Isotopic age determinations on the Cape Granite and intruded Malmesbury sediments, Cape Peninsula, South Africa

Isotopic age determinations on the Cape Granite and intruded Malmesbury sediments, Cape Peninsula, South Africa

The spread of these results is sufficient to suggest that the mineral-ages have been influenced by the loss through migration of radiogenic daughter-...

1MB Sizes 0 Downloads 43 Views

The spread of these results is sufficient to suggest that the mineral-ages have been influenced by the loss through migration of radiogenic daughter-products. The abundance and wide distributiorl of sedimentary xenoliths in the granite showing all stages of assimilation suggests that at least a substantial part of the granite was derived from the invaded Nalmesbury sediments. ‘Detailed geoehemieal


CAPE Table










Fig. 1. The strati~rap~i~


in t~he south-w&cm




studies, involving both major and trace elements, on the Malmesbury sediments (ERrJAN& 1965) and on the Cape Granites and their constituent minerals (KOLBE, The geochemical data which show varying 1965) are at present nearing completion. stages of differentiation of the granites, do not preclude the possibility that the granites were ultimately derived by mobilization of the Malmesbury sediments at depth. The principal purpose of the work reported here was to determine t,he age of emplacement of the Cape Granite using the Rb-Sr method. In addition it was hoped that the isotopic Sr data would t’hrow light on the genesis of the granite: if the Malmesbury sediments were much older than the granites, their later remobilization would have resulted in the formation of a granite with an unusually high primary EW7/Sr8” ratio. In fact the observed ratio is normal at ~0.71, and subsequently an effort was made to verify, at least approximately, the assumed great age of the ~falmesb~~ry by nleas~lreme~lts made on tot,al-rock samples.

The geology of the Cape Peninsula and neighbouring areas is shown in the simplified map (Fig. 2) indicating sample locations. Details of sample localitics, brief summaries of the petrography and descriptions of the mineral fractions are given in the appendix. The Malmesbnry formation, viewed as a whole, consists dominantly of blue arenaceous clay-slates or fine-grained argillaceous quartzites that alternate with groups of ph~~llite, quartzite and felds~athic grit, with occasional conglomerate, limestone and chert. Calcareous rocks are rare in the south-western area but some


age determinat~ions


on the Cape Granit-c:

int,erbedded bands of crystalline limestone occur in the north and east. SCHULTZ (1946) considered t,hat the sediment,s, which attain a total thickness in the Malmesbury dist.rict probably in excess of 20,000 ft, were derived from an eIevatcd contincnt,al area composed mainly of pre-Cambrian granit,ic rocks lying to t,he north








L-l t s-3

Fig. 2. A siltl~~li~~(~ geological





1 mile* i

map of %Lpart of the Cape Pcninstrla, loealit,ies.



and north-east of the developing Malmesbury geosyncline. The sediments are generally strongly folded dipping at high angles and have undergone at least three periods of metamorphism. An early, low-grade, dynamic regional metamorphism up to, but not exceeding the biotite stage, was followed by the far more localized contact metZamorphism due to the invading granite plutons. Further low-grade regional metamorphic effects are associated with the later Cape orogeny. The sediments in the Cape Peninsula area here considered consists of fine-grained arenaceous strata with interbedded thinner slaty layers, both strata being hornfelsed in the vicinity of the granite. The more abundant arenaceous bands (represented by samples M9-B, NI-15 and M-41) consist mainly of fresh elastic grains of quartz, sodic plagioclase and subordinate microcline in a ground-m.ass composed mainly of somewhat chloritized biotite and sericite. Gradations are common to the interbedded argilla~eo~ls bands (represented by samples M-14,4 and M-40). These are t,oo fine-grained for det:ailed optical mineral deternlillation but show fewer and much smaller elastic grains and also contain clay minerals in their matrix. The geology of the intrusive Cape Granite plutons in the south-western Cape has been described in detail by SCHOLTZ (1946). He concluded that they were probablg related genetically to the granites in the eastern (George) and northern (Kuboos) areas. Petrologically the south-western plutons are very similar to each.



