RbSr age determinations on South African kimberlite pipes

RbSr age determinations on South African kimberlite pipes

40 RB-SR AGE DETERMINATIONS ON SOUTH AFRICAN KIMBERLITE PIPES By H. L. ALLSOPP and D. R. BARRETT Bernard Price Institute of Geophysical Research, Uni...

838KB Sizes 1 Downloads 24 Views

40

RB-SR AGE DETERMINATIONS ON SOUTH AFRICAN KIMBERLITE PIPES By H. L. ALLSOPP and D. R. BARRETT Bernard Price Institute of Geophysical Research, University of the Witwatersrand, Johannesburg 2001, South Africa

ABSTRACT A study has been made to establIsh to what extent reliable age-measurements can be obtained by the R1r-Sr isochron method on phlogopite micas from kimberlIte and its associated inclusions. Essentially linear isochrons were usually obtained for the micas when grouped according to their source. The data for the Wesselton and Roberts Victor piPes are examined in detail to assess the validity of the method. The available evidence suggests that mica which occurs in kimberlIte groundmass ahd associated peridotite and eclogite inclusions yields valid ages for pipe emplacement by the isochron method. Phlogopite nodule samples appear to yield an age that is anomalously high due to radiogenic Sr retentIOn, and an association with carbonatite is implied. Results for the following pipes are reported: Wesselton, DuToitspan, De Beers, Bultfontein and Monastery (mean age 86± 3 m.y.); Roberts Victor (127 ± 3 m.y., possibly less reliable); Swartruggens (147 ±4 m.y.); Gross Brukkaros ( - 84 m.y., not considered reliable).

INTRODUCTION Many of the numerous kimberlite occurrences in South Africa are known to be younger than the Jurassic Stormberg volcanism, since the kimberlite pipes either intrude Stormberg rocks or contain xenoliths of these rocks; one occurrence of the possibly related melilite basalt near Heidelberg, Cape Province, intrudes Lower Cretaceous strata. K-Ar ages for the Stormberg volcanism in the central area of South Africa lie between 190 and 155 m.y. (FITCH et ai., 1971; McDOUGALL, 1963), but relatively few isotopic age measurements have been made on the kimberlite pipes themselves. LOVERING and RICHARDS (1964) reported K-Ar measurements on mica and pyroxenes from eclogite inclusions in the Roberts Victor pipe but obtained inconsistent results. ALLSOPP et at. (1969) reported Rb--Sr ages on three mica samples: two Roberts Victor eclogite micas indicated on age of 105 ± 20 m.y. and one DuToitspan peridotite mica indicated an age of approximately 80 m.y. In general, isotopic age work on total rock samples is more reliable than work on mineral samples, but this is not true of eclogites and peridotites from kimberlites: MANTON and TATSUMOTO (1971) obtained Rb--Sr and U-Th-Pb apparent ages ranging from 1800 to 3200 m.y.; ALLSOPP et at. (1969) and BARRETT (1974b) obtained ages ranging from 1000 to 3500 m.y. from Rb-Sr tie-linesI joining total rock and mineral data. These high and inconsistent ages cannot represent the time of pipe formation, and the above authors suggested that contamination of the samples was responsible for the high apparent ages. BARRETT (1974b) showed by leaching experiments that most total-rock samples contain a readily soluble (calcitic?) component having a Sr-isotopic composition different from that of the 605

