IV.

NOTE MATERIALN SO S FOUN KITCHEA DN I N MIDDE HILLSWICKT NA , SHETLAND, WITH SPECIAL REFERENC LONG-HANDLEO ET D COMBS. MILLEY B N COUGHTREY (STUDEN MEDICINE)F TO , EDINBURGH UNIVERSITY. PLATES XIII-XVIII. The bone implements, broken bones, portions of pottery-ware, and •shells, which are described in the first part of this paper, were found by me whils toua Shetlann n ri o t d durin monte gth Septembef ho r 1870. On the western side of the Mainland of Shetland, in the parish of Northmavine, and jutting out into St Magnus Bay, is Hillswick Ness, a peninsula connected witmainlane th h low-lyina y db g isthmus, washed e owesternth e Atlantin th sidy b e c Ocean e easterth n ,o n sid Uriy eb e Firt Hillswick.r ho centre th thif n eI o s1 isthmus Magnut standS d sOl s Kirk,2 between which, and on the western side of the isthmus, is a small voe, called San e southerdth wict a kd n ; an 3corne f thio r sout n voa -s ei crop of sand, from the base of which the present sea-beach shelves. This portion of the isthmus, like every little plot in Shetland, has had

1 Wick means a " voe" or "bay." 2 Keid's "Shetland,"p. 12. 3 Captain Thomas' Chav Shetlanf o t d Isles. NOTES ON MATEBIALS FOUND .IN A KITCHEN MIDDEN. 119

a special name given to it, viz., West Air; but it forms a part of the town of Hillswick. The outcrop is contiguous to the present sea-beach, and is distant only about 20 yards from the average high-water mark; it is necessarily expose wearine th o dtearing-dowt d gan n actiowavee th f tempestsn so i ; informeds wa an I d 1 that scarcel ywintea a r se t wha passee bu th ty b s encroache lane th t tha da n so t place, breaks dow washed nan s away more ooutcrope r th lesd thi s beef so ha sAn n. goin within go memore nth y of e oldessomth f eo t inhabitants. An idea of the savage force of the waves on this coast during storms may be conceived, when I state that at one place the waves uptear and upheave masses of rock, many of which are fully a ton in weight, and "carry them like so many pebbles to the shore behind." a Again, near Stennis fishing station,3 whic closs i h Hillswicko et e th , groun yard0 fro o 4 s e clif r mr th sdo fo f edg plentifulls ei y bestrewn with large pebbles, broken portions of rock, &c., which have been washed over the granite cliffs, which, are at least 30 feet in height. At the time I saw the out-crop quits wa e t i eviden, assisted ha t n n tha di e han ma th tf o d breakin poine on g t down t presentei t i a r fo , gooda d vertical section, which had been made sometime previously, for the purpose of forming a saw-pit. By means of this section, the various layers composing the outcrop were easily recognised, and were presented in the following order, commencing at the top (measurements are given in inches):— Loamy soil from which grass grew, Average depth, 12 . . „ 0 3 Yello . „w beac h. sand, Shell layer, , . . . „ „ 8 Large and small pebbles embedded in sand, ,, „ 27 The average height of the upper layer above the level of the sea was 7 feet. The layers varied in thickness in different parts, and were not quite horizontal dippet bu , d somewha south-weste th o t layere Th .s were undulated, and on account of progressive subsidence having been going on were slightly terraced. The shell layer was divided into two parts,

Reve th .y JameB 1 s Sutherland whoo t , m I no w expres sincery sm e thanks. 1 Vide "Art Rambles in Shetland," byReid, pp. iyan<120, and Sketches; also, Dr Hibbert's "Shetland, . 527pp " , 529.

Vide Reid, loc. . (Cannon tit.,20 . p ) (1.) 3 0 12 PROCEEDING THF SO E SOCIETY, FEBRUARY 13, 1871.

—one, the thicker portion^ about 8 yards In width, which merged into a thinner but more compact stratum of shells, that extended on either side for about 20 yards. The average thickness of the compact shell- layer was 2 inches; whilst the depth of the thicker part varied, being 10 inches in the centre, and 2 inches at its sides, thus tapering off from centre to sides; and on tracing it inland by digging, it slowly decreased in width and depth, and ended at the distance of 5 yards from the cut . ' • edge. It was in the thicker part of the shell-layer, of apparently artificial for- mation, that the above-mentioned' materials were found. The compact shell-layer presente usuae dth l segregated appearanc ancienn a f o e t sea- margin. In some parts three terraces were distinguishable, as if three of these ancient sea-margins were present instead of one; but these appear- ances could easil e accounteunderworkine yb th y b r e waterdfo th f go , ane neares sea-margid th ol te strith f npo havin gbity I sunb . t kbi think that the evidence was stronger in favour of depression than of upheaving. In the .other layers I observed no peculiarities. Shells.—The shells were chiefly large forms of the following edible mollusca :—Most abundant—Ostrea edulis, Linn. ; Patella (vulgata ?), Linn, (depresse othed dan r varieties); Cardium edule, Linn; Mytilus edulis, Linn. ; Littorina littorea (Turbo littorea of Linn.) Less abundant —Buccinum undatum, Linn.; Tapes virgine Venur o a s virginea, Linn. shelle Th s generally wer vera n eyi friablsofd an t e condition. Pottery.—TSo entire vesse founds piecee wa n lTh i aboue .sar n te t number; they var n size,y i e coarsenes n thicknessth i , e n i th . d of s an , material from which they have been.made. -Som nearle ear y J inch thick, and others only £ inch. Some are made of fine material, others contain coarse grains of quartz, &c. They evidently do not all belong to the same vessel. They have been subjecte actioe th o dfiref t n o , man thef yo m still retaining soot or soot-staining externally, whilst two pieces have in their concavity a thick crust, of a dirty black colour. This crust is composed chiefly of soot; and it is undoubted evidence of the fire having come into contact with the interior of the vessel or vessels, of which these two pieces form parts. On the external or convex surface of one of the pieces are some rude ornamental lines; these consis f straigho t t lines crossine gon another d forminan , g rectangular spaces. The prette yar y symmetrical, NOTE MATEBIALN SO S FOUN KITCHEA N DI N MIDDEN1 12 . d appeaan r to have been mad y someb e sharp-pointed implement, and not by the finger-nail. This ornamentation only extends over a portion of the piece, therefore it is probable that several such ornamented patches may hav eexterioe beevesselsth e samth n f e no o f o rTh e e. piecon s eha margins it s quite smooth; thiformes sha moute vessele dth th f ho . The pottery f whico , h thes e portionsar e , mus lookee b t d upo culinars na r yo other utensils, rather than urns. Bones, consisted chiefl f splintereyo d long bones, portion f ribso d san vertebras, short tarsa carpad an l l bonesscapulaew fe a d oss, an ,a innorni- nat f variouao s animals; besides, there whol werw fe ea bones pord an ,- tions of the skulls of various animals, several lower jaws and teeth. They belon ovisd gan .chiefls bo Ther o yt e als ar eo present adul yound an t g teet f Genusho elaphus e o boneth tw (staf r so o red-deer)r ge o on d an , Cervus capreolus (roe-deer) e boneTh f bird. o sd fishe an s s were also found. Since some authors have remarke absence dth f certaieo n portionf so e skeletoth n generall thesn i y e f interesshell-patcheso e b o t ty ma t i , state in a tabular form the number of the different bones present; and I have done so, keeping separate the whole from the broken bones.

Table Bones.of 18 broken portions of cranial bones. 13 inferior maxillae, with teeth, Chiefly of the sheep, the red-deer, and roe-deer. 4 entire first cervical, and 3 entire second cervical vertebrae. 15 broken dorsallumba3 d an , r vertebras, Chiefly sheep and red-deer. 5 broken scapulae; 2 gnawed,1 All of sheep, except one, which was of red-deer. 7 broken parts of the pelvic bones ; 3 gnawed. 9 humeri, broken ; 1 bearing marks of cutting; 1 gnawed. 15 radi ulnaebeed d an iha n6 ;split . Tibiae, 2 whole; 19 broken and split; 1 belonged to red-deer. Tarsal and carpal bones, 5. Metacarpa phalangeald an l , 22 , 8 of which were partly split and gnawed. 122 PROCEEDINGS OF THE SOCIETY, FEBRUARY 13, 1871. broke4 10 n portion ribsf so gnawe5 ; d bearin;6 g mark cuttingf so . honew fe birdf t suffisbrokeno fisheso d A fews o no s san s t - a ,d bu , an , cient to determine species.

