1329550692cooke RA bubye limpopo

South African Archaeological Society Report on Archaeological Sites Bubye/Limpopo Valleys of Southern Rhodesia Author(s...

0 downloads 117 Views 2MB Size
South African Archaeological Society

Report on Archaeological Sites Bubye/Limpopo Valleys of Southern Rhodesia Author(s): C. K. Cooke Source: The South African Archaeological Bulletin, Vol. 15, No. 59 (Sep., 1960), pp. 95-109 Published by: South African Archaeological Society Stable URL: http://www.jstor.org/stable/3886562 Accessed: 15/08/2010 04:12 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=saas. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

South African Archaeological Society is collaborating with JSTOR to digitize, preserve and extend access to The South African Archaeological Bulletin.

http://www.jstor.org

REPORT ON ARCHAEOLOGICAL SITES BUBYE/LIMPOPO VALLEYS OF SOUTHERN RHODESIA C. K. COOKE,F.S.A. Bulawayo INTRODUCTION

While with the Rhodesian Schools Exploration Society during May 1958, a number of sites were found in the area of the Mateke Hills. These give a picture of Stone and Iron Age culturesfrom a large partof the south-easterncornerof SouthernRhodesia. The sketch-map(fig. 9) shows the area covered by this survey and designates the sites located. These, where of sufficientimportance,are describedin this report. Many camps and factory sites of the Stone Age were located, but the EarlierStone Age is absent and the Later Stone Age is very poorly represented. This aspect is discussedin the conclusions. The country where these sites were found is inhospitable, having little permanent water except during the rains. Despite the abundance of cover from trees and grass, the heat, even duringthe colder months, is intense, the thermometerrising to over 1050 F. just before the start of the rainy season. In the Mateke Hills, permanentwater is available and large herds of elephant roam along the streambeds laying waste any crops which are planted near the perimeterof the hills, making human habitation precarious.

winds. The country generallyis Mopane-Combretum woodland with Acacia in or near the stream-beds. The amount of debrispresentmakes it almost certain that this was a home/factorysite. Method of Collecting Instructedby the writer, the schoolboys collected over an area stepped out to cover approximately 10 yds. by 10 yds. All artefacts were brought to a centralplace where they were roughly sorted, and all the debris and unworkedmaterialrejected.A count was then made of the waste flakes and broken cores, these too had to be left becauseit was not possible to carry away more weight in the Land-Rovers. The following artefacts were brought back for further examinationat the main camp, and are now in the National Museum, Bulawayo. Handaxes 2 bifacial 1 unifacial on flake Picks 2 carinated 2 plain, sharpenedpebbles 1 plain, sharpenedflake Cores 5 large tortoise 5 tortoise under 5 in. diam. Side Scrapers 1 on pebble 1 on broken core 1 on 3 in. flake Notched Scrapers 4 thick flakes 1 on rough pebble I on 2 in. flake 5 miscellaneouschopping tools on flakes 1 chopping tool: plain butt flake

The countryconsistsof fairlyopen valleysmargined by hills of graniteand syenitewith occasionaloutcrops of sandstonenearthe escarpment.All the riverswhich run into the Limpopo have cut shallow courses through the escarpment,down to the deep valley. Practically all the Stone Age and Iron Age sites were located away from the deep valley of the Limpopo River on the drierbut slightlymore healthy areas of the escarpment. Malariaand tick-bornediseasesare very prevalent in the damperpartsof the whole area,makingagriculture and human settlement almost impossible. I am very grateful to the Rhodesian Schools ExplorationSocietyfor permissionto use the material collections and details from the Map made on this expedition. (Mateke Expedition 1958, duplicatedfor Flakes 7 plain butt, with bulb removed private circulation.) Flakes Rejected: 250. STONE AGE SITES: BUBYE/LIMPOPOVALLEY Cores Broken Rejects: 60-mainly proto-Levallois. 1.

