Analysis of the Ptolemaic Geography of Scotlakd

Home , Cowal, Duncansby Head, Dunnet Head, Epidii, Geography of Scotland, Inchkeith

198 PKOCEEDINGS OF THE SOCIETY, APRIL 12, 1875.

II. ANALYSIS OF THE PTOLEMAIC GEOGRAPHY OF SCOTLAKD. WITH TWO MAPS. BY CAPT. F. W.L. THOMAS, K.N., F.S.A. SCOT. (PLATES V., TI.)

When making researche se Pictis intprobablth e th f oo he towerag e s which hav thicklo s e y studde e shoredth islandd e nortsan th d f han so west of Scotland, I was led to examine that part of the geography of Claudius Ptolemy which relate modeo t s m Scotland. Many yearI o sag attempted something of the kind, hut the disagreement between the edition wane d accesswhicth f o sufficient s o ha t d I han , t leisure, com- t asidei y la . pelle o t Since dm e the (partialle nth y completed) editiof no " Claudii Ptolemseii Geographice," by Dr Wilberg, has been published, formed fro e examinatiomth f manno y Greek MSS. comparisod an , n with several printed editions. In what follows I have accepted Wilberg's Latin text ,exceptions w witfe ha , which wil explainee b l d furthe. on r importance Th a documen f eo t whic admittes i h o havt d e been com- piled befor hardln eca A.D e overrated0 yb 12 . althougr fo ; h Tacitud san Pliny have both made mention of Caledonia,—the former of whom, has written e mossomth f t o ebeautifu l passage n histori s y concerning it,— yet there is no other detailed description than that of Ptolemy; from whose tim knowledge eth countre th f eo y became lesd lesssan , untiln i , 1525, we have the figure of Scotland reduced to three sides of a parallelo- gram, wit t JohnstoS h n (Perthe north-westh n i ) t corner d Argatan , t (Orkneys) removed intwestere oth n ocean. cursorA y inspectio Ptolemy'f no s " Descriptio f Albiono n " shows that the larger features in the figure of the island are there indicated, and man yt wit nameme h e littlsar e chang t f formgreaeo Bu .t errore ar s soon recognised; the fundamental latitude, that of London, is 2i degrees too e greatdistancTh . e between Londo d Yorknan , towns which then werw asno e centre f civilizationo s greao to one-thiry s b t i ,s tru it ef o d length e wholTh f Norther.e o n Britai cantes ni d fro norte mth h towards easte Hebridee th th , e turnear s d fro e soutmth h toward e westth d s an , e e Solwalengtth th f o hy Firt s exaggeratei h do nearl t fro 0 0 m7 y20 miles. Yet furthen o , r examination gulfse th , , estuaries, rivers headlandd an , s ON THE PTOLEMAIC GEOGKAPHY OF SCOTLAND. 199

usually succeed each othe thein i r r natura lnamee orderth oftee d sar an ,n so little changed as to be readily identified. And in England the names and position comparee townf b so n ca s d wit Antonine hth e Itineraryd an , thus lead up to their identification. In Scotland we have no such corroboration; the sole ancient authority, except a few names preserved by Plin Tacitusd yAnonymouan e th s i , s Geographe f Ravennao r e h d an , scarcely supplies more than a list of names of places, certainly in great profusion, but in very loose order, and strangely corrupt. Upon the coast t weri s ea upowe ear a roan d which leads from placa f placei o et d an , nam distancr eo e canno e reconcileb t d e passedwitb hn ca fac, t whili t e stile wear l sur proceedinf eo g onwar righe th tn d i direction . No wito s t h the towns; the positions, though sometimes noted to five miles, are seldom noted nearer than to ten; and a comparison in England with the Antonine Itinerary shows that if the Itinerary had been lost, we should o trus t r theihav fo d t reha identification much mor o verbat e l resemblance se geographicathath o nt l position e Ptolemais giveth a s n i n c Tables. In the following pages so much of Wilberg's text is reproduced as relates to Scotland. The headings are translated into English. The first column contains references to the notes. The second, the Latin names of places according to Wilberg's text, except in two instances, which are explaine e notese thirth d fourt Th n an di .d h columns contaie th n Ptolemaic latitudee d longitudese d th sixtfift an th e san h har n i d an ; distanc bearind positioe an eth f go precedine n on fro e mth , computegit d fro latitudee mth longituded san meany sb Traversa f so ee Tableth n I . seventh e bearinth , s gcorrectei knowr fo d n errore rightth ,o t .viz. 0 9 , eighte Th h column contain modere sth n positionse namth f eo e giveth n ni Ptolemaic Tables. These have been mainly identifie compariny b d e th g Ptolemaic distances with the actual distances from place to place, measured with compasses upo a modern e northernth maps d westerA an n. n islands have been my home for many years, and having a personal acquaintance with three-fourths of the coasts of Scotland, I am able to investigate this part of the subject with some degree of confidence. It will be seen in the notes that there is often a very close agreement between the Ptolemaic distances and the actual ones, but this must be regarded as accidental, for his unit of distance is, with few exceptions, virtually 0 20 PROCEEDING SOCIETYE TH F SO , APRI , 1875L12 .

