HOWER, PREHISTORIC STRUCTUR PAPN O E A WESTRAY9 30 .

IV.

HOWER PREHISTORIA , C STRUCTUR PAPN EO A , . BY WILLIAM TRAILL, C.E., J.P., F.S.A.ScoT., AND WILLIAM KIRKNESS, E.S.A.ScoT.

This repor illustrates wa t . silenmm td 6 film1 witfee 0 f o ht.40 Sandstorms have preserved for Scottish another pre- historic settlement in Orkney. The site under review lies on the west side of the small island of , an island which has already contributed much to our knowledge of a long-forgotten past. Evidence e occupatiooth f thif no s settlemen bees ha t n know islandere th o nt r sfo a long time as each succeeding storm sends huge seas battering the face of the cliff, and carrying away parts of the foreshore on which it is situated. The name Hower -was given to the site by the . The word is the plural of "how," a mound, because at one time the site was occupied by more than one mound. To-day the islanders speak e Kna th e f sit Howers o pth a fe , showing that they have forgotten the original meaning of the word. Knap means a cap, and one could e moundsth f o p . t e knaca spea th no r f o po k Encroachmen a showese e s thau dth ty b unlest s excavatios wa n done very soon, a survey of this site would be impossible, as it would be entirely swept away. e masonrSanth f a do o covereyt p to e dth depth of some 8 feet. After this was cleared, the work of excavating was comparatively easy. The inside of the building was full of sand, which was very quickly disposed of by throwing it over the cliff. The excavation revealed two buildings lying parallel to each other, the larger being situate e othere sout th lyind th f an o ho ,dt g wit s maihit n axis N.WS.Ed an . (fig . .Bot1) h building constructee sar dry-builf do t rubble masonry. The inside of the larger building somewhat resembles the outlin figura f eo eighf eo aree plan n o taTh .withi thice nth k enclosing wall is 32 feet long and 16 feet wide, and is divided transversely into two parts or chambers by stone slabs set in alignment but having intervals between them. The side slabs are bonded into the lateral walls, and while these slabs stand as high as the remaining height of e wallth , approximatel feet5 y , they appea havo t r e been broket a f nof slabe Th sp (fig . to formin .2) e th e middlgth e portio e partitioth f no n 310 PROCEEDINGS OF THE SOCIETY, MARCH 8, 1937.

complete ar onld ean y measur feee2 inche4 t s high (fig plae . 3)nTh . show s kerbstones in the middle of the partition which may be the remains of a smal e nortl lobbyth n he inne O th sid . f reo chamber fee2 , t 3"inches abov e flooraumbryn a eth s i , fee2 , inche2 t sfoo1 wid inche2 d t ean s

Fig. 1. Plan of Structure at Hower.

high, of the usual type found in early buildings. On the south side e wal closd th thin i lan o et s chambe saddle-quera r founds o nwa Tw . rubbers largee on , r tha othere nth , whic beed hha n used with this quern, were got on the floor which at this part was covered with ground shells of shellfish. The entire side of the south wall of the outer chamber of this building was occupied by a low platform 4 inches high. It had been divided by thin stone slabs and may suggest bed accommodation. The plan show shora s t lengt foundatiof ho ne outsid e th wal th n f o le o HOWER, PREHISTORIC STRUCTURE ON PAPA WESTRAY. 311