1ST INTERMEDIATESrrE: SITUATION 'GILCHRIST'S FOLLY'

This site is 10 miles from the Limpopo River, on a tributaryof Bubye River. Description of Site The artefacts of the 1st Intermediate Stone Age

Material All made from Kalahari sandstone, the natural rock of the Limpopo River escarpmenton which the site is situated.All implementsappearto be abraded by wind and sand. They do not have the typical rolling associated with water wear.

are on a sandstone outcrop in very tangled bush Discussion country, about 200 yds. away from and 100 ft. above Little has been describedpreviouslyfrom near this an unnamedtributaryof the Bubye River, sheltered area. The site described is very rich, and further by koppies from the prevailing south-east winter collections over a larger area might result in further 95

'f;X~~~~~'

~~~~~~~~~~~2

1

I

A-

I

13T) ~~~~~~~~~~~~~~~~~~~~~~~011 FiG.

96

1

tool types being found and allow a more comprehensive view to be made of the industry present. But it is evidentfrom the materialexaminedthat this industry is Sangoan, and taking the material used into consideration it is akin to the Southern or Bembesivariant.Apparentlythere are no cleaversin this assemblage, Jones'0 says 'cleavers are absent, though a solitary example was found' when writing of the type site at Bembesi. The rolling or abrading is not constant, some show the arrises still fairly sharp whilst others are rounded. This may indicate varyingages for some of this material;on the other hand, material sheltered behind rocks and other implementsmay account for the differences(figs. 1 and 2, nos. 1 13 and 1 11). The wear is consistent with wind abrasion.

the fine implements found in other sites2 are not present.It is shown, however, that when the makers used chalcedonyand induratedshale that they were capable of producing implementsof the finest type. The nearest to this site so far describedare the sites from Lake Ngami and Lake Dow, Bechuanaland.67 The points, made mainly on side flakes, are akin to the shale points from Amadzimba Cave in the Southern Matopos.9 At this site, despite relatively abundant artefacts of the 1st IntermediateStone Age and Middle Stone Age, no trace could be found of either the 2nd Intermediateor the Later Stone Age. Both Magosian and Wilton can be found within 10 to 20 miles but none was located in the near vicinity.No trace could be found of the Earlier Stone Age; this may be accounted for by the fact that anywherewithin 20 2. MIDDLESTONEAGE SITE-'GILCHRIST's FOLLY miles of this river system would most probablyhave Description of Site been very heavilyinfestedby both anophelesmosquiLocated within 100 yds. of the Sangoan Site toes and tsetse flies making human life unbearable. describedabove, but on the exposed hillside facing This may have applied to the other periods as well, south-easttowards the river-bed. althoughthe climate duringthe Magosianperiod was Collecting:Method as describedabove. probably similar to that at the present time coming after the period of erosion in the Stillbay which had Summary of Artefacts greater rainfall. Points 4 Unifacial from 1+-3 in. 3.

Anvils

THE BUBYE HOMESTEADSITE 2ND INTERMEDIATE

STONEAGE: MAGOSIAN

1 plain Flakes

Descriptionof Site

16 plain butt 12 miscellaneous 12 faceted 4 single facet Rejected on site: approximately300 flakes.

This site is located in typical Mopane country and

is located on a large area of sheet erosion, sheltered from the north-eastby the hillsideon the top of which stands the remains of an old ranch house. Running up this hillock is an outcropof mudstoneshale which wasthe materialusedfor practicallyall the implements found.

Cores

12 miscellaneous 9 flat or disc 20 high-backedtortoise-7 unstruck 5 polyhedral 4 large blade

Method of Collecting All artefactswere collected over an area of 70 yds. by 30 yds., which was the size of the eroded land. The artefacts listed below show the complete assemblagefrom that area.

Crescents

2 1J-2j

inches.

Scrapers

End 10 from i in. to 3 in. Side 21 from i in. to 3 in. Notched 11 from i in. to 4 in. Burins

1 gouge shaped

Saws

3 convex Material

Sandstone, shale, chalcedony and quartz. Mainly however of the local sandstone which, unlike the implements in the Sangoan, is unrolled and sharp.

Summaryof Artefacts Cores 11 flat-tortoise 2 flat-disc 1 very thin disc 3 micro-disc 8 polyhedral 7 high-backedmicrolithic 4 blade 3 micro-blade TOTAL39

Scrapers 6 end The implements are very similar to assemblages I micro-end from other parts of SouthernRhodesia, and for that 12 notched reason no illustrations have been given. Some of 2 micro-notched Discussion

97

_iG.

2

---

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~-

7)fI9

g

FIG. 2

98

3 8 7 5

compound (side and end) side nosed compound (miscellaneous) TOTAL44 Points

4 Unifacial (on side flakes) 1 bifacial (triangularin section) TOTAL5 Drills

I micro Anvils

1 bi-polar (diorite) Burins

1 single blow 1 bec-de-flute Crescents

2 (1 microlithic) Flakes with secondary working

1 plain butt 2 faceted butt 21 with butt removed = (some Tabalabat technique) TOTAL24 Flakes unworked

11 faceted butt 2 plain butt 3 single facet 12 miscellaneous 3 blade TOTAL31 TOTALARTEFACTS 149 Materials

Mainly mudstone shale, with some ironstone, quartz and chalcedony. Discussion

