volcanological and structural aspects of the venetia kimberlite

STEPHAN KURSZLAUKIS AND WAYNE BARNETT

SOUTH AFRICAN JOURNAL OF GEOLOGY, 2003, VOLUME 106, PAGE 145-172

145

IntroductionTectonic control on kimberlite emplacement is integrallylinked with the understanding of the physical structureof the lower lithosphere. Currently accepted modelsfavor mantle hotspots (e.g. Crough et al., 1980),continental rifting (e.g. White et al., 1995), or shallow-angle subduction (e.g. McCandless, 1999) as triggers forkimberlite generation and ascent. It is generally acceptedthat prominent deep crustal discontinuities act asfavorable ascent paths for the magma towards thesurface (e.g. Friese, 1998). However, few publicationsactually document how the near-surface crustalweaknesses affect the shape and the positioning ofpipes within clusters, or even the effects of the localstructures on volcanism.

Deakin and White (1991) describe the emplacementof the Argyle lamproite during local extension related totranscurrent shearing within the Halls Creek MobileZone of Australia. White et al. (1995) proposed a strongrelationship between kimberlite occurrences on theAfrican, Australian and Canadian cratons, with deep-crustal mobile belts undergoing transtension or being re-activated as a landward extension of transform faults.

Detailed work on local structural geology in theCanadian kimberlite fields shows the clear associationbetween kimberlite cluster trends and pipe shapes withupper crustal discontinuities (e.g. Lake TimiskamingStructural Zone - Sage, 1999). The Jericho pipes of theSlave Craton are located along an east-northeast strikinglineation, while the distribution of kimberlites in theDiavik area follows a northeast-trend (Cookenboo,1999). Card et al. (1997) relate an orthogonal system(northeast and northwest trending) of arch-style upliftsof pre-existing structures to the location and timing ofkimberlite magmatism on the Canadian Shield.Kurszlaukis et al. (1998b) and Kurszlaukis and Lorenz(2000) described a distinct elongation of manykimberlite pipes parallel to pre-existing faults and jointswithin the country rock of the Gibeon Kimberlite Field.Kurszlaukis and Lorenz (1996) also report a directcontrol of the stress field on the shapes of diatremessurrounding Gross Brukkaros, a carbonatitic domestructure in southern Namibia. Zhang et al. (1989) andDobbs et al. (1994) describe a close structural control onthe location and outline of highly eroded pipes in theShangdong Province, China.

Volcanological and Structural Aspects of the Venetia KimberliteCluster – a case study of South African kimberlite maar-diatreme

volcanoes

Stephan KurszlaukisDe Beers GeoScience Centre, PO Box 82232, Southdale 2135, South Africa

e-mail: [email protected]

Wayne BarnettVenetia Mine, Geology Section, PO Box 668, Messina 0900, South Africa

e-mail: [email protected]

ABSTRACTThe paper summarises an investigation of the volcanology of the Venetia kimberlite cluster and the country rockstructures in the immediate vicinity of the pipes. Our study shows that the pre-existing country rock structure andthe stress field prevailing during emplacement control the position and morphology of the pipes. The pipes wereformed along pre-existing zones of weakness, and the pipe shapes are elongated in the direction of fractures andfaults. The highest degree of vertical and lateral pipe shape irregularity is found in the vicinity of country rockstructural weaknesses. The understanding of this relationship between pipe geometry and country rock structure cansignificantly affect mineral resource models.

The most important volcanological feature of the cluster is the presence of bedded fragmental rocks filling themain pipes K1 and K2. These rocks are readily interpreted as volcaniclastic in origin, yet are petrographicallyidentified as typical Southern African tuffisitic kimberlite breccias (“TKB”). These observations contradict thehypothesis that “TKB’s” are intrusive rocks, which are expected to be massive and unstructured in nature.

The presence of accretionary (armoured) lapilli, soft sediment deformation, and signs of reworking within thevolcaniclastic deposits suggest a high moisture content of the volcaniclastic material during and after emplacementof the pipe. Monolithic talus fans as well as contact breccias represent significant facies types within the Venetiakimberlite cluster. The contact breccias show signs of downward movement along the pipe walls, suggesting aconsiderable volume deficit at depth. These volcanological observations need to be accounted for in a comprehensiveemplacement model for the Venetia kimberlite pipes.

The observed structures, internal textures and facies suggest that the Venetia pipes were emplaced in severalviolent volcanic phases, separated by periods of relative inactivity.

Data from the West Eifel Volcanic Field, Germany,also demonstrate a close relationship of the style oferuptions with structural inhomogeneities of the countryrock. (e.g. Büchel, 1984; 1993; Büchel et al., 1987;Lorenz, 1973; 1985; 1986; 1998; Lorenz and Büchel,1980). There, maar-diatreme volcanoes are formed byphreatomagmatic activity on the intersections of mostlybasaltic dykes with local water bearing faults or joints,especially but not exclusively underneath valley floors.If meteoric water is not available, scoria cones areformed by magmatic activity, without diatremes. TheWest Eifel Volcanic Field represents a classic example ofhow pre-existing country rock structures influence andcontrol the emplacement behavior of rising magma inthe uppermost crust. For that reason, the internal andexternal geology of a volcano must be carefully mappedand evaluated to understand its emplacement.

The kimberlite occurrences in the Limpopo Belt,including the Venetia cluster, follow a strong east-northeast trend through the Belt. Watkeys (1983a) usedaerial photograph interpretations to show the strongassociation of the Venetia and Beit Bridge kimberliteswith east-northeast striking lineations. The actualposition of the Venetia pipes is related by Watkeys(1981) to a fracture and fault striking north and northeastrespectively from an extension of the northwest-trendingSiloam fault zone. These faults, and the nearbyintersection of the east-northeast striking Dowe-Tokwefault with the Siloam fault zone, potentially provided acrustal opening for ascent of the kimberlite magma.Watkeys (1981) also drew attention to the northwesttrend of the Venetia K1, K2 and K3 pipes.

Although the authors intentionally do not suggest anemplacement mechanism for the Venetia pipe cluster, itis of importance to understand the principles of the models to be able to follow the discussion of thevolcanological features presented in this paper. Over the last two decades two principally different geneticmodels were developed to explain the emplacementmechanism of kimberlite pipes.

The one model, termed the “magmatic model”,operates with an assumed high volatile content in akimberlite magma rising to shallow pre-eruptivesubsurface levels. Fluids are concentrated at the tip ofthe rising kimberlite magma column and are initiallyprevented from breakthrough to the surface by thepresence of a sealing cap rock(s), such as thick basaltlayers or impermeable sandstone beds. The fluids arerammed into the country rock surrounding thekimberlite and fragment these country rocks. With the breakthrough to surface, a fluidisation front movesviolently down the magma column, fragmenting andmixing kimberlite magma and country rock xenolithsfrom different lithologies. The consequent rock is calleda “Tuffisitic Kimberlite” (“TK”) or “Tuffisitic KimberliteBreccia” (“TKB”), depending on its xenolith content.These “TKB’s” are by definition intrusive rocks and musttherefore be unbedded (Clement and Reid, 1989).“TKB’s” are highly specific rocks, which, with few

exceptions on a worldwide scale, apparently occur onlywithin the larger kimberlite diatremes of southern Africa.“TKB’s” exhibit characteristic petrographic features, likethe presence of pelletal lapilli and abundance ofmicrolitic clinopyroxene. This emplacement model (andits variations), is discussed in the publications ofClement (1975; 1982), Hawthorne (1975), Skinner andClement (1979), Clement and Skinner (1985), McCallum(1985), Clement and Reid (1989), Field et al. (1997),Field and Scott Smith (1999), Scott Smith (1999), andfurther references therein.

The second, contrasting model uses the interactionof the hot magma with external, meteoric water as theenergy source for the fragmentation of magma andcountry rock. This model is termed the“phreatomagmatic model” (Lorenz, 1975; 1979; 1985;1986; 1987; 1998; White, 1991; Lorenz and Kurszlaukis,1997; in preparation; Lorenz et al., 1999a; b). The modelwas developed contemporaneously with the magmaticmodel and is generally accepted among scientists toexplain the emplacement of non-kimberlitic maar-diatreme volcanoes world-wide. The model is supportedby experiments (e.g. Zimanowski et al., 1991; 1995;1997; Lorenz et al., 1999b; Morrissey et al., 2000)providing a well constrained understanding of thephysical processes involved (Büttner, 1997; Büttner andZimanowski, 1998), and observations in nature (Kienleet al., 1980; Ort et al., 2000; White and Houghton, 2000).Experiments on remelted kimberlite show that veryviolent explosions can be generated with this magmatype as well (Kurszlaukis et al., 1998a; Lorenz et al.,1999a; b). In the phreatomagmatic model, most of thethermal energy of the magma is converted into highlyenergetic, supersonic shock waves, which are able tofragment not only the magma but also the country rock.The smaller portion of the thermal energy is used totransfer meteoric water into steam, which transports thepreviously fragmented magma and country rockparticles to the Earth’s surface.

Although both models do not exclude processes ofthe opposite model, the principal difference remains theenergy source required to form a deep diatreme in the country rocks from a rather narrow feeder dyke: in the magmatic model it is the expansion of juvenilevolatiles, in the phreatomagmatic model it is the thermalenergy of the magma converted to shock waves andsteam.

This paper presents geological and structuralobservations of the Venetia kimberlite pipes at thecurrent mining level and within drill cores, which haveto be explained in the development of any pipeemplacement model for Venetia.

Geological setting The Venetia kimberlite pipes are located within theLimpopo mobile belt, which is an ancient collision zoneof the Kaapvaal and Zimbabwe craton. At the presenterosion level the country rock in the Venetia areaappears as a complex arrangement of Proterozoic

SOUTH AFRICAN JOURNAL OF GEOLOGY

VOLCANOLOGICAL AND STRUCTURAL ASPECTS OF THE VENETIA KIMBERLITE CLUSTER146



147

Figure. 1a to c: The surface outlines (678m asl), pipe positions and a simplified map of the internal geology of the pipes and dykes of

the Venetia kimberlite cluster (modified after Seggie et al., 1999, Colgan (pers. comm.) and Skinner (pers. comm.)) is shown in (a)

(b) highlights the pipe outlines at 678m (surface), 630m and 606m asl. (c) shows the most important faults, shear zones and joint sets in

the Venetia area together with the pipe outlines at 630m asl.

gneisses, amphibolites and metasediments of theLimpopo Group. In the Venetia area, the LimpopoGroup is believed to have a thickness of about 10kmand is overthrust onto Archean rocks of theZimbabwean craton (Pretorius, 1996). The overthrustmay be related to the Limpopo orogeny (Barton andPretorius, 1998).