606

H. L. ALLSOPP AND D. R. BARRETT

unleached samples; this is also suggestive of contamination. In contrast, although fresh kimberlites (BARRETT, 1974b) are insufficiently enriched in radiogenic Sr to yield precise ages, such rocks do indicate an age of the right order of magnitude. The present study was undertaken to establish to what extent reliable ages could be obtained by the Rb-Sr isochron method on phlogopite micas from kimberlite and its associated inclusions, and to establish whether significant age differences exist between different pipes. Micaceous samples of kimberlite and associated peridotite and eclogite nodules were selected mainly on the basis offreshness; details are listed briefly in Appendix 1. The mica occurs in kimberlite mostly as individual flakes and discrete booklets occasionally ranging in size up to 10 mm; it may occur as phenocrysts or xenocrysts. The phlogopite nodules, which are clearly xenoliths and up to 200 rom in size, consist of coarse aggregates of mica with irregularly oriented flakes. In peridotite nodules the mica occurs as individual flakes in a roughly equigranular matrix. Mica in the eclogites may present a similar aspect, but generally it occurs as smaller flakes bordering other minerals, mainly gai·net. ANALYTICAL PROCEDURE AND RESULTS Mineral separation was carried out by conventional techniques, and mica concentrates at least 98 % pure were obtained in most cases. Note, however, that alteration (mainly to chlorite) and permeation by calcitic material (in the case of the phlogopite nodules) is discounted in assessing mica content. The micas were dissolved in HF and HCI0 4 , and the Rb and Sr concentrations (considered accurate to within 1 %) determined by conventional isotope dilution techniques. Sr 87/Sr86 ratios were mostly determined on tracerfree aliquots; these data are gene{ally precise to ±0.0003--o.oo08 (10-), but for some samples less precise data (±0.001) were obtained. Eimer and Amend standard Sr yielded Sr 87 /Sr 86 ratios of 0.7081 ± 0.0003 (la). The data were plotted on isochron diagrams and straight lines fitted by the method of YORK (1966), allowance being made for the less accurate data by means of proportional weighting factors. Suites of micas from different pipes and from different host rocks were studied and the data are listed in Appendix 2. Isochron diagrams for the various suites of micas are shown in Figs. 1-7 and the age and primary Sr 87/Sr86 data are summarized in Table 1; the uncertainties listed in this table and elsewhere in the paper are 2a values. In calculating the ages a minority of samples were rejected due to the data points lying more than five standard deviations from the mean line, the data points for all rejected samples being plotted as open symbols. The quantity of mica available was often insufficient for representative splitting TABLE 1. SUMMARY OF AGES AND PRIMARY Sr 87 /Sr 86 RATIOS -----~._--

Source Roberts Victor eclogite nodules Wesselton kimberlite groundmass Wesselton peridotite nodules Wesselton phlogopite nodule DuToitspan, De Beers and Bultfontein peridotite nodules Monastery kimberlite megacrysts Swartruggens kimberlite groundmass Gross Brukkaros groundmass

Age (m.y.)

127

±3

83 ± 4 84 ± 4 114 ± 8 90 ± 3 90 ± 4 147 ± 4 -84

0.7081 0.7073 0.7089 0.7101 0.7071 0.7046 0.7066

± 0.0011

± 0.0020 ± 0.0018 ± 0.0006 ± 0.0014 ± 0.0016 ± 0.0017

RB~SR

AGE DETERMINATIONS

607

and no measurt'ments were duplicated. Thus experimental error is a possible cause of scatter, but certain sample groups, e.g. Roberts Victor, show greater scatter than others and this can only be attributed to geological factors.

DISCUSSION Isotopic inhomogeneity between different coexisting minerals has been reported for presumed mantle-derived rocks. KUDO et al. (1972) and HARRIS et al. (1972) consider that these inhomogeneities existed in spite of inferred high temperatures in the source region; on the other hand, ALLSOPP et al. (1969), MANTON and TATSUMOTO (1971) and BARRETT (1974b) consider that contamination is the operative factor. Since isochron linearity demonstrates original isotopic homogeneity, a most important result is that the various groups of micas do define essentially linear isochrons. From the standpoint of original isotopic homogeneity, the validity of the method is established; note, however, that the xenolithic phlogopite nodules appear to be an exception as further discussed below. The isochron age would be expected to refer to the time of pipe formation since rapid adiabatic cooling of the kimberlite is postulated (KENNEDY and NORDLIE, 1968). Furthermore, the pipe dimensions are relatively small and BARRETT and BERG (1974) show that the kimberlite emplacement temperature, at least at the levels now being worked, was probably < 500°C. From the age determination point of view the textural nature of the mica, and whether it is primary or secondary, is largely irrelevant. All primary mica would be expected to yield the age of pipe emplacement, and secondary mica, if formed at the time of or prior to pipe emplacement, would be expected to yield the same age. However, primary and secondary micas may well differ in their primary Sr 8 7/Sr 86 ratios, and all data from a given suite of micas may not define a single isochron. Other factors that could give rise to scattered points on an isochron are the incorporation of older mica in later kimberlite intrusions, and secondary alteration involving percolating ground water and/or carbonatitic fluids.