structure, with a pavement of heavy slabs opposite the main entrance. destroyeo s s ha a d se thiErosio e sth featury nb e tha s purposit tt no s ei ascertainable. e buildinEntrth obtaines o yt gwa stone-coverea y db d passage-way fee6 , t long fee3 , t fee4 wide inche8 d t an , se highth n i , centrwese th f t entrance o end Th . e doorway shows some exceedingly interesting features. The door-checks are formed of stone slabs set on end, whic passage-wa e placee reae th h ar f th r o t da y instea nearef do e rth middle where one would expect to find them. Apparently the door has been of light framework, because the openings in the stone jambs, which were made to receive the cross-bars used for keeping it- shut, are very small. The communicating passage, which gives direct access to the other chamber, is also stone-covered. It measures 8 feet long fee5 , t throug soute wale th hf th h o l chambe 3 fee d t an througr e walth h l fee2 e north 6 s th ti n t i o ; e on e oth f inches wide, and 3 feet 4 inches high. The jambs consist of upright slabs, but the door-checks are at the other end of passagee th , thus showing that this entry could only be locked from the building on Fig . Hower2 . : Slab bonded into waln o l the north. The smaller building, which is north side of larger building. 30 feet long and 12 feet broad inside the walls, is somewhat similar in shape to the other. Its walls are not so high and, generally, the struc- ture is in a more ruinous condition, but from the character of the dry-stone masonry and other details it may be regarded as contemporary. Two parallel rows of slabs set across the structure have partitioned it into three compartments maie Th n .doorwa alss yoha bee nshore builth t ea t end. When excavated, this doorway was found to have been blocked up, as was the case with the rear of the doorway leading directly into e adjoininth g building. This main doorwa d checkyha s consistinf o g upright slab notee b s y dsimilaothee doorso thama th t tw ro a I t r . piece of cetacean bone was used as building material in the wall of this chamber at the doorway of the communicating passage. thire inmosr Th do t compartmen possibls i t mose yth t remarkables a , the greater part of the wall had been faced with slabs set on edge with other slabs projected from the mo fort ma serie f cubicleo s r stallo s s round the end. Filling the entire south wall of this chamber, three 312 PROCEEDINGS OF THE SOCIETY, MARCH 8, 1937.

small cupboards of masonry, approximately 1 foot high, had been constructed in the thickness of the wall 1 foot 6 inches above the floor. No indicatio purpose th o t thesf s eno a e stall cupboardd san revealeds swa . Two pits were found in the floor of this chamber, of which one in the front of the cupboards, measuring approximately 12 inches by 10 inches,

Fig. 3. Hower: Larger building from the south-east.

contained som n immatura e bone d a youn an f o sx e o g sheep, alsoa hammer-stone. This pit was covered with a stone slab. The rough formation of stones round the other pit, which measures 2 feet by 14 inches, suggest a well, as the enclosed space is continuously wet, winten i d ran contains clear water. Some bone werx o f shee o sed pan go thi n i t talso spi , whic similarls hwa y coveredflooe th rn nead o an t ,i r level \vere two horn cores of oxen. e middlth n I e compartment o cupboardtw , r recesseo s d beeha sn e soutbuilth hn i t wall foo1 , inche6 t s abov floore d slabeth d an , an s very rough masonry of a later occupation formed a partition 2 feet 6 inche norte wale th fron n sth i n h spaclo f e sideo t Th e. behind reminded HOWER, PREHISTORIC STRUCTUR PAPN O E A WESTRAY3 31 .

one of the so-called bed recesses found elsewhere in Orkney; but here the space was filled with earth and debris. From the amount of masonry which the spade has laid bare, this site has obviously been one of considerable importance. In considering e structures e th date evidenc th f s th begi o eu o t t ns le ,a e wit e hth buildings d comparan , e the world-famouw m no wit e hth s settlemenf o t , als Orkneyn oi . differena Skar n i s ai t geographical setting. A large sandy bay was selected by this colony of workers, and it would appear that this locatio s necessarnwa r theiyfo r work s despita , e eth sandstorms "which repeatedly buried their buildings, generation after generation struggled to clear out the sand. Sometimes they found re- building an easier way, but they continued to live on the site. There suco in e smasonr h th e builderst Howersanda o th t y f s yo yA ba ., does it give us any clue in our attempt to establish a date for Hower? The buildings at both places show rounded corners, but the type of masonry at Hower is of a much superior quality to that of Skara, and no feature as elaborate as our door-checks was designed by their archi- tectshave W e. referre cetaceae th o dt n bone built unde e linteth rf o l our doorway in the smaller building at Hower. This feature was also founwala t n Skaraa i dl t othebu , r builder prehistorin i s c times used this material similarly, as in a very early building at in Howern i s finA e .w d, saddle-quern buildinge th n si t Skarasa , where four were discovered. But many such querns have been reported from entirely different sites. In the Bronze Age foundry at Jarlshof, Shetland, Dr A. O. Curie, who conducted the excavations there, found similar querns suggested an , d that these might have been use r predfo - parin e clagth y employe fashioninn di g mould r castinfo s g metale W . found ground shell abundancn i s e flooth rn eo wher e saddle-quereth n stood at Hower, and Dr Callander, in examining the pottery found at Hower, detected similar pulverised shells in the shards found on that site .claiy Thumma e thasw t these querns wer treatmene useth r dfo t of clay use n earthenwarei d . Comparin e toolsgth a lisf ,whic o t s hi appende o thit d s paper e hammer-stoneth , d stonan s e anvilse th , hammers and borers of bone, and indeed all the movable equipment foun t Howerda parallelee b n brochssaie e b samca th ,e d n n di Th eca . of the pottery. It is also interesting to note that at Hower there are large box-like structure r cubicleso s ,againsd formeen n f slabo to dt se s e wallth . Such constructions were quite prominen e secondarth n i t y buildings at the brochs of Midhowe and Aikerness. We finsituatioe dth t Howena r very puzzlingsandyw site no s e,i Th . t thougbu h founful e sanf b o building le o ddt th erectee clayd sar re ,n do 314 PROCEEDINGS OF THE SOCIETY, MARCH 8, 1937.