At this obvious factory site, the assemblageof tool types is very similar to the phase at Sawmills,1' but would on the face of thingsbe an early Magosian.4 However,such a statementcannot be proved because the materialused was not suitable for the production of cores and therefore many implements are made directly from pieces of laminatedshale giving to the assemblagea very crude appearance. However, the microlithic tools made of quartz and chalcedony could be mixed with material from Khami5 and never be recovered.The crescentsshow the ear, the cores of the harder materials show evidence of bipolar strikingwhile the recoveryof a typical pitted anvil shows that this technique was well developed (fig. 3, nos. 1-31). Once again we have the MiddleStone Age methods predominatingover the intrusiveblade/burinculture, and anothervariationin the long line of intermediates from the Hom of Africa to the Cape of Good Hope. 4. LATER STONE AGE: MARUNGUDZI,LUCHENDI RIVER

At these two sites on high ground above the river valleys, small assemblages of typical Wilton tools were 99

recovered. Both were small hunting-camp sites showingvery little sign of long occupation. Crescents backed blades, a few cores and waste flakes made of quartz were all that could be found after intensive searching. There is insufficientmaterial to warrant any illustrationor description.In this area of poor material it is possible that these nomadic hunters had to rely more on bone and wood than was usual in other areas. DIscUSSION Stone Age The river system north of the Limpopo bounded by the Bubye River was inhabited by humans from 1st IntermediateStone Age times up to the Later Stone Age, but shows no sign of any Earlier Stone Age. Thereis little doubt that heavy rainfallduringat least part of the Earlier Stone Age made the valleys veryunhealthy,both to humanand animallife, owing to the presenceof the mosquito and tsetse fly. Jones and Bond at Sawmills and Khami give differing interpretationsof the climatic conditions during the 1st Intermediate Stone Age. 1"i' It does appear however that the climate may have been wet but graduallygettingdrierduringthis period. If this is so, it would, because of the reduction of water supplies in the healthier high veld, have brought these 1st Intermediate into the valley following the game migrationsto the water. The 1st IntermediateStone Age sites are restricted to the escarpment. It is thereforepossible that the swampy lowland adjacent to the rivercontainedwaterand animallife duringthe period under discussion but the lack of implements suggests that it was too unhealthy for human habitation. Middle Stone Age sites are found in the valley as well as on the higher ground, but the large factory sites and home sites also appear almost exclusively on the higherplains and on and around the koppies. The 2ndIntermediateStoneAge is only represented by one site very near to permanent water on the higher ground; this site was undoubtedlya factory site being situated next to the supply of the main materialused for the manufactureof the implements, a soft mudstone shale. The materialavailablewas not suitablefor the fine implementsof the Middle Stone Age and the later industries.This fact may have had a strong influence on the hunterswho would not live for long away from the source of their hunting materials.The fact that there are a few implementsin the Magosianmade of materialwhich does not occur in the vicinity shows that they did carry weapons from a long way off. There is no evidence to show that they carried raw materialwith them to this factorysite whichmay also have been an annual hunting-campduring the dry season when animalsmigratedto the biggerriversfor water.The area,althoughoccupiedfor so long, shows no sign of extensivesettlementor prolonged occupation at any one site. Health as well as shortage of suitable fine materials were probably the main reasons for man not living in greaterconcentration

& $

'

16 17'

"'9~~~~~~~~~~~4

0-

19

-

--

-i->tv2

4)

'-

27 FIG. 3

in the area. Malaria is still a common disease in the vicinity of permanent water.

I would estimate the age of the chief informant, Chinana, as between 60 and 70 years. The water was cloudy with kaolin, but otherwisefairly wholesome. THE IRON AGE A large slab of granite was pushed over the opening No very early sites were found, the earliest being every night to prevent the elephant from using the at Zwergibe Waterhole, where a few sherds akin to supply and polluting it (fig. 8). From the amount of broken pottery (fig. 4 nos. 31 Leopard's Koppie Ware13were found. This site is very interesting from a geological point of view and 32), Middle Stone Age flakes and animal bones because this perennialwaterhole is in the apparently scatteredaround this water supply, it appearsthat it solid granite,being a naturalcircularhole about 2 ft. has been used for a very long time both by humans in diameter and 4 ft. deep, fed by an underground and also possibly by animals before them. When I stream. Local Africans told me that, except in the visited this site during May 1958, there was no other driest of seasons, the watersupplywas constant.They water available,except deep in the sand of the Bubye said that it had never dried up during their lifetime. River about 8 miles away. The rainy season usually 100