The basement rocks form a synclinal structuredipping east-west, with the kimberlites of the clusterlocated in the fold axis of the syncline. The country rockappears to comprise two main packages of rocksbelonging to the Limpopo Group. The first package isexposed in the open pit and is closely associated withthe kimberlite contacts at the current depth of exposure.It can be described as a high metamorphic gradegneissic package (upper amphibolite facies) thatdominates the core of the synclinal structure withinwhich the pipes are located. Biotite schist and biotitegneiss predominate, with common lenses and layers ofamphibolite. Quartzo-feldspathic gneiss also makes up asignificant proportion of the rock types. The secondpackage can be described as a meta-sedimentarypackage. It occurs in the outer limb enfolding thegneissic package in the syncline. This packagecomprises marble and limestone, fuchsitic quartzite,phyllite, calc-silicate rocks, and lenses of bothamphibolite and banded iron formation. The kimberliteswere emplaced into this second package at depth, but inthe current exposures the metasediments are in directkimberlite contact only on the northern edge of K2 andall around K3.

The two packages appear to be tectonicallyjuxtaposed by shearing with a structurally complexcontact zone of cataclasite, hydrothermally altered schistand isoclinal folding in between. Retrograde mineralgrowth is abundant throughout both packages of rocks.The most common metasomatic process wasepidotisation giving the altered rock a greenishcolouring, particularly around the shear zones. Thishydrothermal alteration is most distinctly developedwithin faults and shear zones, giving evidence tosignificant quantities of water that migrated along thefaults in the geological past. The lithology of the twopackages at Venetia Mine has not yet been clearlyassociated with the lithology of the Beit Bridge Complexcharacterising the rocks of the Limpopo Group aroundMessina (e.g. Watkeys, 1983b; Hoffman et al., 1998;Kröner et al., 1998). There are also two varieties ofintrusives that predate the kimberlite. A 25 to 60m thickdolerite sill is intruded at a depth of about 250m belowpresent day surface. The dolerite shows strong evidencefor hydrothermal alteration, particularly along joints.Dykes occasionally extend from the sill upwards. At leasttwo phases of pegmatites also intruded the country rock.

Little is known about the erosion rate in the Venetia area since the time of emplacement of the pipes ~519 Ma years ago (Phillips et al., 1999). The contemporaneous River Ranch kimberlite ~45km tothe east-northeast shows a high abundance of crater

facies blocks (Muusha and Kopylova, 1998), suggestinga relatively low erosion rate since emplacement. TheOaks pipe (of similar, but slightly younger age; Phillipset al., 1999), which is situated ~125km to the west, isidentified as an exhumed root zone. However, since theoriginal size and emplacement history of these pipes isunknown, no further conclusions should be drawn fromthese observations. At Venetia, grey and red shale aswell as vesicular basalt xenoliths found in the kimberliteof the various pipes suggest that a sedimentarysequence, probably of the Soutpansberg and Waterbergformations, overlaid the Proterozoic basement at thetime of emplacement (Bumby et al., 2001 and referencestherein). It is, however, difficult to estimate the erosionrate, since it is not easy to link these xenoliths to the pre-existing stratigraphy of the Precambrian WaterbergGroup. Brown et al. (1998) suggested a denudation ofthe Kaapvaal craton of 3.4 ± 1.4km due to lithosphericuplift about 90 Ma ago. However, since CarboniferousKarroo sediments do crop out close to Venetia thiserosion rate would only have affected sedimentsdeposited after the emplacement of Venetia and is thusnot relevant for the erosion history of the pipes. Theerosion of the Venetia pipe cluster must have happenedbetween its emplacement and the onset of the Karoosedimentation, and, to a minor extent, during the LateTertiary and Quaternary.

The Venetia kimberlite cluster consists of 14 bodiesthat outcrop in an area of about four km2 (Figure 1a).The largest and central body is “K1” which covered anarea of about 12ha at the present surface. K1 is kidney-shaped, about 650 by 200m in size, and elongated east-west. Most of the pipe is infilled by fragmentalkimberlite, but localised hypabyssal kimberlite intrusionsoccur at the margins as well. A country rock brecciaoccurs in contact with K1 at the southeastern margin andis delineated by prominent fractures at its western and eastern side (Figure 1c). The breccia occupies a150m deep and 100m wide space within the countryrock. An increased number of country rock xenolithswithin the pipe adjacent to the breccia testifies to thesyngenetic relationship of the breccia with the pipe.

The second largest pipe of the cluster is K2, whichoccurs to the west of K1 and is separated from K1 by athin country rock barrier of about 200m thickness. K2 isroughly oval in shape, with dimensions of about 250 x200m and covers an area of 5ha at the present surface.Towards its western side K2 is occupied by a brecciathat is very rich in country rock clasts. The central partof the pipe is filled by ”TKB”, but at depth, andespecially towards the southern and western margin,hypabyssal kimberlite is present.

Structural geologyThe synform structure in which the kimberlite pipeshave been intruded plunges at about 31° eastwards inthe vicinity of the pipes. It is a similar-type fold vergingsouthwards with the northern limb vertical tooverturned. The closure is clearly visible in




SOUTH AFRICAN JOURNAL OF GEOLOGY, 2002, VOLUME 105, PAGE 145-160

149

aerial photographs to the west of K2. Mapping(Doorgapershad et al. 2003) shows that the synform alsocloses to the east and is therefore cross-folded. Remnantprimary sedimentary structures in the meta-quartzitesindicate the synclinal nature of the fold. In the pit and incross-sections of the geological model it is clear that theK1 pipe is situated just to the south of the fold axis andwas therefore emplaced within the southern limb(Figure 1c). K5 appears to be on the hinge zone of thefold, although it is unclear what the relationship is atdepth.

There are three primary directions of faults exposedin the Venetia farm area, which are striking east-west,northeast and northwest. The Lezel Fault is part of thenortheast striking Mutshilashokwe Fault Zone (Watkeys,1981) and intersects the open pit and eastern edge of theK1 kimberlite pipe (Figure 1c). The position andgeometry of the other satellite pipes east of K1 appearto be closely related to the Lezel Fault (describedbelow). The Lezel Fault has horizontal striations on thefault plane with 100 to 150m of displacement evident onaerial photographs. Seepage observations in the pit andborehole pump tests show that the north-east strikingfractures and faults are hydrologically active.

The approximately east-west striking faults (Figure 1c) also show horizontal striations on the faultplane. Actual variation in strike is between east-northeast and west-northwest. This trend may be relatedto the east-north-east trend of the Dowe-Tokwe Faultabout 7km south of Venetia. The K1 elongation and theshapes of K4 and K7 are discussed below in this paper

in the context of the east-west fault orientation. It is clearfrom the tectonic history of the Limpopo MetamorphicBelt that the faults may have been re-activated numeroustimes (Watkeys, personal communication, 2001). Thegeometry of the kimberlites indicates that the faultsexisted prior to kimberlite pipe emplacement, but thestrongly sheared nature of the kimberlites along the Lezel Fault shows that further displacementdefinitely occurred after kimberlite emplacement.

A prominent northwest striking fault (an extension ofthe Siloam Fault Zone) is present 10km south of themine. Watkeys (1981) postulated that the northweststriking fault, northward trending fractures and thenortheast striking faults have controlled the position ofthe Venetia kimberlite cluster.

There are numerous joint sets exposed in the VenetiaMine open pit. The sets do not appear to be axial planaror otherwise related to the major folding event. Aerialphotograph interpretations suggest that some of thejoints shear through the kimberlite, indicating that thesesets either formed or were re-activated afteremplacement. The aerial photograph interpretationsuggests strikes for at least four joint sets (also M. Watkeys, personal communication, 2001). Detailedline sampling in the open pit by geotechnical engineershas distinguished up to eight sets developed in thecountry rock (see Barnett, (2003) for a detaileddescription). Figures 1c and 2 show the four mostdominant joint sets.

The J0 joint set represents the best-developed jointset, in the form of a mechanical parting of the

Figure 2. Contoured stereonet (poles to planes) showing the four most well-developed subvertical joint sets at the mine. These joints are

controlling the shape of all the pipes. The figure is an amalgamation of four stereonet plots, each contoured individually as indicated in the

inset table.

metamorphic layering. The lineation, most stronglydeveloped on aerial photographs, strikes northeast,almost parallel to the Lezel fault and occasionally in east-northeast directions. The J2 joint set is sub-verticaland locally very well developed in the pit. Another setstrikes north-south in aerial photograph but appears torotate towards the north-northeast in the vicinity of K1and K8. This is the J1 joint set. In aerial photographs J1clearly forms in en-echelon patterns that elongatetowards the northwest, typical of tension joints formingunder north-south compression. In such a model the lesscommon, “overlapping” north-north-east joints may havebeen pre-existing, but also reacted in tension to thesame stresses.

The aerial photographs also show that there is atleast one set striking west-northwest, and anothertowards the north-west. J4 is the west-north-west jointset, also sub-vertical and varying towards east-west. J6strikes towards the northwest and on average dips 50o

towards the southwest. Observations in the pit suggest astrong genetic relationship between J6, minor NWfaulting and the east-west faulting (Barnett, 2003).Northwest striking lineations connect K1, K2 and K3along strike (Figure 1c). The other sets clearly exposedin the pit are never well developed and dip atintermediate angles (see Figure 1c).

The J1, J2 and J6 joint sets are poorly developedwithin the kimberlite and all have sub-horizontalstriations. In places these joint surfaces show two sets ofstriations, with plucking indicating an opposite shearsense. The kimberlite must have been partially shearedby at least two tectonic events. This is not surprisingconsidering that the deposition of the Karoo withingraben basins into the Limpopo Group basement andthat the tectonic break-up of Gondwana post-dates thekimberlite emplacement.

Judging from the above descriptions, it is concludedthat the origin of the joints is likely to be strongly linkedto the formation and re-activation history of the faults,and has a strong control on the pipe shapes. In general,the pipes show the highest degree of irregularity withrespect to pipe outlines and depth development in thoseareas where the pipes are in contact with zones ofweakness within the adjacent country rock.

Pipe shapesPipe shapes appear highly variable. In plan they includespherical, and oval to irregular shapes. In general, thepipes seem to be elongated into west-northwest to east-southeast or northeast-southwest directions (Figure 1).One of the most irregular pipes in the cluster is also thelargest one (K1). The infill of K1 seems to consist of atleast two major units, one occupying the eastern andone the western part of the pipe (“TKB East” and “TKBWest”; Seggie et al., 1999). Recent petrographic workconducted by Skinner (personal communication, 2000)suggests a more complicated picture with at least threedifferent units of fragmental kimberlites. At 400m depththere is still no indication that the pipe splits into

separate roots, which would be one of the possibleexplanations for the complex pipe shape, the multitude of volcanic units and the variations indiamond grade.

The complex shape of K1 seems to be controlled bythe country rock structure. K1 is generally east-westelongated, but has its highest degree of irregularity at itseastern end, where it extends towards the north-east. It is in this region that the Lezel shear zone strikestowards the northeast. A splay of this fault passesdirectly through the centre of the northeast extension ofK1 (Figure 1c). The general east-west elongationcorrelates closely with the east-north-east strike of jointset J4, which is developed in the country rock west ofK1. The Lezel fault clips the southeastern corner of K1and then passes east of the main kimberlite body.