other but elsewhere, especially in the Saldanha Bay area, a younger intrusive Numerous xenoliths in all stages of assimilation are quartz porphyry is found. found in the granites giving clear evidence of contamination. The generally coarsegrained porphyritic granit#e plutons almost invariably grade upwards into a medium-to fine-grained hoodfacies clearly representing (on geochemical grounds) a far more highly differentiated granite t’ype. In the Cape Peninsula area considered here, the normal very coarsely porphyritic granite (samples 110, 108 and 104) forms the overwhelming bulk of the exposed pluton and t’he finer-grained facies probably constitutes 1 per cent or less of the area exposed. Samples 136, 145 and 147 represent t,his fine-grained facies as exposed near the Granite-Malmesbury contact a.t Kloof Quarry, wyllies Siding (WALKER and MATHIAS, 1946, p. ,501). Here the normal coarse Cape Granite passes gradationally into a finer-grained porphyrit)ic granite which itself grades into a micro-granite and terminates in a Ohin pegmatitic vein Ohat intrudes t,he Malmesbury country rock. The three last-named sa,mples are thus clear1 y of t,he same age but could be significantly younger than t,he normal coarse granite represented by samples 110, 108 and 104, since their highly differelttiated nat,ure relative to the normal granite is well demonstrated, for instance, by lower K/Rb, K/G and K/T1 ratios and very low Ba and Sr content (KOLBE, 1965). -4nother sample of particular interest is located near the world-famous Sea Point contact described by WALKER and MATHIAS (1946). A sketch map of the area is shown in Fig. 3. Thin veins of aplite are found cutt,ing the granit,e and Malmesbury sediments near the contact. These veins are petrologically very similar to a large cross-cutting mass of aplite intruded into the Malmesbury hornfels represented by sample 156. Further evidence of post-granitic or late-stage granit,e nchivity is afforded by veins and nests of pegmat,ite found in the normal coarse granite. RESULTS AND DISCUSSION A description of the chemical and mass-spectrometric techniques used for the Rb-Sr work is given by COMPSTON etal.(1965). A mixed SP-Sr*4 spike was used, and the Srs7/Srs6 ratios of all the samples except the plagioclases were calculated from spiked runs. All the Sr isotope ratios listed in Tables 1, 2, 4 and 5 were normalized to conform with the SP*/Sr86 ratio of 8.340, as discussed by COMPSTO-U et ~11. Using HF and HClO, it was not possible to dissolve completely some of the total-rock samples: with the granites significant amounts of tourmaline, together wit,h traces of zircon and apatite, remained in certain cases, and with the argillaceous sediments significant amounts of carbon remained. The carbon was most#ly oxidised by strong HCIO,, but in all cases residues were removed by centrifuging. The Sr measurements were made by the static method using a Reynolds-type massspectrometer, as described by MCDOUGALL (1964). The potassium concentrations were measured flame-photometrically. The following physical constants were used: Rb--Sr method:

J./?= 1.39 x IO-l1 yr-I,


;lp = 4.72 x lo-lo yr-I, Ilr = 0.584 x lo-loyr-I, K40 = 1.19 :,: 1O-4 mole/mole Ii.



= 2.600, Sr88/Sr86 = 8.340

Isotopic age determinations


on t,he Cape Granitst-


1120 Table



1. Cape granit,es;

Rb (ppln)

Srx (ppm)


93.2 93.3

108 A 13 110 104 A B 136 A I3 147 A H 145

261.6 271 295.1 296.9 307.5 304.1 294.7 293.6 290

and P. KOLBE Xb-Sr

total rock data

0~7701 0.7719 0.7781 0.7859 0~7830 0,8418 0.8453 1.1181 1.1171 I.1602

ti.077 8.059 9.119 9.516 9.431 17.296 17.166 52.66 52.3% 59.07

85.6 89.2 90.5 51.1 60.9 16.1 16.1 14.1

Sormalizing* fact)or ( ‘jO)


O~?O 0.40 0.25 0~50 0.34 0.37 0.39 0.2-I0.33 0.16

* This column lists the factors by which the measured Srss/Srs4 ratios escced thr \-altos computed (using the known SS/S4 ratio of the spike); these data arc used to normalize the ot,hcr factor is tlctlrrct~l Sr isotope ratios. With unspiked runs (Tables 2, 4 and 5) the normalizing from the measured Sr8*/Srs6 values.