terminates at the end of March in this area, the averagerainfall being 10-12 in. a year. The Iron Age wall builders had also found this supply and built a small circular enclosure on a near-by koppie. As only unidentifiablepottery was found at this site, it is impossible to date it or to connect it with any other sites. Thereare numbersof walledhills in the area which appear to fall into two periods, the early Nguni Raidingperiod of the late eighteenthcenturyand the nineteenth-centuryMatabele period; they appear to be refuges, not home sites, nor sites of true fortifications.The main ones are:

DESCRIPTION OF THE POTTERY

Cliinana'sOld Kraal The potteryfrom this site consists of beakerbowls, deep bowls and pots with wide mouths; none of the pots shows long or constrictednecks. The fabric is fairly thin, well baked, with very little coarse grit added to the clay. The clay howeverwas not of very good quality, and therefore the pottery is not very hard despite the fact that it is burnt right through. The decorationconsistsmainlyof fine incisedpatterns made while the clay was fairly dry, but before firing. The patterns consist of rim decoration of crosshatching, bands of cross-hatchingenclosed in two parallel lines, chevron patterns, bands of triangles Marungudzi Geologically,the hills which form the ring structure some with the bases at the bottom, others with known as Marungudzi are very complex, the ring alternate reversal of the hatched triangles. All the of syenite surroundinga flat crater-likedepression. decorationcould have been carriedout with a small This depressionhas been a refugefor many years, the iron arrow or spearheadusing either the blade or the inhabitantsleaving many signs of Stone Age as well tang. No graphiteor otherpaint appearsto have been as Iron Age occupation. The southern side of this used (fig. 4, nos. 1-20, 22, 23). depressionshows another half-moon-shapeddepres- Dawn Mine Hill-Ruin sion which formeda sanctuarydifficultof access, but The greater part of a pot of very unusual design into which in times of war both animalsand humans (fig. 7, no. 15) was found associatedwith the walling. took refuge. To make the entrance more difficult, This was a large vessel naturally reddish in colour a series of stone walls have been built across the from the firingof the clay; well burnt,thin in section; defilewhich is the only means of entry.Beads, pottery clay with coarse grit, no burnishor graphite.It has a and the barrel of a muzzle-loadinggun circa 1850 narrow constricted neck, rim rounded and upright. make it almost certain that this refuge was occupied Neck diameter 62 in.; body immediately below during the Matabele raiding periods of the early shoulder 14 in. diameter.The design consists of four nineteenth century. lozenges of incised patternspaced equallyaroundthe neck from the shoulderto within an inch of the rim. ChumbangulaHill The design resembles one of the divining-bones The walling on this hill is of the same general (hakata)used by the people who occupiedthe Khami type as Marungudzibut appearsto havebeen designed Ruins,13and attributed to the Rozwi people. The to make the way up the hill difficult. This indeed fact that this ruin is adjacentto an ancient coppermay have been a strong-point which, if necessary, working makes it possible that the occupants were could have been defended by determinedmen with of the Lemba tribe. Few other sherds were found, stones, spears, bows or guns. The walls arevery low, all of which were unidentifiable.It is noteworthythat not exceeding 2 or 3 ft. in height now. There is no a sherd with part of a similar design was found at evidence to show that they were ever higher, there ChumbangulaHill. being no heaps of stone or signs of collapsed walls ChumbangulaHill on this hill. Associated with ruined low walling, 41 sherds of Chinana'sOld Kraal decorated pottery, and 4 undecorated rim sherds An African named Chinana said that four genera- were collected from rock clefts and surface sites. tions of his people had lived at this kraal site before It was possible to reconstructa numberof pots from the place was abandoned about 1896. It was these remains (fig. 5 nos. 4-13; fig. 6, nos. 1-5; abandonedbecauseof worn-outlands and a diminish- fig. 7, nos. 1-15), the beaker bowl being the coming water supply. There are large herds of elephant monestshape,but in additionthereare shallowbowls, in the area which were also a menace to agriculture. deep bowls, pots with constrictedopenings,pots with long necks, globular pots and one pot with a wide Dawn Hill Mine Koppie flared neck. These pots do not appear to have any A small ruin of rough stone shaped like a 'figure kinship with those from Chinana's Old Kraal, but of eight' which consists of two small enclosureswith they show some resemblance to pottery of other separateentrances.The only feature of this ruin is a traditions, namely bandware from the Rhodesian grain pit built of stone, this forms a stone-linedhole Stone buildingcomplex,Mapungubwein the northern 3 ft. in diameter and 4 ft. deep in the middle of a Transvaal,and pottery from old kraal sites at Chibi thick wall. Remains of millet seeds were found at (unpublishedreports, K. R. Robinson), but it is not the bottom of this pit, no otherseeds nor any artefacts possible to draw any definite conclusion from the being found. availableevidence.It is safe, however,to state that this 101

12~~~~~~~~~~1

Ict

p

,I)~1 24 25 2627

-

A^.__||

A,-

30

K

:X~~~~~~~~~~~~~~~~~~1 1821

FIG. 4 102

1

3

", N, e-v''%'~

o MS

*--_C4

*a

m)

O'r.

