Uncovering one of the largest Mesolithic Sites in the UK
For detailed analysis select lithic type
Information recorded for all findsYear, site, field number, grid reference, date when found and finder’s intialsAttributes recorded for all the lithicsMaterial: (almost exclusively flint), + notes on quality, inclusions, flawsColour + notes (see General Results, Colour)Burnt: yes (cracked or crazed), slightly burnt (red tinges), or noCortex: yes / no, colour, chatter marksSize: Using a graded series of circles Weight: weighed to 0.01gm (Scout Pro up to 200gm)Type: based on Torben BallinTypesFlake: a piece removed from a nodule or core with one identifiable ventral surface (including ‘chips’, flakes smaller than 10.0mm)Blade: flake with length equal to or twice the maximum width, parallel edges and ridgesCore: a piece with only dorsal surfaces produced by removal of three or more flakesSplit pebble: with fewer than three flakes removedTool/Retouched (scraper, arrowhead, microlith, microburin)Indeterminate fragment: unidentifiable as either a flake or a coreChunk: a larger indeterminate pieceOther attributes recorded in addition where relevant
Almost 10,000 flints were collected during this project. I used Microsoft Excel for cataloguing, recording attributes and processing the data.
FlakesWhole, almost whole,Platform remnant width, depthPlatform preparation: abrasion, heavy abrasion, trimmedTermination: feathered, hinged, plungingCortex, postion and percentage of dorsal surfaceRidges: number and positionLength, at right angles to platform, max width, max thickness Fragment: Proximal, medial, distal, proximal+medial, medial+distal, splitAs above, where possible, but not measured, and cortex recorded as yes/no
BladesWhole, almost wholeAs for whole, almost whole flakes where possibleFragments: length equal to or twice maximum widthProbable fragments: length more than two thirds maximum widthPossible fragments: length less than two-thirds msaximum width
Single platform coresWhole/almost wholeLength, at right angle to the platformPlatform width parallel to the facePlatform depth at right angle to the faceNumber of removalsType of removals (flake, blade, long flake)Cortex amount and positionPercentage of platform edge workedLongest face lengthReason for abandoningAngle between platform and faceMorphology (e.g, pyramidal, cylindrical)Damaged, brokenAs for ‘whole, almost whole’ where possibleMore than one platformAs for ‘other cores’ (below)Other coresNo definite platformLength (longest dimension)Width (second longest dimension at right angle to length)Thickness (third longest dimension at right angles to width)Number of removalsType of removalsCortex, amount and positionMorphologySplit pebblesAs for ‘other cores’ where possible, and/or brief description
Reference for the Lithics Analysis pageTorbenBallin2013.NethermillsFarm,Banchory,Aberdeenshire.TheLithicAssemblage.46 pages.
Bothbroadbladeandnarrowblademicrolithsarepresentinthisassemblage,including unidentifiablefragmentsnotdealtwithinthisreport.Widebasedtrianglesandobliquepoints comprisethebroadbladetechnology,whereascrescentsandedge-backedmicrolithsarethemain representativesofthenarriowbladetechnology.Thewide-basedtrianglesfallintotwogroups, large and small.
Broad Blade Technology (?early Mesolithic)Threeobliquepointsand15wide-basedtriangleswerefoundduringthefieldwalkingat Crathes. All have the proximal end removed Thedimensionsandretouchattributesoftheobliquepoints,twofromNMWandonefrom NME, agree very well with those from the JBK excavation (See tables and graph).
Figure M1 shows the dimensions of the wide based triangles from all 3 sites (a) and (b) oblique points from NME and NMW.
Preferenceforretouchalongtherightedge,andthemedial(probablyextendingdistally)sectionofthe bladebeingusedareinlinewiththosefromtheJBKmaterial.Oneofthelargeoneshadtheremainsof a notch from the microburin technique.However,whencomparingthetwosampleswiththosefromtheJBKdig,theaveragedimensionsof thetwogroupsfromeachsamplediffer:theOFARS‘large’arelarger,theJBK‘small’aresmaller;the JBK ‘large’ are nearer in length to the OFARS ‘small’ (see Fig. M2 (a). Theaveragelength/widthratiosofthetwosetsofsmallandlargetrianglesseemtobesimilar(Figure M2 (b).
Figure M2 shows the average dimensions of the OFARS (HS) and the JBK dig (JK/TB) triangles (a), and (b) the linear regression at SE 95%.