Sample 104 Biotit.e Biot,ite K-feldspar K-feldspar Plagioclase Plagioclase 110 Biotite K-feldspar 108 Biotite K-feldspar

2. Cape granites;

Rb (ppm) A* 13 A B A Bt

1133 1157 444.1 440.7 32.8 32.7 1115 415 1048 393

SrAV(ppm) 2.68 2.88 149.1 149.0 68.1 6.7 122.8 3.28 140.2




Srai/Srs6 9,507 9.278 0.7745 0.7754 0.7263 0.7244 4.301 0.7840 7.679 0.771%

‘li Generally, samples numbered A and B means that dllplicate In this par&ular case, however, A and B refer to slightly different 1’ Sr8’/Srs6 measured from an unspikcd Sr run.

Apparent age (m.y.) 518 932 540 551

539 546 ,544 545

Sormal izing fact,or (O<,)! 0.17 0.28 0.21 0.76 0.08 -0.06 --0.06

measurements wtrc made. fractions (see appcndis).

Rb and Sr measurements normally agree to better than 1 per cent, and Ar measurements to better than 2 per cent. Taking into account all factors the uncertainty in the value used for the decay constant of R8bs7 apparent ages are thought to be accurate to within 2 per cent, and K-Ar ages to 3 per cent.


K and except Rb-~Sr within

1. The Cape


Table 1 summarises the Rb-Sr total-rock measurements on the Cape Granite. The usual isochron approach may be applied to a group of data only if the samples are known to be a chemically heterogenous group of rocks all of the same age. The

Isotopic age determinations on the Cape Granit,e

plotted against Rbs7/Srs6for all Fig. 4. The Cape Granite isochron: SrS7/Srs6 of the t,ot,al-rock samples and some of the constituent minerals.

three samples of the fine-grained, differentiated facies satisfy this condition although they may be slightly younger than the normal granite samples. However, if a regression line is fitted to all the points, as shown in Fig. 4, it is noted that there is no systematic divergence from the line and it is concluded that there is unlikely to be a significant difference in age between the two facies. Further support for the procedure of grouping all points is derived from the mineral-fraction results discussed below. Duplicate analyses were made on four of the six total-rock samples, and in the least-squares analysis of the data the duplicates were regarded as separate datapoints. From the slope and intercept of the regression line obtained it is calculated that the age of t’he granite is 553 * 8 m.y., and the primary Sr8’/Srs6 ratio (Xi) is 0.710 f 0.003, where the errors quoted are at the 95 per cent confidence limit. Although the Rb/Sr ratios of the three samples of the normal granite do not differ sufficiently from one another for the age and Ri to be determined for these points alone, this information can be deduced from the mineral-fraction results if it’ is assumed that, since crystallization each mineral has remained a closed system with respect to Rb and Sr. The relatively coarse-grained minerals of the normal granite are the most likely to have fulfilled this condition and measurements were thus made on K-feldspar and biotite fractions from each of 108, 110 and 104, as well as on plagioclase 104. These data, summarised in Table 2, are also mostly shown in Fig. 4.


H. L.





Taking Ri as 0.710 (the total-rock value) the apparent ages of the K-feldspars and biotites (except 104, biotite A) are found not to differ significantly from the t,otal-rock age. 104 biotite A, which contains slightly more chlorite than sample B, gives a significantly lower age and this indicates that the chloritized sample has leaked radiogenic Sr. Further evidence of Sr migration is afforded by t,he trend of all other mineral points. Figure 4 shows that none of these deviates significantly, but it is noteworthy that the ages of the Rb-rich minerals? biotite and K-feldspar, are lower whereas the single plagioclase age is higher than the total rock age. It, is concluded that a mild metamorphism has influenced the mineral-ages causing the apparent ages of the K-feldspars and biotit(es to be somewhat lower than the true ages. The observation, therefore, that minerals from the normal coarse granite give ages apparently concordant with the total-rock age supports, but cannot be considered conclusive evidence for, the supposition that all the granit)es, both coarse- and fine-grained, are of the same age. K-Ar measurements were made on the two biotites 104B and 108 and the results are shown in Table 3. The results on the two samples are in agreement and Table


LSamplo 104 Biotite J-08 Biotito

3. Cape granites;




data O. Ail correction

Apparent age (my.)