excepwes e wher d earlien th a en t f a te o rthe p kitchento y e stanth n -do midden. Examining the section of sand left after excavation, it was found that at a level covering the top of the walls, vegetation had grown to a depth of about 4 inches. The line of vegetation could be followed alon entire gth e section. Stones foun thin di s line wer t water-worneno . A closer view showed stone fragment profusionn i s contoue Th . thif ro s site at this period must have differed greatly from that of to-day, as a thicknes fee8 san coverf f w o tso dvegetationo e sth n whic have hw e noted. A fragment of black pottery, possibly of an earlier type than that got in the buildings, was recovered from the midden. Indeed the buildings t thia s parmiddee th tf wero nn earliep a useeto foundey e db r th n o d people. From the amount of bones found, the occupants must have used cattle as food for a considerable time. Shellfish were also a favourite article of diet. After one storm ten years ago, a pit, 4 feet deep, was seen full of limpets. Shells of razor-fish were also got in abundance. But e findinth f largo g e quantitie f oyster-shello s a sit n eo swher s i t i e now impossible for them to grow suggests that erosion by the sea havy ma e bee formidablo ns o havt s ea e altere e wholdth e outlinf o e the land at this site. A large land-locked bay might have contained suitable breeding beds for oysters, and would also explain the presence muco os f h sand. s knowni s a e twi r th A, nfa s building t Howea s r hav o exacn e t parallel in Scottish archaeology. The occurrence of door-checks in all doore th believes si masonruniquee e exceedingln b th a o f d t o an ,s yi y high standard for prehistoric buildings. As the pottery and many of the artifact e similaar s thoso t r e people brochss a th useth f d y o edb an , e architecturasomth f o e l feature e see b t n somHoweni a so t f o e ar r these structures, a contemporary occupation is suggested.

We acknowledge our indebtedness to Dr J. Graham Callander, Director of the National Museum of Antiquities of , for much valuable assistance Societe th o t ;Antiquarief yo Scotlandf so grana r fo , t which enabled us to finish the excavation; to Mr Charles S. T. Calder, A.R.I.A.S., F.S.A.Scot., for his assistance; to Miss Margery I. Platt, Repore M.Sc.. Animae th D th r . n fo A ,o t r l BoneM o Shellsd t san d an ; Lacaille, F.S.A.Scot. Repors Flintse hi th r n fo ,o t .

LIST OF ARTIFACTS. Poin f tongue-shapeo t d piec f cetaceao e n bone e undeth , r side smoothe rubbing.y b d , 4^ inches long, 1^ inch broad inc1 , h thick. HOWER, PREHISTORIC STRUCTURE ON PAPA WESTRAY. 315

Scapul f sheepao , imperfect, measuring 6^ inches long, 2^f inches broad, spin d rubbeunden o an e f rdof sidt downcu e . Poin f finelo t y made bone pin, 3-g- inches long, point lyjt inch .a pointe f o Shar d dpen bone implement, 3^ inches long. Finely pointed bone borer made from splinter of bone. Hammer of cetacean bone, imperfect at one end,

j Inches.