The pipe outline of K1 seems to be controlled notonly by pre-existing faults within the country rocks, butalso by the stratigraphy of the adjacent basementlithologies. The folded stratigraphic layering plungeseastwards at ~31°. The western K1 boundary also dips eastwards at a relatively shallow angle apparently following the country rock layering closely inplaces.

A number of small kimberlite bodies and dykes haveintruded along the Lezel fault plane. From the south theK1 pipe has a hypabyssal intrusion extendingsouthwards between the two splays of the fault zone.Further east are the K17 and K10 dyke-shaped bodies.K10 is clearly emplaced on the Lezel fault, is stronglysheared by a re-activated movement on the fault, and ishighly altered due to groundwater flowing along thefault. North of K10 is K16, that strikes northwards awayfrom the fault plane, possibly along the J1 joint sets. It isclear from Figure 1b that both K16 and K17 arekimberlite bodies that do not extend to the presentsurface. North of K16 is K6, which is clearly elongatedalong the strike of the Lezel fault plane. The northern tipof K6 intersects the synclinal fold axis with the contactedges being strongly controlled by the foldedmetamorphic layering as it intrudes partially along theaxis. K6 elongates northeastwards and coalesces with K5at depth. Near the surface K5 is a wider, irregularkimberlite body that is partially elongated towards theeast-northeast. K5 appears to be on the intersection areaof the two northeast trending fault splays and the west-north-west striking Gloudina fault. Approximately 20mof the eastern edge of the K5 kimberlite is stronglysheared, again showing that the faults have been re-activated. Exploration drillholes indicate that dykes,which intruded along the Lezel fault plane, connect K17,K10, K16, K6 and K5 at depth.

K4 is a very irregularly shaped kimberlite pipe thatextends strongly along an east-south-east trend awayfrom K6. Protrusions appear to extend along north-eaststriking joint sets away from the main body, but theoverall shape is likely to be controlled by a west-northwest trending fault. A drillhole situated east of K4was intended to intersect such a fault. The drillhole did





151

intersect strongly sheared gneisses and schists andyielded significant groundwater.

A set of joints strikes northwestwards from K1intersecting K2 and then K3 along the same trend. K2 isan internally complex pipe that appears to be dividedinto a distinct eastern and a western half. The easternside has been documented as a “TKB” facies and thewestern half as a hypabyssal breccia facies with atransitional segregationary textured kimberlite at thecontact between the halves (Figure 1a; Seggie et al.,1999; Skinner, personal communication 2000). However,investigations in the open pit and of drill cores reveal acrudely bedded nature of the “TKB” infill. K2 lies on theintersection of a strong northeast trending fracture/faultset (J2) and the northwest striking joints (J6). In addition,there is an eastward striking fault clearly visible on thesoutheastern edge of the pipe. This fault has caused the irregular shape on the southeastern corner. A distinctextension of the K2 pipe expands northwestwardstowards K3, obviously following the north-west-strikingjoints. The extension is filled with country rock breccia(see observations in section 5.1 and Figure 5) and beingnearly non-existent at surface it increases in size towardsthe northwest with depth.

K3 occurs at the end of the northwest trending jointset as a very irregular pipe outcropping at surface witha protrusion extending towards the west. This protrusionappears to shorten and narrow towards depth, althoughmore detailed exploratory drilling is required to confirmthis. The overall east-west alignment of the protrusion isthe result of two trends, a northeast trend where aprominent northeast striking fracture intersects K3, anda joint controlled side-step of the protrusion towards the northwest. The position of K3 is coincident with theintersection of the northeast fracture and the northwesttrending joints.

K8 is an unusual hypabyssal breccia pipe with astrong north-northeast alignment since it lies alongnorth-northeast trending fractures. The full geometry ofthe pipe is not known at this stage but it is clear that theK8 kimberlite is surrounded by a wide (perhaps as muchas 20-30m) and irregular envelope of country rockbreccia. Very little is known about the K11 kimberlitefound north-northeast of K1. Pipes K7 and K12 appearas separate bodies at the present surface, but about 20mbelow the surface they merge into one body that iselongated towards the northwest, again following anorthwest trending fracture or fault. This same faultforms the southern contact of a prominent rectangularbody of country rock breccia that is found immediatelyeast of K12 and extends northwards into the K1 pipe(see also section 5.1). The eastern and western edges ofthis breccia body are extremely sharp and defined bysouth-south-west striking J1 fractures. The K12 and K7bodies have not yet been modelled to the depthsindicated in Figures 1b; c.

The above observations highlight the importance ofthe influence of the following structures on the Venetiapipes positions and shapes: [1] northwest striking minor

faults, fractures and joints (including J6), [2] northeaststriking faults and joints (J2), and [3] north-south tonorth-northeast striking joints (J1). As mentioned above,J1 forms as tension fractures in en-echelon envelopesthat imply north-south compression. Following thismodel through implies dextral re-activation of northweststriking minor-faults, and sinistral re-activation of north-east striking faults. The combination of these two faultdirections respectively with north-south tension jointsduring north-south compression would producedilational jog on the faults and areas of crustal extension.A swing in the direction of compression towards thenorthwest at the time of kimberlite emplacement wouldopen up the north-west faults (such as the Siloam fault)allowing the migration of kimberlite to the surface (M. Watkeys, personal communication, 2002). The actualgrowth of the kimberlite pipe shape during volcanismwould then be governed by joint-controlled tensionalfailure and slumping of the pipe sidewalls.

Geological observations in the kimberlitesObservations in the open pitsCountry rock dominated breccias occurring along themargins of the pipesGeological observations in the pits indicate a generallysharp contact of the kimberlite pipes with the countryrock. The country rock appears only locally brecciatedby tectonic fracture zones (see section 4) or byvolcanological processes (e.g. contact breccias related toroot zones). This is especially true for K1, where thecountry rock at the contact is only brecciated in onespecific area at the southwest side of the pipe. Asmentioned in section 4 this country rock breccia isconfined by fractures and faults, which not only defineits sharp boundaries to the east and west, but also thesouthern side of the breccia furthest from the pipe(Figure 1c). The latter roughly northwest trending faultdefines the breccia’s extent towards the south and wasexploited by various intrusions of hypabyssal kimberlite,forming the kimberlite bodies K7 and K12. At thecontact with the breccia, the ”TKB” of K1 shows a highcontent of xenolithic basement fragments, which arethought to derive from the breccia and which weremixed into the “TKB”, suggesting a syngeneticgeneration of the breccia and K1. The country rockblocks within the breccia reach sizes up to 10m (Figures3a; b), but most fragments are less than 0.5m indiameter. The disorientation of the country rock is notvery obvious, but a downward slumping of the brecciatowards K1 is evident. The matrix of the breccia seemsto consist of finely brecciated country rock without anykimberlite present. The cement is secondary carbonate.Near the southern margin of the breccia, both a stronglycemented country rock breccia and a poorly cementedcountry rock breccia are found in the same exposurewith a sharp contact between them (Figure 3b). This suggests that the emplacement of K1 extended overtime, since the initial breccia was consolidated and cemented before renewed volcanic activity



Figure 3a. The photograph shows the contact breccia at the south-western margin of K1. The breccia is clearly confined by faults

(red lines). Wall height about 40m, view towards south. The inset highlights the sharp eastern contact of the breccia with the country rock.

View from the bottom of the pit upwards.

Figure 3b. The picture delineates the southern margin of the breccia sliding into K1, close to the contact to K12 (view towards the north).

The breccia shows a sharp contact to unbrecciated country rock on the left hand side. A shear zone along the contact suggests movement

of the breccia. The central part of the breccia appears compact and solidified, while the breccia above and right of the stippled red line

was re-activated after consolidation. A downward movement of this breccia is indicated.

occurred, causing subsidence and re-brecciation of thematerial.

A part of the northern contact of K4 with the countryrock consists of a monolithic breccia, which is exposedto a width of ~25m and to a height of ~12m. The clastsizes range from millimetres to a few decimetres.Elongate clasts are strongly orientated along dislocationor shear planes that dip towards the centre of K4 (Figure 4). The contact with the unbrecciated countryrocks is sharp and exploited by a late hypabyssalkimberlite intrusion. Some thin hypabyssal dykes alsointrude the breccia near the kimberlite contact. Weinterpret this breccia as a contact breccia, a feature thathas also been observed in a number of root zones ofkimberlite pipes within the Kimberley field in SouthAfrica (Clement, 1982; Clement and Reid, 1989) andelsewhere in the world (Lorenz and Kurszlaukis, inpreparation).

On the northwestern side of K2 a kimberlite breccia,with a country rock xenolith content exceeding 80%,protrudes in the direction towards K3, following north-west-striking joints (Figure 5). The breccia extends overthe complete height of bench four (12m) and over awidth of about 40 m. The breccia forms an extension ofthe K2 pipe that grows larger and further from K2 withincreasing depth. The full depth of the pipe extension iscurrently unknown, but the cave-like nature of theextension with an overlying hangingwall is clear from

mapping. The lateral contact to the unbrecciated countryrock on bench four is sharp and clearly defined. Withinthe kimberlite breccia, large blocks of country rocksappear tilted away from the sidewalls into the brecciazone (as is suggested by the rotated appearance of thefoliation and joint bounded edges of the blocks) andcollapsed from the hangingwall of the cavity. A finer-grained breccia is more abundant in the lower half of thebench four exposure where it is observed draping overand lapping onto the larger xenolith blocks. This finer-grained material (clasts from 5 to 30cm) is bedded on adecimetre scale, manifested in a crude rounding ofclasts, clast orientation and a variation in kimberlitematrix (Figure 5).

We interpret this kimberlite breccia as a cavity opentowards the pipe, which was subsequently filled up byvolcanogenic as well as collapsing roof and wall rockmaterial. It appears that after the cavity was formed, alarge block of country rock collapsed and tilted from theside towards the centre of the void space and that onlythen was the cavity filled up with finer material (thebedding of this material is lapping onto the large block)before the roof of the cavity finally collapsed (Figure 5).Late hypabyssal kimberlite intrusions (with variablecontents of country rocks) exploited the northwesternmargin of the breccia-filled cavity. We regard this unique feature as being related to the emplacement of K2.



153

Figure 4. A profile through bench 3 in K4 shows the progression of a contact breccia from the country rock contact (right) towards the

central zone of the pipe. Hypabyssal kimberlite intrusions exploited the contact zone. The orientation of clasts and shear zones becomes

increasingly steeper with distance from the country rock, a sign for increased subsidence. View towards the west-north-west, bench height

about 12m.