1. ASSESSMENT OF METHOD To further assess the validity of the isochron method for kimberlitic micas the results of detailed work on the Roberts Victor pipe (eclogite nodule samples) and on the Wesselton pipe (kimberlite groundmass, peridotite nodule and phlogopite nodule samples) are first considered. Post-emplacement alteration of kimberlite and its inclusions might be expected to affect different samples to differing extents, depending on grain size and other parameters, and to give rise to scattered points on an isochron diagram. BARRETT (1974a) suggests that alteration has significantly influenced the Rb-Sr characteristics of Roberts Victor eclogites. Eleven different mica concentrates, of different degrees of purity, were analysed from a typical, partially altered, eclogite (RVl). The internal isochron so defined (Fig. 1) can be regarded as a tie-line between pure mica and a hypothetical Sr-rich contaminant. The data for micas from two other eclogites and one kimberlite mica (RV30) lie on the same isochron, and it would appear that most of these micas and the contaminant were in isotopic equilibrium at the time of pipe emplacement. The evidence suggests validity for the isochron age

608

H. L. ALLSOPP AND D. R. BARRETT

R~BERTS VICT~R MICR~

0.770

~

z"-

Q)I

0.760 Q)Z

0.750

"r-

~n

.~

to

CD

cr: (f)

0.7l10

II )!(J

CD

cr: (f)

0.730

~I ~7

0.720 AGE = ( 127 :t 3 ) MY INTERCEPT = 0.7081 :to.0011

0.710

5

10

IS

35

30

25

20

lIO

RBB7/SR86 FIG.

1. Isochron diagram for Roberts Victor eclogite nodule micas with one ground mass mica (sample 21). The internal isochron for sample RVI concerns points 12, 14-20.

O. 780 0.170 O. 160 to

CD

a: (J')

0.150

rto

0.740

"-

a:

~

WESS KIMB MICRS

&

i

i-

~

r

~1



(J')

0.730 ~I

0.720

n

4

AGE = ( 83 INTERCEPT

0.710

10

20

30

liD

1: II J MY 0.7073 1:0.0020 50

50

70

RBB7/SR85 FIG.

2. Isochron diagram for Wesselton kimberlite groundmass micas.

609

RB-SR AGE DETERMINATIONS

WESS PERro MICRS &

0.840 0.820

w a::: (f)

0.800

CD

"r-

CD

0.780

3

a::: (f)

0.760 0.740

~5

AGE = ( 84 t 4 ) MY INTERCEPT = 0.7089 to.0018

0.720

60

40

20

RB87/SR86 FIG. 3. Isochron diagram for WesseIton peridotite nodule micas.

MICA

N~DULE

MICAS

0.730

12 /

O. 725 w a::: (J")

O. 720

([)

0.715

CD

"r-

a::: (J")

"

/

/

"

/

'Wessel ton

n,. ~.

/

/-Du Toitspan

O. 710

"

0.705

/

/~15

AGE = ( 114 t 8 ) MY INTERCEPT = 0.7101 to.0006 2.0

4.0

6.0

8.0

10.0

12.0

RB87/SR86 FIG.

4. Isochron diagram for WesseIton (points 1-16) and DuToitspan (points 17-20) phlogopite nodule micas. The internal isochron for the non-calcitIc sample WB concerns points 5-13