Some of the bones are entire, others splintered, broken, and cracked, as extractioe th r fo f i f marrowno . possesw Nofe a t s teeth marks. There e severaar l other mark f cuttingo s , that must have been producey db human agency. Two pieces of bone bear marks of fire; one piece, about an inch long (a portion of a rib), is charred at one end, and the rest is fire-cracked. The other piece is also about one inch long, is conical, and looks like the core of a sheep's horn. It is burnt quite black into char- coa f Kiitimeyer'sI l 1 observation e correctb s I ,woul d infer from them alone that all the above bones are those of domesticated animals. Bone-Needles, Awls, $c.—One narro rounded wan d bone implement, 3-2 inche f n aboulengthi o s e thicknesd th t an , a quill f o s , broken at both ends, but narrower at one end than the other. It appears to be part of a needle. Two awls, made of splinters of the long bones of some animals, very rude, each having one end tapering, rounded off, and pointed splinteA . f boneo r , wite extremiton h y ground dowa o t n bevelled edge; probably been use scrapinr dfo g purposes. Portioe th f no ulna of some animal, with one end sharpened or pointed, hut broken: thibeard en s s mark f cuttingso . e tibisheepa Parth f af o o t, broken above the middle; the splintered end has been ground down and polished, til simplla e blun remaind en t transversa bonse s ;th eha e mark neae th r articular end; this is either a scratch or a cut.2 Several other splinters of bone which may have been used as awls, but these, though pre- served, I consider " doubtful," and have therefore placed them among the broken bones. Perhaps the most important and interesting objects found in the shell-patch were four of the long-handled combs, made of bone. They lay quite close together. These are described at length hereafter. What was this shell patch! Was it in connection with a Pictish broch r simplo , n isolateya d shel a broce faclf th o moundth o t s A ? having been found at this place in the memory of man, evidence is

1 See Lubbock's " Prehistoric Times," 2d edit., pp. 195 and 196. Josepr M h Andersen considers this implemen havo t t e been use a.puncl)s da r o , 2 for some similar purpose, since it resembles very closely such implements. NOTE MATERIALN SO S FOUN KITCHEA N DI N MIDDEN3 12 .

positionwantingw lo t fro s Bu mit , . from the presenc a kin f f o o de stone in certain adjacent buildings similar to what had been used in making a Pictish broch about four miles distant, near Priest-Houland (which I examined), and to such as exists in the Clikamin broch; from face th t tha l long-handleal t d combs hitherto found abov e Firteth f ho Forth have been found in connection with brochs; also from a want of completion chaie inth .brochsf no , unless thi s includedsi , I thins i t ki most probable tha broca t oncd hha e existed here, but, owing partlo yt the ravages of man and of the sea, the stones have either been washed awa remover yo r buildinfo d g purposes r botho , ; and l tha haval e . w t e left is the shell .patch or kitchen-midden of such broch. That buildings once existed there, old inhabitants can affirm; but they say these were fishermen's cottages or huts, and that they were dismantled by the sea many b d . an I ought also to mention that the old kirk of St Magnus is supposed to have existed at a spot not more than 30 to 40 yards south of this patch thad burial-grouns an ,it t d approached even nearer. e Therb n eca no doubt that these buildings d kirk,an bot ,t werhhu e entirely uncon- nected with the shell patch.

THE LONG-HANDLED COMBS. . I now proceed to give the results of a detailed examination of these implements. Although numbers of combs of this form have been found and are duly reported in the proceedings of the various Antiquarian and Geological Societies, yet in no place is there given a detailed description of them. Thinkin presene gth t fin gooda d opportunit contributinr yfo a o gt more exac minutd an t e knowledg f theso e e combs I ,hav e endeavoured to supply that want by giving a somewhat minute and detailed descrip- tion, embracing not only their general form but also certain peculiari- ties in form; the materials from which they have been made, their orna- mentation, comparison with other combs, and interesting data connected with their manufacture, use, and distribution, in the hope that such a description may be of service, not only for future comparison, but in affording another link of evidence that may aid in computing the age of e Pictishth . brochsfurnishinn i d an , a littlg e more insight regardine gth habits and state of civilisation of those who used these implements. 124 -PROCEEDINGS. OF THE SOCIETY, FEBRUARY 13, 1871. The general description is founded upon the characters of the combs in the possession of this Society, hut I have not scrupled to avail myself of the information derived from hooks and plates, in all cases acknow- ledgin source informationy gth m f eo . I wish especially to refer to the very great obligations I am under to Mr Joseph Anderson, keeper of the Society's museum, for his courtesy in allowing me to examine the various specimens of the combs in the museum, anmucr dfo h valuable information concerning the othersd man ; indeed, it was partially through his advice and that of Professor Turner that I was induced to write this paper. These combs hav gived eha theo nt m various names, accordine th o gt place in which they -were found-^-such as Pictish broch or burgh combs, Kitchen-Midde Hut-circld nan e combs, long-handled Koman Camp combs, bu I thint mose kth t expressiv mosd ean t genera l i steral thaf mf o o t Long-Handled Combs. The made yar e eithe f bono r r deer'eo s horn, more frequentle th f yo latte r; Kelle figures madrha e f don yew-woodo e sai,e whicb o dt y hma belon thio gt s clas s nearlcombsi f so t I inchey5 . s lon 2y ^gb wide; teets it h 1numbern 0i 1d £ an ,inc h long. General Form.—Elongated, posses expandeo tw s d extremitiea d san constricted s thic i middlbluntd d an ke othe en r bodyth ,eo e s i r on ; bevelled and toothed. The body and the blunt extremity together form, the handle, and to toothee th d extremit havI y" e applie tere dth m of'dental end. surfaceo handle tw Th s eha s whose margin e roundear s f morof dr o e less in different specimens; one surface generally possesses a variable amount of convexity, whilst the other is often flat. convee Th externalr xo surface correspond exterioe come th th o st f - o r pact outer wall of a long or flat bone, or the hardened external layer of flae r internalhorno th t t bu ; surface present charactere sth longia f o s- tudinal section through the cancellated texture of a bone or horn. e comThth ef bbluno isd usuallen t thickess yit t part ,mosn i an , dtis specimens, slightly concave from corner to corner, but rounded off from surface to surface.

1 Lake Dwellings, plate 5, fig. 21; text, page 34. NOTE LONG-HANDLEN SO 5 12 D COMBS. • The above may be taken as a typical form of the handle, but there arc some combs which vary somewhat. Peculiar Forms.—In some the concavity at the blunt end is present to markea come d th degre b n i r Benjami foun s M e a y db n Neeve Peact ha Kettleburn, 1854n 1amongsi w no , Kettlebure th t n museume relicth n si , handle whicn th i f lobedistae o s e i h th d den l lik fisha taie f eth o l. Another Caithness broch yielde dcoma b whose handle, from neae th r teethe baseth f ,o s tapers sharpl a point o yt , whic s surmountehi a y db large circular knob. This was presented to the Society by Mr George Innes 1783n i , . Madsen2 mentions severala simila f d o figure an e ,re on th fors o mt above, but much smaller in dimensions. The Eoman camp combs e similae above3ar th o t r, except that they generalle ar y surmounte squara y db e knob. Mr Farrer* found some of these combs with cylindrical handles at a broc Burrayn hi , Orkney. combe Th s foun Kent'n di s cavern similae ar 5 forshoe-hornn a ri o mt , broad at the dental end and narrow at the blunt extremity. On thf eo e comb sr PetrifounM y n Lingroei b d w broch possessea s similar obtaine e form on doe o s e , sth d froe circulamth r buildin t Uisga t r GordonM y b . Several of the combs possess one or more holes in the handle near to the distal end, probably for the purpose of attaching it to the person by means of a thread or cord, so that it may be carried about by its owner. Such a practice is common among the Esquimaux, North American Indians, and Maories. Examples of these may be seen in the case of one of Kent's cavern specimens comb e botn th i ; f ho s Eev e founJoassr th brocM .a y db n 6i h 1 A Eeport on the Ancient Remains of Caithness. By Joseph Auderson, Loc. Sec. of Anthropological Society. Memoirs of Anthropological Society, vol. ii., p. 227. 1865-66. Also Proceedings, vol. 6, p. 264; also plate XIII. fig. 2. 2 Afbildninger af Dtmske Oldsager og Mindesmierker ved A. P. Madsen, p. 6, Hefte xiii; also plat . figeI . 1 . 3 Proceedings, vol. vii., p. J24. 4 Ibid. vol. ii. p. 6, and page 157. : 5 Reports Brit. Ass. Committee for 1865-67. 6 Viile Joassr M ' Pape Archaeologin i r a Scotica, vol. v . 126 PROCEEDINGS OF THE SOCIETY; FEBRUARY 13, 1871. near Dunrobin, Sutherlandshir come th ebn i r Gordon • founM y db n "i 1 a circular building at Uist" (a broch) during 1853; in a comb found at Moosedorf—in this com holesbo coma thertw n e i ;b ear foun t Nussa d - dorf , a bronzals n i o e comb foun n Terramarai d e lasTh t .thre e ar e mentione figured dan Keller.y db 2 The dental end and the teeth.—Thes& present great variations in all the combs which I had the opportunity of examining. In those made from sections of deer's horns, or from splinters of the long or flat bones, more especiall thosn yi e made fro e bonemth f whalesso e dentath , d en l is bevelled on both sides, but more on the flat or soft side than on the convex side thosn I . e possessing cylindrical handles, wher obliqun a e e sectio s beenha n made throug "e tyne,th e shaf hth f o t" bevellins gha only been made from one side, and that agrees with the section. The bevelling generally commences on the surfaces of the handle, from on einc e a-hald incon h an ho t f fro teethe basee mth th f so . dividee b teete y Th hdma intmarginaa d ocentraa an t l se lone ; each attachen toota s hha d thic ka free-pointe basd ean d apex. e Thosth f o e central set each possesses four surfaces and four borders; two of the sur- face externae sar internad an l l corresponding respectively wit e exterhth - nal and internal surfaces of the handle. The remaining two surfaces are opposine th laterar go l surface f eaco s h toothe fouTh r . borders bound these surfaces, and are best marked near the bases of the teeth; they are rounded off near the apices. This is well seen in the larger Kettleburn comb. marginae Th . l n teethnumberi o ,tw , agree generally wit e centrahth l set, except that they are commonly more massive, and taper more than the other teeth from base to apex, and are usually a little shorter. Each tooth tapers laterally from bas apexo et als d oan , from externao t l internal surface, but the tapering in all the cases except one, was greater e latteth n ri diameter tha e former nteete th Th h. var n numbeyi n i r different specimens, from six to sixteen. They vary also in size and form, some being shor cuneiformd an t onld an ,y one-fourt incn a f hn i ho length, others long and parallel, extending to If inch. e interdentaTh l notches (betwee e teeth nth e variabl ar ) n widthi e , 1 Proceedings, vol. iii., pp.122-124. 2 Keller's Lake Dwellings. NOTES ON LONG-HANDLED COMBS. 127