Figure 6 shows the view facing directly away fromthe centre of pipe K2 towards the northwestern breccia-filled cavity (see above) very close to and parallel to thecountry rock contact. The face can be considered toshow the “entrance” of the breccia-filled cavity into the pipe. It shows the same breccia facies exposed in thewest and southwest side of K2 (see below). However,bedding within these breccias is less clear than in theother breccias described from K2. The volcaniclasticbreccia shows signs of slumping and the boundariesbetween “packages“ of different layering appear to becontrolled by soft-sediment faults. In general thelayering of the breccia dips away from the country rockcontacts on the east and west and towards the centre ofK2. A “U”-shaped channel is visible just right of thecentre of Figure 6, which must be channelling volcanicbreccia in or out of the cavity.

Country rock dominated breccias occurring within thepipesThe western portion of K2 comprises monolithic countryrock dominated breccias that are exposed on bench 5 over the entire western half of K2 (Figures 7 and 8).The clasts consist entirely of local basement lithologies.The basement clasts often show a high sphericity andare relatively well sorted with clast sizes mostly between5 and 50cm, although blocks up to 4m were observed.The smaller clasts frequently show concave percussionmarks, which seem typical for rounding in the vent(Figure 9) (Lorenz, 1971a; b). The matrix of this brecciais made up of finely comminuted and highly altered

clastic and kimberlitic material (Figure 10) andsometimes shows internal bedding (Figure 11).

The breccias are characterised by their internalstructure. They show clear bedding and orientation ofelongated clasts dipping with an angle of about 30 to 35°towards the centre of K2 (Figures 7 and 8). Bedthicknesses are between several decimetres and about10 metres. The beds are not planar, but often wedge-shaped. They are defined by differences in clast sizesand sorting and can be clast- or matrix-supported.Sequences of beds form fan-delta-like, 10 to 30m thickpackages, with a discontinuous onlapping of bedsbetween the packages (Figure 7). The approximately150m long face of bench 5, where these bedded brecciasare exposed, comprises about 5 of these packages(Figure 7).

We interpret these bedded breccias in K2 as talusfans filling up the crater floor from the side by rockavalanches. The talus fans are clearly on lapping awayfrom the crater sides towards the pipe centre. The talusfans closest to the southwest and western country rockcontacts appear virtually devoid of kimberlite matrix andcontain the largest and most angular joint-boundedcountry rock fragments. These talus fans are visuallyidentical to slope failure “fans” at the base of the openpit of the mine. It is perhaps not surprising that theslopes south-west of K2 are for geological reasons the least stable in the open pit, with up to 60% of theslope’s area having undergone failure (Barnett, 2003).

However, the monolithic nature of the brecciasrequires a special explanation, especially since the TKB,



Figure 5. The picture shows the breccia on the north-west side of K2, which is interpreted as a collapsed cavity. The large block in the

central area probably originates from the left side wall and rotated into the open space of the cavity, which was later filled with finer

grained, bedded material. View towards the north-north-west, person as scale.



155

Fig

ure

6 (

7).

The

cut

thro

ugh

the

talu

s fa

ns

is p

erpen

dic

ula

r to

the

tran

sport d

irec

tion (

looki

ng

from

the

centre

of

K2

tow

ards

the

north-w

est). Si

ngl

e flow

s an

d a

“U

”-sh

aped

chan

nel

just

rig

ht

of

the

centre

are

visi

ble

. So

ft s

edim

ent def

orm

atio

nal

fea

ture

s an

d s

lum

pin

g ar

e ev

iden

t. Ben

ch h

eigh

t is

12m

.



Fig

ure

7 (

8).

The

amal

gam

atio

n o

f photo

grap

hs

show

s ta

lus

fans

filli

ng

up K

2 from

the

side

conta

ct (

righ

t). At le

ast five

diffe

rent fa

ns

can b

e m

apped

, so

me

of

them

with

dis

cord

ant co

nta

cts.

The

talu

s

fans

are

ove

rlai

n b

y “T

KB”

(lef

t). The

num

ber

s def

ine

azim

uth

and d

ip o

f th

e la

yering.

Vie

w tow

ards

the

south

.



157

which is filling the central parts of K2, seems to overlythe fans and contains xenoliths of the Waterbergformation, which are absent in the talus fans. We suggestthat these monolithic breccias are the result of a seriesof very localised crater sidewall collapses. The processcould have been enhanced by the instability ofpreviously brecciated basement country rock. Remnantsof contact breccias would be an example of suchfragmented zones in the country rock, as well as pre-existing shear zones or just irregularities of the initialpipe shape which are levelled to a more stable rockmechanical slope. The cavity observed at the contact ofK2 to the country rock (see above) could be the remnantof such a collapse and removal of brecciated material.Since the contact of the talus fans towards the countryrock is not well exposed, it is not clear, whether the fansare in situ or whether they subsided into lower portions

of the diatreme. In any case, they must have beenformed within the basement level. If the breccias are in situ and not subsided, then primary volcanic andsecondary sedimentary processes must have beenoriginally very deep and subsequently filled up thecrater. This would also mean, that the “TKB” overlyingthe talus fans, which shows a mixture of xenolithlithologies including sedimentary clasts which overlaidthe basement (missing in the talus fans), is either areworked pyroclastic rock, i.e. a mass flow fed from thetephra ring, or a primary pyroclastic rock derived fromvolcanic activity post-dating the deposition of the talusfans, either directly from K2 or from a neighboured pipesuch as K1. The crude to fine bedding observed in the“TKB” would support the latter origin. Regardless of the origin of the “TKB” overlying the talus fans, it cannotbe an intrusive rock.

Figure 8 (9). The profile from bench 3 in K2 is situated just opposite the outcrop shown in Figure 7. The figure shows a series of talus

fans sedimented from left to right, some of them with a slightly discordant on-lapping. The fans also show distinct differences in country

rock xenolith contents and the lithologies of the blocks: some fans consist nearly entirely of amphibolite blocks, others of quartz-feldspar

gneiss fragments. View towards the North, bench height is 12m. See also person as scale.

Kimberlite dominated breccias occurring withinthe pipesBedding within “TKB’s” has been observed in variousforms in the open pits of both K2 and K1. Althoughwidely unstructured, the “TKB” in K2 overlying the talusfans shows local areas with very distinct bedding on adecimetre to metre scale (Figure 12). Similar to the talusfans, measured bedding orientations dip at shallowangles (less than 25°) towards the K2 centre, suggestingsubsidence or differential compaction.

Bedding may also have been present on a largerscale, but it was not entirely clear from the outcrop inthe open pit of K2 whether the observed colourdifferences were due to original bedding or if they areonly an alteration effect. The latter also applies topossible bedding in K1, were nearly horizontaldifferences in colour, hardness and alteration wereobserved. However, single loose blocks removed fromtheir original position during mining activity showedclear bedding on a decimetre scale. Unfortunately theexact source of these blocks could not be established,

but they certainly originated from bench 10, the deepest mining level in Venetia K1 at the time ofobservation.

At bench 10 close to the northwestern margin of K1,numerous allochthonous crater facies blocks have beenfound embedded within superficially unstructured“TKB’s”. These blocks are red-brown in colour, which isin contrast to the grey colour of the enclosing “TKB”.The blocks are up to a few metres in size, but an exactsize is difficult to determine, since the larger blocksespecially, have only very vague, fluidal outlines.Bedding and lamination as well as soft sedimentdeformation textures at the interface between beddingplanes are evident within the crater facies blocks. In thefiner-grained beds mud clasts were frequently present,and some of these mud clasts were also found in the coarser-grained beds. This suggests reworking in thepresence of water, which is also confirmed by the softsediment textures of the bedding interfaces (see alsosection 5.3.3). The fluidal outlines of the crater blocksalso suggest that the crater blocks were only partly



Figure 9 (10). The cobbles and pebbles in the talus fans show a high degree of sphericity, but also distinct and often concave chip marks

(arrows). Pipe K2, detail of one of the talus fans shown in Figure 7.



159

Figure 11 (17). The core section shows bedded kimberlite located in void spaces and cracks between country rock xenoliths of the talus

fans in K2. Note the grading and the orientation of elongated clasts. Depth increase from top to bottom, left to right. Pipe K2, drill core

RDH1a.

Figure 10 (16). The core sections show fragmental kimberlite which is intercalated in a series of large xenoliths forming bedded talus

deposits. The kimberlite shows an abundance of pyroclasts rimmed by a magmatic selvage (arrows). Depth increase from right to left, top

to bottom. Pipe K2, drill core RDH1a

solidified at the onset of renewed volcanic activity andthe “TKB” also must have had some moisture content.The crater blocks are thought to be not in situ but tohave downrafted towards deeper levels of the pipeduring ongoing volcanic eruptions, together with thehosting “TKB”.

Hypabyssal intrusionsThe central areas of the pipes filled with “TKB” andpreferentially the margins of the pipes were locallyexploited by hypabyssal kimberlite intrusions. They areat least in part late dyke intrusions, as dykes wereobserved at the southeast side of K2 in contact to thecountry rock. The latter dykes have a thickness of up to0.5m and are usually relatively carbonate-rich, olivinemacrocryst-poor kimberlite intrusions.

The genetic relationship of massive, rathercarbonate-poor, olivine macrocryst-rich hypabyssalintrusions at the south side of K2 and in areas along theeastern margin of K1 is not clear. In the magmaticemplacement model, they are interpreted as intrusionspredating or syngenetic with the “TKB”s in the centre of

the pipes, since they may show features interpreted asbeing transitional towards the fragmental “TKB”s (i.e.increase of microlitic clinopyroxene and decrease ofcarbonate; Skinner, personal communication 2000).However, the genetic relevance of these features is stillunder discussion. Some massive hypabyssal intrusionsshow a sharp contact to the adjacent fragmentalkimberlite and their intrusion probably post-dates thefragmental kimberlite. In general, hypabyssal kimberliteis more frequent in smaller bodies and in deeper levelsof the pipes, which coincides with the observations indrill cores (see section 5.2).

Floating reefsIn a large area within the western portion of K4,massive, olivine macrocryst-rich hypabyssal kimberliteintrudes large blocks of basement filling the deeperlevels of this pipe (Figure 13). The basement raftsappear to block the pipe, since below the house-sizedblocks only hypabyssal kimberlite occurs, but above andtowards the centre of the pipe a “TKB”-looking rockprevails. We consider the large country rock blocks



Figure 12 (12). The photograph shows clear bedding observed in a bench wall close to the centre of the open pit of K2.



161

filling K4 as part of a large, hardly disintegrated floatingreef, which probably subsided only a relatively smalldistance from its original position. However, furtherdrilling is needed to clarify the structure of this pipe inmore detail.

Observations in drill coresDuring a drilling program in 2000, the first author hascarefully logged a number of drill cores. All cores drilledinto TKB exhibit some degree of bedding.