610

H. L.

ALLSOPP AND D. R. BARRETT

of 127 ± 3 m.y.; it should be noted, however, that two samples (RV4 and RVL2) yielded data that lie well off the isochron and, as petrographic assessment of these samples shows they are not significantly more altered than RV1, the isochron age is open to some doubt. From the Wesselton data two significantly different ages were obtained: the kimberlite groundmass (Fig. 2) and the peridotite nodules (Fig. 3) yield indistinguishable ages (83 ± 4 m.y. and 84 ± 4 m.y. respectively), whereas the phlogopite nodule age (Fig. 4) is 114 ± 8 m.y.* The differences in the ages obtained from groups of micas from a single pipe presents a problem and possible explanations of this are now discussed. (i) Multiple intrusions A possible explanation of the discrepant ages is that there were different periods of kimberlite intrusion in the Wesselton pipe, with the phlogopite nodules being largely or entirely confined to one particular intrusion and the peridotite nodules and micaceous groundmass samples to another. This possibility cannot be assessed since the original location of most of the samples is unknown, but R. CLEMENT (private communication) considers that the hypothesis is feasible in that phlogopite nodules tend to be associated with particular intrusions. A further collection of samples from Wesselton was made to test the hypothesis. At two localities phlogopite nodules and micaceous peridotites and kimberlites were collected in close proximity to one another. None of the phlogopite nodules could be dated because of their high common Sr content (as further discussed below), but two kimberlite micas from the one locality (WK9/1 and WK9/2) as well as one questionable** kimberlite mica (WK/10) from the other locality yielded data lying on the 83 m.y. isochron. (Three peridotite micas from the same localities had low Sr 87 /Sr 86 ratios and could not be dated reliably.) However, it was shown at both of the localities studied that phlogopite nodules do occur within 83 m.y. old kimberlite, and the evidence is against the hypothesis of detectably different ages of intrusion. (ii) Radiogenic Sr retention Assuming the phlogopite nodules to have crystallized prior to pipe formation, radiogenic Sr may have been partially retained in the samples in spite ofthe inferred high temperature of the source region. This would be feasible if the phlogopite exists at depth in the form oflarge masses such that the distance between the interior of a particular mass and a suitable sink for generated radiogenic Sr is too great for isotopic equilibrium to be established by diffusion. A common feature of the phlogopite nodules is the presence of fine-grained calcitic material; of seven nodules examined visually, all but two contained calcite, and measurements were made on three such samples, including one from DuToitspan. The data for these samples (Fig. 4) are not colinear with that of two non-calcitic nodules from the same pipes, a higher primary Sr 87 /Sr 8n ratio being indicated for all the calcitic samples, except W4(B), a possibly contaminated sample taken from the outside edge of a nodule. Analyses were made on different fractions of calcitic mica (WAl-WA4 in increasing order of purity), but even the purest sample had ~ 4000 ppm Sr and an only slightly higher Sr 8 7I Sr 86 ratio. Sample W4 was leached for 16 hr in cold 2N HC1, and ~ 80':~ of the Sr dissolved; although the leach solution had a lower Sr 87 /Sr 86 ratio (0.7152) than the residual mica * The results for the calcItIc phlogoplte nodules are discussed later. ** The sample IS probably a basic mclusion. the bIOtIte It contams should

formatIOn

nevertheless date the time of pIpe

RB-SR AGE DETERMINATIONS

611

(0.7186), the difference is small and it is clear that the mica itself was permeated by calcitic material. Electron microprobe studies of sample WA (F. R. BOYD, private communication) show a widespread but non-uniform distribution of calcium. The Sr content of the calcitic samples is very high (2000 to 48,000 ppm) compared with 500 ppm for average calcites (BROOKINS, 1967). The high Sr content suggests a carbonatitic affinity, since carbonatites typically have 1000-10,000 ppm Sr (BROOKINS, 1967), but the Sr 8 7/Sr 86 ratios are too high for carbonatites. These contradictory facets can be reconciled if carbonatitic material was present in the phlogopite nodules prior to pipe formation: as noted above, radiogenic Sr diffusion would occur at the inferred temperatures and in this instance only very short diffusion distances are involved. The high Sr 8 7/Sr 86 ratios of the calcitic samples is diagnostic of such diffusion, and provides evidence that the phlogopite existed at depth prior to pipe formation. A relatively shallow depth of origin may, however, be implied by the incomplete isotopic homogenization. Noting that the non-calcitic Wesselton data all pertain to different portions of a single nodule (WB), the linear isochron could represent a tie-line between pure phlogopite and a minor amount of Sr-rich contaminant. The relatively high primary Sr 87/Sr 86 ratio (0.7101) suggests that partial isotopic equilibration has occurred, and that the calculated age is intermediate between that of the phlogopite crystallization and pipe-formation. Since only two samples of the DuToitspan non-calcitic nodule were analysed (see Fig. 4), the indicated age is not reliable; however, it is significant that an anomalously high age is again indicated. In view of the lack of evidence for detectably different ages of intrusion, the hypothesis of radiogenic Sr retention is considered feasible, and it is probable that age measurements on phlogopite nodules are unreliable. The available evidence suggests, however, that the mica which occurs as dispersed flakes or small booklets in kimberlite groundmass and in associated nodules yields valid ages for pipe emplacement by the isochron method. Accordingly the best estimate of the Wesselton age is 84 ± 3 m.y.