ten miles; so that without any error in his notation two places may be on nineteer eo n miles apart. The note e mainlar s yt propose geographicalno s o wa entet d t i r fo , upo histore countrye nth th f yo . Two maps have been made to accompany this memoir. The first shows the Ptolemaic positions and Latin names as found in Wilberg's text,—wit e exceptions*th h f Eggarhicin"o r Engarhicinafo a d Loggan , i fluvii ostia for Longi fluvii ostia; and the position of Varar sestuarium is taken fro a mvariou s reading e point Th e merel.ar s y connectey b d straight lines, by which the shape of the country is shown. secone I havp th n edma I made much greater alterations plais i t I n . thae erro th tf directioo r n begins fro e Ptolemaimth c position Vedrse fluvii ostia ; I therefore traced on tracing paper the map of Scotland from e positiofirse th th n t o map nd Vedran , a centre s a , turne e tracinth d g from the east to the north. This brings Scotland into its natural position, when the proportions of the coast-line are seen to agree quite sufficiently with fact, except on the north side of the Solway Firth, i.e., from the Mull of Galloway to Carlisle. I have filled in this piece of coast by reducing the Ptolemaic distances proportionally, and joining Novantarum promontorium to Itunse fluvii ostia. When the tracing of Scotland was turned north and south, the position of Monaceda was found to approximate to that of Mona. Now, the Ptolemaic map shows an island off the Mull of Galloway, which must have been intended for the Isle of Man, and another between England and Ireland, which is in the position of the Isle of Man; so that both positions are identified wit e namesh th Man t o doubn bu , t belong respectivelo yt Man and Anglesey. From, several considerations I conclude that Ptolemy has made use of pre-existing maps to form his tables (of descriptions as he calls them), and I have been led to doubt if the length of the longest day was really measured at the places named; rather, that the connection e resulth f s calculationo i t o s regarda tw r e t a leastfa th o, f s s a , Scotland. Another result of turning the tracing of Scotland to the north, is to bring the first Ebuda (Outer Hebrides) into its natural position; and by turnin tracine gth g half round, wit firse hth t Ebud centrea s a othee th , r Ebudse fall into their natural positions with sufficient exactness. PTOLEMAIE ONTH C GEOGRAPH SCOTLANDF YO . 201

The northern isles are in the same way brought into their proper places by using Cape Orcacentrea s thin a sI .s manne a restorer hap s ma d been formed, showing the true amount of geographic knowledge in the second century when freed from obvious errors. It will be seen that many of the names have been shifted; this is shown whe e Latinth n name reasonth s writte i es inkr d A sfo re . n ni so doin notee givee th gar sn i they need notf*be repeated here. My friendly neighbour, theEev. Charles Darnell,M. A., besides translat- ing the headings of the chapters, &c., has given me efficient assistance whenever good scholarshi wanteds pwa . In conclusion, I would point attention to the large amount of philo- logical, historical, and geographical material which has been saved to us by the industry and genius of the Alexandrian philosopher.

BOO. KII CHAPTER I.—THE POSITION OP THE BRITISH ISLAND, HIBERNIA. Firstof Europe.Map (Here follows the geography of Ireland, which is omitted as not necessary to the present inquiry.) Above Hibernia some islands are situated which are called Ebudae, five in number whicf o , h that whic situates hi d wese mosth calleds i to t t —

Bearing Cor- Notes. NAME. Long. Lat. D. or rected Modern Name. Course. Bearing.

A. 1. Ebuda. 15-0 62-0 ... Outer Hebrides. The , towardnexit o t t easte sth als s i , o 2. Ebuda, 15-40 62-0 19 E S Skye. next 3. Ehicina, 17-0 62-0 38 E S Egg- next 4. Maleus, 17-30 62-10 17 N55 E s35E Mull. next 5. Epidium, . 18-30 62-0 30 S70E s20w Islay. And towards the east from Hibernia are these islands — 6. Monaoeda, . 17-40 61-30 ... . . > ... Isle of Man. 7. Mona insula, 15-0 57-40 ... Isl Manf eo . 8. Edri ins. deserta, 15-0 59-30 ... 9. Limni ins. deserta, 15-0 59-0 i 2 20 PROCEEDING SOCIETYE TH F SO , APKI , 18/5L12 .

