732
Dawson 1992) suggests kimberlite as a possible candi-
date, a possibility supported by the high-K, Ti, Mg, Ca,
and OH nature of the magma, inferred from the min-
eralogy of the veins.
References
Allsopp HL, Barrett DR 61975) Rb-Sr age determinations on South
African kimberlite pipes. Phys Chem Earth 9:605±617
Bailey DK 61982) Mantle metasomatism ± continuingchemical
change within the Earth. Nature 296:525±530
Bailey DK, Lloyd FE 61975) Light element metasomatism of the
continental mantle: the evidence and the consequences. Phys
Chem Earth 9:389±416
Belousova EA, Grin WL, Pearson NJ 61998) Trace element
composition and cathodoluminescence properties of southern
African kimberlitic zircons. Mineral Mag62:355±366
Boctor NZ, Boyd FR 61981) Oxide minerals in a layered kimber-
lite-carbonate sill from Benfontein, South Africa. Contrib
Mineral Petrol 76:253±259
Boyd FR 61990) Mantle metasomatism: evidence from a MARID-
harzburgite compound xenolith. Carnegie Inst Washington
Yearb 90:18±23
Boyd FR, Pearson DG, Nixon PH, Mertzman SA 61993) Low-
calcium garnet harzburgites from southern Africa: their relation
to craton structures and diamond crystallisation. Contrib
Mineral Petrol 113:352±366
Carswell DA 61975). Primary and secondary phlogopites and
clinopyroxenes in garnet lherzolite xenoliths. Phys Chem Earth
9:417±429
Dawson JB 61984) Contrastingtypes of mantle metasomatism? In:
Kornprobst J 6ed) Kimberlites II: The mantle and crust-mantle
relationships. Elsevier, Amsterdam, pp 289±294
Dawson JB 61987) Metsomatised harzburgites in kimberlite and
alkaline magmas: enriched restites and ``¯ushed'' lherzolites. In:
Menzies MA, Hawkesworth CJ 6eds) Mantle metasomatism.
Academic Press, London, pp 125±144
However, inversion of the REE data for the BD3024
average zircon, using the zircon/liquid partitioning data
of Hinton and Upton 61991), indicates that the zircon
must have been in equilibrium with a magma which was
highly enriched in the LREE 6La 3,600 ´ chondrite),
with a very steep slope towards low HREE contents 6La/
LuCN ratio ꢀ1,600, i.e. >three orders of magnitude).
Kimberlites, both group I and group II 6orangeites), al-
though having LREE enhancement and steeply sloping
patterns 6summarised by Mitchell 1986, s1995), contain
lower amounts of all the REE than this calculated
magma. The only magmatic rocks which consistently
match these calculated parameters are carbonatites 6e.g.,
see Woolley and Kempe 1989). However, carbonate is
rare in the veins which 6if the veins formed from a car-
bonatitic melt) implies that they are the result of open-
system crystallisation, and that the vein phases are the
early precipitates platingthe vein walls. The lost liquid
residue of this magma must have had a very low viscosity
to enable it to migrate onwards before the freezing of the
very thin veins. This again points to carbonatite, as
currently it is the only magma type known to possess
such very low viscosities 6Treiman 1989).
Dawson JB 61999) Meltingand metasomatism in spinel peridotite
xenoliths from Labait, Tanzania. In: Gurney JJ, Gurney JL,
Pascoe MD, Richardson SH 6eds) Proc 7th Int Kimberlite
Conf. Red Roof Design, Cape Town, pp 164±173
Dawson JB, Hawthorne JB 61973) Magmatic sedimentation and
carbonatitic dierentiation in kimberlite sills at Benfontein,
South Africa. J Geol Soc Lond 129:113±131
Dawson JB, Hill PG 61998) Mineral chemistry of a peralkaline
combeite lamprophyllite nephelinite from Oldoinyo Lengai,
Tanzania. Mineral Mag62:179±196
Dawson JB, Smith JV 61975) Chromite-silicate intergrowths in
upper-mantle peridotites. Phys Chem Earth 9:339±350
Dawson JB, Smith JV 61977) The MARID 6mica-amphibole-rutile-
ilmenite-diopside) suite of xenoliths in kimberlite. Geochim
Cosmochim Acta 41:309±323
Dawson JB, Smith JV 61988) Metasomatised and veined upper
mantle xenoliths from Pello Hill, Tanzania: evidence for
anomalously-light mantle beneath the Tanzanian sector of the
East African Rift Valley. Contrib Mineral Petrol 100:510±527
Dawson JB, Smith JV, HervigRL 61977) Late-stage diopside in
kimberlite groundmass. Neues Jahrb Mineral Monatsh
1977:529±543
Dawson JB, Smith JV, HervigRL 61980) Heterogeneity in upper-
mantle lherzolites and harzburgites. Philos Trans Roy Soc Lond
A297:323±331
Dawson JB, Steele IM, Smith JV, Rivers ML 61996) Minor and
trace element chemistry of carbonates, apatites and magnetites
in some African carbonatites. Mineral Mag60:415±425
Delaney JS, Smith JV, Carswell DA, Dawson JB 61980) Chemistry
of micas from kimberlites and xenoliths ± II. Primary- and
secondary-textured micas from peridotite xenoliths. Geochim
Cosmochim Acta 44:842±857
In the case of the alteration rims around ilmenite
6Fig. 6), as in the case of the zirconolite rims on zircon,
there must have been interaction with a Ca-rich medium
to produce the perovskite. Nonetheless, although crys-
tallisation from REE-enriched melt appears to be valid
for zircon, later crystallisingphases such as zirconolite,
apatite, calcite and perovskite do not have the same
REE contents as their analogues in carbonatites.
A simple model might be that a calcite-rich kimberlite
intruded minor fractures in the harzburgite palaeosome,
inter alia causingCa metasomatism of the palaeosome.
After initial precipitation of diopside, zircon, phlogopite
and ilmenite, most of the residual liquid migrated,
though interstitial serpentine, calcite and apatite may
represent small amounts of nonmigrated residual liquid.
Such a model, drawingin part on the zircon and ilmenite
alteration, implies evolution and substantial changes in
the magma chemistry during cooling. In this context, it
is pertinent to note that pre-emplacement changes in
kimberlite magma chemistry have been previously re-
ported from the Kimberley area in the fractionated
carbonate-rich kimberlite of the Benfontein sills 6Daw-
son and Hawthorne 1973).
Acknowledgements Paula McDade and John Craven are thanked
for their help with the analytical work, and the polished sections
were prepared by Mike Hall. Richard Hinton and Steve Haggerty Edwards D, Rock NMS, Taylor WR, Grin BJ, Ramsay RR
gave useful advice on zircon-liquid partitioning and oxide miner-
alogy, respectively. Helpful reviews were provided by Steve
Haggerty and Terry Williams. We are also grateful to Graham
Pearson for providingprepublication data on his MARID zircon.
The electron- and ion-probe facilities at the University of Edin-
burgh are supported by N.E.R.C.
61992) Mineralogy and petrology of the Aries diamondiferous
kimberlite pipe, Central Kimberley Block, Western Australia.
J Petrol 33:1157±1191
Erlank AJ, Waters FG, Hawkesworth CJ, Haggerty SE, Allsopp
HL, Rickard RS, Menzies MA 61987) Evidence for mantle
metasomatism in peridotite nodules from the Kimberley pipes,