Drill cores from Pipe K2In general, the bedding in K2 is less apparent than in K1and shows greater bed thicknesses than the upper partsof the cores drilled into K1. In RDH2, the drill holedrilled through the “TKB” of K2, two major units couldbe differentiated. However, bedding planes weredifficult to define (the transition zone between the twomajor units was about 6m thick), although differences inxenolith abundance, -size and -type between and withinthe two major units were clearly visible (Figure 14).Thus, a statistical approach by point counting was usedto confirm the inhomogeneous nature of the “TKB”(Figure 15). Especially at a depth between 40 and 46mthere is a clear increase in xenolith content producing aclast-supported texture within this depth sequence. Thexenoliths consist of a well-mixed assemblage of mantlenodules together with gneiss, amphibolite, dolerite,rhyolite and shale clasts (i.e. a mixture of local basementand overlying sedimentary lithologies), most of themrelatively well rounded and with average clast sizesbetween 5 and 10cm.

Drill hole RDH1a is located in the western part of K2and was drilled into the talus deposits described insection 5.1. Hannweg (1998) and Seggie et al. (1999)noted that the western margin of K2 consists of acountry rock breccia which grades into a 50 to 90m widearea of transitional, xenolith-rich hypabyssal kimberlite.Therefore drill core RDH1a consists of fragmentalkimberlite and hypabyssal kimberlite with variablexenolith contents (up to >80% country rock), countryrock breccias with minor kimberlite matrix and massivecountry rock blocks several tens of metres in section.The latter blocks may show local brecciation and smallintrusions of xenolith-poor hypabyssal kimberlite (<15%xenoliths) and xenolith-rich hypabyssal kimberlite (>15 % xenoliths). The margins of these blocks are oftenhighly brecciated and there seems to be a gradualtransition of the blocks into the adjacent country rockbreccias. Late crosscutting kimberlite dyke intrusions arealso present and consist of nearly aphanitic, highlysegregationary kimberlite.

The hypabyssal kimberlite intersected by RDH1a isolivine-macrocryst rich and shows a fine-grained, darkgreen to brown groundmass. It contains rare autoliths ofa previous hypabyssal kimberlite intrusion and thegroundmass is occasionally segregationary in texture.There is a high probability that several differenthypabyssal kimberlite batches were intersected by

RDH1. These batches show slight differences in colourand olivine-macrocryst size and abundance, but due tothe high content of country rock xenoliths it is hard to distinguish any boundaries.

The matrix filling the interstitial spaces of the countryrock breccia drilled by RDH1a consists of fragmentalkimberlite with juvenile pyroclasts. Finely beddedfragmental kimberlite with a well developed inversegrading is locally present interstitially between largerxenoliths. These inverse graded beds represent eitherprimary bedding of volcaniclastic kimberlite materialdeposited on the surface of the talus fans or post-depositional filter sedimentation of interclast void spacesby a water-mud mixture. A similar process is suggestedby core sections which show a distinct vertical alignmentof platy xenoliths within hypabyssal kimberlite but anabsence of larger olivine macrocrysts (which are usuallypresent), which may have been lost due to a filterprocess of the kimberlite (see above) intruding theclosely-packed xenolith breccia.

In general, the abundance of xenoliths seems toincrease with drill core depth. This could be due to theprobably constricting shape of the pipe (reduction involume) and that the drill core actually approached thecountry rock contact with depth, although it was drilledslightly dipping to the east, away from the contact. Mosthypabyssal breccias contain a wide range of often well-mixed xenolith basement lithologies (mostlyamphibolite, biotite schists and gneisses). Sedimentaryxenoliths or dolerites (as they have been found in the“TKB’s” of drill hole RDH2) were not found, confirmingthe observation in the talus fans within the open pit (seesection 5.1). The shape of the basement xenoliths ismainly angular to sometimes subrounded, and the clastsize is highly variable from a few mm to several metres.No correlation between the degree of rounding, the clastsize, and any particular lithology was observed. Somewell-rounded xenoliths (up to 20cm) have beenobserved from 124m to 126.5m depth.

Drill core DDH195 supports the principal structure ofpipe K2: TKB-like, fragmental material fills most of thecentral and eastern portion of the pipe and is underlainto the west by very country rock rich material. Thehypabyssal kimberlite locally intruded into the latterbreccias is widely of segregationary nature, butfragmental (bedded) volcaniclastic kimberlite can alsobe found in between country rock blocks. These countryrock rich basement breccias are interpreted as talus fanssuccessively filling the crater and which are overlain bya later massive pyroclastic or resedimentation event (seesection 5.1).

Drill cores from Pipe K1Drill holes from pipe K1 were either drilled intofragmental kimberlite or into the marginally occurringhypabyssal intrusions in the southeast and northeast. K1was originally subdivided into the East and West “TKB”(Seggie et al., 1999), but the petrographic differencesbetween these two “TKB’s” are only minor. However, all

cores drilled into fragmental kimberlites show more orthe less distinct bedding, from a local centimetre scaleup to several tens of metres in bed thicknesses.Especially the uppermost portions of the drill coresreveal relatively well sorted volcaniclastic rocks withnormal graded, 3 to 8m thick beds. Bedding planes orinterfaces between different beds are usually not verydistinct or show soft-sediment deformation textures. If distinct interfaces are present, they often show a dipbetween 30° and 80°.

Towards deeper sections of the drill cores the bedstend to become thicker and more difficult to distinguish– bedding planes are almost always obscured. However,the beds are still characterised by frequently cleardifferences in xenolith abundance, size and types. Theabundance and size of olivines may also vary clearly.Bed thicknesses are frequently between several tens ofmetres up to about 80m. Local orientation of olivines

suggest shear movement during deposition orcompaction, which could diffuse or destroy originalbedding features. The thickest unstructured section ofabout 80 m in a drill core was encountered in the coresection from 235.5m to 317.1m in drill hole RDH6, whichappears as massive, TKB-like rock. Although there areminor lithological differences in this section, thedifferences are too vague to suggest a furthersubdivision. This massive section is underlain by adistinctly different, fine grained material, suggesting thatalso the overlying massive bed is still of volcaniclasticorigin and is either a single unit of a massive back fallor mass flow or a combination of several units whichhave lost their primary bedding structures due to relativemovement and compaction.

Many beds, including those beds situated in deeperlevels of the pipe, often show fine-grained tops withthicknesses between a few dm and two metres. These



Figure 13 (13). On bench 3 in K4 highly fractured amphibolite is intruded by at least two phases of hypabyssal kimberlite (light and dark grey

veining suggest that the entire block is not in situ and is in fact a floating reef. The dashed line in the plan view (inset) shows the delineation of



163

tops are usually well sorted (clast sizes < 2cm), normallygraded, and show planar and cross bedding on a cm-scale (Figure 16). Only a single bed shows inversegrading with an increase of the matrix content towardsdepth (188.95 m to194.3m, RDH6). Isolated, irregularpatches of fine-grained material (originally ash layers?)can be frequently found within the massivevolcaniclastic layers. They usually show soft sedimentdeformation textures (Figure 17), testifying theirunconsolidated nature and the presence of moisture stilltrapped within them during compaction. It is difficult tointerpret the origin of these finer-grained layers as they may represent either resedimented material oforiginally graded pyroclastic or epiclastic beds, the fine-grained products of elutriation ash clouds or basesurge deposits.

The normal colour of the drill cores is light olive togrey, but the fragmental kimberlite may show intensely

red and brown coloured sections, often with very sharptransgressions from the still dominating olive colours.There is no direct evidence that the colour change isattributed to a lithological change (bedding), but the redareas seem to incorporate a slightly higher proportion ofred mud clasts. These mud clasts are often internallyhomogenous, but may also show fine, partly convolutelamination or may contain themselves finely brecciatedsmaller mud clasts. Occasional small olivinesincorporated in these mud clasts suggest a direct geneticrelationship with the kimberlites of the cluster.

The fact that these mud clasts as well as theallochthonous crater blocks (see section: 5.1) areembedded within more massive volcaniclastic unitsclearly suggests the presence of various phases ofvolcanic activity and inactivity during the life span of theVenetia volcanoes. The crater blocks are fine-grained,laminated and may show soft sediment deformation

colours). Note the difference in the direction of the foliation, which reflects a country rock fold. Increased fracturing and intensive kimberlite

the floating reef towards the in situ country rock.



Figure 14. missing

Figure 15 (15). Percentage of olivines (> 2mm), xenoliths (> 2mm), and groundmass/matrix (all constituents < 2mm) over depth.

The analyses were done using a line scan about all three metres in distance with a ruler over an area of 1m, using the cm-lines as reading

points. Pipe K2, drill core RDH2.

RUNNING HEAD SANDY CORRECT


165

textures. This and the presence of mudclasts suggest ahigh moisture content of the sediment during depositionand possibly even reworking under the influence offreely flowing surface water. The fine grain sizes suggesta relatively low relief gradient on the crater floor, i.e. thatthe crater was either already filled up and levelled withthe surrounding land surface at a certain stage within thevolcano’s life span, or that this fine-grained materialexplicitly derives from a central playa or crater lakeenvironment. Fragments of earlier mud clasts alsosuggest reworking of earlier consolidated material,supporting the temporal gap between the volcanicphases. This means that groundwater must have beenpresent in these phases of volcanic inactivity at a veryhigh structural level within the pipe and on the maarcrater surface. Since these mud clasts and allochthonouscrater blocks are incorporated into massive volcaniclasticlayers with a certain degree of moist content of theirown (see above), groundwater must have been alsopresent during the onset of the volcanic event producingthe host layers.

Clear evidence that external water – or at least a highmoisture content - was present during the volcanicactivity of Venetia and that the intersected rocks are in

fact products of a volcaniclastic process and notintrusive rocks like a “TKB” sensu strictu, is the coreintersection between 27.6m and 33.7m of drill core RDH7 (“TKB” East in K1). This section starts (from the top) with finely laminated beds of very fine-grained material (Figure 18), which towards depthbecomes more and more interbedded with lenses, andlayers of medium- to coarse-grained sandstone. Thetopmost laminae show distinct convolute beddingincorporating some microfaulted blocks of the samelaminated material. This suggests initial consolidation ofthe rock (probably during a drying process form thesurface), but also that water was still trapped in deeperlayers during the onset of the drying process. Theconvolute deformation is either a result of sliding of thestill water-rich sediment or the sedimentation of the nextvolcaniclastic layer on top of the drying surface that mayhave squeezed the water out by its weight (Figure 18).With increasing depth the laminae disappear and thelayer becomes more and more coarse-grained andmassive. After about 3.7m (at –31.3m depth in drill holeRDH7) the bedding planes turn from ± horizontal toalmost vertical, and a few dm deeper the verticalbedding structures are replaced by vertical orientation of

Figure 16 (19). A fine-grained top layer of a bed shows some cross- and lensoid bedding. Pipe K1, drill core DDH203, depth increase

from top to bottom, left to right.

olivines parallel to vertically oriented, very fine-grainedthin tracks, either representing water escape structuresor shear zones. The last dm of this zone arecharacterised by homogeneous fragmental kimberlite,which shows a sharp and steep contact to the nextunderlying graded bed.