CHAPTE. RII POSITIOE BRITISE TH TH F NO H ISLAND, ALBION. Europe.of FirstMap The Descriptio e nortth f hno [west] side, above oceae whicth ns i h whics hi called Ducaledonian. 55— Bearing Cor- . Notes. NAME. Long. Lat. D. or reeled' Modern Name. Course. Bearing. B. 1. Novantarum Chersone- sus, and the promontory of the same name, 21'0 61-40 ...... Mul Gallowayf lo . 2. Kherigonius sinus, 20-30 60-50 52 s 1E 7 3 w7 s Loch Ryan. 3. Vindogara sinus. 21-20 60-30 31 S50B N40E Ayr Bay. 4. Clotae aestuarium, 22-15 59-40 57 s 29 E N61E Clyde. 5. Lemaannonius sinus, . 24-0 60-0 56 s69E N21W Loch Fyne. 6. Epidium prom.. , 23 '0 60-40 50 K37w s 53 w Mull of Kintyre. 7. Loggi fluvii ostia, 24-0 60-40 29 East North Loch Linnhe. 8. Ityis flu. ost., 27-0 60-40 87 East North Loch Alsh. 9. Volsas sinus, 29-0 60-30 59 s 80 E N 10 E Loch Inver or Loch Broom. 10. Nubaei flu. ost.. , 30-0 60-30 29 East North Cape Wrath. 11. Tarvedum sive, . 31-20 60-15 E 4 39 6 s N21E Dunnet Head. Orcas cautes, Description of the west [south] side, to which are adjacent both the Hibernian Ocea [thad f nan Vergiviu o t s ; after 0. l. Novautaram Cher. prom. 21 '0 61-40 2. Abravanni flu. ost., 19-20 61-0 63 s 5E 1 9 w3 s Cree. 3. lenae faest, 19-0 60-30 32 s 19E 1 w7 s Fleet, 4. Devae flu. ost.. , 18-0 60-0 w 4 35 4 s E 5 4 s Dee. 5. Novii flu. ost., . 18-20 59-30 E 3 28 1 s N72E Nith. 6. Ituuae aest, 18-30 58-45 E 4 56 s N84 E Eden. 7 Moricambe aest, . 17-30 58-20 41 S 5,2E w8 3 s Morcambe Bay. (Here follows Westthe Coast of England.)

D. 1. A. description of the next sides, which look between the winter sunrise and the south whico t , Germae hth n Ocea s adjacentni . After Tarvedum promontorr yo Orcas, of which mention has been made. D. 2. Virvedrum prom., 31-0 60-0 18 s 34w s 56 E Duncansby Head. 3. Verubium prom., 30-30 59-40 25 s 37w s 53 E Noss Head. 4. liae fluvii ostia, . 30-0 59-40 15 w s Helmsdale Eiver. 5. Eipa alta, 29-0 09-40 30 w Caithnessf so d Or . 6. Loxae flu. ost., . 28-30. 59-40 15 w s Cromarty Firth. 7. Varar aest., . 27-30 59-40 30 w s Moray Firth. ON THE PTOLEMAIC GEOGRAPHY OF SCOTLAND. 203

Bearing Cor- Notes. NAME. Long. Lat. D. or rected Modern Name. Course. Bearing.