2. DATA FROM OTHER PIPES The Premier pipe in the Transvaal is known to be pre-Cambrian (ALLSOPP et al., 1961), but as far as is known other pipes are Jurassic to Cretaceous in age. To ascertain whether age differences occur, a number of other pipes were investigated': (i) De Beers, Bultfontein and DuToitspan

Micas from peridotite nodules from these spatially closely associated pipes yield an essentially linear isochron (Fig. 5). However. the calculated age of90 ± 3 m.y. is dominated by data pertaining to a single De Beers sample and no definite conclusion is possible as to whether the three pipes are of the same age. The Wesselton pipe is also in close proximity to these three pipes and the age of 84 ± 3 m.y. for Wesselton overlaps at the 95 ?:, confidence level with that of the other three pipes. It is concluded that these four closely associated pipes are of essentially the same age; the mean age is 86 ± 3 m.y., and the mean primary Sr 87 /Sr 8h ratio is 0.7080 ± 0.0012. (ii) Monastery This pipe is situated approximately 250 km east of Kimberley, and contains particularly fresh megacrysts of mica in the kimberlite. Five samples were studied, different mineral

612

H. L. ALLSOPP AND D. R. BARRETT

PERIOClTITE MICRS

0.800

1"

0.790 0.780 0.770 lD

CD

cr

,r-

.7

O. 760

(f)

CD

cr (f)

0.750 O. 7110 0.730 0.720 0.710

~

-.

AGE = ( 90 t 3 I MY INTERCEPT = 0.7071 to.OOlll

10

20

30

50

40

70

60

80

RB871SR86 FIG. 5. Isochron diagram for De Beers, Bultfontein and DuTOItspan peridotite nodule micas.

MelNRSTERY MICRS

0.780 0.770 0.760 co

0.750

co

0.7YO

CD

a: en "r-

a:

9

en

O. 730 0.720 0.710

5

AGE

~

(90

INTERCEPT

:t

~

Y

) MY

O. 70~5 to.0016

-.1__ --L_ _~ _ ~_ _- - ' - - _ - - - - ' - _ - - ' 10

20 30 RB87/SR86

YO

50

50

FIG. 6. Isochron dIagram for Monastery kimberlite megacryst micas.

613

RB-SR AGE DETERMINATIONS

fractions being used in three of them, and the isochron from the total of ten points (Fig. 6) yields an apparently reliable age of 90 ± 4 m.y. As this overlaps with the Kimberley-area age it appears that widely separated pipes may be associated in time. (iii) Swartruggens This pipe is situated approximately 400 km north-east of Kimberley. Five samples were studied and, although the isochron (Fig. 7) is largely dominated by a single point, good overall linearity was obtained. The calculated age of 147 ± 4 m.y. is considered reliable; a significant age difference for this pipe is therefore established.

SWARTRUGGENS MICAS 1.000 0.950

to

CD

0.900

~

i

......

0.850 0.800 0.750

AGE • ( lij7

t

ij •

I'll

INTERCEPT. 0.7066 to.0017

20

m

1

60

80

100

120

lijO

R887/SA86 FIG.