4x7 496







also agree reasonably with the value given by ALDRICH rt al. (195,8), but are considerably lower that the Rb-Sr ages on the identical samples. The fact that K-Ar ages on the two biotites are identical, indicates that metamorphism was widespread and not confined to proximity with intrusions of aplogranite or pegmatite. The Pb207-Pb206 age of 530 t 40 m.y. for the zircon given by ALDRICH et al. (1958) is consistent with the Rb-Sr total-rock age reported here, and the discordant pattern of the uranium-lead and thorium-lead ages is also consistent with the supposition of a later metamorphic event. On the other hand the value of 600 m.y. given by ALDRICH et al. for the Rb-Sr age of the biotite is slightly greater than the ages given here, and the reason for this is not understood. 2. The aplogranite One total-rock sample of the aplogranit,e (156) was analysed by the Rb-Sr method and measurements were also made on the K-feldspar and plagioclase which, t,ogether with quartz, virtually constitute the whole rock. These data are shown in Table 4 and also in Fig. 4. With only one total-rock point available there is no way of demonstrating that this rock has remained a closed system with regard to Rb and Sr. Therefore the results in this section should be regarded with some caution, even though the rock appears quite fresh and unaltered. The total-rock point differs significantly from the other granite points, and this result could be interpreted in several ways, the most probable being:


age determinations


on the Cape Granik

(1) The aplogranite may have been derived from a different source from the other granites, in which case its Ri ratio may also have differed. In particular, if R, = 0.700 for t’he aplogranite the age would be 550 m.y., the same as Ohat of the main granite body, or (2) The aplogranite may have been derived from the same source and had the same Iii ratio. In this case the aplogranite is significantly younger at 500 4: 15 m.y. The second possibility seems the more likely, firstly because a value of Xi as low as 0.700 has not been reported previously for a Cambrian granite, and secondly Table 4. Aplogranit,e; lib S;tlnpl(~ 1.56 Total rock Kfi~ltispar I’lagioclasc


227.9 227.8 702 IO.9


Hr8i/[email protected]

(PPm) X 13


32.7 33.2 66.7 13.1

0.8488 0.8499 0.9208 0.737

Apparent Sormalizing aqx (1n.y.) * factor i”,,) 304 503 ,509

0.19 0.18 0.23 -~0.30

* Assmning A2, := 0.7 10. t srs’/Srs6 measnrcd from an unspikrtl Sr run.

the aplogranite mass must be younger than the main granite if it is considered t,o be a counterpart of the thinner, petrologically very similar, aplite veins which cut the main granite. In passing, an age of 500 & 15 m.y. for the aplogranite would be consistent with an association between the occurrence of daughterproduct diffusion in the main granite, and the intrusion of the aplogranite. The evidence of the aplogranite minerals also favours interpretation (2). These data points, together with that of the total-rock, define an “isochron” with an Ri ratio o’f ~0.716, i.e. in excess of the corresponding ratio of the main granite. It might be argued that, in this fine-grained rock, sufficient diffusion of radiogenic Sr from t#hemicrocline to the plagioclase could have occurred to produce such a pattern irrespective of the actual initial ratio. However if use is made of the following data : (a) the K-feldspar and plagioclase each comprise 30 per cent of the rock, (b) the plagioclase fraction analysed is 50 per cent pure, (the remainder being quartz) and the K-feldspar fraction is 100 per cent pure, (c) the deviations of the K-feldspar and plagioclase points from the maingranite isochron, an order-of-magnitude calculation shows that the amount of radiogenic Sr apparently gained by the plagioclase exceeds by a factor of 4 that apparently lost by the K-feldspar. Since, apart from the K-feldspar, there is no other significant source of radiogenic Sr, this suggests t,hat the observed isochron does not result from mineral-to-mineral diffusion; certainly, diffusion on the much larger scale required to support the hypothesis of an Ri ratio of 0,700 is implausible.


,?. Mabmesbury sediments &4lthough t’he Ri ratio of the Cape granite is not anomalously conclusions can be drawn as to the genesis of the granite without

high, no firm knowledge of

H. L.


and I’. KOLBE


the age of the Malmesbury sediments, and direct dating of these sediments was attempted. Strictly, the total-rock approach to geochronology is valid only for a suite of samples originally homogeneous in isotopic composition. When applied to elastic sediments linearity of t*he data-points would not be expected since such sediments normally contain significant amounts of inherited radiogenic Sr; in c~onseyuence there would be variations in Ri from sample to sample. However work by C~MPSTON and PIDGEON (1962) and by WHITXEP and HURLEY (1964) showed that in some cases meaningful results can be obtained from shales. In



Sr87/Sr86 plotted against Rbs7/Srs6 for all isocbron: total-rock samples of the sedimonts.