some being so narrow as to scarcely admit the large blade of a pocket penknife, others wide enough to allow a blade three times the above thickness to pass between the teeth; the width of the notches in speci- mens doe t depen numbee sno th teete girte mucn o do th s th s rf hn a o h o of teeth present at the dental end; some are less than one-eighth of an inch in width, others one-fourth of an inch. The " mode of implantation," if such a term is here allowable, is also important. In the typical specimens, the whole series of teeth, from their base theio st r apices, usuall quite yar e parallel y f theri an d s ei an , convergence, it is so slight as to be with difficulty perceptible. In a few, the series diverge so that the width of the dental end at the apices of the teet greates hi re basesthath t na . Bot convergene th h divergend an t t forms have their analogues in the two kinds of combs used by the Esquimau presene th f xo t Hillswice dayth f . o Againe kon combsn ,i e th , teeth ait obliquel handlee se axie th f th s o e o y t ; this also occur othen si r combs somrudn e I th .f e eo specimens, fro teetbasee e mth th f hso grooves proceed along the bevelled portion of the external and internal surfaces of the handle ; these grooves are continuous with the interdental notches, where they are best marked, but they become shallower and more indis- tinct at the a-dental end of .the bevelled part of the handle. They are slightly narrower in width than the interdental notches; they vary in length, not only as compared with one another in the same comb, but also n differeni t combs d the e moran , yar e deeply indente e internath n i d l cancellated or softer surfaee of the comb, than in the external harder e combseee shelth b n man I nf faina y o .l yma t ridg ridgesr e o th t a , base of the groove, dividing it into two or more secondary grooves, and most commonly when these secondary grooves are present, they are not parallel to one another. Again, in many combs, and especially in those that possess the grooves, opposine th g surface teete th hf so have oblique marks, running witha greater or less amount of obliquity, from one side at the base, towards a portion of the opposite side near the apex; those commencing at the base nea e internath r l surface, crossing those which e basebegith t ,na near externae th l surface vicd an ,e versae interdentaTh . l notche f combo s s possessin above gth e groove obliqud an s e marks, have, withou a singlt e exception, clogged and unfinished bases;'a small conical projection of 8 12 PKOCEEDING SOCIETYE TH F SO , FEBRUARY 13, 1871. bone having bee e cutting ne base th lefth n f eaci tsn o i , h interdental notch betwee teethe nth . This contrasts strongl e welyth wit ln i finishe he whase e d w t combs, .possessing sharp and clean teeth without marks, interdental notches well t a handledown cu d an , , whose bevelle s devoii d f grooveso en d . If careful examination of the teeth of these combs be made, there will, in almost all specimens, be found transverse markings on the surfaces of the teeth, nearer to the apices than to the bases; these are marks caused by the use of these combs, and will be considered in another part of this paper. Microscopical Structure.— I hav t Aseye only examine structure dth e come oon f b unde microscopee rth indebtem a Stirlingr ;I M e th o dt f ,o Anatomical Museum, Edinburgh University, for having kindly mounted for me the sections. They were taken from one of the Hillswick combs, and compared with some sections made from the bones of a long-armed Cape whale, which, from having been subjected to similar influences, and being in a similar condition, permitted a comparison being made. The late Mr Quekett, the distinguished histologist, in a note to the " Archaeological Journal," about the comb found at Kettleburn, said that tha spheroidao t combtw d an , l bone balls, which were found along with it, were made from the lower jawbone of a whale. Now, fro examination ma thesf no e section microscopee th y sb , though I could not positively assert that the comb examined was made from a whale's bone, still less tha t cami t e particulafroy man n ca r boneI t ye ; say, that the appearances presented by the different parts of the osseous tissunome th simila e n bear i thoso rt e see n sectioni n s made froe mth boneCape th f eso whale. With this exception, tha decae th ts gon yha e on for a longer period in the comb, we have the canaliculi more molecular, anlaeunse dth e somewhat more blurred fro depositioe mth earthe th f no y particles Haversiae th t ye ; n system groupine th ,e lacunse e th th d f go an , character canaliculie th f so , have same th ' e general appearanc bothn ei . Manufacture.—-In treating of the mode in which these combs have been made, and of the kind of implement, stone or metal, that may have "been employe foro dt teethe mth I shal, l limit mysel bringino t f g forward certain facts, and of drawing provisional inferences from these facts. littlThere b n e eca doub t tha bonee tth deer-hornr so t ofs ou whic h these NOTES ON LONG-HANDLED COMBS. 129

combs were made must have beecomparativela n i y freshe stateth r fo , structure of horn or bone becomes more and more brittle, in proportion animae asth l matte losts r i weathering r o , bone d e eitheol ag f si y rb d ;an had been used, in addition to the greater difficulty of cutting long teeth without breaking them, we would expect cutting grooves, where present, with chequere broked dan n margins I hav y ecase an ,th n e i whict no s hi seen. Onl kindo ytw s of bones were used fla,e viz. th tlon e bonesd ,th gan . In the former, there is a thick, compact, hard, and ivory-like portion exter- nallycancellatea d an , softed dan r texture internally thosn i d e an ,comb s made from splinter lonf so g bone compace sth t tissue predominatesn I . the tabular bones, we have a much thinner compact portion, but a more closely cancellated centre thosn i d e ;an comb s made from these flat bones, cancellatee th d texture predominates, though her hare eth d compact por- tiobees nha n taken advantag . eAlthougof h ther e threar e e kindf o s horn t onlkine ye , yon d seem havo st e been used, viz., that belongino gt ' Cervideaa," and when the horns of these animals have been stripped of "e velvet,th " the have nw e essentiall ybona y structure, consistina f go spongy centre, the spaces of which gradually become smaller and smaller, as it merges into the compact tissue of the horn, there being no sharply defined line between the compact and the cancellous texture. The bone composing these horns approaches somewhat in properties the bones of young animals, whic e softehar d bonesr ol tha e .th n Hence y thima s account for the fact, that horn seems to have been employed in preference boneto manin , y cases sinc eit coul morbe d e easily cut. Concerning kine instrumenth f do t employe thein di r manufacture I ,hav e como t e the conclusion that saws, or some implement with a filing or serrating edge, were used in cutting the teeth of these combs, for the following reasons:— If the grooves running from the bases of the teeth along the bevelled e comparth e f examinedbb o t t wil,i e seeb l n thabottoe th t f eacmo h groove is equal in width to its "lips," whereas, if a knife had been employed e groov th e bas th f ,o ee would have been much narrower a crostha e lipsd nth san , sectioe groovth f no e woule dth hav d eha lettee forth f mro V.e limbth , s beine e groove th sidee th gd th f o san , angle of junction the base. Besides, if we attempt to cut a new bone, and more especially a brittle old one, with a knife, the groove made is VOL. IX. PART I. I 0 13 PROCEEDING SOCIETYE TH F SO , FEBRUAR , 1871Y13 .

neither so well' defined, nor so clean as one made by a saw, the compact tissue splitting and cracking, owing to the great pressure of the knife's edge, and the lips of the groove being broken and jagged. . The hardness and compactness of the bone or horn cut, the characters of the above grooves, alsopposin e thosf oth o n eo g surface teethe th charactere f o s,th s cuttinge ofth markbrokee d th san n so n bones othed ,an r implements that have been found along with these combs, the comparative width of the interdental notche e typica mostth n si f o l combs e presencth , f flino e t saws among the e lakfindth en i sdwellings , where some combs belonging to this class hav ee presen been'foundf th sawo n i e s t us day n e i ,th , cutting the teeth of combs and other things,—all tend to show that saws have been employed. I am aware that as yet no saws have been found in the brochsj and as for flint implements, the generalle yar exceptione yth , though some were found in the black mould over the stalagmitic layers in Kent's Cavern, in conjunction with comb thif so s class I als. o know that saw metalf so f o , the bronze and iron periods, are very scarce, if any at all exist, and that the interdental notche f somso e narrocombo s preclud o t e s war sa e ideeth a of stone saws in some cases having been used. Yet it must be borne in mind, that if metal saws were employed, and I'believe they were, it is more than likely that they were iron ones, for there is no doubt but that e earliesomth f o er dweller e brochth f so s were co-temporary wite hth Eoman invaders. Consequently f suci , ha period has elapsed since these saws were made or employed, or even if only a cen- tury had elapsed, knowing as we do the physical propertie influence th ironf so d an , e and conditions which any iron implement, especially one. fine enoug o makt h e th e groove interdentad san l notche comba f so , Pig. 1. Natural Size. 6 ' . ' varioun o e cuttinge oth rse s bonee w s s found amon e brocgth h relics, must have been subjecte o sinct d s eit deposit I ,nee d have little fea f beino r g stigmatise rass da assuminn hi g that suc implemenn ha uses thein dexistwa d i di d t r an ,manufacture . No s suci r assumption a h n unfounded conversane on ,y sincan o tt e with the subject, it will be seen that the very nature and disposition of NOTE LONG-HANDLEN SO 1 13 D COMBS.