Fig . Howe4 . r : Hammer Cetaceaf so n Bon Deer-hornd ean .

measuring 4^ inches by 2^ inches by lf-^ inch, and nearly half of another of deer-horn measuring 2§ inche inches2 lond gan s broae On d (fig. 4) . hammer-stone measuring 5-^ inche y 1b ^s inc y 1^hb - e ovainchOn l . stone with indentations on top and bottom and sides. Stone, 6-J inches by 5 inche -y |s b inch , with sharp inworked saddle-quere edgeOn . n madf eo irregular-shapen a d stone measurin foo1 feeinches2 6 y t feeg 3 b y t b t . Two rubbers use r samefo d . Nine flakes forme y splittinb d a gthi n water-worn stone, showing the hollow percussion on one side and end. Anvil stone mad flattenea f eo d conical stone measurin incheg3 2-y b fs 316 PROCEEDINGS OF THE SOCIETY, MARCH 8, 1937.

inches by 2|- inches, with picked indentations on top and bottom... Oval, pebble, 5|- inche 2-y fsb inche 2y ^sb inches bottod , abraidean p mto n do and ends. Irregular shaped stone measuring 7^ inche 6y sb j inchey sb 2f inches, with deep hollow picked on top—perhaps a lamp. Irregular shaped stone, 5^ inches by 4j inches by 2|- inches, with large hollow picked on top, small one on under side. Two spindle-like stones with word en frictiony nb e on , measuring Qj inche d 5^san inche lengthn si ,

1 Inches

Pig. 5. Hower: Spindle-like Stones. e seconth d imperfec wore (figd th . Severat n5) en .a t l piece pumicef so , largest piec inchee4 inche3 y s2^b y sb inches. Pottery.—Three fragmentm ri wald an s l fragment vessea d f re o sf o l clay, containing broken oyster- and limpet-shells (chiefly oyster-shells).

Rim, flatopn o t ,thick YSwald an ,l f inch thic partsn ki . Found n

n middei n used durinmc gprevioua s occupation ,f dar o smal km ri warel , very slightly round on top, with small projection to inside, ^ inch thick, wall § inch thick.

REPORT ON THE FLINTS. By A. D. LACAILLE, F.S.A.Scot. Flakes and Spalls. 1. Althoug t struckhno lone th , g edg s beeeha n utilised, probabls ya a knife. . Thes5 d "seconde ean ar , 4 , 2 3 r ."third "o " flakes; saythao t s i t, they are not the first removals from a raw nodule, but have HOWER, PREHISTORIC STRUCTUR PAPN O E A WESTRAY7 31 .

been detached either from a core or from an implement in the making. t firsA tappear 5 sigh . No t s fractured t inspectiobu , n shows break to be a hinged one. The partial surface alteration is interesting. ) sho5 Ale abovwo th t l norma2 ( e l phenomen e intentionath f ao l fracture of flint, but it is not possible to assert what percussor was used, i.e. stone, wood, bone, or metal.

Scrapers. Possibl e chieye th th f s poini 8 f interes o d te scraper th an , n 7 i t , 6 s amount of "wear to which they testify. Viewing them technologically, 6 and 7 are particularly remarkable, No. 8 less so, as they consist of naturally fractured piece flinf so f convenien o t t size, whose shape readily permitte adaptatiof do round-edges na d scraper mose th ty sb elementar y of dressing e ogivaTh , .althoug l6 . endingNo f ho much worn, strongly suggest gravea s r extremity oppose scrapere th t o thidt bu ,s cannoe b t stressed. 7. This steep end-scraper shows extremely delicate pressure trimming. 8. Shows wel e amounth l f surfaco t e alteratio f thino s brown flint, the corpus exposed by primary dressing, which has been sufficien givo t t e necessary working edge. 9. Thick spall humanly struck, but is not an implement. 10. A fragment of a flake, and smalA . l 11 fragmen implemenn a f o t t secondarily trimmed. As regards the period, it is regretted that there is nothing whatever whabo t ytel n s whica lca t culture hon e specimeneth s belong. Little can be said typologically, and the only guide will be the other relics with whic flinte associatedhe th ar s . t reallno s yi possibl t monetarI a x fi o et y value upon these flintss A . specimens of implements they are extremely poor, but as examples of artifacts worked in the simplest way possible they are scientifically very interesting.