5. The Malmesbury

of the four

cases reported

by these authors

(two on unmetamorphosed


and one on a meta-sediment)

the data-points were found to be surprisingly colinear, although there was in each case some indication of regional variations in Ri. In the fourth case, on another unmetamorphosed shale, gross Ri variance was found. Measurements were made on five total-rock samples of Malmesbury sediments and these data are summarised in Table 5. From the isochron presentation of Fig. 5 it, is apparent that all the data-points are closely colinear, and in contrast Table 5. Malmpsbury; Rb @pm) I.D.* M-15




M-9H &I- 14.1 M-40

140 195 199

Al-L1 f

Rh-Sr data




90 102 148 194 198

218 117 130 93 64 2010


* X-ray fluorescence data reported by ERLANK t Only the isotope dilution data aro used. t Brs7/Sr*6 measured from an unspiked run.


218 116 130 89 63

1.069 2,390 3,097 6.014 9.038 cr.002 (1965).


Sormalizinfi factor (I,,)



0.7306 0.7387 0.7602 0.7873 0.702x

--0.36 -~0,09 0.2-L 0.28 --0.14

lsot,opic age dctorminations


on the C’apc (:ranitc

with the previous work on sediments no _F1,variance is suggest.ed. Possible reasons for the good linearity in the present case are discussed below, but accepting for the moment the ~r~~~ facie interpret~atiol~ that the data ~define a real isochron, a least-squares analysis gives an age of 595 5 45 m.y. and an R, rat,io of 0.713 I& 0.003, where the errors quot’ed are the 95 per cent confidence limits. The calculated


















Fig. 6. A comparison of the measured K, ratios wit,h the Sr8’/Srs6 ratios of t’he I’iketberg lim&one and of 600-m-y.-old sea-wat,er. All the Sr isot’ope ratios are normalized to conform wit,h a W8/Srss ratio of 8340.

is acceptable on geologic grounds and accords with the view of SCHOLTZ(1946) that the Cape granites were intruded only short81y aRer the depositi of t,he Malmesbury formation. It is unlikely that the observed linearity is fortuitous, and more than one mechanism may theoretically have caused the linear trend. Much of the discussion of these mechanisms hinges on a comparison (see Fig. 6) of R, with the Srs’/SrS” rat’io of 600-m.y.-old sea-water. The sea-wat,er ratio is estimated as 0.705 f 0.003 from the limestone measurements reported by HEDGE and W_~LTHALL (1963), recalculated t,o conform with the Sr**,&+ ratio of 8.340. From the Malmesbury isochron an Ri value of 0.712 t 0.003 is obt~ained, a value that is certainly higher t,han that of the sea-water. Three possible meehanisms will be considered. age

1. The sediment,s may have all been derived, without loss or gain of RF and Srt from a single source terrain which was quite young during the formation of t$he Melmesbury geosyncline. Apart from the fact that no source region of the required age is known it appears improbable that fine-grained argillaceous sediments could


be formed medium.