This core section is interpreted as a (resedimented)fine ash layer, probably reflecting a period of relativelow activity in the life span of the volcano. The craterenvironment is supported by the relatively well-definedgraded beds overlying this section (see above), whichare less massive than the beds beneath this section. Softsediment deformations, convolute bedding and thelensoid appearance of mm-thick coarser layersfurthermore suggest the presence of a high moisturecontent or even of free surface water. The scenario ofsedimentation of a fine elutriation ash cloud settlingonto the maar crater floor (into a shallow crater lake?)with subsequent resedimentation seems plausible.Microfaults (see above) on top of the layer and a clast ofthese laminated sediments found in the graded bedsabout 14m higher in the succession also suggest initialconsolidation (drying) of the rock before the next layerwas sedimented on top of it.

Drill holes drilled into the adjacent “TKB West” showa very similar picture of bedded volcaniclastic material,often with normal grading and signs of reworking,especially in the finer grained upper parts of the beds.The bed thicknesses vary from several metres to fewtens of metres. Grading is restricted to few dm at the topof the respective beds, otherwise the beds appearmassive with only gradual differences in xenolith orolivine size or abundance. The fine grained tops of thebeds sometimes show fine planar bedding andlamination or even cross bedding, suggesting reworkingor possibly a base surge origin.

The rocks of the drill cores are frequently cut bymostly thin, kimberlite dykes. The dykes are normallyfine grained to aphanitic, are segregationary in textureand show a sharp and steeply dipping contact to the

surrounding fragmental rocks, which often reveal acontact metamorphic change in colour. Some of thedykes consist of a multitude of different kimberlitephases and show an enrichment of carbonate towardsthe margins of the dyke (often more on top of inclineddykes). Locally, strong flow alignment of olivines isevident. The most striking feature of most of the dykesis the presence of carbonate quench crystals, oftenforming a spinifex-like texture in the rock.

The last 30m of drill core RDH6 are made up byolivine macrocryst-rich hypabyssal kimberlite. Theinterface to the overlying volcaniclastic unit is sharp anddoes not show signs of extensive alteration. Thehypabyssal rock is locally segregationary in texture, but acontinuous transition from the coherent hypabyssal rockinto the fragmental “TKB” was not observed. This shouldhave been evident if the hypabyssal and the volcaniclastickimberlite were cogenetic (e.g. Clement, 1982).

Observations in thin sectionsKimberlite samples from Venetia have been identified ashypabyssal and crater facies rocks, but also as typicalsouthern African “TKB’s” (Colgan, personalcommunication 1982; Field and Scott Smith, 1999;Skinner, personal communication 2000; own data).

Hypabyssal kimberlitesHypabyssal kimberlite samples are in many cases veryolivine macrocryst-rich and unusually closely packed –sometimes even grain supported. They may besegregationary in texture, although most hypabyssalkimberlites show a homogeneous groundmass.Segregationary-textured kimberlites (e.g. Clement, 1982)are often better developed towards the contact with“TKB’s”, but sometimes hypabyssal kimberlites can stillshow a non-segregationary, homogeneously-texturedgroundmass immediately adjacent to the “TKB” contact.However, all contacts between the hypabyssal and thevolcaniclastic kimberlites observed in drill holes weresharp and not transitional.



Figure 17 (20). The photograph shows a drill core section which is characterised by irregular patches of fine grained material (highlighted

by black dashed line) embedded within a coarser grained host rock. Pipe K1, drill core RDH6, depth increase from top to bottom, left to

right.



167

Late intruding kimberlite dykes, ranging in thicknessfrom a few cm to several metres, frequently cutfragmental as well as hypabyssal kimberlites. Thesedykes often show a concentration of larger olivinemacrocrysts within their central areas. However, slightlyinclined dykes may show normal olivine gradationtowards the lower edge due to gravitational settlement.Some dykes show signs of multiple injection phases.The contact with the intruded host rock is usually fine(r)grained.

The relative abundance of olivine-poor (nearlyaphanitic), very carbonate-rich, mostly thin dykelets ormarginal sections of thicker dykes which occur indeeper levels of the Venetia pipes (see above) is afeature peculiar to Venetia. These carbonate-kimberlitesoften show a spinifex-like carbonate quenchcrystallisation forming a comb-layered texture. Thequench crystals are up to 2cm long and usually grewperpendicular to the cooling front from the contact tothe country rock into the still hotter central portions of

Figure 18 (21). Soft sediment deformation on top of the finely laminated unit. Depth increase from left to right. Pipe K1, drill core RDH7.

Figure 19 (28), sample 173/50/001/1148: Thin section photograph from an allochthonous crater block from bench 10, north-west side of

K1 open pit. A deformed patch of fine-grained sediment is embedded within well sorted coarser-grained material, which consists mainly

of angular quartz grains and clay-altered olivine fragments. PPL.

the dyke. However, sometimes the carbonate quenchcrystals occur in isolation within the dykes and are flow-aligned throughout the intrusion. Along the contactbetween these carbonate-rich late dykelets and theircoarse-grained groundmass host rocks, abundant coarsespinel is found.

The more massive hypabyssal kimberlite intrusionsoccasionally show carbonate-filled probable amygdales.Before the void space of the vesicles was filled withcarbonate, relatively coarse-grained microliticclinopyroxene was growing from the wall of the vesicleinto the vacant spaces from the sides.

“Tuffisitic Kimberlite Breccias”A truly interesting aspect of the volcanology of the

Venetia pipes is the presence of microscopically typicalSouthern African “Tuffisitic Kimberlite Breccias” withinmacroscopically clearly bedded volcaniclastic material.In thin section, most of the samples representingfragmental kimberlites have been identified as “TK’s” or“TKB’s” by Colgan (personal communication 1982),Field and Scott Smith (1999); Skinner (personalcommunication 2000) and own observations.Petrographically the rocks are typical southern African“TKB’s” and comparable to those found in the Kimberleypipes (Field and Scott Smith, 1999). However, judgingfrom drill hole logs and the depth of these samples, theyderive from clearly bedded volcaniclastic material or

originate from already mined sections of the pits. Sincemore or less obvious bedding was found in all drill coresdrilled into fragmental kimberlite (especially in the upper parts of the cores), it is assumed that theoverlying, mined-out sections of the pipes also showedmacroscopic evidence of bedding. Even the moremassive bottom parts of the drill cores show signs ofbedding (although by far less obvious than in theirupper parts). The deepest drill hole information ofbedded cores from K1 is at –350m (RDH7), drilled fromthe bottom of the pit at –80m from the present surface.Together with the amount of erosion sinceemplacement, which is estimated to be between 300 and600m, the thickness of bedded material would thus liebetween 700 and 1000m.

By this observation the samples cannot qualify astuffisitic kimberlite breccias sensu strictu, since these areby definition intrusive rocks and as such are expected tobe massive and unbedded in nature (Clement and Reid,1989). One could argue that these “TKB’s” are actuallyderived from the transition zone between an intrusiveTKB potentially present in deeper levels of the pipe (notyet intersected by drilling) and the overlying zone ofmassive back fall (or mass flows) filling the vent.However, the latter would then be of volcaniclastic andnot of intrusive origin and the nomenclature forpyroclastic or epiclastic rocks must therefore be used.Judging from the development of the volcaniclastic



Figure 20 (29), sample 173/050/K001/0183, drill core DDH010, K1 open pit (already mined area): The thin section photograph shows

accretionary (armoured) ash grains, which testify to water within the eruption clouds at Venetia. XPL.



169

material present in cores towards depth, the authorsexpect that the fragmental kimberlite becomes more andmore massive and unbedded in nature. Whether thismaterial should then be regarded as a true TKB in thesense of an intrusive rock is highly questionable. A lossof bedding, which is in the constricting lower portion ofa pipe difficult to preserve in any case, could easily beexplained by primary volcanic processes (inflation anddeflation of the lower pipe portions by explosions in theroot zone; Lorenz and Kurszlaukis, in preparation) or bysecondary processes due to e.g. differential compaction.

Volcaniclastic kimberlitesAs mentioned earlier, the volcaniclastic kimberlitesfound in the pits and drill holes of the Venetia clustershow a great variety of sedimentary features includinglamination, cross bedding, normal and inverse grading,sorting, soft sediment deformations etc., and most ofthese features can also be observed in the thin sectionsof the respective samples (see Figure 19). As mentionedabove (sections 5.1 and 5.2), isolated blocks of wellbedded and even laminated material can be foundwithin the coarser bedded volcaniclastic material fillingthe vents. These isolated epiclastic blocks generallycontain a high abundance of well sorted, mostly angularto minor subrounded quartz grains. Sandstone clasts ofsimilar material were not observed. The relativeangularity and sorting of the quartz grains, and theabsence of sandstone clasts of the same material suggesteither the presence of a large amount of unconsolidatedsand on top of the paleosurface at the time ofemplacement or a high fragmentation energy duringvolcanic activity which was able to fragment the countryrock down to small grain sizes. However, these quartzgrains were not observed in the bedded “TKB”-likematerial, which enclose the epiclastic blocks.

Mud-rich layers interbedded with the quartz-richepiclastic beds occur, which may contain occasionalquartz and altered olivine grains. The most frequentconstituents of these mud-rich layers are reworkedgranule- and pebble-sized rounded mud clasts of earlierconsolidated mud-rich layers. These clasts, and the factthat the isolated blocks of this epiclastic material arelocated within the more massive volcaniclastic material,suggests that the emplacement of Venetia was a long-lived process with considerable periods of inactivitybetween renewed volcanic activity, i.e. the emplacementwas not a single event.

Another interesting feature of Venetia K1 is thepresence of occasionally abundant accretionary(armoured) ash grains and lapilli, which are best seen inthin section (Figure 20). These pyroclasts were found in samples from drill hole DDH 10 drilled into thecentral part of K1. Unfortunately they derive from analready mined-out level (core interval from 78.5m to103.5m in the -62° inclined hole) and potential bedthicknesses within this interval are not known. The ashgrains and lapilli have diameters between 0.5 and 2.5mmand often have an altered olivine (and rarely a country

rock fragment) as kernel. The kernels are surrounded byup to three, highly altered and very fine-grained rimsconsisting of serpentine, clay minerals and carbonate(recrystallisation products). Within these rims, very smallphlogopite laths are concentrically aligned around thekernels. On occasions, the outermost rim is partiallymissing or seems to be squeezed into intergranularspaces. The abundance of the pyroclasts varies from thinsection to thin section, but may reach a grain-supportedtexture (> 50 % of the rock volume).

These very specific pyroclasts are usually formedwithin a moisture-rich eruption cloud and testify to thepresence of water during eruption (Schumacher andSchmincke, 1991; 1995). They are usually associatedwith a phreatomagmatic eruption process.