the materials found in these brochs would favour the rusting away of any iron implement deposite n themr i dindependen Fo . e moisturth f o t e which naturally must exist in and about buildings lying in a low situa- tion, either built on the arms of the sea, or partly in some inland loch, or midse th somf n o i t e undrained mossbo r morassgo have walle w , eth f so such buildings partly broken down huge th , e stones jumbled againse on t another, leaving underneath or between them little crevices, in which carbonic acid and moisture may collect; we have a soil infiltrated with, and indeed partly composed of, organic substances, such as broken and splintered bones e remainth , f feast o sd slaughterings an s , certainle yth most favourable conditions for producing rust. Indeed, Mr Farrer found, in the broch of Burray, Orkney, a small por- tion of a double-margined comb, with remains of an iron knob or rivet projecting fro centrs m it alse H ei obtaine(fig. 1) . d fro same mth e place some button bonef so , with iron knobs projectin themf o fro d t an mgou ; another anothet brocgo e hh r double-margined comb, also rivetted with iron. These were found along with long-handled combsbotn i d h casean , e sth iro onls ni y rusteendse th t d.a Here, althoug imperfece hth t conducting power of the bone may have interfered with the species of voltaic action which is produced in the rusting of iron, it would not wholly prevent it, yet, by protecting the shaft of the rivet, it did tend to preserve the iron. l Again r JohSi , n Lubbock2 mentions about several implements that were found at Thorsbjerg, in Slesvick, in which " the iron has been almost entirely removed," owing to chemical action, and the handle of bronze and wood are perfectly preserved." But if the negative evidence of the us f iroeo n t permissiblesawno s i s l ,al thet a t ndoee i t see m no smo t improbable thae margith ta shel f no l might have been used; indeed, ther lote naturaf ear so l articles which would subserv purposee th l eal f so a saw whicht ,bu , unti I lhav e further investigate subjecte dth I forbea, r mentioning. That the cutting implement was made of metal is rendered more facte likelth , y ythab almosn i t t every find bronz irod enan articles have been foiind associated with these combs. As yet, the chief difficulty facte th lie , n si thaalmoss i t ti t impossibl separato et e what implements

1 "Miller's Elem Chemistryf o . 580d editionan h . 9 4t , 57 . , parpp . ii t -»-« Prehistoric Times,. 9 editiond , "2 8 . pp , 132 PROCEEDINGS OF THE SOCIETY, FEBRUARY 13, 1871. belongs to one or what to another of the successive generations who occu- pied these brochs. Concernin e mannegth f cuttinro g them, thery ema same th e b difference t no f opiniono s . Mr Joseph Anderson1 says, " the idea of the artist who fabricated the comb from a shank bone was evidently to make a rude imitation of the human hand, wit s fingerhit s outstretched, t certainli d an " y does look t abouse . e h it t w ho e se s au thif bees d i t casee sha n Le th .

Fig . One-hal2 . f Natural Size.

Regardin e instrumength t LainProceedingr figuree th M y n gi b d f so f thosSocietye o th e eon combs 2a s partly made t wouli , d appear fro, mit carefua d an l e studotheth f yro combs, thae handlth t formeds ei d an , the dental end bevelled to some extent, before the teeth are sawn or cut; whed an n thes cute ear , the sawe yar n from both surface e comb,th f so 3 just in the same way that a tyro-carpenter is often seen using a saw of a certain breadth, with a back to it, to cut through a log of wood, the dia- mete whicf ro mucs hi h greater tha breadte nth e saw th f . ho This sawing from both sides would account for the oblique grooves on opposing sur- teete faceXVIIIth L f h(P o s . commencin, fig2) . g interna e botth t ha d lan external surfaces, and also for the grooves on both surfaces of the bevelled portion of the comb, being deepest near the bases of the teeth. somn I e combs, afte above th r e rough cuttin dones gi , the usee yar d with- furthey ouan t r finishing n othersi ; e interdentath , l notche e sawar s n clean, that is, the small portion of bone at the base of the teeth, which is left 1 Anthropological Soc. Memoirs, vol. ii. 1865-66, p. 228. 2 Vol. vii. p. 83. Vide also pp. 67 and 68, or fig. 2 above. 3 This was suggested to me on reading about '' a deerhorn handle rudely sawn at one end and rounded at the other, the sawing having been done from opposite sides," found in a Plot's house at Old Stirkoke—Vide Anderson, vol. ii. "Anthropological Memoirs," 1865-66 . 231p , . NOTES ON LONG-HANDLED COMBS. 133 in the unfinished specimens, is removed, so that the plane of the base of the interdental notc t righa hs i t angle internae th o externast d lan l surface.1 The grooves upon the bevelled portion are effaced by further bevelling, the margin of the teeth are rounded off, and the comb is ornamented or polished. The handle is hollowed out by means of shavings cut2 away by a knife or some sharp-edged implement, and in some grinding seems to have bee e Kettleburth n resorten I . ndto Comb (fig. 134p , .3 )a straight line had evidently been cut before the teeth were made, on a level with the intended bases of the teeth, showing, probably, that these people mapped out the length of the teeth before they began to cut them. Ornamentation f theo Combs.—The ornamentatio f theso n e combs varies consists—1d an , f singlO . r paralleo e l straight lines,3 either run- ning directly across the body of the comb, or crosssing one another, or both combined; found chiefly in broch combs. 2. Of concentric circles,4 either groupe r distinctdo , sometimes confluent thosn i s a e, found Roman i n parallecampsf O . 3 .l arranged zigzag lines Kent'n i s ,a s Cavern specimens carvef O . 3 d figures. somn i s a e, whic hn Mexicoi wer t go e , Britise th n i h w Museumno ane dar . The Eev Joasr M . s Carn-liate founth n di h Broc hcomb,a whicn 6i e hth external surface possesses two arcs of a circle running crossways, the ends of each arc being at the margin, of the surface, and their convexities facin another e beinblune c gth on ar t e ga t endon , dentae othee th ,th t ra l othee end th opposite f th r;o betwee d d an e en c extremite ar n th e on f yo arc, stretches a straight line, which is broken at its middle by a hole. The ornamentation generally, but especially that of the broch variety, e conveth n o x s i surface handleth f d moro e an , e frequently grouped neae dentath r l end. Exception occur—foo sd r example e largth en i , Kettleburn Comb ; and in No. II. of the Hillswick Combs (PI. XIV. figs. 1 and 2) there is slight ornamontal marking on the flat side also. I fancy, perfectlt thougno m a y hI sure, that ornamentatio botn no h e sideth s si rule rather than the exception, in the Roman camp variety, and those true

1 For explanation, vide PI. XVIII. figs. 1, 2, 3, and 4. 'l "Wel markinge l th see handlee y nb th n so , vide addenda. 3 PI. XIII PId .. an fig XIV, 1 . 2.d ;figs an als1 . o . 134 p fig , ..3 * PI. XIV. fig. 3; PI. XV. figs. 8, 9 ; also PI. XVI. fig. 4. 5 PI. XVI. fig. 1. 134 PROCEEDINGS OF THE SOCIETY, FEBRUARY 13, 1871.

long-handled comb se Swis ^oth f s lake dwellings e brocth n hI variety. , in all the combs I have seen, the markings are imore or less rude and asymmetrical, d thean y appea havo t r e been made wit hknifea n i r fo , some, but especially in the Hillswick Comb, No. II., the knife evidently slippe s f makino ha t dac gwhilsstraigha e th n i t t lineanothed an , r line has been made at an angle with the previous line. Again, in the above comb, where we have a series of three closely situated parallel lines, crossing one another diagonally, and these again between two series of three lines, which run transversely from margin to margin of the comb, in some of the series only two lines are present, and a faintly marked portion of a third. In another series it commences with four linesj but soon the fourth line is given up, as if it had been an abortive attempt.

Fig. 3. One-half Natural Size.

Kettlebure th n I n Comb (figt consisti ) .3 s eithe f paralleo r l lines run- ning acros come sth t righba t angle s lonit go s t portionf axiso r ,o sucf so h lines placed at irregular intervals. Where lines are present on both sur- faces, they are not continuous one with the other, but appear to have been made distinct fromarke reste mTh th .s bear character knifa f so e having been propelled backwards and forwards like^a saw in making them. Some of the broch combs possess here and there a coarse rude mark; s smootonit en o comh s surfacha b shalloo tw e w lines crossing one1 another, partly effaced variete Th . y whose surface ornamentee ar s d with dots and concentric circles occur chiefly in England, and universally in connection with the so-called Roman camps.