some chemical



place with the surrounding

2. The minerals containing the bulk of the Rb and Sr may be predominantly authigenic, though the widespread occurrence of clast’ic grains, including some feldspars and micas, makes this unlikely. Furthermore, according Co this hypothesis the Ri ratio of authigenic sediments should be the same as the Sr87/Sr86 ratio of contemporaneous sea-wat,er (unless the sediments were deposit’ed in isolation from the ocean), yet this is not the case. 3. Arenaceous and argillaceous beds (with lower and higher Rb/Sr ratios respectively) generally alternate throughout the Malmesbury. Homogenization of the Sr isotope ratios over distances small compared with the total thickness of the sediments would in these circumstances eventually result in the Sr isotopic rat>ios everywhere approximating to the average for the whole formation. An I?< value that is higher than the sea-water value would then result from the incorporation of the inherited radiogenic Sr of the elastic grains. The age obtained from the isochron would in this case be t,he time since homogenization, and the sediments themselves could be older. However, if the sediments were much older, Ri would be anomolously high: for example, assuming the 43000-m.y.-age indicated by correlation with the Primitive system and an average Rb8’/WG ratio of 5, homogenization that, occurred ~600 m.p. ago would have given rise t,o an extraordinary Ri ratio of ~0.85. The homogenization itself could be the result of either diagenesis or met,amorphism. Three distinct periods of metamorphism are known. Prior to the intrusion of the granites, and also during the Cape orogeny, the Malmesbury sediments were subjected to low-grade dynamic metamorphism. The thermal metamorphism due to the intrusion of the granites is confined t,o the contact area. Only sample M9-B, at about 2 mile, is located near to the contact, and this sample shows no significant deviation from the least-squares line. It would be surprising if a,ny of these metamorphic events effected widespread homogenization, and a more probable explanation lies in the processes of diagenesis, particularly the circulation of connate waters during the compaction of the sediments. Measurements were made on a sample of Malmesbury limestone, for which the Rb/Sr ratio is virtually zero. Complete isotopic homogenization would be demonstrated if the Srs7/Sra6 ratio of the limestone was the same as Ri, but the value actually obtained is 0.708 & 0.001, about midway between Ri and the value for BOO--m.y.-old sea-water. The limestone in question came from Piketberg, some The distance 70 miles from the area in which all the other samples were obtained. may be a factor in the ambiguous result; alternatively, still accepting the homogenization theory, it is possible that the very high concentration of Sr in the limestone precluded complete homogenization. 1-n spite of the negative limestone result, the homogenizat#ion mechanism is considered the most plausible way of accounting for the linear t*rend of data points, largely because it accounts for Dhe observation that Ri exceeds the sea-water SrS7/‘Sra6 value. However the nature and duration of the postulated homogenization process is uncertain, and it would be most valuable if these results could be

Isot,opic age determinations on the Cape Granite


duplica,ted on a formation of known age. Whichever mechanism was operative it should be noted that the age of the Malmesbury deposition cannot greatly exceed the apparent age of 595 & 45 m.y. obtained from the isochron. If the Cape granite was formed by remobilization of the Malmesbury sediment’s (in effect an extreme case of homogenization as already discussed) t’he Ri ratio of the granite would exceed that of the sediments by an amount proportional to the This age difference is uncertain because age dift‘erence between the two formations. of the relatively high uncertainty in the Malmesbury age and because the validity of the isochron approach is itself in doubt when applied to the sediments. However, the observed R, ratio of the granite (0.710 & 0.003) does not differ from that of the Madmesbury (0.712 & 0.003) at the 95 per cent confidence level, and it follows that the remobilization hypothesis is feasible, even if unproved. The latest-Precambrian to early-Cambrian age indicated here for the Malmesbury also throws light on its possible correlation with the pre-Cape and pre-Nama rocks in the Bitterfontein-Van ahynsdorp area of southern Namaqualand. SANSEN (1960) considered the paragneiss in the Bitterfontein area to have formed by granitization of what he considered “Malmesbury System” sediments. However, NICOL~~YSEN(1962, pp. 594-595) reported age measurements from Namaqualand and concluded that the paragneisses must have formed prior to 1050 m.y. ago. Therefore, if JANSEN’S field interpretation is accepted, the granitized pre-Nama sediments must represent a formation considerably older than the Malmesbury. For instance, BRINK (1950) correlated the pre-Nama paragneisses from about 13 miles north-east of Bitterfontein with the Kheis System. AI~PENDIX: SAMPLE LOCALITIES, PETROGRAPHY AND MIP\‘ERALFRACTIOR’S (a) Granites 104 (GAlII4). contact (Fig. 3).*

Sea Point


380 feet

SW of the granite-migmatite

Coarsely porphyritic biotite granite containing phenocrysts of microclinemicroperthite up to 5 cm long, set in a matrix of quartz, microperthite, plagioclase and biotite. The average grain-size of the matrix is about 4 mm. The microcline is only slightly altered but some zoned plagioclase (oligoclase) crystals show saussuritized cores. Some biotite flakes are strongly chloritized. The granite contains small amounts of muscovite and a few relatively large crystals of strongly pinitized cordierite. Further accessories include zircon, apatite, brown tourmaline and ore. The approximate modal composition, based on the mesonorm (BARTH, 1959) and analyses of K-feldspars and biotites, shows the rock to contain quartz 270/b, microcline-microperthite 25O/“,, plagioclase 28% and biotite 8%. Thle following minerals were separated: (a) Ill:icrocline (-99% pure) showing a few grains of quartz. (b) Plaqioclase, cont’aining ~5% quartz. (c) Biotite used in two fractions. (A) Biotite, containing ~5% the extent’ of -200/O.