REPOR THN O TE ANIMAL BONES y MARGERB . . I YPLATT , M.Sc., Royal Scottish Museum, . A collection of skeletal material excavated by Mr Traill and Mr Kirkness on this northerly placed island of the Orkney group was sent 8 31 PROCEEDING E SOCIETYTH F O S , MARC , 19378 H .

Royae th o t l Scottish Museu identificationr mfo . Wit exceptioe hth f no fragments of a whale's skeleton and a few bones of sea-birds, the whole of the remains represent animals domesticated by man. The relics are extremely well preserved, many have been burie n fini d e white sand d consequentlan y hav a eclea n white appearance e boneTh .s found throughou e coursth t f excavatioeo whicf n(o h this collectio s onlni ya small part) were exceedingly numerous, and probably represent the remains of the food animals of an early human settlement. One frag- ment of a human skull, included among the rest of the bones, has been kindly examine y Professob d . LowA r f Aberdeeo , n Universityo wh , states that, althougfragmentaro to s i t hi classifyo yt nevertheless i t i , s not unlike the human skulls found in other Orcadian structures of datee variouTh . s specie f animalo s s e generafounth n i d l material are recorded below in order of numerical importance.

Ox. The bovine remains are an extremely numerous and interesting collection of bones. More than half are from very young animals. Unfortunately most of the relics from adults of large size are broken and r thi fo ,o recors t f littl reasono e s dus ei t theii , r measurements. Some fragments excee sizn dcorrespondine i eth g bonepresent-dae th f so y ox, and approach closely the dimensions of those of the European bison now extinct. Notable fragments of massive size are the proximal portio radiusa f no sacruma , lumbaa , r vertebra patellaa d an , . This feature is also borne out by the many enormous horn-cores of a breed of very large regretteb eo t cattles i do cas t n I tha .en i twer e th e large horn-core d thaan t s p theicompletti e r th ful o t el size therefore can mattea onl e yb f conjectureo r smale Th l. table below showe th s e horn-coreth siz f eo Pape th f aso Westray bree largf do e cattl coms ea - pared wite horn-corth h e measurement e Uruth f so s from Scottish localitie preservew sno Royae th n di l Scottish Museum. Scot. Urus (B.S.M.). Papa . C Westray . B .A. Minimum distance between horn-core . s 24-7 22-7 19-8 23-. cm 5 Distance between horn-cores on level of ligmentum nuchse .... 27-5 27 19 25-4 ,, Circumference of base of horn-core . 33-1 35-3 19-3 26 ,, Other complete bones of the bovine skeleton are not so large as those mentione d e remainin r otheaboveth o f e d belono r on an , go gt smaller breeds indicated by the horn-cores. Characteristics in these, HOWER, PREHISTORIC STRUCTURE ON PAPA WESTRAY. 319 denoting cattle of different types, are seen; one kind is not represented in Britain to-day, but was found in a on , also of course of prehistoric date. This horn-core of cornute shape belongs to the Bos

frontosus (Nilssori)1 type, being of rugose texture and with a thick collar at the base of it which characterises this variety. Its circumference at the base is 24-2 cm., and its length along the outside curvature 25-3 cm. A third breed present is reminiscent of the modern Kerry cattle, the core having a well-defined upward and outward curvature. Measurements of this are as follows:— Minimum distance between base horn-coref so s . 16-cm 4 Distance between tips of horn-cores . .34 ,, Circumference of base . . . . .18 ,, Length along outer curvature. . . . 26-3 ,,

A remaining type is typically that of Bos longifrons (Owen), with short curved horns separated by a high convex forehead. These cores are typically oval in transverse section for a greater part of their length, and in general size resemble the last-described breed (Kerry). Measure- ments are:— Calculated distance between base f horno s - cores (about) ...... 16-cm 8 Circumference of base . . . . .18 ,, Length alon g. 22-oute, , 2 . r curvatur . e

All these types probably indicat earln ea y domesticatio f cattleno , biit to assign to them their actual age is quite an impossibility.