Addenda, comb 9. NOTES ON LONG-HANDLED COMBS. 135 Where, Found.—As generally indicated in my preceding remarks, these combs have been foun brochn di similar so r buildings kitchen ,i n middens or shell mounds, in bone caves (as Kent's Cavern), and in chalk pits ; and their characters differ wit charactee hth place th f ero where they were found. Comparison with Analogous Implements.—The true typical long-handled combs do bear some resemblance to those at present in use among bar- barous^ and semi-civilised tribes, such, for example, as the North American Indians, Hottentots e Eskimo th layind d An an g, . aside differth e - ences in finish and ornamentation, as being too difficult to determine what is peculiar to these tribes and what has been contracted by contact and intercourse with civilised races wile w , l consider their agreements disagreementsd an . formo Thertw f Eskim o se ar e o n combsi whic e e wholth hon , e series of teeth converge from base to apex, the other in which, they diverge, that is, the apicilar end of the dental portion is much wider tha s basilait n e convergenr th end n I . t teethed comb, whicy b s hi e mosfath r t commo e two th e teet f th e largeo ,n har d mor an r e tenu- interdentae th d ousan , l notchewideo n e r ar sbetwee e apicee nth th f so teeth than the betwixe yar basese tth . The comb itsel thins i f , wided ,an surfaceo tw quite s shortit ar s d e flaan , t (vide PI. XVII 14)d . an figs .3 1 . In the second variety of the Eskimo comb the body, the handle, e tenuit e samee th teet th t th the d s f hi bu ,yo an y diverge BO thae th t interdental notches are much wider between the apices than at the bases interdentateethe e oth f Th . l notche true typicar th eo f so l long-handled comb als oabove diverge th teet e n notei th e t caseshs ar a , bu ; , long, thin, and of equal thickness throughout, but possess the characters which I have before enumerated; then betwee e brocth n h comb d thesan s e combs there is a great difference in the general form and in the charac- ters of the body. e combs'iTh n use amon Norte gth h American Indians approach more in form those found in the Danish shell mounds. They are smaller in size, and comparatively narrower than the Eskimo combs. Mr Edward S. Stevens1 likenese referth o st s existing between these long-handled combs and some implement e Eskimsth r "scrapiny n receni b fo o e us t g fat,

i Flint Chips, p. 64. I 6 13 PROCEEDING SOCIETYE TH F SO , FEBRUAR , 1871Y13 .

&c., from the back f skinsso e Eskimth ; o tool made sar f woodo e , with the sharp claws of birds lashed to them." He says,—" In the Christy Museum there are examples of these; in the same collection there is a Basuto tool, used for a similar purpose, the short thick teeth of which are of iron, bound to a wooden handle with twisted fibres." Here the likeness exists only in the general form; for, as before stated, the broch comb made sar e fro msingla e piec f boneo e ; whilst thes e madear et no , onl f differeno y t materials t als f bu severa,o l piecese SocietTh . y possesse mose s th on f te o typica l Eskimo combs,Britise th d h1an Museum possess all the forms of the above combs. Difference lika f eo s natur thoso et e subsisting betwee Indiae nth d nan Eskimo combs separate the Danish combs from the typical long-handled combs. Such as those found in the Danish shell mounds, and figured Worsaa6.y b 2 Som f theso e e wide ar ed approac an , n charactehi e th r ordinary Roman or Saxon single-margined combs; others are more allied to the Eskimo and Indian combs. The latter are chiefly bronze combs. Professor George Stephens, in Ms grand work,3 mentions and describes a bone comb whic s foun mossa hwa n di , along with some dozene th f o s same form. Thea two bone combs were found at Ickleton and Chester- ford, along with some Eoman coins4 (Plate XVII. fig. 12). Mr F. W. Wakeman, speakinn 5i Irise th hf go antiquitie Saxoe th f so n period, also mentions and figures several of the Irish combs; these are much orna- e abovmentedth l e Al combgeneran .i e sar l ford characteme an th f o r teeth halfway between the Eskimo combs and the ordinary Saxon single-margine museue d th comb n I m. attache Joho dt n Knox's house, in this city, are two wooden combs identical in general character with those figured by Stephens ut supra.6

1 Proceedings, vol. v. p. 126 ; also PI. XVII. fig. 13 of this paper. 2 Afbildninger fra Det Kongelige Museum for Nordiske Oldsager, p. 15, fig. 63 ; , figs45 . . 179-181p . Northerd 3Ol n Runic Monuments, Scandinavi Englandd aan , . par305p . ti . 4 Archseological Journal. 17 ,. volp . .vi 6 Collectanea Antiqua, by Charles Eoach Smith, Archfeological Jour. vol. iii. p. 43. 6 I am indebted to Dr John Alexander Smith for the knowledge of the above two combs. Where they came from is unknown, but it is supposed that they came from som thef eo ' tribes aroun Calabard dOl . NOTE LONG-HANDLEN SO 7 13 D COMBS. Among the combs found in Swiss lake dwellings, those two figured by Keller,lookee b Ixiii. L y P dn ma , 1i upo beins na g more connected with the Eskimo combs than the typical long-handled comb, the only marked difference between them and the Eskimo consisting in the fact that the f severamade o yar p eu l piecee interdenta th f bone o sd an , l notches are much narrower. The Moosedorf yew-wood comb,2 the three bone combs found at JSTussdorf,3 and the bronze comb of Terramara,4 also the cast6 of a bronze comb figured and described by Keller, I regard as being intimately connected, both by general form, characters of teeth, sized an , wit long-handlee hth d combs. Similarly, though much smaller n sizei , those combs mentione y Madsen,b d 6 some possessing three, others four, others seven teeth, and of which he gives a figure of one,-— these, too, mus lookee b t d upo varieta e long-handle s th na f yo d form. Most of the above combs have evidence of use in the presence of small transverse grooves on the surfaces of the teeth, similar to what I pointed out as existing among many of the broch combs. f placo t o mentiont e ou e b t , thano y t ma certaie so-calle t th I f no d

back-scratchers hav certaia e n resemblanc broce th ho t e combs; indeed, some bronze, articles, e abov 7 calleth y eb d e name th n i , whicw sa I h antiquarian departmen r WilliaSi f o tm Brown's Museum, Liverpoole ar , identica n generai l n theii l ford r man teet thao t h t comb whics wa h presente Societ e limer th M o d1783t y n s yi b . The typical long-handled combs have peculiarities in form and orna- mentation, whic coincidente har t onlno , y wit kine hth placf do whicn ei h they were found, but also in their distribution. The English combs are as distinct from the Scotch combs as the latter are from the Eskimo. 1 Loc. cit. . 229p , pi. . Ixiii(Worsaa15 .d figsan alss 6 . eha o figured these.) 2 Loc. cit.p. 34, pi. v. fig. 21. 3 Loc. cit.p. 114, pi. xxviii. fig. 8. Vide also addenda of this paper, and 4 Loc. cit. . 299p , pi. Ixiii. fig. 7 . PI. XVII. 5 Loc. cit. . 299p , pi. . 9 lixg fi . « Afbilduinger, &c., loc. cit. part xiv. pp. 6 and 7 ; vide also PI. XVII. fig. 4 of this paper. f opinioo m a nI 7 that these e combswilb o l t f properltur i t , nou y investigated, especially so, when one sees the marked difference between them and the back scratchers of the Japanese, the South Sea savages, and other barbarous tribes; indeed, the. back scratcher e Sout th savagea f so hSe vere ar sy lik Eskime eth o skin-scraping tool thr o ,e Basuto implement. 138 PllOCEEDING SOCIETYE TH F SO , FEBRUARY 13, 1871. I shall not enter upon the question of the relation of these combs to the single. and double margined combs; let it suffice that I think they belonge mann di y same caseth eo st people, were use differenr dfo t purposes. broce th l h Al comb similae sar forn i r thao mt I havt e detailee th n di typical description. Those found in so-called Roman camps are more or less knobbed, and differ from the broch forms besides by having flat surface d sharan s p square margins d similaran ; differences exiss a t regards their ornamentation. It is rather a curious coincidence, that the same kind of ornamentation whic founs i h d universall e Eomath n yo n combsd concentrian t , viz.do ce ,th circles s beeha , n discovere exiso t d t on many of the stones or slabs found ia the Pictish brochs and houses, and yet, so far as I know, no combs with this ornament have ever been

Fig. 4. Natural Size. foun thes n di e sculpture th e f brochso w dfe . slaba t Againf o sno n o , Scotland representations are present of double-margined combs, the domestic animals, birds, fishes, &c.; and the brochs, when excavated, yield the bones of the very same kind of creatures that are figured on the coincidence th stones t Bu . e goes still furtherr JohD r n fo ,Stuar t points out e sculpture, th tha n o t d slab f Scotlanso d geographicae "th l distri- butioe symbols—thth f o n e com f tho e b d com (figan mirrod ) an b.4 r (fig. 5)—is as follows :— NOTES ON LONG-HANDLED COMBS. 139

Mirror and Comb Comb. together. " Between Elvers Forth and Dee, 8 6 Betwee nSpeyd Elveran e , sDe 7 7 North of Spey, 5 4 South of Forth, 0 0" So likewiseknowledgey m e bes e typicaf th o th t o t ,l al l, long-handled combs of the broch form occur north of the Forth. greaA t variet f useyo s have been assigne theso dt e comb variouy b s s authors. Montfaucon thought they might have been use instruments da s of torture. A glance at the general form and chaiacters of the markings on the teeth of the combs will soon dispe ferociouo s l idean sa . Nos rha that very commonly thought opinion, that they were employe n dressindi g and ornamenting pottery, much more support tha e weanth k leg suppof so - sition to rest upon. Again confoundiny b , g them with those implements in use among the .Eskimo, North American Indians, Fig. 5. and certain of the South Sea savages, for taking off the. fat from the inside of skins and furs, and which consisn oblona f o t g piec f woodo e, it ,f o havin d en g e lasheon o dt in such a manner that they may project, three or more birds' claws, many hav ebelievo t bee d nle e thae long-handleth t d combs were used for the same purpose.1 That such is not the case, a careful consideration of what I have already said when comparing them with the above imple- e implementSomth f o e se Fijian useth y thir db fo s s purpos e made th ar e f o e 1 teeth of animals lashed to a piece of wood from the cocoa-nut palm. Their back- scratchers are also like the above. Mr Williams, the missionary among the Fijians, mentions that " their priests posses amons a s e sacregth d insigni f theio a r officea , long-toothed comb." I have shown some of these to the Society; they are made of several slender piece cocoa-nuf so t wood, lashed togethe y fibreb r f grasseso s . Some are made from reeds. Vide PI. XVIII. fig. 11. 0 14 PROCEEDING SOCIETYE TH F SO , FEBRUARY 13, 1871.