AlanSaville(perscomm)suggestedthatthesmalltrianglefromMCwasanintermediateform. Doesthisrepresentacontinuumofsizesincontemporaneoususe,oranindicationofprogress fromthebroadbladeindustrytothenarrowbladeovertheyears?JBKwasoftheopinionthatthe OFARSobliquepointswereonthesmallsideandreferredtothemasreminiscentoftheEarly Mesolithic rather than definitely diagnostic (DES ref).
Narrow blade technology (?Later Mesolithic)Narrowblademicrolithswerefoundatallthreesites(TableM3);almostallareeithercrescents, edged‘withahighlyregular’convexcurvature(seeTB)oredge-backedmicrolithswithstraight edges,backedononeside,withtheproximalendremoved.Acoupleof‘crescents’,notincluded, were backed on the chord with the curvature on the blade edge.
Theaveragedimensionsofthecrescents,areinverycloseagreementwiththeJBKcrescents,butthe rangeinlengths,isslightlyless,maybearesultofasmallersample.Theedge-backedmicrolithsare almost as close, average (slightly smaller) and lengths again have a slightly smaller range. Inviewofthesmallsample,andthepieceswithindeterminateretouch,theleft/rightretouchratiosof the OFARS narrow blade microliths are not too dissimilar to the JBK material.
Figure M3 shows the dimensions of all narrowblade microliths from MC, NMW and NME (a) edge- and backed and (b) crescents.
Other microlithsAlongnarrowblade‘point’backedalongalltherightedge17.0x5.3x1.3waspickedupfromNME.Another,21.0x6.2x2.0collectedfromNMWwithleftretouchseemedtobeintermediatebetweena broadbladeobliquepointandthelongnarrowbladepoint.Themedial/distalsectionofthebladewas used for both.
Table M1: Number of microliths found at each site
Wide based triangles, large
Wide based triangles, small
Table M2: Main attributes of all wide based triangles and oblique points
Average Length (mm)
Average Width (mm)
Average Thickness (mm)
Range of lengths (mm)
Large wide-based triangles
22.6 - 35.5
Small wide-based triangles
16.7 - 18.0
17.2 - 20.1
Table M3: Main attributes and the length range of all the edge-backed microliths and the crescents
Numbers of platform cores from the three sites in relation to burning and colour
Attribute / Site
Total number of platform cores
ThetwoscatterplotsbelowshowthedistributionofplatformcoresovertheNethermillsEastsite(NM 3 and NM 4): Upper: all platform cores, Lower: Single platform cores
Burnt cores were found at all three sites. Thecoloursofthecorestosomeextentdoreflectthoseateachsite,withbrowndominantat NMWandNME,andgreyatMC.HoweverthepercentageratioofgreytobrownatMCismuch higher than that of the whole assemblage.
(This page is still in preparation)
FLAKES AND BLADES
Flakesandbladescomprisethemainpartofthis collection of flints from Crathes, but most are broken.
The pie-charts below are based on the above tables.FiguresFB1(a)andFB1(b)representcolumns2-7toshowtheratioofflakestobladesateach cortex percentage at the two sites.
Table FB2: Nethermills East NM 3 and 4 CR 09
Flakes + Blades
0 - 100 (i.e. all)
95 - 100
66 - 94
33 - 65
5 - 32
0 - 4
Table FB1: Nethermills West NM 1 CR 08
Flakes + Blades
0 - 100
95 - 100
66 - 94
33 - 65
5 - 32
0 - 4
Flakes and blades: percentage cortexTheamountofcortexvariesfrom0-100%,andpieceswereassignedtotodifferentdivisions accordingtothepercentagepresent.Therewereextremelyfewwithaverysmallamount,andthe 0-4%seemedamoreappropriateplaceforthemthaninthe5-32%.TablesFB1andFB2givethe number of flakes and blades, and blade percentages in each cortex category at the two sites .
One flake with indeterminate cortex/flaw was included only in the 0 - 100% category
Three flakes did not have the cortex recorded, and have been included only in the 0 - 100% cortex category.
Flakeswerecataloguedas‘whole’,‘almostwhole‘(iewithfairlyminimaldamage),orfragments.Fragmentswerefurtherdescribedasproximal(p),medial(m)distal(d),proximal+medial(pm)or medial+distal (md).Blades were categorised in the same way, but blade-like fragments were identified as follows:blade fragment: length >/= 2 x widthprobable blade fragment: length >/= 1.5 x widthpossible blade fragment: length >/= 1 x widthOnly ‘whole‘ blades and flakes have been considered in the data below.