D. 8. Tuaesis aest., 27-0 59-0 43 s 22wE 8 S6 Spey. 9. Celnii flu. ost., . 27-0 58-45 15 S E Cullen. 10. Taezalorum prom., 27-30 58-30 22E 6 S4 N44 E Kinnaird Head. 11. Devae flu. ost., . 26-0 58-30 46 W S Dee. 12. Tava aest., . 25-0 58-30 31 t-W S South Esk. 13. Tinae flu. ost., 24-0 58-30 31 'w S Tay. 14. Boderia aest., 2230 58-45 48 N?2w 8 18W Firth of Forth. 15. Alauni flu. ost., . 21-40 58-30 30 s 60w s30 E Tweed. 16. Vedrae flu. ost., . 20-10 58-30 47 w s Tyne. (Here follow positionsthe East the on Coast England.)of E. 1. The Novantse dwell alon norte gth h [west] side, belo Chersonesue wth same th ef so name, among whom are these towns : — 2. Lucopibia, . 19-0 60-20 40 N60E N30w Glen Luce or Wigton. 3. Rhetigonium, 20-10 60-40 40 N60B N30W Cairn Ryan. 4. eas e Belowhic] th of t Selgovse o wh[t th e ; among who these mar e towns: 5. Carbantorigum, . 19-0 59-50 61 s 35w s 55 E Perhaps by the Dee 1 6. Uxellum, 18-30 59-20 34 s 28E 2 w6 s Birrenswark ? 7. Corda, 20-0 59-40 49 N66E N 24 w 8. Trimontium, 19-0 59-C 51 s 38w s 52 E Eildon Hills 2 9. From these toward e easth s t [north] e Damnonith , e nortth o ht i [west]; among towne whoth e sm ar : — 10. Colania, 20-30 59-10 47 N 77E N 13W 11. Vanduara, . 21-40 60-0 61 N 55 B s55w Ayr? 12. Coria, 21-30 59-20 40 s 7w s 83 E 13. Alauna, 22-45 59-20 39 B N Stirling. 14. Lindum, 23-0 59-30 13 N39E N 51 W Callender? 15. Victoria, 23-30 59-0 34 s 28 E N62E Near Perth. 16. The Otadini to the south [east] ; among whom the towns : — 17. Curia, 20-10 59-0 103 West South OTweede nth . 18. Bremenium, 21-0 58-45 E 260 6 s N30E Rochester. Afte Damnoniie th r , fro ease mth t [north] t morbu , e toward norte sth h [west], from the promontory of Epidium almost to the east [north], the 19. Epidii ; after whoe mth Kintyre. 20. Cerones ; next toward eaststhe the , About Loch Oreran. 21. Creones ; nexe th t Garmoran. 22. Carnonaeae ; thee nth About Loch Carron. 23. Carini ; and most to the east [north] ; and furthest away, Lord Reay's Country. dwell the 24. Cornavii, Caithness. EYo Gule m th f Lemaannoniu e o estuarf th s a f y o r fa s a s Varar, the 25. Caledoni wes e abovthed ] th i of e tan ;o m th e[t Athol to Mid-Ross. 204: PROCEEDING SOCIETYE TH F SO , APRI , 1875L12 .

Bearing Cor- Notes. NAME. Long. Lat. D. or rected Modern Name. Course. Bearing.

E. 26. Caledonian Forest ; from which towards she eist [north] the East sideof Drumalban. 27. Decantae; after whice hth Black-isle, Cromarty, 28. Lugi, borderin Cornavie th n go i ; abovd an e wese ftoth t of] East Ross and Suther- the Lugi, the land. 29. Smertae; but below [to the east of] the Lugi, the About Loch Shin. 30. "Vacomagi; among whom are the towns : Speymout o Blait h r Athol.1

31. Banatia, 24-0 5930 • • l Near Blair Athol. 32. Tamea, 25-0 59-20 33 s72 E N18E Near Euthven, Bade- noch. 33. Alata Castra, 27-15 59-20 70 B N Burgh Head. 34. Tuaesis, 26-45 59-10 19 s 59w S 31 E Fochabers ? Below [to the east of] these, towards the west [south], the 35. Venicones; among whom this town Angus and Mearns. 36. Orrea, 24-0 58-45 90 74w|s E 6 s1 Forfar 2 Towards the east [north], the 37. Taezali ; and the town Buchand an r .Ma 38. Devana, 26-15 59-0 7E N8 2| 7 N12W Aberdeen. Again, below [soutSelgovae Otadinie th th ] hd of ebotan o t , h seas stretce hth 39. Brigantes ; with whom these towns : — (Here follow the Towns in England.) The islands [named belowe ]ar adjacent to the island Albion, at the promontory Orcas : — 40. Ocites insula, 32-40 60-45 Fair Isle. 41. Dumna insula, . 30-0 61-0 80 fr79w ... Orkneys. nortAbove th whichf ho o e[t aboue ]ar t thirty 42. Orcades ; of which the middle is Foula. situated, . 30-0 61-40 4 Nort| 0 . h.. | And much higher than that is 43. Thule ; that part of which, which reaches farthest west is Shetland. situated in depress '29-0 63-0 85 Nl9w Thelfarth'est east, 31-40 63-0 73 E N The farthest north, 30-20 63-15 39 N67w s 23w farthese Th t south, 30-20 62-40 35 s E And the middle, . 30-20 63-0 .20 N w PTOLEMAIE ONTH C GEOGRAPH 5 SCOTLANDF YO 20 .