~~~5

6

;

X

~

'Ii FIG. 5

/77777 :X

-

I~~~~~~~~~~~~I

FG __

w

6

_


15~~~~~~~~~~~~~~~~~~~

_

/

14

,0

cm

-

13

ING CMiS FIG. 7

105

pottery does not appear to be derived from, or be ancestralto, the band and panel ware which can be found in the Umanendau Ruins near Beitbridge,at Khami"3and Dhlo Dhlo Ruins.3 Other cultural material was recovered, namely a spindle whorl made from a piece of decorated potsherd (fig. 5, no. 14), centrallybored with hour-glassperforation, ostrich eggshell beads similarly perforated, and a piece of copperbanglemadeof a thin stripof flattened copper wire wound on to a core of grass. Not all the pottery listed was unburnished,but none of it showed any sign of graphite.Three sherds were coloured with a reddish pigment before firing.

been disturbed by other tribes raiding for crops, cattle and slaves. It is probable that many tribes occupied the area, completelyor as conquerorsfrom time to time, thereforethe pottery traditionsmay be mixed. It is of interestto note that in the more open country to the west there are cultural remains of other people who never penetratedinto the Mateke Hill area. These people can be associated with the stone-wall builders of the Khami Ruins period, which has been dated at about 300 years ago.13 A comparisonof the pottery from Chinana'sOld Kraal with that from ChumbangulaHill makes it doubtful if there is any connexionbetweenChinana's

I_3 D~PSRArirSr'

SOTIlvl1 C

FTr.

OFp FIG. 8

All decorationwas incised on a fairly dry clay before baking. Nearly all the rims show a flattenedsection which is unlike other pottery recovered from the area.One interestingsherd(fig. 7, no. 12) of a different material was recovered; this is part of a carinated shallow bowl, baked to a reddishcolour, made from fine clay baked very hard. The materialused appears to differ from any of the local clays used in the manufactureof pottery.The decorationis the same in technique,consisting of a line of cross-hatchinghalf an inch below the rim, and another similar line just above the shoulder or carination. This assemblage of pottery could possibly be a combination of many traditions, because Chumbangula Hill was, accordingto Africantales, a sanctuary from raiding tribes for many generations. High up on the Mateke Hills, of which Chumbangulais one, is a rock which when struck gives a clear bell-like note.'6 This was used to warn the workers in the fields that raidingpartieshad been sighted.This gong was used as an alarm for the women and children to seek shelterin the hills and to drive the cattle over the ill-definedpaths and up the stream-bedsinto the sanctuary of the hills. DIsCUSSION-IRON

ancestorsand the wall builderson ChumbangulaHill. Chinana'speople did seek sanctuaryin the hills. He was, by his own statement, born below the walled hill, but not on it. Fragmentarypottery found in Chumbanyokarock shelterbelow the hill shows great similarity to the sherds from the old kraal site. From surface collections only, it is not possible to be certain of obtaining a complete picture of these cultures, particularly because very few beads and only fragmentaryiron implements were recovered. Because there is little evidence available, dating is a very difficultproblem, but it does appear that there are at least three separatecultures of the Iron Age in this area, the latestbeing Chinana'sOld Kraal from 1860 onwards, Chumbangula Hill probably as earlyas 1760,and the dressed-stonebuildingperiod of 1660 to 1820. The middle date is very tentative and based on native tradition and not on sound archaeologicalevidence. The absenceof graphitein all the potterydescribed is noteworthybecause all the modern pottery in the district is heavily painted with graphite which is obtainable locally from a surface seam near Chiturapadzi(see sketch map, fig. 9).