7. Isochron diagram for Swartruggens kimberlite groundmass micas.

(iv) Gross Brukkaros This pipe, situated in South West Africa, was not studied in detail. A calculation using the data points for two samples (a mica and a calcite) indicates an age of ~ 84 m.y. Since the mica and the calcite may not be cogenetic, this procedure is of doubtful validity, and the age is not considered reliable. 3. SIGNIFICANCE OF THE PRIMARY Sr 87 /Sr 86 RATIOS Primary Sr 87/Sr86 ratios for the various micas are listed in Table 1, and, in principle, these ratios should indicate whether the micas are primary or secondary. This aspect is discussed by BARRETT (l974a); in brief, isotope ratios compatible with primary mica occur for all peridotites and for the Monastery megacryst micas. The isotope ratios for other groups are either intermediate or indicative of secondary mica.

614

H. L. ALLSOPP AND D. R. BARRETT

APPENDIX 1. SAMPLE DESCRIPTIONS From each of the samples listed below the freshest flakes of mica were selected by hand-pickmg. DuToitspan

(KImberley area): DPA, pendotlte nodule with fresh mIca; 198, phlogoplte nodule, mainly fresh; 199, phlogopite nodule permeated with calcItic matenal throughout. De Beers (KImberley area): JJG 299, DBX, DB7, 4AB 116/, peridotIte nodules containing mainly fresh mica petrographically assessed as primary. Bultjontem (KImberley area): Butt Oi, pendotlte nodule, mamly fresh and pnmary mica Wesselton (KImberley area): W7 and WP (/-7), pendotite nodules, mamly fresh; WK (/-10) groundmass mIca whIch appears fresh although the kImberlite groundmass is much altered; WA, phlogopite nodule permeated with calcitic material throughout; WB, phlogoplte nodule, mainly fresh; W4, phlogopite nodule with mmor amount of VIsible calcIte, collected at same site as WK9. Roberts Victor (Orange Free State, ~ 80 km east of KImberley): RVL3, RVL20, RV 4, RV 1, RV L2, eclogite WIth much kelyphitic material; RVL20 and RV4 contam very fresh mica. others somewhat altered; RV30, groundmass mica, descnptlOn as for Wesselton. Monastery (Orange Free State, - 240 km east of Kimberley): B,SS PA, B,SS PB, BDI381, BD138~, BD14lO, megacryst micas. All very fresh (and probably primary). Swartruggens (TransvaaL -400 km north-east of Kimberley)' Swartruggens (/ -6), groundmass mIca, description as for Wesselton. Gross Brukkaros (South West Africa, 600 km north-west of KImberley): groundmass mica and calcite from a single hand sample, J F51; fresh.

APPENDIX 2. Sample

J- -

Rb, Sr AND Sr 87 /Sr 86 DATA FOR ALL SAMPLES Rb (ppm)

--- - - - -

-

Gs~sr86

(Sr 87/S r 86)p

(ppm) -

-

----

---

A. Roberts Victor mIcas (Fig. 1) I. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21.

RVL2 RVL3

(1)* (2)* (I)

(2) (3) RVL20 RV4 RVI

(I)

(1)* (2)* (3)* (1)* (2)* (3) (4)* (5) (6) (7)

(8) (9)

(10) (11) RV30

(I)

688.1 581.6 203.3 196.6 181.5 203.1 285.5 274.6 435.7 395.5 591.7 582.5 584.4 564.7 559.1 590.7 350.5 577.2 442.3 405.8 601.6

48.25 58.15 2986 303.2 445.3 280.2 36.89 39.22 38.64 77.92 67.38 46.82 47.33 51.73 48.61 49.14 246.9 100.2 137.6 226.3 54.86

41.42 29.02 1.967 1.873 1.178 2.095 22.40 20.28 32.65 14.72 25.46 36.16 35.88 31.71 33.41 34.93 4.104 16.69 9300 5.185 31.86

0.7628 0.7534 0.7102 0.7110 0.7105 0.7120 0.7266 0.7351 0.7307 0.7452** 0.7456 0.7689** 0.7660** 0.7648 0.7665 0.7676 0.7152 0.7387 0.7244 0.7183 0.7661

Notes: The numbers immediately preceding the sample names are those used to identify the pomts on the various isochron dIagrams. Rb and Sr are total abundances; Rb 87 /Sr 86 is the atomIc ratio; (Sr 87 /Sr 86 )p IS the present-day ratio normalized to Sr 88 /Sr 86 = 8.375. Samples indicated * aJe omItted from the Isochron calculatIOns as they have uncertainties in (Sr 87 /Sr 86 )p greater than 50' from the calculated isochron. Samples mdlcated ** are precise to ± 0.00 I.