speamentso t d k an ;mor e dogmatically mentioy ma o I n, tw thae th t kinds of implements are so distinct in form and use from each other, that I only compared the tham o s e accuse I b might t f havinno o dt g over- looked them. The Maories carry wooden combs about as mere ornaments, and some have even considered the long-handled combs were put to the same use. Suc I hdoubt t ther bu ;s mori e e probabilit opinione th n yi , that they were used for combing the shaggy locks of their owners, since combs allied to them and the common Eskimo combs were in use a few years back in reducing to order not only the hair of the living, t alse scalbu oth p slaie lockth nf so warriors , whil scalpe eth undere sar - going the preparatory processes to fit them for decorating the person of the slayer. And I exhibit a photograph, vide PL XVIII. fig. 8, of a girl of the Seshaaht tribe, representing the manner in which such a coml) is worn on the head. The females of that nation are evidently behind the age, while the females of our nation have become more than imitators of the practice in vogue by the simple Eskimo girl. That such an use is not improbable I do not deny, yet the other two uses which have been assigned to these combs seem to my mind more feasible, viz., either that they were employed in keeping disentangled the threads of rein-deer sinew whils e combinth t r plaitinfo f woor go o , glit flax. Dr Eobert Brown, a well-known traveller among the Eskimo and the Indian tribes of North America, kindly gave me several facts about these combs in a letter, part, of which I now copy :—" I am strongly in- cline believo f thidt o se Shetlaneus thae th t dseparato t com s e wa be th strands while making sinew thread ; the tendon, in order to make it, is steeped in water, then flattened out by beating, and the whole mass separated into threads by tearing out with a comb like this. This is the methoe amonus e Eskimon th i gd e ISTortth , h American Indians, and- indeed all tribes which now use sinew thread." " The combs found in the Danish kjokkenmoddingt, " [the italic intent l minee al s ar d o t ]san e "ar purpose e Eskime th samth s s a e o com r separatinfo b e sineth g w threads; and to make the likeness more remarkable, at the base of the teeth could be distinctly seen the transverse markings caused by tugging threae th d mass from sid o side,,iet n orde separato t r fibres.e t th e Bu " fro ma carefu l examinatio mark the teet the the weaof of nsof h on r NOTES ON LONG-HANDLED OOMBS. 141 long-handled combs, I think they were employed rather in combing fleece, flax, or hair, than rein-deer skein. The transverse marks on the surfaces o wels wido s r l t woulmarkeno e no teete on o th e ds f a dhar expec t if they had been used to disentangle coarse sinew threads. Then, as Dr Brown remarks, not only the Eskimo combs, but all the Danish shell-mound combs, have these transverse marks of wear near the base, whils almostypicae n i t th l tal l long-handled comb oppore sth I hav d -eha tunity of examining, these marks are nearer to the apices than otherwise; and, according to sketches, such seems to be the case in that one figured by Madsen severaSwisn e i th d f s an o , formele combsth n I r .case e th , threads watere beinth n gi , ther s nothinewa opposo gt thrustine eth f go the comb well down among the thread mass, since there would' be little fear either of damaging the teeth or of scraping them against any hard substance; but in the latter, whether the comb was used for combing wool, flax r hairs o quiti , t i , e evident either that o taket cars t ewa no n bury the teeth too deeply in the material, or that such material was n sufficieni t no t thicknes alloo t se teet wth o becomht e immersep u d to their bases, and the points of the teeth, if opposed to anything, it must have been some soft substance. The marks on the long-handled combs, however, make it apparent that, whateve e materiath s sufficienf wa ro l s usedwa t ti , lengt alloo ht o t wt i wind round each tooth thao s ,n somi t e e comcases th d been f ha bi , , employed much more ende th teete ,f som so th f heo would have beet ncu oif from wear. From the results of experiments I have made with combs on wool, from the fineness of the transverse grooves of wear on the teeth, fro positioe mth f suco n h grooves frod e microscopian ,m th c anatomf yo wool, I am inclined to favour the idea that these combs were used chiefly for combin r cardingo g woo r fleeceo l . !N"o s hand-wooi r l combing r JosepnewM r fo ,h Anderson pointet ou d e Senchuth an i passag e s m Mor e ancieno t eth , t law f Irelando s n i , which the following words occur:—" For the wool bag, i. e. the bag whicf r peso foot. e he combt e . ,sh t hi ou a , materiale s sth whicha e hsh , e combinth tha , is t g connection i bag. t Bu " n with thi t shouli s e db remembered that no combs of the long-handled form have yet been found in Ireland, and the only combs approaching this form are some which resembl eSteven y thosb t ego s fro ma mos Denmarkn si . Again, 142 PROCEEDINGS OF THE SOCIETY, FEBRUARY 13, 1871.

Archbishop Potter, in his " Archjeologia Grseca," says, that Julius Pollux enumerates at great length the furniture and utensils of the Grecian women's lodgings, among which are instruments for spinning and weav- ing, basket e wooth r l smeasurefo s (ayw6i<;, e A.eia)th r comba o , (faviov)a , car r combindfo g wool. &c , Among the many interpretations of the comb symbol on the sculptured slab stones is that in which it is said to indicate that a wool-comber sleeps below e oldeTh r. Latin dictionarie f abouo s sixteente th t h cen- tury afford additional evidence that combs were used in weaving. Thus we have " Carmmare, to comb, or 'card wool; 'to heckle flax or hemp.'1'' Pecten lanaris, a weaver's slay ; a wool-comb; a wool card; a hatchel." " Carpere lanam, to comb wool, &c." Macaulay, in his " Lays of Ancient Borne," tells how Virginia warbled the lines of the good old song—

" How for sport the princes came spurring from the camp, And found Lucrece combing the fleece under the midnight lamp." In-concluding, allow me to state, that the study of these combs points plainl factse th o y—t . Tha1 Eomanse th t e brocth , h dwellerse inhabitantth d e an th , f o s Swiss lake habitations, used tools possessin a remarkablg e similaritn yi for maked mprobabld an an e sam, th r ey fo purpose , whethe t thino s r ro similarity was. produce othey b d r things, than tha l weral t e subjecteo dt similar conditions of life, and consequently compelled to resort to similar method similao t d san r tools difficuls i t i , sayo t t . 2. Tha make t leasth a t f thesn o ei t e combs e Eomath , scarceln ca n y be said to be better than the* broch dweller; for although the Eoman combs, both by their general finish and their special ornamentation, indi- cate generally the use of more complex instruments, and probably of a higher grade of intelligence to produce them; yet there are some of the broch specimens, which, if ornamental forms and marks be left e questionth ouf o te equall ar , s wela y l forme d quit s an dwela e l finished as the Eoman combs; not that I wish to say the broch dwellers were wanting in aesthetic culture, but that their idea of symmetry and beaut s marredywa evidentl wany yb propef o t r instrument exhibio st t i t objectively. 3. Tha varietiee th t thesf so e broch combs, formboto t s ha , finishd an , NOTES ON LONG-HANDLED COMBS. 143

eithee rudenessdu successivo e rt b y ma , e generations occupyin broche gth s t differena t times shows i s e a findmany , b nth hav o f t sy o e often been the case, or to different classes occupying the broch at the same time. 4.. There can, however doube faco n th te f b ,o tthat o product , e these long-handled combs, and also double-margined combs with rivets of iron, or single-margined combs .with handles like that foun t Kettleburnda , there must have been some of the generations who possessed greater manipula- tive skill and more delicate instruments than what the explorers of brochs have yet discovered, or we are in the habit of giving them credit for. e sam thay r juspasth d ou en wa t i t An t e progresideath f o s s these people had made in the arts of life required modification, it may be sub- sequently found thae intellectuath t l acquirement accomplishmentd san s of the broch dwellers were neither so few nor so simple as has hitherto been supposed. I am sorry that excess of my ordinary duties and ill health together prevented me using up all the material I have by me, and completing the paper as I wished, but I hope to do that by a future day. 3?or very kind and copious replies in return to my queries, my gratitude followine th o t e gisdu gentlemen:—Eobert Brown, Esq., Ph.D., F.E.G.S., Edinburgh . WA ;. Franks, Esq., A.M.,&c. Britise th f o ,h Museume th ; Eev. George Gordon, LL.D., Cor. Mem. S.A. Scot., Birnie; Eev. Mr Sutherland of Northmavine, Shetland. For the loan of sketches—Eev. Mr, Joass, Cor. Mem. S.A. Scot. f Sutherlandshireo , ; George Petrie, Esq., Cor. Mem. S.A. Scot., Kirkwall. And for assistance in obtaining materiale th s froe kitchen-midden—Mmth r John Anderson, merchant, Hillswick. 144 PROCEEDINGS OF THE SOCIETY, FEBRUARY 13, 1871.