AGE

There is little doubt that the area under discussion has been under occupationfor a very long time; but it has always been an area in which the people have 106

CONCLUSIONS GENERAL Archaeologically, this was one of the completely unexplored areas of Southern Rhodesia. This preliminary survey shows that the area was inhabited for

4Fu~h~

~~~~~ ,MA Pu

2

WA~ gW

~~

f1Ae

os Q

'.

S/A

L

A

u U^Di

-'0-

c

,

rI!G!Nr.

RX

.:

N-

\

_ -

<

9

Tll %?~~~~~ eIO

zi 1

S

'

K

'"SO~T

LC>/ FO~~~~~~~~~ I_ T R,-

-

F- gA

.

IN E D

r

.

;

107~~~~~~~~~\ ,- -

~v

K

S CALE

IlC)T MIOLP-5. oC.rE _

Z

-3C HIKWAAR:I-.

0|

LIlMfPOPO

FIGo.

9

107

PI

VE7F,\.

,

a verylong period,but did not carrylargepopulations in the Stone Age or during the Iron Age. Water on the top of the escarpmentwould neverhave been too plentiful; waterholeswere of the greatestimportance to a residentpeople. It is apparentthat both the Iron Age and the Stone Age occupants lived within easy reach of the two areas where water, even in bad seasons, is available. During the 1st Intermediate Stone Age, man appearsto have been forced to live closer to the main rivers.In this districthe settled on the escarpmentaway from the swamps but within easy reach of water and food supply. There seems to be some difference of opinion over the climatic conditions pertainingthen, but there are no signs of human occupationduringthe 1st IntermediateStone Age nearerto the LimpopoRiver than the edge of the escarpmentat the points examined,or at sites 50 or 60 miles upstreamfrom the sites discussed(Robinson, unpublished, 1960). Many areas of erosion were examined without success. Some of these are very extensive.The tributariescuttingdown the escarpment to the west of the area examined(Robinson-reports to Historical MonumentsCommission)have yielded Sangoanimplementsof a similartype. Although odd tools have been found nearerthe riverthere has been no discoveryof a home or factory site other than on the escarpment. The absence of any signs of the EarlierStone Age is of interest, the reason being in all probability climatic. Jones"' gave the climatic fluctuation at Sawmills. This shows that the Earlier Stone Age period was very wet, and therefore the area under discussionwas in all probabilitytoo wet and unhealthy for human occupation. This was followed by the drying-up period of the 1st Intermediate. Bond demonstratedfrom the Khami deposit' that a period of heavier rainfall was present at the end of the Sangoan, diminishing slightly in intensity until the end of the Middle Stone Age. From the 2nd Intermediate Stone Age onwards, except for minor fluctuationsin both directions,the climaticconditions appear to have been very similar to those obtaining today. The distributionof these later industries, in the Limpopo valley, shows that for at least part of the year it was possible for man to live in the valleys as well as on the escarpment.The limiting factors were twofold, lack of permanentwater except in the large rivers and the lack of materialsuitable for the manufactureof fine stone implements.The sites of

Sawmills and Khami are within 200 miles of the Mateke Hills. It may be presumed,except for variations similarto those whichare knowntoday, that the conditions were much the same during the period under discussion. The country is inhospitable,and although capable of carryinglarge numbersof elephant and antelope, has never been suitable for agriculture.Therefore the Iron Age people who lived here appear to have used the hill areas as sanctuaries from migrant war-partiesand slave-hunters;it is evident that they never lived, until present times, anywhere except within a very short distance of these hideouts. It is recordedin SouthernRhodesianhistory that a native commissioner,cut off from his station at Chibi, lived in a cave in the Mateke Hills during the Matabele Rebellion in 1896, but so far, except as ranch land, this area has not been occupied by Europeansat all and by Africans only to a very limited extent. This