615

RB-SR AGE DETERMINAnONS

-

Sample

-]

(p~~~~7jSr86

(Sr 87 jSr 86 )p -----

B. Wesselton kimberlite groundmass micas (Fig. 2) 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. II.

WK1 WK2 WK3 WK4 WK5 WK6 WK7 WK8 WK9

* (1) (2)

WKIO

586.8 661.0 485.2 489.4 473.8 952.8 709.9 201.7 732.8 613.4 570.0

39.80 67.61 114.4 281.8 174.8 42.94 48.57 101.6 41.75 59.48 116.1

42.75 28.33 12.27 5.022 7.835 64.57 42.42 5.747 51.00 29.90 14.20

0.7462 0.7380** 0.7196 0.7158 0.7158 0.7822 0.7545 0.7240 0.7662 0.7430 0.7217**

C. Wesselton peridotite nodule micas (Fig. 3) I. 2. 3. 4. 5. 6. 7. 8. 9.

(I) (2)

W7 WPI WP2 WP3 WP4 WP5 WP6 WP7

226.5 143.2 715.0 857.0 1066 798.1 247.3 231.7 242.3

238.4 363.3 38.75 35.88 107.7 20.28 207.3 168.9 266.9

2.746 1.139 53.66 96.74 28.67 115.2 3.448 3.967 2.614

0.7127 0.7116 0.7750** 0.8226 0.7375 0.8423 0.7128 0.7130 0.7070**

D. Phlogopite nodule micas (Fig. 4) I. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

(I) * (2) * (3) * (4) * (la) (I b) (3a) (3b) (4a) (4b) (8a) (8b) (9) (A) * (B) * (C) * (I) * (2) * 0) * (2) *

WA

WB

W4 198 199

108.4 464.8 463.1 487.9 414.9 412.3 448.6 368.9 420.5 423.8 488.8 474.9 429.8 461.0 450.1 450.9 424.5 462.0 422.3 473.5

47610 12310 13160 4059 413.2 600.0 267.7 303.7 505.7 291.3 235.3 117.5 196.7 4288 1995 2493 387.5 224.0 9356 9370

0.0066 0.1091 0.1017 0.3474 2.903 1.987 4.487 3.512 2.404 4.207 6.006 11.70 6.321 0.311 0.652 0.523 3.167 5.966 0.131 0.146

0.7119 0.7127 0.7127 0.7127 0.7145 0.7139 0.7173 0.7156 0.7135 0.7176 0.7195 0.7285** 0.7202 0.7148 0.7064 0.7170 0.7135 0.7196** 0.7138 0.7140

E. Dt' Beers, Bulifontein and DuToitspan peridotite micas (Fig. 5) 1 DPA

(I)

2. 3. Bult Di 4. JJG 299 5. DBX

(2)

6. DB7 7. AAB 1161 8.

0) (2)

9

(3)

10.

(4)

363.3 350.1 247.9 2309 265.8 239.7 551.1 554.9 340.6 313.0

232.3 234.5 239.6 206.4 332.5 177.9 36.53 161.7 12.76 12.91

4.521 4.317 2.990 3.233 2.311 3.894 4379 9.921 77.84 70.65

0.7122** 0.7128** 0.7106 0.7120 0.7109 0.7110 0.7565 0.7174 0.8025 0.7988

=r

616

H. L. ALLSOPP AND D. R. BARRETT

Sample

Rb (ppm)

Sr (ppm)

_ _ _--L

_

~

F. Monastery kimberlite megacryst micas (Fig. 6)

I. Biss PA

(1)

2.