of impurities,


quartz, but chloritized

* ‘GA’ numbers refer to Aust’ralian Sational University classification.








pure, but containing only ~10% chloritized material. (B) Biotite, also “95% 156. (GA1 153). Sea Point contact. Sample of the discordant aplogranite mass of about 300 feet diameter intruded into the Malmesbury hornfels, due NE from t,he contact and just outside the migmatite zone (Fig. 3). This very fresh rock is equigranular in text’ure with grain-size 1 mm, and is The approximate mineral composed almost entirely of quartz and feldspars. composition is as follows: Quartz 36%, K-feldspar 2976, plagioclase 34%, muscovite and biotite O.5o/o or less. Separated mineral fractions are comprised of K-feldspar of -980/b purity and a plagioclase concentrate containing about 50% quartz, but virtually no K-feldspa,r. 108. (GA1113).

Small ravine,

120 feet above High Level Road,


This normal, coarsely porphyritic biotite granite is mineralogically very similar to 104, and composed of quartz, microcline-microperthite and plagioclase in about Ferromagnesians are mainly biotite together with small amounts equal proportions. Other accessories include small amounts of of muscovite and cordierite (pinite). both brown and blue tourmaline, apatite, zircon and ore. The rock is relatively unweathered, the K-feldspar showing only normal slight clouding. Biotite is slightly chloritised in places. The K-feldspar separated is w98o/o pure and the biotite fraction contains between 30/ and 5% chloritic alteration product. 710. (GA1152).

Coastline, Llandudno

Bay, Cape Peninsula.

INormal coarsely porphyritic biotite granite, slightly weathered as shown by some sericitization of the K-feldspar and development of chlorite around the borders of biotite flakes. The rock contains normal proportions of the minerals quartz, microcline-microperthite, plagioclase and biotihe with subordinate amounts of muscovite and a few euhedral pinitized cordierite crystals. Accessory minerals minerals make up include tourmaline, apatite, zircon and ore. Ferromagnesian 14oj,, by volume. The separated K-feldspar fraction was 9856 pure and the biotite used contains ~3% chlorite in the form of composite grains. 136. Centre of west face, Kloof

Quarry, Cape Town.

Fine grained porphyritic muscovite-biotite granit,e showing an average grainsize of the matrix of 0.3-0.4 mm. Phenocrysts of microcline-microperthite up to 2 cm. long are most conspicuous but all major constituents occur porphyriticallp as well as in the groundmass. Muscovite is more abundant than biotite and accessory minerals include occasional idiomorphic crystals of cordierite, zircon, apatite and brown tourmaline in higher than normal proportion. The rock is somewhat altered showing chloritization of some biotite, and clouding of feldspars. The approximate mineral composition is: quartz 32%, microcline-microperthit’e 27 %, plagioclase 28 %, muscovite and sericite 8 “/, biotite 4.8%. 147. Kloof

Quarry, 200 feet north of 136.

Fine-grained granite with roughly equigranular texture; the average size of the grains of quartz, microcline-micropherthite, plagioclase and muscovite being

Isotopic age det,erminations on the Cape (iranit?


0.6-0.8 mm with very few small phenocryst,s of K-feldspar, quartz or biotite. The feldspars are somewhat altered and the few flakes of biotite present are strongly chloritized. Brown tourmaline is relatively abundant. The major minerals present 28%, plagioclase 37%, muscovite and are : quartz 2!10/,, microcline-microperthite sericite 4.6:/,, and biot,ite 1.4%.

I-15. (GA1051).


Quarry, 65 feet north of 136.

Fine-grained granite with roughly equigranular texture very similar to 14i in mineralogy but slightly finer-grained wit,h an average grain-size of 0.5 mm. It also shows only very few small phenocrysts of feldspar or biotite of up to l-2 mm diameter. Again muscovite (3 “/o by volume) exceeds biotite (2 o/o) and brown, green and a little of the blue variet,y of tourmaline is present,. The plagioclase and microcline-microperthite (containing 17 “/o plagioclase) are slightly altered and some biotit*e flakes, containing very small zircons, are chloritized.