Sheep. t thiA s site remain f sheeo s alse par o numerous animald an , s repre- sentin l stagegal growtf so h from tiny lamb adulto st s with long slender lowe ra well-wor jawd an s n dentition occurvariete Th . y represented has massive horn-cores and resembles the mouflon stock. Two skull fragment hornlesa f o s s adult sheep probably indicat e ewe eth f thi o s s breed. The horn-cores are strongly pitted, almost triangular in section, with the "widest flat surface seen from the posterior aspect. They curve away from each angln othea f about eo a r t 45°, trending upwardd an s outwards, then slightly backward downwardsd san . Horn-core f verso y

young animals show the same characteristic angularity. Scottish1 Naturalist, Jan.-Feb. 1933. 320 PROCEEDING E SOCIETYTH F O S , MARC , 19378 H .

Pig. Relics of this animal are but scanty. The material submitted here contained only largpara f o te skull a smal, l scapula well-wora d an , n last molar tooth t woulI . d seem, therefore, that bothd ol youn d an g animal e representedar s , and, fro e abrasemth d conditio e singlth f neo tooth wila , semi-wilr tha n di o bease s th t wa t d state.

Whale. Many large fragments of bones of a whale'(species unknown) occur.

Birds. Indicated by the presence of stray bones, the birds represented are: the eider-duck (Somateria m. mollissima (L.)), the cormorant (Phalacro- corax . carboc (L.)) e commoth , n gull (Larus . canus,c L.)e ganneth , t (Sula bassana, L.), and a swan, probably the Whooper (Cygnus cygnus).

Shells. Shellfish of various kinds had probably been utilised as food, according to the numerous shell remains present. These include winkles (Littorina littorea . obtusataL (L.d an ) (L.)), oysters (Ostrea edulis, L.), mussels (two kinds: the edible one Mytilus edulis, L., and the horse-mussel Modiolus modiolus, L.), one small "buckie" (Buccinum undatum, L.), a small cowrie (Cyprcea europcea, Mont.), and lastly a shell allied to the venus shell (Dosinia exoleta (L.)). Include materiae th n di l just recorde portione th s di f huma"o n skull referred to already in the introduction.

Bones from a large hole in the floor of Chamber 3 smallerin building. Here sheep and ox remains are represented in equal proportions. Those of the former are from young and matured animals. One complete metacarpal has on the surface innumerable tooth impressions resembling thos dog a bonelatte e o f eth o N .f rso have been foun thin di s material. bovinw fe e e relicTh thin si s sectiol froal me nar immatur e beasts. 1 32 DONATION E MUSEUMTH O T S .

Bones from a small hole in Chamber 3 smallerin building. e mosThesth r t efo par e remainar t ribsverf w o sfe ,y A youn . gox a metatarsal terminad an , l facefemua f o t r alone represent immature sheep.

n conclusionI indebte Williar m a M I , o mt d Trail r forwardin!fo g the skeletal material found at Hower for examination here, and for kindly allowin Royae gth l Scottish Museu incorporatmo t e e theth mn i cabinet collection f sub-fossiso l material from Scottish sources.

MONDAY, 12th April 1937. SIR GEORGE MACDONALD, K.C.B., President e Chairth n i ,.

A Ballot having been take followine nth electes gwa dFellow:a — Mrs H. NUGENT YOUNG, 10 Onslow Court, Drayton Gardens, London, S.W. 10.

The following Donation Museue th o t s m were intimated thankd an , s Donors:e voteth o dt — . CARFRAEA . J y (1Quee3 B ,) n Street, Edinburgh. Two Old Dutch Grate Tiles taken from the Duke of Gordon's House on Castlehill, Edinburgh ,pulle s whewa t dni down. Majoy (2B ) r HARR . HBBDENYH , M.C. f ,o . Collection of Neolithic and Iron Age Pottery, and other Objects found at , Orkney. See previous communication by C. S. T. Calder and J. Graham Callander, LL.D. (3) By Dr W. A. MUNRO, F.S.A.Scot. Two fragment a star-shape f so d Bea f bludo e vitreous paste, found on Dryburgh Mains, Berwickshire donore th y ,b . y (4WILLIAB ) M BROOK, F.S.A.Scot. Twenty-six Communion Token f variouso s parishes. 1 2 VOL. LXXI.