DatabelowcomparethetwomajorCrathessites.NethermillsWest(NMW)NM1and NethermillsEast(NME),NM3and4,butinformationfromNM2(theeasternendofNMWhas not yet been included.
Figure FB2 (a) and (b) represent columns 3 - 4 to show the relative numbers of flakes and blades within the different categories of cortex percentages at the two sites.
FigureFB3(a)and(b):graphsshowingthelength/maximumwidthofallwholebladesand flakes indicating the variations in size at the two sites.
Flakes and blades: Length / Maximum width
Allinalltherelationshipbetweennumbersofflakesandbladeswithcortexpercentagesarevery similar at Nethermills West and Nethermills East.
Figure FB4 (a), (b) and (c): bar-charts showing the number of blades at each maximum width.
Atbothsites,bladeswith0-4%cortexhaveanaveragemaximumwidthof10.4mmandadipin numbers at 7 mm; fairly similar overall, but with some differences at 8, 10 and 14 mm.
(This page is in preparation)
Balbridie A large Early Neolithic timber hallSouth of the River Dee near the B9077 and about 2 km west of the A957.Excavation directed by Ian Ralston and Nicholas Reynolds 1977-80I enjoyed working as a volunteer on this excavation too.Ian Ralston and Nicholas Reynolds, Interim Report 1981: Balbridie, Excavations 1977 - 80. Excavations at Balbridie Farm, Kincardine and Deeside District (NO 733959)
EARLY BRONZE AGE
This barbed and tanged arrowhead from Nethermills West, was found by D D Sabnis Jr in 1980. Another one from the Milton Cottage field shows activity continuing into the Early Bronze Age.
NEOLITHIC ACTIVITY NEARBY
Left Aquickroughsketchtogiveanindication ofthedirectionoftheflakescars.Theone ontherightisopposed,andthetwoon theleftaren’texactlyinthesame direction as each other.
RightCrathes,NethermillsEast(NM3and4).Graph showinglength/widthofwholeandalmostwhole bladeswith0-4%cortexandfeathered terminations.ThepossibleLUPbladehasbeen added for comparison
ThispossibleLUPbladewasbrokenatthe proximalend(nobulbofpercussionpresent), andwithapartialhingetermination/steppedor broken.Itisalsobrokenattherightproximal edge,somayoriginallyhavebeenwiderhere.It does taper from proximal to distal end.Oneofthethreeflakescarsisopposed–froma core with opposing platforms?Length:77.2 mmMax width20.6 mmMax thickness 5.5 mm
Possible Late Upper Paleolithic (LUP)
Theabovefind,noticedwhenthiscollectionwasreturnedtousaboutthreeyearsago,wasanunusuallylong, large flint blade when compared with other blades from the Crathes site.
AfterreadingthatLUPbladeswereverylong,andcoresoftenhadopposingplatforms(TBallinandC Wickham-Jones2017),IemailedphotosanddetailstoCarolineasshehadseenthisbladeearlierat ‘MesolithicDeesideDay’andhadsuggesteditmightbeLUP.SheforwardedmyemailtoTorbenBallin,and the reply came back that it was ‘definitely a possible Late Upper Palaeolithic blade’.
A find from the Sabnis Children’s Collection 1978-1980
ReferenceTBallinandCWickham-Jones,SearchingfortheScottishLateUpperPalaeolithic:Acasestudyfrom NethermillsFarm,Aberdeenshire.JournalofLithicStudies.PublishedbytheSchoolofHistory,Classicsand Archaeology, University of Edinburgh (2017) vol. 4.
Several pieces diagnostic of the Neolithic were picked up during the OFARS fieldwalking at Crathes
Warrenfield A second large Early Neolithic timber hall, and a Mesolithic pit alignment with later use in the NeolithicNear Crathes Castle, about 2 km northwest of NethermillsExcavation directed by Hilary and Charles Murray 2004-6Hilary K Murray, Charles Murray and Shannon M Fraser, 2009 , A Tale of the Unknown Unknowns: A Mesolithic pit alignment and a Neolithic timber hall at Warrenfield, Crathes, Aberdeenshire.
Early days during the excavation at the Warrenfield site
After having worked on the Balbridie excavation, I was very pleased to join the Warrenfield dig as well.