(6) Malnaesbury sediments M9-B.

Sea Point contact;

on beach,

150 feet N of the Sea Point pavilion.

Fine grained but arenaceous, quartzitic type of hornfelsed siltstone. The thermal effect of the nearby granite is shown by slight spotting owing to the development of some poikiloblastic cordierite, but ERLANK (personal communication) considers this rock to be unaffected by metasomatism. MI-I-A.

De Waal Drive Quarry, Cape Town.

Fresh shale (slate) showing some recrystallization. M-15. Same locality as M14-A. This sample of hornfelsed siltstone represents the more arenaceous type of Malmesbury sediment in the area. 31-40. Bellville Quarry on national road from Cape Town to Paarl, boundary between Parow and Bellville (Grand National Quarries).

near the

Sample is a fresh shale (slate) similar to M14-A. M-41.

Same locality

Hornfelsed JIL-I.


as M-40. a fresh, arenaceous

Quarry of the Portland

type similar to M-15.

Cement Company

at, De Hoek, Piketberg.

This locality is about 70 miles norbh of the Cape Peninsula Limestone,

fairly pure CaCO,.

would like to esprcss samples together witjh X-ray Compston for a measurement on tho limestone forming some mineral separations, to J. Cooper


the Malmesbury


our sincere thanks to A. Erlank Tvho provided fluorescent data. WC are also indebted to W. samplo, to R,. Rudowski and H. Berry for per-

for making the flame-photometer measurements, and to RI. Vernon for assisting with chemical processing. We wish to thank Dr. L. 0. xicolaysen

and Dr. IV. (lompst,on for critically roading the manuscript,. One of us (H. L. A.) is indebted to the Australian National University for the award of a Visiting Fellowship.

REFERENCES ALnRICH 1;. T., \VETHERILL G. TV., DAVIS G. L. and TILTON G. R. (1958) Radioactive ages of micas from granitic rocks by Rb--Sr and K-Ar methods. 2’raw. Amer. (&oph~~s. Cn. 39,

1124-l 134.





and P.


I~ARTH T. F. W. (1959) Principles of classification and norm calculation of metamorphic rocks. J. Geol. 67, 135-152. 13~1~~ W. C. (1950) The geology, structure and petrology of the Suwcrus aroa, Cape Province. Ann. Ulziv. Stellenbosch 26, A, 97-221. COMPSTON W. and PIDGEON R. T. (1962) Rubidium-strontium dating of shales by the total rock mothod. J. Geopkys. Res. 67, 3493-3502. COMPSTON W., LOVERING J. F. and VERNON M. J. (1965) Rubidium-stront’ium age of the Bishopville aubrite. Geochim. et Cosmochim. Acta (in press). ERLANK A. E. (1965) The geochemistry of the Malmosbury Formation, South Africa. Doctoral thesis, Univ. Capetown. (In preparation.) HEDGE C. E. and WALTRALL F. G. (1963) Radiogcnic strontium-87 as an index of geologic processes. Scielxe 140, 1214-1217. JAKSEN H. (1960) Explanation of Sheet 253 (Bitterfontein). Geol. Survey, DepC. of Mines, South Africa. KOLBE P. (1965) The geochemistry of the Cape Granites, south western Cape Province, South Africa. Doctoral thesis, Australian Nat. Univ. (In preparation.) MCDOUGALL IAX (1964) Potassium-argon ages from lavas of the Hawaiian islands. Geol. Sot. Amer. Bull. 75, 107-128. XICOLAYSEN L. 0. (1962) Stratigraphic interpretations of age measurements in Southern Africa. Petrologic Studios: Buddington Volume, 1962. Geol. Sot. Amer. 569-598. SCROLTZ D. L. (1946) On the younger pre-Cambrian plutons of the Cape Province. Proc. Geol. Sot. S. Afr. 49, 35-130. WALKER A. R. E. (1929) The Sea-Point granite-slate contact. Int. Geol. Coxgr. XV (S. Afr.) Guide A3. WALKER F. and AfaTHIAS M. (1946) The petrology of two granite-slate contacts at Cape Town, South Africa. Quart. J. Geol. Sot. Lo&. 102, 499-521. WHITNEY P. R. and HURLEY P. M. (1964) The problem of inherited radiogenic strontium in sediment,ary age determinations. Geochim. et Cosmockim. Acta 28, 425-436.