cave, known as Chumbanyoka, yielded a few potsherds(fig. 4, nos. 33, 34) and some blue annular beads. The area has only been examined superficially; the conclusions reached must thereforebe tentative and subjectto correctionwhen more work has been done. Thereis a close resemblancein the generalassemblages to the same series of stone industries from other parts of Southern Rhodesia; unfortunately it has not been possibleto comparethe Sangoanwith that from PortugueseEast Africa and the Transvaal, but it undoubtedlyfits in with the southernvariants where the makers have been forced to use similar material for the manufactureof their implements. The Middleand LaterStoneAges are, withinmaterial variations, the same as those from such sites as Bambata,12 Khami,6 and other sites examined within Southern Rhodesia bearing a strong resemblanceto sites described from South Africa. The 2nd IntermediateStoneAge is nevera constant varying from site to site and from area to area. Material has had a great influence on the site described,showingthat crudityin manufactureis not alwaysthe criterionof any industry.If the microlithic elements made of quartz and chalcedony were removed, the remaindercould be thought to be a very early form of Stillbay, whereasit is in fact the final expressionof that long and developing Middle Stone Age industry.

REFERENCES 1 2

3 4 5 6

7

Occ.Pap. Nat. Mus. S. Rho.,no. 21A, 44-60. Bond, G. (1957). 'The Geology of the Khami Stone Age Sites', River, Bechuanaland', Bond, G. and Summers,R. (1954). 'A late StillbayHuntingCamp Site on the Nata 35. IX, Bull., S. Afr. Arch. Caton-Thompson(1931). TheZimbabweCulture,Oxford. a furtherexamination',S. Afr.Arch. Bull.,V. Cooke, C. K. (1950). 'The Middle Stone Age Site at Khami: Stone Age and proto-historic',Occ. S. Rhodesia; at site Khami, Waterworks 'The K. C. (1957). Cooke, 1-43. 21A, no. Rho., S. Pap. Nat. Mus. Cooke, C. K. ibid. no. 7. Bechuanaland1960',S. Afr.Arch.Bull. Cooke, C. K. and Paterson,M. L. 'A StillbaySite from LakeNgami, (In press.) 108

8

Cooke, C. K., and Paterson,M. L. (1960). 'Stone Age Sites-Lake Dow Area, Bechuanaland1960',S. Afr. Arch. Bull., XV. 9Cooke, C. K., and Robinson, K. R. (1954). 'Excavationsat AmadzimbaCave, Matopo Hills', Occ. Pap. Nat. Mus. S. Rho., no. 19, 669. Jones, N. (1938). 'The Bembesi Industry',Occ. Pap. Nat. Mus. S. Rho., no. 7, 7. 1' Jones, N. (1924). 'On the palaeolithicdeposits at Sawmills,S. Rhodesia',J.R. Anthrop.Inst., . 54, 176. 12 Jones, N. (1940). 'BambataCave: a Re-orientation',Occ. Pap. Nat. Mus. S. Rho., no. 9, 11. 13 Robinson, K. R. (1959). KhamiRuins.Cambridge. 14 Robinson, K. R. ibid. 15 Robinson, K. R. ibid. 16 Robinson, K. R. (1958). 'VeneratedRock Gongs and the presenceof Rock Slides in S. Rhodesia', S. Afr. Arch. Bull., XIII, 50. 17 Robinson, K. R. ibid., no. 8.

STONE IMPLEMENTS FROM PRETORIA CITY PAVEMENTS D. KEENAN-SMITH Johannesburg During recent visits to certain Pretoria suburbs, notably Sunnyside and Arcadia, I have recovered numerous stone artefacts on the wide and exposed pavements. These have been made largely from quartziteand induratedshale and relate to the late Early and Middle Stone Ages. Of the latter cultural horizon (MSA), a numberof excellentStillbaypoints have been found together with Pietersburgelements. The accompanyingfiguredepicts a fine exampleof the highly evolved handaxeswhich appearedtowards the end of the Early Stone Age. It is made of quartzite, is unrolledand is patinateda rich brown colour. It appears to have been made from a core and, in the absenceof stratificationhas been assignedby the writerto the StellenboschV phase. This artefactwas found at the surface of a freshly turned flower-bed (shortlybefore, in situ?) which bordersthe pavement at the comer of Walkerand JoubertStreets,Pretoria. This note is intended to remind the amateur archaeologist that archaeological remains may be easilyrecoveredeven whiletakinga casualwalk down a pavementin a built-uparea.

109

6~~~~~~~6M

I)

\

Late Early Stone Age handaxefrom PretoriaCity