NOTES. Ebudce.—Pliny names "30 Haebudes " and Ptolemy has "'E/3ovSfish and milk. There is but one king for all the islands; for they are all separated from each other by narrow channels. The king has nothing of his own, everything belongs to the community; he is compelled to act justly by fixed laws, and that avarice may not turn asidm hi e from truth e learnh , s justic reasoy b e f povertyo n , o that s i t say, he has nothing belonging to himself, but is supported at the public expense. He has no wife of his own, but from time to time he takes any fancya woma s wifs whor hi ha efo .r e nmfo h wisThereforo n hs ha e eh for, nor hope of, children." This description appear e genuineb o t s . Althoug e inhabitantth h s may not have been altogether unacquainted with corn, it is certain that, until the introduction of the potato, the staple food was milk and fish. And with regard to the peculiar celibacy of the king, the reports which reached the Roman authors may be compared with the fact that no Pictish king was succeeded by his son ("Chron. Picte and Scots"); and also s showina , e permanencgth f institutionso e , that those provinces of Scotland which have remained more essentially Pictis notee a har r dfo certain laxit conducf yo thio t t s day. The importance of the Ebudee in ancient times is proved by the gigantic megalithic monuments in Lewis, as well as by the scores of Pictish towers d retrenchean d islet d pointan s s whic e scatterear h d throughoue th t islands. VenetiaA n editio f Ptolemy'no s Geograph f 156yo 2 translates "Insulas Ebudee" by " Silvestres insu," but, I fear, without any authority. Ptolemy describes the Ebudse as lying above, that is, to the north- 206 PROCEEDINGS'OF THE SOCIETY, APRIL 12, ward of, Hibernia, and places them in an east and west line. In so far as two of them, Epidium and Maleus, can be readily identified, it appears that whil e mainlanth e f Scotlano d s been.turneha d d froe nortmth h toward e east th e sEbudc th , e have been turned fro e soutmth h towards s bee e westha n th t I alread. y explaine broughe b they w ydma ho t into their natural position. . Ebuda.—The1 . A firs e Outetth Ebudw r no Hebrides s ai , callen di Gaelic Innis Fada, Innis-fhada,d an i.e., Long Islan whicd :an h appears in a Latinised form in the " In'chades, Incades " of Giraldus Cambrensis, for Inchf hade fhe s ;th bein surmisege b sileny Gaelicn i ma td t thatI . ' e ancienth Ebud s twa e aOute namth f o re Hebrides e (whicb y ma h retaine fore f Uistth mo n di , Veist, Huwyste, Yiest, lust, Guiste, Ivist, Ywest, Wyst, Ewist, West, Ywst, Vist, Wist, Oest, Ewyst, Ewyast, Weist, Vyisfc, Evist, Ywyst, Vyst, Tyrvist) d thae namth an ,t e from the bees mha n extende whole th o dt e group. . JSbuda.—2 . A Another Ebuda, place y Ptolemb d 9 m1 y. east, i.e., south of the first. The island intended is Skye. the " Soetis" of the Eaveuna Geographer, and the " Scia insula" of Adomnan; for Sgiath, i.e., wing. By a strange mistake the name of this island has wandered far awan islana o y t f Scotland do e nort ofth d f en h , wher writtes i t i e n 'OKITIS, 'Oxi/nys, Ocetis, but which in one edition of the Geography is written SKITIS, i.e., Skitis mistake Th . f omicro o esigmr fo ) a(o n(cr s i ) very n easysomi s a e, Greek alphabet e onlth s y difference betweee nth two is that the latter is not entirely closed. The distance from Uig in Skye, the embarking place, across the Minsh, abous i t 17 miles. . . Khicina.—3 . A The Ptol. distance fro secone mth d Ebud Khicino at a is 38 m. east, i.e., south. In one copy 'PiiaVa is written 'Eyyapi/ceva, and this leads me to conclude that the island Egg, the " Egea insula" of Adomnan and the " Eilean Eagach," of modern Gaelic, is meant. It may be true that in ordinary Greek Eyya would be pronounced Enga, but if thwrittee eb namislann o a t .Green n f i d o ed ha kcalle lettersg dEg I , presume it would just be Eyy. "Ehicina" has usually been identified with Eachrin d thian ,s vie ws strengthenei face th t y thadb n Pliny excludes "Eicina" fro Hebridese mth explanatioe Th . thas nnamee i th t s of two distinct islands have been confused together. ON THE PTOLEMAIC GEOGRAPHY OF SCOTLAND. 207