ADDENDA. NOTES OP THE COMBS IN THE SOCIETY'S MUSEUM. COMB 1. Long-handled Comb found in ruins of a Pictish broch in Caithness, by G. Innes, Esq 1783n i . . For f handlmo e usual, except body much constricted endd a an ,n si knob. Made fro mlona g bone. Ornamentation none. Scratches frof i s ma , gnawing, on convex side. Length of comb, 4J in.; girth at dental end, 3j in. Diameter of knob ; girt, In constricteft i h a d part ,. ITeethin f number—n i 8 , 5 present par,3 - tially broken; set obliquely to axis of comb. Length of teeth | ,t oin.1 , middle ones longest. External teeth, usual form. Interdental notches, Jth in. wide. Cut- ting groove t basesa s . Indicatio a narro f no w instrument having been used. Border teetf so h near bases have mark weaf so r (PI. XIII. . fig3) . COMB 2. Com Burgf bo Burgarf ho , Evie, Orkney, 1825.— Vide Archasologia Scotica, vol. iii. , pi fig. 1831. v 44 . . 3 Founp , d along with human skeleton and deer'a par f o t s horn. Mad f horno e , typica formn i l . Length, 4| in.; girt t mosa h t constricted part, 2J in. t dentaa ; l end, 3^ in. t bluna ; t end ,. (barely)2in f . Ornamentation (vide fig.). Teeth, 9 in number; typical in form. Length of each, | in.; have marks of wear on exterior (PI. XIII. fig. 1). COMB 3. A. H. Ehind's Kettleburn Comb. One side convex, with ornamental marks; other concave at dental end and cancellated; made from along bone. Length, 6Jin.; girth at dental end, 4i in.; at blunt end, 2J in. (good). Teeth, 6 in number; an external one is broken away; they are large, clumsy, and wide apart; remaining external one is longest. Cutting grooves at bases. One of cutting grooves passes through one of the transverse ornamental lines. Ornamentation on convex side has near dental end two complete transverse lines and six incomplete ones. Inter- l concavna e surfac twos e completha eon ; e nea e teeth th d another an ,littlea t a r distance incomplete. Stone vessels, implements, bronze tweezers, iron weapons, &c., found along with it; also next comb. Proceedings, vol. i. pp. 264 and 269; and . 134.p vol . ii . Menis. Anthropolog. Soc. vol. . ii227p . . Archasolog. Jour. vol. x . . 223p , contains Quekett's examinatio f boneno s foun t Kettleburna d e saysH . , " This comb, and some spheriodal bone balls which were found along with it, were formed fro lowee mth r jawbon whalea f eo ; several small portion whosf o s e osseous structure were likewise found in unmanufacturena d state, (Vide woodcut, fig. 3, page 134). NOTE LONG-HANDLEN SO 5 14 D COMBS.

COMB 4. Same reference s lasa s t comb. Foun t Kettleburnda Benjy b , . Neeve Peacn hi 1854, in a Pictish broch. Body has a constricted middle, and a fish-tailed distal end. Mad f lono e g bon f somo e e animal. Compact wal bonf o l e = convex surface. Cancellated texture r= concave surface. Very little cancellated texture left. Length, 4J in. Distance between lobes at distal end, 2 in. Girth at dental e teetth l hend Al excep, brokee . 3|in on t n away. From character f stumpo s s interdental notche wide sar e (PI. XIII. fig. .2)

, ANCOME7 , . D6 8 , S5 Foun Jamey db s Farrer, Esq. ,e broc M.P.th Burray f hn o i , , Orkney, along with a portion of a double-margined comb. Among other articles found were stone implements, bod f vertebro y whala f ao e hollowed, bone scoops, bones with holes drilled through them, bone pin , a parbon f o te wheel pard f lowean ,o t r jawbonf eo whale.—Proceedingsa , . 6volp ; . als.ii . 157op . Com (thes5 b e number e affixear s combo t d n Museum)i s . a Thiver s i ys rough and rude comb, made from shaft of a deer's horn, which is naturally grooved longitudinally, is rounded and rough bevelled at the dental end from one side; is a tyne obliquely cut. Extreme length, 4| in. Girth of body, 4J in.; at distal end, ditto t dentaa ; l . endin Teeth f ,3 n number i 9 , , shorwedge-shapedd an t ; broad at base or attached part, with a free round apex; hardly J of an in. in length grooves of cutting well seen. Teeth present near apex; transverse shallow grooves as if worn; but in an indifferent state of preservation (Plate XV. fig. 1). Com . Betteb6 r finished than Com . b5 Mad hornf eo , also grooved longitudinally; from roundness and size most probably red deer. Extreme length, 6 in.; girth of body and end,3| in.; girth at bevelled dental end, 3J in. Body is round, not constricted; obliqun dentaa d een l section mad ee other froth e sid mo on et ; can- cellated structur ee compactcoarseth d an ;, thin. Teeth irregularlt y no cut o d , possess any transverse marks, but cutting grooves, seen on the surfaces of the handle for an inch or more, are 11 in number (PI. XV. fig. 2). Comb 7 is round, and made from horn. Length, 4J in.; girth at body and end, 3J in.; girt t dentaha l end, 3J in. t expandi ; s near denta slightlyd en l . Bevelled . 7 Teeth n numberd i an 0 1 ,6 simila a o present6 t ; n y i wa r , rest broken; teeth are sharp, from cancellated texture having been worn away. Surfaces of body bear mark f cuttino s g (sluices have been shaved off). Dental surfac transversa s eha e ornamental mark . fig(PIXV . 3).. CombS. Made of horn, longitudinally grooved. Length, 44 in.; girth between the ends, 2f in.; at tail end, 2J in.; at dental end, 2J in. External surface of comb smooth; internal surface rough (cancellated). Comb tapers from teeth to distal end. Teeth, 7 in number—4 present, 3 broken—are narrow and tenuous. Cutting grooves present to the extent of 1 in. Transverse marks of wear near the ape teetf xo h . fig(PI. XV .4) VOI;. IX. PART I. K 146 PBOCEEDINGS OF THE SOCIETY, FEBBUARY 13, 1871.

COMB 9. Made of bone, probably whale. External length, 6 in.; girth at constricted part, 2J in.; girth round dental end, 4 in.; tail end, 8J in., expanded. Comb has both surfaces flat, margins rounded e smootheon , r tha e otherth n ; d notchetaien l d obliquely. Teeth, 7 in number; vary 1J in. to If of an inch in length; average girthn i . e flatin ar ; , 1 strong,-and parallel; central teeth longest; interdental notches wide. Smooth surfac f comeo neas b ha line o r dentatw s d crossing-onen l e another; thes prette ar e y broa shallowd dt an know No . n where foun XIVL (P d . fig. 6). . COMProceedings10 o BN , vol. iii. 122.pp , 124. Gordon's Comb, foun circulaa n di r buildin t Uistga , mad f horno e , grooven i d longitudinal direction ; length, 4 in.; girth at one end, 2J in., diverges to 4| in.; hole diametern i . , Jin , near distal end teeth8 . , which diverge very much; teeth, f in. Jong; girth of teeth at base | ,in. ; all taper t opointa . All teeth rounded; interdental notches wide 5 to J; teeth present transverse markings near apex-; one of teeth evidently constricte twina f ofted i es ha nda woun come d rounTh b . dit possesses the same curve as the tyne it has been made from. Among things found were—a copper needle e throug, ey forme n a t i hneaf wire e o dd on ran , end a quantit; f deer'yo s horns a huma; n thigh bone; thirt r fortyo y vertebra) of whales flattene d markedan d , eight cut a bon; e articled , an flat eaca d t en h round in the middle, 8 or 9 in. long (lost); bone lid of a small box; six large black stone dishes, all about 2^ in. thick, and varying from 1 ft. 8 in. to 10 in. in diameter (PI. XIII. fig. .4) . COMHILLSWIC11 . BNo K (No. 1) . Typica forlin m ; Sin . long; girt dentahat l end, 3£ in. tai at ; Teethl in. end 3f , , 10 in number; middle ones longest, have marks of wear near apex (PI. XV. fig. 11). . COMHILLSWIC12 . BNo K (No. 2) . Typica formn i lt ornamentebu , d. lon; 5 Jin g; girt t taiha l end in., 3 t dentaa ; l end, 3f in. Teeth have been 14 in number, only 4 remain; length, f in., set close together with very narrow interdental notches. Evidences of the handle havin n instrumena gy b bee t ncu t lik ea knif e seenconven ar eO . x surfacf eo handlornamentae th e ear l lines depicte PIn di . XIV. fig . .2 Some of these serief so lines commence with four, but two invariably run into one another, and only three lines are seen on coarse inspection. The rough flat side of handle has, within three lines of the bases of the teeth, a complete broad transverse mark, evidently made wit hcoarsea r instrument tha e othenth r lines. Between thibased san f so teet founs hi d another coarse groove, only running abou . tthroug in J h froe mth margin of the comb (PL XIV. fig. 2). COM . 13B.No HILLSWICK (No. 3) . Is made of bone. More flat, and possesses coarser teeth than rest of Hillswick combs, of which it. is the third. Handle oblique as'regards teeth ; its distal end, NOTES ON LONG-HANDLED COMBS. 147

