopside the Bobbejaan mine. Unpubl. MSc thesis, University of the
assemblage from Yakutian diamond. Dokl. Akad. Nauk SSSR Orange Free State. South Africa, 88 pp.
192, 134 – 137 (in English). Viljoen, K.S., Robinson, D.N., Swash, P.M., Griffin, W.L., Otter,
Sobolev, N.V., Kharkiv, A.D., Lavrent’ev, Y.G., Pospelova, L.N., M.L., Ryan, C.G., Win, T.T., 1994. Diamond- and graphite-
1973a. Chromite – pyroxene – garnet intergrowths from the Mir bearing peridotite xenoliths from the Roberts Victor kimberlite,
kimberlite pipe. Geol. Geofiz. 12, 15 – 20 (in Russian). South Africa. In: Meyer, H.O.A., Leonardos, O.H. (Eds.), Kim-
Sobolev, N.V., Lavrent’yev, Y.G., Pokhilenko, N.P., Usova, L.V., berlites, Related Rocks and Mantle Xenoliths, Proc. 5th Int.
1973b. Chrome-rich garnets from the kimberlites of Yakutia and Kimb. Conf. Spec. Publ.-CPRM, vol. 1A/94, pp. 285 – 303.
their paragenesis. Contrib. Mineral. Petrol. 40, 39 – 52. Wang, W., Sueno, S., Takahashi, E., Yurimoto, H., Gasparik, T.,
Sobolev, N.V., Pokhilenko, N.P., Yefimova, E.S., 1984. Diamond- 2000. Enrichment processes at the base of the Archean litho-
bearing peridotite xenoliths in kimberlites and the problem of spheric mantle: observations from trace element characteristics
the origin of diamonds. Geol. Geofiz. 25, 63 – 80. of pyropic garnet inclusions in diamonds. Contrib. Mineral.
Sobolev, V.N., Taylor, L.A., Snyder, G.A., Sobolev, N.V., Pokhi- Petrol. 139, 720 – 733.
lenko, N.P., Kharkiv, A.D., 1997. A unique metasomatised xe- Webb, S.A.C., Wood, B.J., 1986. Spinel – pyroxene – garnet rela-
nolith from the Mir kimberlite, Siberian platform. Russ. Geol. tionships and their dependence on Cr/Al ratio. Contrib. Mineral.
Geophys. 38, 218 – 228. Petrol. 92, 471 – 480.