ConclusionsThe detailed study of the Venetia kimberlite pipe clusterclearly illustrates a distinct influence of structuralfeatures of the country rock onto the pipe shapes. Theshapes and positions of the pipes are widely controlledby pre-existing inhomogeneities in the country rock.The pipes are elongated and follow the strikes of shearzones, faults and other inhomogeneities within thecountry rock. Some of these country rock structureswere re-activated after emplacement of the kimberlite. Anumber of pipes (K4, 5 and 6, as well as parts of K1) areeven located on the intersection of faults. The pipeshapes are then irregular or lobe-shaped and follow thelocal trends in the country rock (see K2 and possibly K1and K4). At least some of the country rock breccias,which had marginally slumped into the pipes, are boundby older faults. Due to the close relationship of countryrock inhomogeneities with pipe shapes it is of greatimportance to map the immediate country rock with thesame accuracy as the internal structure of the pipe,especially for mining purposes.

The Venetia kimberlite pipes have been regarded astypical southern African pipes, i.e. they are large andpresumably deep pipes occupied by a fragmental rockwhich had been termed “Tuffisitic Kimberlite Breccia”,or “TKB”. The origin of this rock had been thought to beintrusive, and by that it should be unbedded andmassive.

Our work has shown that bedding is present on amicro-, macro-, and megascale throughout the entirebody of fragmental kimberlite down to a depth of~430m from the present land surface. In general,bedding seems to become less obvious towards depthand the beds become more massive, often up to severaltens of metres thick with only thin, fine(er)-grained andpartly reworked or primary pyroclastic tops. Althoughthese bedded rocks are clearly of volcanogenic origin,they still classify as classical “TKB’s” in thin section.

Bedding, signs of reworking in between and on topof the massive “TKB” layers, and the presence of mud-clasts and quartz-rich layers within fine grained and finebedded crater-facies blocks found as isolated blocks inthe younger volcaniclastic kimberlite hosting them as

well as the presence of talus fans suggest time spans ofrelative volcanic inactivity in between very violentphases, contradicting the presence of a single, highly-energetic eruption.

Water was evidently present during emplacement ofthe Venetia kimberlite pipes. This is testified by thepresence of accretionary (armoured) lapilli, possiblelake beds, soft sediment deformation textures and crossbedding observed in the reworked, fine-grained layersoverlying massive beds. Isolated crater-facies blockswith unknown genetic relationship to the hostingvolcaniclastic rocks clearly show bedding (lamination),sorting, the presence of mud clasts and soft sedimentdeformation textures. All these features suggest a highmoisture content of the volcaniclastic sediments duringand after emplacement and possibly even free flowingwater on the crater floor.

Collapse structures as seen in the slump breccias atthe south-west side of K1 and other contact breccias,sidewall oversteepening, a possible floating reef in K4,and the collapsed roof of a possible cavity at the north-west side of K2 all suggest a volume deficit in deeperlevels of the pipe. The nature of these apparentlyslumped country rock breccias around most of the pipesand W. Barnett as part of a doctoral thesis isinvestigating the evidence for their origin.

General problems are the different size and thedifferent structural levels of the pipes within the samecluster of the same age at the same erosion level. At K4,a relatively small pipe, a contact breccia is present,which is rather typical for root zones or lower levels ofa diatreme. This is also the case at K8, a xenolith-richhypabyssal intrusion which is surrounded by anenvelope of highly brecciated country rock. This pipemay well be interpreted as an exhumed root zonesurrounded by contact breccias. In adjacent K1, thelargest pipe of the cluster, bedded volcaniclastickimberlite is present over 350m deeper than the contactbreccia in K4. At the same depth level talus fans withinK2 suggest sedimentation onto the crater floor, whichare overlain by “TKB”-looking rocks.

A fundamental problem is the presence ofpetrographically identified “TKB”s. In thin section theyfulfil all requirements for typical “TKB”s (Colgan,personal communication, 1982; Field and Scott Smith,1999, Skinner, personal communication 2000, and owndata), but they derive from bedded sequences and evenoverlie crater floor talus deposits in K2. We regard theserocks as volcaniclastic deposits and not as intrusiverocks.

A further open question is by which kind of processthe volcaniclastic layers filling the pipes were formed.They are either pyroclastic or reworked epiclastic andmay either be related to the same volcano into whichthey were deposited in or related to a neighbouringvolcano in the same cluster (for the latter case onlyrather thin beds are expected) (Cloos, 1941; Lorenz,1982). It is also not clear to which extend the diatremewas evacuated by a potentially vent-clearing process, or

whether the maar crater was relatively shallow and filledup by the renewed sedimentation of volcaniclasticmaterial during ongoing subsidence (Lorenz, 1985;1986).

The authors stress that the information provided inthis paper is the state of knowledge at the time ofsubmission of the manuscript. We are aware, that withthe continuous mining of the pipes, more featuresrelevant for developing an emplacement model will be discovered. However, in any model that will bedeveloped to explain the emplacement of the Venetiacluster, all of the volcanological and structural featuresdescribed above must be taken into consideration. Theauthors suggest that the observations presented in thispaper are more easily explained by a volcanicemplacement history that was dominated by episodes ofphreatomagmatic eruptions.

AcknowledgementsThe authors thank the De Beers Consolidated Mines Ltd.for the support provided and the permission to publishthe data. We would like to thank all our colleagues whocontributed in discussions to the final stage of this paper.Venetia Mine, its staff and the Venetia mine guest houseare thanked for their outstanding hospitality andsupport. Special thanks goes to Peter Davis for hisvaluable support during the drill core-logging program.The paper benefited from the helpful and constructivecomments of J. M. Barton Jr., M. Field, C. Hatton, V. Lorenz, C.B. Smith, J. Stiefenhofer and F. Winter.

ReferencesBarton, J.M., Jr. and Pretorius, W. (1998). Crustal xenoliths in Venetia

kimberlite pipes indicate a dÈcollement at ~10km beneath the Central Zone

of the Limpopo Belt, South Africa. South African Journal of Geology, 101,

323-327.

Barnett, W.P. (2002). Geological Control on Slope Failure Mechanisms in the

Open Pit at Venetia Mine. South African Journal of Geology, South African

Journal of Geology, 106 __-__.

Brown, R.W., Gallagher, K., Griffin, W.L., Ryan, C.G., de Wit, M.C.J., Belton,

D.X. and Harman, R. (1998). Kimberlites, accelerated erosion and evolution

of the lithospheric mantle beneath the Kaapvaal craton during the mid-

Cretaceous. Extended abstracts of the 7th International Kimberlite

Conference, Cape Town, South Africa, 105-107.

Büchel, G. (1984). Die Maare im Vulkanfeld der Westeifel, ihr

geophysikalischer Nachweis, ihr Alter und ihre Beziehung zur Tektonik der

Erdkruste. Dr. rer. nat. thesis, Faculty of Earth Sciences, University of

Mainz. 385 pp. Please spell out abreviations

Büchel G. (1993). Maars of the Westeifel (Germany). In: J.F.W. Negendank,

and B. Zolitschka (Editors): Paleolimnology of European Maars. Lecture

Notes Earth Sciences, Springer, Berlin, Heidelberg, Germany, 49, 1-13.

Büchel, G., Jacoby, W., Lorenz, V. and Zimanowski, B. (1987). Geophysical

aspects of the formation of Eifel maars. Terra Cognita, G 11.54, 365.

Bumby, A.J., Eriksson, P.G., van der Merwe, R. and Maier, W.D. (2001). The

stratigraphic relationship between the Waterberg and Soutpansberg Groups

in Northern Province, South Africa: evidence from the Blouberg area. South

African Journal of Geology, 104, 205-216.

Büttner, R. (1997). Molten fuel coolant interactions in entrapment

configuration. PhD thesis (unpublished), University of Würzburg,

Germany. 73 pp.

Büttner, R. and Zimanowski, B. (1998). On the physics of thermo-hydraulic

explosion. Physical Review E, 57, 5726-5729.

Card, K.D., Sanford, B.V. and Card, G.M. (1997). Controls of the

Emplacement of Kimberlites and Alkalic Rock-carbonatite Complexes in





171

the Canadian Shield and Surrounding Regions. Exploration Mining

Geology, 6, 285-296.

Clement, C.R. (1975). The emplacement of some diatreme facies kimberlites.

Physics and Chemistry of the Earth, 9, 51-59.

Clement, C.R. (1982). A comparative geological study of some major

kimberlite pipes in the Northern Cape and Orange Free State. Ph.D. Thesis

(unpublished), University of Cape Town, South Africa, 406 pp.

Clement, C.R. and Reid, A.M. (1989). The origin of kimberlite pipes: an

interpretation based on the synthesis of geological features displayed by

southern African occurrences. Proceedings of the 4th International

Kimberlite Conference, Perth, Australia. Geological Society of Australia,

632-646.

Clement, C.R. and Skinner, E.M.W., (1985). A textural-genetic classification of

kimberlites. Transactions of the Geological Society of South Africa, 88, 403-

409.

Cloos, H. (1941). Bau und Tätigkeit von Tuffschloten. Untersuchungen an

dem Schwäbischen Vulkan. Geologische Rundschau, 32, 709-800.

Crough, S.T., Morgan, W.J. and Hargraves, R.B. (1980). Kimberlites: their

relation to mantle hotspots. Earth and Planetary Science Letters, 50, 260-

274.

Cookenboo, H.O. (1999). History and Process of Emplacement of the Jericho

(JD-1) Kimberlite Pipe, Northern Canada. Proceedings of the 7th

International Kimberlite Conference, Cape Town, South Africa, 125-133.

Deakin, A.S. and White, S.H. (1991). Shear zone control of alkali intrusives;

Examples from Argyle, northwestern Australia and Yengema, West Africa.

Proceedings of the 4th International Kimberlite Conference, Perth,

Australia, 251-257.

Dobbs, P.N., Duncan, D.J., Hu, S., Shee, S.R., Colgan, E., Brown, M.A., Smith,

C.B. and Allsopp, H.L. (1994): The geology of the Mengyin kimberlites,

Shandong, China. In: H.O.A. Meyer and O.H. Leonardos, (Editors).

Proceedings of the 5th International Kimberlite Conference, Araxa, Brazil,

1991. 40-61.

Doorgapershad, A., Barnett, M., Twiggs, C., Martin, J., Mellonig, L., Zenglein,

R., Klemd, R., Barton, E. S., Barnett, W. and Barton, J. M., Jr. (2003) The

geology of the area surrounding the Venetia kimberlite pipes, Limpopo

Belt, South Africa: A complex assembly of terranes and granitoid

magmatism. South African Journal of Geology, 106, __-__.

Friese, A.E.W. (1998). Structural control on kimberlite genesis and crustal

emplacement within South Africa and the Kaapvaal Craton during the

Cretaceous. Extended Abstracts of the 7th International Kimberlite


Field, M., Gibson, J.G., Wilkes, T.A., Gabakotse, J. and Khutjwe, P. (1997).

The geology of the Orapa A/K1 kimberlite Botswana: Further insight into

the emplacement of kimberlite pipes. Russian Geology and Geophysics, 38,

24-39.