concave t righa t , tcu i s angle t unevenno axio t sd f handle o san , . Length, 6J in.; girt t distalha in.4 , t dentala ; , 4J in. Teeth, probabl numbern i 3 y1 ; marka of 10 left, 2 only unbroken, these J in. long. Remnants of teeth show that they must have been set obliquely, and irregularly as regards handle, not parallel like the iron barpalisada n si e railing somt ,bu e pointin one.directionn gi , other slightla n si y uniforf o different t no m d thicknessan , . Cutting grooves well seen. This comb is remarkable from its appearance and a comparison with other combs ; also bearing in founs minwa dt d i •wit h three typica well-finished lan d combs almosm a ,I t tempted theoriso t e concernin productions git firse th t n placeI . , whether from necessitr yo voluntarily obliqun a , e splinte bonf ro e seem havo st e been selected fabricatoe Th . r of the comb failed first in the parallelism of the teeth; second, in the parallelism observed between apices of teeth on the one hand, and margin of distal end on other. The margine t nequallhandlth e th no f e so ey ar hollowe t indicati n d outCa e eth . work of a beginner ? or was it intended for one of the lower classes of that age ? (PI. XV. fig. 10). . COM14 . BNo Fourt f Hillswicho k lot, alse largestth o . Typica n formi l , except distad en l slightly thickened, 6J in. long; girth at distal, 4. in.; at dental, 4J in. Teeth, 16 in number (largest number yet found in any broch comb). They are £ in. in length; only 4 are complete, and these exhibit easily seen transverse marks of wear near apex. Handle of comb has marks both of old and recent shavings. Near dental end on flat surface is a deeply cut groove. Sections of this comb were made and examined microscopically, and showed well the histological characters of bone, espe- cially resembling that of whale (PI. XIV. fig. 4. Dotted line at side of plate marked a—b indicates length of this comb).

STUAET'S COMB, No. 15. Proceedings, vol. vi. pt. ii. p. 402. Mad f boneo e . Body, typica forn i l m ; bot h side e slightlar s y convex, one side especially smooth and polished; this surface also presents the appearance of having had very thin slice t of t fshercu i thered ethed an an ,n thes t surfaceecu s polished. otheo tw rr o abrasion e Theron e f thieo ar s s surface thae evidentlar t y of more recent production. Length of comb, f in. Teeth, 14 in number, short, thin, flattened from side to side, and well finished at their bases. Average length, £ in., though som e mucar e h shorter entir5 ; e ones only remain whicf o 3 , h stand in series, and are of different sizes, so that the 8th tooth from one side (or 7th from the other) is shorter than 7th tooth on same side (6th'on the other), so likewise the longes i h 7t r coune tha (orw h nf 6t i t, them froe othemth thah r e 6t side nth e th , cuttino 5th)N . g groove presente there sar ar r e no indication, thef so m ever having been presen d thean t n effaced t ther bu ;e lot f transversear o s e groove f weao s r presene teetth hn i ttha t remain e stumpe broketh th d f so nan , ones; these ear especially outside foun onle th th n ydf o e o externa l tooth that remains traceo n , f so such grooves being seen on the opposite or internal lateral surface of such tooth, and as regards the remaining teeth and the stumps, they are not only worn more on one 148 PROCEEDINGS OF THE SOCIETY, FEBRUARY 13, 1871. side t havbu , e these groove weaf so rsame onlth n ey o latera l surfacoutside th a a e e tooth, very few grooves of wear being seen on the opposite lateral surface. These grooves are present in greatest numher nearer the bases of the teeth than the beaapicese ar t d marke ,an marginse th n do , which are'betwee polishee nth d surfacf eo the comb and the lateral surface; in other words, on the part of the teeth in which the compact harder bone tissue is found._ This points to a view that I have regard- ing these comhs when used, viz., that the compact or hardest part of the comb is that which has been opposed to the wool or flax (Plate XVIII. fig. 5).

PETKIE'S. COMB16 . No , Body typical in form; distal end approaching fish tail form; one lobe broken away; surfaces flat. Margins of handle have appearances of having been sliced or shaved in manufacture of comb, especially so near dental end. Teeth broken away from stumps evidens i t i ; t they have beenumbern i n8 , irregularly cut f variou,o s thicknesses werd t chieflan , ecu y froe surfacmon f combo e , from cutting grooves being founcharactere th side e dd onlon an , n thesf yo so e grooves themselveso Tw . stumps, one belonging 'to an external tooth, and the other belonging to one of the central set, have transverse grooves not produced by wear; this is the only comb on whic hI hav e seen suc "laterahe th cut n so l surface teete th hf ; the so muc e yar h deeper, and possess all the characters of a cut, as distinctive from a groove of wear (PI. XV. fig. 6). . BALFOUKD . COM17 . , EsqBNo . Is made of bone. Is remnant of a very strong comb, 6 in. long. Handle has one marginss oit f teethe d somth an f , o ,e broken away. Seven long, clean verd an ,y strong teeth remain. Interdental notches comparatively narrow t cleanbu , . Teeth 2 in. long. Dental surface has a sharp, transverse line at very bases of teeth. Handle hacke d shaveddan , especiall s remaininyit gexaminine marginon y An .g this comb would nevea momen for r t doubsaw t a tha teet by its beet hhad ncut (PI. XV. fig. 5). . ROMACOM18 . NBNo CAMP COMB. Found, along with an iron dart-head, in a "Roman Camp" at Hamhill, Somer- setshire, by J. P. Irvine, Esq., F.S.A. Scot. (Proc., vol. vii. part ii. p. 424). Made of horn ; 6£ in. long ; girth at dental end, 2| in. ; round the body, 2J in. Handle lon narrowd gan ; margins very round; enda kno y st blunb b a t end. Teethn i 0 1 , number perfec4 , t only f inc, n lengthhi , with sawing marks present t basA f e.o teeth is a transverse line, which appears to have been the guiding line in cutting the lengths of the teeth (vide PI. XIV. fig. 3).

COMB No. 19. ESKIMO COMB. Proc., vol. v., page 126. From Repulse Bay, Hudson's Bay, lat. 66° 22'. Handle and teeth almost equal anothee toon lengtn i r hhandl; almoss ei t square, margin sider so squarf so e being concave. Both surface teeth 2 smoothe 1 ar ss . Ha Eac. h tooth, excep extere th t - NOTE LONG-HANDLEN SO 9 14 D COMBS.

nal ones, is long, and slightly narrower at apex than at the base. Point is bevelled from surfac o surfacet e . Margin f teeto s h have been shaved off. Marginal teeth much thicker and stronger. Bases of teeth mostly without groove of cutting. Middle teeth longest, not because they project beyond the others at end, but becaus basee th e linconcave s si f th e o e . Teeth converge, hence more difficulo t t f handlo d forsmaln a i e en combf t m th o bont lcu t A rings .ei cu meany b , f so hunge b y whic. ma Diametet hi f holeo r . , JHandlin e ornamented like panel oa doorf . Comb evidentl one w d possessinne ,an ya traceo gn r markso weaf so r (vide PI. XVI. fig 3).

EXPLANATIO PLATESF NO . Plate XIII. Fig . Evi1 . e Combaddenda2 . No , . Typical. Pig . Peach'2 . s Fish-tail Combaddenda4 . No , . Fig. 3. Innes's Comb, No. 1 addenda. Fig. 4. Gordon's Comb, No. 10 addenda. This comb is like the Kent's Cavern specimens, according to description in Brit. Association Reports..

Plate XIV. Fig . Par1 . Hillswicf o t k addenda2 Comb1 . No , . Smooth surface. Fig. 2. Part of Hillswick Comb, No. 12. Rough surface. Fig. 3. One form of so-called Roman Camp Comba, No. 18 addenda. Fig. 4. Part of Hillawick Comb, No. 14 addenda. Fig. 5. Part of Comb, No 9 addenda. Plate. XV Figs. 1, 2, 3, and 4 refer to Farrer's Combs, No. 6, 6, 7, and 8 in addenda. Fig. 1 exhibits very well the cylindrical form of their handles. Fig. 6. Balfour's Comb, No. 17 addenda. Figs an6 . . Petrie'd7 s Comba addenda6 . 1 . FigNo , ..6 Fig fros i msketc7 .a y hb Mr Petrie. Speltisbure th Figs i 8 . y (Dorset) Roman Com Britisn bi h Museum. Fig. 9 is the Danebury (near Stockbridge, Hants) Comb in British Museum. Figs. fro9 d m an roug 8 h sketche Augy sb . W. Franks, Esq., British Museum. . Hillswic11 d an Figs k0 1 .Combs addenda3 1 . 1 .1 No . Fig , ;No fig10 .. .11 addenda. Plate XVI. . Carnliat2 d Figsan 1 . h Combs. From sketche Revy sb Joassr .M . Fig . Eskim3 . o Comb varietye on , . Vide addenda, Comb. 19 . Fig. 4. Is a Roman Camp form, in British Museum ; where got unknown. From a sketch by Mr Franks. 150 PROCEEDINGS OF THE SOCIETY, FEBRUARY 13, 1871.

Plate XVII. Fig. 1. Bone Comb. Figs. 2 and 3. Bronze, Worsaae. Vide p. 136

Fig. 10. A portion of a comb found on the shores of Cheshire. Vide Hume's work. Fig. 11. A Fijian Priest's Comb. Fig. 12 BronzA . e Pompeian Comb. Ceci Piccoli Bronz l Museide o Borbonico. Ta-vola viii. figs. 49 and 60. I would class them with single-margined Saxon comba.