Tlie appearance of Egg is most remarkable, the top of the hill which forms the island being scarped, forming a basaltic cliff. This is called the "Score (or Scaur) of Egg;" but the name is a pleonasm, the "sgat- hadh" being, in fact, the " eag." From Uig, in Skye, to Egg, is about 40 miles. . Maleus.—4 . A The Ptol. distance from Ehicin Maleoo t amile7 1 s si N. 55 E.; i.e. S. 35 E. From Egg to the north end of Mull is 17 miles. Therreasoo n s ei doubo nt t that MoXeos, MaXatos "Malee th , a insulaf "o Adomnan, is represented by the modern Mull; -written " Mulle, Mule, Moyle" in charters, and " Myl" in the Sagas. The ancient form was probably "Maol," Gaelic = bald, barer WalteSi . r Scott comparee th s south side of Mull to a spot which "sets human industry at defiance" (p. 281, Lockhart's Life). It is only fair to add that although the head of Benmore is bald enough, the lower part of the island is of average fertility. A. 5. Epidium insula.—From Maleos to Epidium ins. the Ptol. dist. is 30 miles S. 70 E.; i.e. S. 20 "W. From the Eoss of Mull to the middl Islaf eo abous yi miles0 name 3 t Th f thieo . s island, Epidiums i , evidently borrowed froe neighbourinth m g Epidium promontorium. 1 " Ilea insula s firs"i t mentione Adomnay db n (unles e e Birilth th s f o a Eav. Geo. represents Islay), and-under the Gaelic form of " Ila" it occurs in the additions to the " Historia Britonum" (p. 23, Chron. of Picts and Scots) etymologs ;it obscures yi . The difference of latitude between the middle of Lewis and the middle of Islay is 145 miles, while the Ptol. dist. between the first Ebuda to Epidium ins. is but 99 miles; this discrepancy may be reconciled by supposing that the original authority gave the distances from the shore of one islannearese th o dt t nexte shorth f .eo A. 6. Monaoeda; the "Manna" of the Eav. Geo.—The Ptol. dist. from M"ovantarum prom, is 109 miles S. 85 W.; i.e. S. 5 E. The place intended is the Isle of Man, of which the nearest point is but 13 miles south fro e Mulmth f Galloway o l e largTh e. erro s referabli r e th o t e false theory causin e Mulgth f Gallowao l o represent y e northmosth t t point of Albion. When the whole body of Scotland is turned hack into its natural position, Monaoeda comes very near to Mona; from which may be inferre drepresentn tha Isle Ma f th te o s both Monaoeri Monad aan . 8 20 PltOCEEDING SOCIETYE TH P SO , APRI , 1875L12 .

. Mona7 . A insula.—This island, placed nearle middle th th n yf i o e Irish f cours Seao s Isle i ,f Maneth o e . t Ptolem(SeBu e) ynot6. , eA shows no island where Anglesea is, for which the explanation may he that having authorit islando tw r nearlf syo fo same yth e name havingd an , , in error, found positions for them, he has rejected what would have been a third, Monafacn i t t Monaoedhu ; a should have been applie Mano dt , and Mona to Anglesea. A. 8. Edri insula; the "Andros" of Pliny?—Indeterminate; perhaps the Wicklow Bank. A. 9. Limni insula,; the "Limnus" of Pliny.—Indeterminate; perhaps the " Silura insula" of Solinus, and the " Sylina insula" of Sulpicius Severus, i.e. Scilly. WEST COAST OF SCOTLAND. B. 1. Novantarum Gher. prom.—The description of the British island Albio commences ni promontore th t da e l^ovantaru th f yo m Cher.—the NeuavTa>v Xepcrov-qo-ov a/cpo f Marcian—whicvo s readili h y identified with Mule th Gallowayf lo Ptole Th .. latitudlighthouse th s a 61-4s ei d an e 0; is in 54'38 the absolute error of latitude is 422 miles. But the Ptol. latitud f Londoeo 54.0s ni thao Ptole s ,th t . differenc f latitudo e e from Londo milesreae 0 22 Th l differencs .ni f latitudo e e betwee t Paul'S n s anMule dth f Gallowayt mileo 3 lerroe 3 no 7 miles th s f i so r18 d s i an , e differencexcessiveth t Bu f longitudo e. s i enormousle n i errory , amountin longitude ° 45'Th 5 . o g t f neao e r places t thaa , t early time, could only have been known from measured or assumed distance; but this large originae errodoubtless th wa n ri t l authorityno s causes i t hu ,d by piecing together portions of the coast on a false theory. The west side of Scotland is made to bend from the Mull of Galloway towards the east, instead of towards the north, and the position of the Mull is exaggerated until it becomes the northmost point in Britain. B. 2. Rerigonius sinus.—The form of the name is now Loch Eyan. The shape of the coast leaves no doubt as to the place intended; but it is Novantf o place ) Wmile7 E. 2 1 d5 .3 . . 7 (i.e.promsS entrance . S Th . e to Loch Eyan is 21 miles to the northward of the Mull, but measured along the coast the distance is between 30 and 35 miles. B. 3. Vindogara sinus is the name of the bight or bay into which the ON THE PTOLEMAIC GEOGRAPHY OF 9 SCOTLAND20 .