(2)

3. 4. Biss PB

5. 6. 7. BD 1381

8. 9. BD 1382

(3)

(1) (2) (3) (1) (2)

*

to. BD 1410

666.8 732.0 748.6 880.5 624.2 800.8 734.0 715.6 677.4 654.9

276.1 250.1 153.4 180.3 402.7 242.3 36.05 36.10 53.41 38.75

6.982 8.462 14.12 14.12 4.480 9.556 59.22 57.66 36.80 49.08

0.7134 0.7143 0.7220 0.7200 0.7121 0.7156 0.7780 0.7770 0.7519 0.7595**

G. Swartruggens kimberlite gr9undmass micas (Fig. 7) 1. Swartruggens (1) 2. 3. 4. 5. 6.

(2) (3) (4) (5)* (6)

537.2 373.1 565.2 484.4 303.2 512.6

12.88 199.2 148.9 135.7 62.00 82.89

123.5 5.416 11.00 10.33 14.18 17.93

j

0.9594 0.7170 0.7305 0.7307 0.7468 0.7432

H. Gross Brukkaros groundmass mica and calcite Mica Carbonate

JF51 JF51

524.1 ~ 175.4 5.37 4377

---------'------

8.639 0.004

0.7124 0.7041

--_.

ACKNOWLEDGEMENTS We are grateful to Professor L. O. Nicolaysen and Dr. J. Kramers for helpful discussion, and to the following persons who supplied samples used in this work: R. Clement, J. B. Dawson, A. Bisschoff, A. van Zyl and 1. Ferguson. We are also indebted to Anglo American Corporation for their co-operation and for financial support.

REFERENCES ALLSOPP, H. L., BURGER, A. J. and VAN ZYL, C (1967) A minimum age for the Premier kimberlite pipe yielded by biotite Rb-Sr measurements, with related galena Isotopic data. Earth Planet. SCI. Lett. 3, 161--6. ALLSOPP, H. 1., NICOLAYSEN, L. O. and HAHN-WEINHEIMER, P. (1969) Rb/K ratios and Sr-isotopic compositions of minerals in eclogltIc and peridotitic rocks. Earth Planet. Sci. Lett. 5, 231-44. BARRETT, D. R. (I 974a) The genesis of kimberlite and associated rocks: Sr-Isotope eVidence. Phys. Chem. Earth 9, this volume, pp. 637-53. BARRETT, D. R. (1974b) Ph.D. thesis in preparation, University of the Witwatersrand. BARRETT, D. R. and BERG, G. W. (1974) Complementary petrographic and Sr-Isotopic ratIo studies of South African kimberlites. Phys. Chern. Earth 9, thiS volume, pp 619-505. BROOKINS, D. G. (1967) The strontIUm geochemistry of carbonates m kimberlite and limestones from Riley County, Kansas. Earth Planet. Sci. Lett 2, 235 -40. FITCH, F. J. and MILLER, J. A. (1971) Potassium--argon radioages of Karroo volcamc rocks from Lesotho. Bull. Volcanologique, XXXV, 64-84. HARRIS, P. G., HUTCHINSON, R. and PAUL, D. K. (1972) Plutonic xenoliths and their relation to the upper mantle. Phil. Trans. R. Soc. Lond. A271, 313-23. KENNEDY, G. C. and NORDLlE, B. E. (1968) The genesIs of diamond depOSits, Economic Geology 63,495-503.

RB-SR AGE DETERMINAnONS

617

KUDO. A. M., BROOKINS, D. G. and LAUGHLIN, A. W. (1972) Sr-isotopic dIsequilibrium m Iherzolites from the Puerco Necks, New Mexico. Earth Planet. Sci. Lett. 15,291-5. LOVERING, J. F and RICHARDS, J. R. (1964) Potassium-argon age study of possible lower-crust and upper mantle mclusions m deep-seated intrusions. J. Geophys. Res. 69, 4895-901. McDoUGALL, 1. (1963) Potassium-argon age measurements on dolerites from Anatarctica and South Africa. J. Geophys Res. 68, 1535-45. MANTON, W. 1. and TATSUMOTO, M. (1971) Some Pb and Sr isotopiC measurements on eclogites from the Roberts Victor Mine, South Afnca. Earth Planet. Sci. Lett. 10, 217 -26. YORK, D. (1966) Least-squares fitting ofa straight Ime. Can. J. Phys. 44,1079-86.