Field, M. and Scott Smith, B.H. (1999). Contrasting geology and near-surface

emplacement of Kimberlite Pipes in Southern Africa and Canada.

Proceedings of the 7th International Kimberlite Conference, Cape Town,

South Africa. 214-237.

Hannweg, G.W. (1998). Venetia Diamond Mine. 7th International Kimberlite

Conference, Cape Town, 1998: Large mines field excursion guide. 23-31.

Hawthorne, J.B. (1975). Model of a kimberlite pipe. Physics and Chemistry

of the Earth, 9, 1-15.

Hoffmann, A., Kröner, A. and Brandl, G. (1998). Field relationships of mid-

to late Archaean high-grade gneisses of igneous and sedimentary

parentage in the Sand River, Central Zone of the Limpopo Belt, South

Africa. South African Journal of Geology, 101, 185-200.

Kienle, J., Kyle, P.R., Self, S., Motyka, R.J. and Lorenz, V. (1980). Ukinrek

Maars, Alaska, April 1977 eruption sequence, petrology and tectonic

setting. Journal of Volcanology and Geothermal Research, 7, 11-37.

Kröner, A., Jaeckel, P., Hoffmann, A., Nemchin, A. and Brandl, G. (1998).

Field relationships and age of the supracrustal Beit Bridge Complex and

associated granitoid gneisses in the Central Zone of the Limpopo Belt,

South Africa. South African Journal of Geology, 101, 201-213.

Kurszlaukis, S., Büttner, R., Zimanowski, B., and Lorenz, V. (1998a). On the

first experimental phreatomagmatic explosion of a kimberlite melt. Journal

of Volcanology and Geothermal Research, 80, 323-326.

Kurszlaukis, S., Franz, L., and Lorenz, V. (1998b): Volcanology of the Gibeon

Kimberlite Volcanic Field, southern Namibia. Journal of Volcanology and

Geothermal Research, 84, 257-272.

Kurszlaukis, S. and Lorenz, V. (1996). Volcanological features of low viscosity

melt: the carbonatitic Gross Brukkaros Volcanic Field, Namibia. Bulletin of

Volcanology, 58, 421-431.

Kurszlaukis, S. and Lorenz, V. (2000). Volcanology of the kimberlitic Gibeon

Volcanic Field, southern Namibia. Communications of the Geological

Survey of Namibia (Henno Martin Commorative Volume), 12, 353-358.

Lorenz, V. (1971a). Collapse structures in the Permian of the Saar-Nahe area,

Southwest Germany. Geologische Rundschau, 60, 924-948.

Lorenz, V. (1971b). An investigation of volcanic depressions. Part IV. Origin

of Hole-in-the-Ground, a maar in central Oregon (geological, geophysical

and energy investigations). NASA progress report (NGR-38-003-012,

Houston, Texas), 113 pp.

Lorenz, V. (1973). On the formation of maars. Bulletin of Volcanology, 37,

183-204.

Lorenz, V. (1975). Formation of phreatomagmatic maar-diatreme volcanoes

and its relevance to the formation of kimberlite diatremes. In: Ahrens, L.H.,

Dawson, J.B., Duncan, A.R. and Erlank, A.J. (editors): Proceedings of the

1st International Kimberlite Conference, Cape Town, South Africa, 1973.

Physics and Chemistry of the Earth, 9, 17-27.

Lorenz, V. (1979). Phreatomagmatic origin of the olivine melilitite diatremes

of the Swabian Alb, Germany. In: Boyd, F.R. and Meyer, H.O.A. (editors):

Kimberlites, diatremes, and diamonds: Their geology, petrology, and

geochemistry. - American Geophysical Union, Washington, 354-363.

Lorenz, V. (1982). Zur Vulkanologie der Tuffschlote der Schwäbischen Alb.

Jahresberichte Mitteilungen oberrheinische geologische Vereinigung, N.F.,

64, 167-200.

Lorenz, V. (1985). Maars and diatremes of phreatomagmatic origin, a review.

Transactions of the Geological Society of South Africa, 88, 459-470.

Lorenz, V. (1986). On the growth of maars and diatremes and its relevance

to the formation of tuff-rings. Bulletin of Volcanology, 48, 265-274.

Lorenz, V. (1987). Phreatomagmatism and its relevance. Chemical Geology,

62, 149-156.

Lorenz, V. (1998). Zur Vulkanologie von diamantführenden Kimberlit- und

Lamproit-Diatremen. Zeitschrift der Deutschen Gemmologischen

Gesellschaft, 47, 5-30.

Lorenz, V. and Büchel, G. (1980). Zur Vulkanologie der Maare und

Schlackenkegel der Westeifel. Mitteilungen Pollichia, 68, 29-100.

Lorenz, V. and Kurszlaukis, V. (1997). On the last explosions of carbonatite

pipe G3b, Gross Brukkaros, Namibia. Bulletin of Volcanology, 59, 1-9.

Lorenz, V. and Kurszlaukis, V. (in preparation). Root zone processes and

diatreme growth in kimberlite and other maar-diatreme volcanoes.

Lorenz, V., Zimanowski, B. and Buettner, R. (1999a). Discussion on the

formation of kimberlite pipes: the phreatomagmatic model. Newsletter of

the IAVCEI Commission on Explosive Volcanism, September, 1999; Flagstaff,

Arizona, USA. 11-17.

Lorenz, V., Zimanowski, B., Büttner, R. and Kurszlaukis, S. (1999b).

Formation of kimberlite diatremes by explosive interaction of kimberlite

magma with groundwater; field and experimental aspects. Proceedings of

the 7th International Kimberlite Conference, Cape Town, South Africa. 522-

528.

McCallum, M.E. (1985): Experimental evidence for fluidisation processes in

breccia pipe formation. Economic Geology, 80, 1523-1543.

McCandless, T.E. (1999). Kimberlite: mantle expressions of deep-seated

subduction. Proceedings of the 7th International Kimberlite Conference,

Cape Town, South Africa. 545-549.

Morrissey, M., Zimanowsky, B., Wohletz, K. and Büttner, R. (2000).

Phreatomagmatic fragmentation. In: H. Sigurdsson, B. F. Houghton, S. R.

McNutt, H. Rymer, H. and J. Stix (Editors): Encyclopedia of Volcanoes.

Academic Press, San Diego, U.S.A. 431-445.

Muusha, M. and Kopylova, M. (1998). River Ranch Diamond Mine. 7th

International Kimberlite Conference, Cape Town, South Africa: Large

mines field excursion guide, 33-37.

Ort, M., Wohletz, K. Hooten, J, Neal, C. and McConnell, V. (2000): The

Ukinrek maars eruption, Alaska, 1977: a natural laboratory for the study of

phreatomagmatic processes at maars. 1st International Maar Conference,

Daun/Vulkaneifel. Terra Nostra, 6, 396-400.

Phillips, D., Kiviets, G.B. Barton, E.S., Smith, C.B., Viljoen, K.S., and Fourie,

L.F. (1999). 40Ar/39Ar dating of kimberlites and related rocks: problems and

solution. Proceedings of the 7th International Kimberlite Conference, Cape

Town, South Africa. 677-687.

Pretorius, W. (1996). A geochemical and geophysical investigation of a suite

of crustal and upper mantle nodules from Venetia kimberlite pipes,

Limpopo Belt, South Africa. MSc thesis (unpublished), Rand Afrikaans

University, Johannesburg, South Africa. 275pp.

Sage, R.P. (1999). Project Unit 93-12. Structural patterns and kimberlite

emplacement. Ontario Geological Survey Report, Open File Report, 5937,

224-229.

Schumacher, R. and Schmincke, H.U. (1991). Internal structure and

occurrence of accretionary lapilli – a case study at Laacher See Volcano.

Bulletin of Volcanology, 53, 612-634.

Schumacher, R. and Schmincke, H.U. (1995). Models for the origin of

accretionary lapilli. Bulletin of Volcanology, 56, 626-639.

Scott Smith, B.H. (1999). Near surface emplacement of kimberlites by

magmatic processes. Newsletter of the IAVCEI Commission on Explosive

Volcanism, September 1999; Flagstaff, Arizona, USA. 3-10.

Seggie, A.G., Hannweg, G.W., Colgan, E.A. and Smith, C.B. (1999). The

geology and geochemistry of the Venetia kimberlite cluster, Northern

Province, South Africa. Proceedings of the 7th International Kimberlite


Skinner, E.M.W. and Clement, C.R. (1979). Mineralogical classification of

southern African kimberlites. In: F. R. Boyd and H. O. A. Meyer (Editors):

Kimberlites, diatremes and diamonds: their geology, petrology and

geochemistry. American Geophysical Union, Washington, D.C., U.S.A., 129-

139.

Watkeys, M.K. (1981). Reconnaisance geology of the farms Gotha-673 and

Koningsmark-707, Northern Transvaal. Unpublished report, University of the

Witwatersrand, Johannesburg, South Africa, 18 pp.

Watkeys, M.K. (1983a). The Precambrian Geology of the Limpopo Belt, West

and North of Messina. PhD Thesis (unpublished), University of

Witwatersrand, Johannesburg, South Africa, 391pp.

Watkeys, M.K. (1983b). A Retrospective View of the Central Zone of the

Limpopo Belt, Zimbabwe. Special Publication of the Geological Society of

South Africa, 8, 65-80.

White, D.L. (1991). Maar-diatreme phreatomagmatism at Hopi Buttes, Navajo

Nation (Arizona), USA. Bulletin of Volcanology, 53, 239-258.

White, J.D.L. and Houghton, B.F. (2000). Surtseyan and related

phreatomagmatic eruptions. In: Sigurdsson, H., Houghton, B.F., McNutt,

S.R., Rymer, H. and Stix, J. (editors): Encyclopedia of Volcanoes. Academic

Press, San Diego, U.S.A. 495-512.

White, S.H., de Boorder, H. and Smith, C.B. (1995). Structural controls of

kimberlite and lamproite emplacement. Journal of Geochemical

Exploration, 53, 245-264.

Zhang, J., Yang, G. and Wan, G. (1989): A review of the geology of some

kimberlites in China. In: Ross, J. et al., (Editors), Kimberlites and related

rocks, Special Publication of the Geological Society of Australia, 14, 392-

400. Please provide all editors

Zimanowski, B., Büttner, R., Lorenz, V., and Haefele, H.-G. (1997).

Fragmentation of basaltic melts in the course of explosive volcanism.

Journal of Geophysical Research, 102, 803-814.

Zimanowski, B., Froehlich, G. and Lorenz, V. (1991). Quantitative

experiments on phreatomagmatic explosions. Journal of Volcanology and

Geothermal Research, 48, 341-358.

Zimanowski, B., Froehlich, G. and Lorenz, V. (1995). Experiments on steam

explosion by interaction of water with silicate melts. Nuclear Engineering

Design, 155, 335-343.

Editorial handling: J. M. Barton



volcanological and structural aspects of the venetia kimberlite

Documents