Eivers Boon, Ayr Irvind ,an e debouch Ptole Th .. distance from Eerigon. sinus to Vindogara sinus is 31 miles S. 50 E. (i.e. IS. 40 E.) The true distance fro entrance mth miles0 Locf 3 eo s i h.r ByaAy o nt B.. Clotcs4 cestuarium, i.e. Firt Clyde.—Thf ho e Greek e Latinth t ,no , text of Wilberg's edition has the latitude 58° 40', an error for 59° 40'. This give . (i.e.distancsa E mile9 7 2 5 N. . f s eS o ) fro 6 1E. m Vindogara sinus o spenT . d this distanc t i muse e supposeb t d thae originath t l distance was measured along the coast, in which case 57 miles would reach nea Glasgowo t r . The form of the name has undergone little change in eighteen centuries. Tacitun I s "i "e Clota,t sCledi th d misplaces "i an Irelano de t th y db Kav. Geo. Bede names " Alcluith," i.e. Rock of Clyde, and it is quite possible that under "Oleiclavis, olea clavis, oleclavis," the same is meant e Eavbth y . Geo t I occur. s "Alteluta s , Alclut, d "Stratclutan " " (Welsh); "Alocluade, Ailecluathe" (Gaelic), in the native chroniclers (Chron. of Picts and Scots.) . Lemaannonnius5 . B sinus.—The name intende Locs di h Lomondt bu , the place intende s Locdi h Fyne, near Ardrisheag Ptole Th . . distance from Clotte aert. to Lemaannonnius sinus is 56 miles N. 69 E. (i.e. JST. 21 W.) The true distance from Glasgow, in a straight line, to Ard- risheapassaga miles2 se 4 e s gs i verei th ; y tortuous longes i d ran , than 56 miles. Both Lomon Leved dan n have been showSkenr dialectie M b y o net b c forms of Leman, produced by the different aspiration of the consonants. B. 6. Epidium prom.—The 'EmSiov aKpov is the Mull of Kintyre, one of the most distinguishable points upon the west coast. The Ptol. distance from Lemaan. sinu Epido st ) . W. prommile 0 3 . 5 (i.e. 5 s W i s. , S N. 7 3 The distance from Ardrisheag to the Mull is about 50 miles. This is a good station from which to test the value of the Ptol. position f JSTovanto s . prom d Vindogan , . sinus. The Ptol. distanco t e Novant. prom, is 58 miles; the true distance is 52 miles. The Ptol. distanc o Vindogt e . 0 milessinu5 e trus i sth ;e distanc 1 miles4 s i e. This is sufficiently near, for it may be observed that the positions are seldom noted nearer than to one-sixth of a degree, i.e., 10 miles. Epidiu mprobabls i ytranslatioa native th f no e name. B. 7. Logffi fluvii ostea.—Aoyyv is translated " Longi," but the Greek VOL. XI. PART I. O 210 PROCEEDINGS OF THE SOCIETY, APRIL 12, 1875.

word doe t measno n "long, t i"i s more likely thaKeltie th t c word " loo" is expressed. In the same way Eyya has been translated " Enga," which appears to me very misleading. forminn I opinion g a wha o t s a nt plac intendeds i e t musi , e bornb t e in mind that but four stations are indicated between Caithness and Kintyre. . . The Ptol. distance from Epid. prom, to Loggi flu. ost. is 29 miles east (i.e. north); but as this would only reach to West Loch Tarbert in Kin- tyre, the reading of A is rather to be adopted, viz., long. 25°, which gives 50 miles from Epid. prom. This reaches to the entrance of Liniie- dubh, commonly called Loch Linnhe, betwee e islandth n f Mulo sd an l Suilwhicd stationan ,e hI believth e b . o et B. 8. 'Irvos worajtiov eK/3oAeu,—There are no rivers of any importance between Cape Wrath and Kintyre, the watershed, Drumalbain, being locha nea e coastse e quit th ar rst e man eth bu , riverinf yo aspectn i e . The Ptol. distance from Loggus flu. to Ityis flu. is 87 miles E. (i.e. N.); but adopting the corrected longitude (25°) for Loggus flu. the distance is miles8 5 frod an ,m Epid. prom. 116 miles. This reache Soune th o t sd of Skye, in which a group of lochs or estuaries meet. The only name beginning with a vowel is Loch Alsh, which the Highlanders call Loch (prons Oi r . o Aish s havy Ai ,ma eOish) e here—aW . s with Hebrides, Islay—a permanent chang ''Irtioen i y fro d spellingr ma s m ba e th r o ; hav same th bee et nA tim writte. t musi eA e remarker b t nfo d that the spelling of A, "Injov, is curiously near that of Loch Etive,—and the name e frequentlar s y misplacede Gaelith t c Bu for. f Locmo h Etive appear "loe b co st n'Eitch, Eitche." vol, (Pp32 . i "Trans., .20 . Gaelic Soc.") B. 9. Volsas sinus.—The Ptol. distance from Ityis flu. to Volsas sinus . (i.e.E e tru 0 mile 9 8 E.N0 Th 5 e . 1 .)s S i sdistanc e from Loch Alsh to Loch Inver (Assynt miles6 5 s )i . M. has.'OvoAas s usualli . M yd an , right. The position is either Loch Inver or Loch Broom. I suspect " Assynt" (Easan, fro e waterfallsmth s concealei ) d unde e forth f rm o "Volsas." B. 10. Nabd flu. ost.—The Ptol. distance from Volsas or Tolas sinus to Nabei flu. ost. is 29 miles E. (i.e. N.) This is the distance from Loch Inver to Cape Wrath, but the name is evidently the equivalent of ON THE PTOLEMAIC GEOGRAPHY OF SCOTLAND. 211

the modern Navir. Here, then bees ha , n confusion; the position thas i t of Tarvedum prom., of which the name has been carried to the next station, while the latitude and longitude of the Navir have been entirely omitted. Cape Orcaf courso s i s e Dunne o namestw e t ,th Head d an , Orca Tarvedumd san , have been confounded together distance Th . e from Loch Inve Capo t r e Wrat thio miles8 t 2 s d s hi poinan , distancee th t s fro Mule mth f Gallowao l accurate yar degrehava o t o et t e no ebee n expected.- The name Tarvedum, the Tauroedunum akron of Marcian, appears to be continue Parfn di . Farrif. Parwe modern ;i n Gaeli cBarb n "A h (pron. Tin Bar-ruv);" in Lewis, " An Garbh" (pron. Un Car-ruv),1 .unless the words represen Norse th t e "Hvarf" = turning-point Northmee th t Bu .n f alterino y wa hag a dname o suit s t their fanc s modery(a n sailoro d s still); witness Liodhus for Leoghas, and Skid" for Scith. Fro e abovmth e , stationPtolem10 . e B ,y th bendo t e coas ° th s 21 t eastward, an additional proof that the station is Cape Wrath; but this change of bearing is not sufficient, for he should have turned it 90°. The consequenc f thio e s derangerroo t s i re succeedineth g positionn o s north-ease th t coas Scotlandf o t . . Tarvedum11 . B sive Orcas cautes; Orcae th s akro f Marcian.—Ano s noted above, this station is Dunnet Head. The Ptol. distance from Nabius flu. to Orcas cautes is 43 miles 8. 49 E. (i.e. N. 21 E.) The distance from Cape "Wrat Dunneo ht t Head miles8 ie Navis4 th o t ;r (pron. Noar) it is only 25 miles.

SOUTH-WEST COAST OF SCOTLAND. Having described wha n erroi t norte e supposeh rth e hb coaso t sf o t Scotland, the Geographer returns to the Mull of Galloway, and describes the north shore of the Sol way Firth and the west side of England. In error he represents the south coast of Scotland to be continuous in e samth e direction wite westh h t coas f Englando t , i.e.,s runnina g north instead of west. But besides the error of direction he exaggerates the length excessively, and for which he could have had no authority in numbers onln ca y I suppos. a sketce coas th d ef t ha o htha befor e h t e him havind an , g fro mfalsa enearlf o theor mile0 p y ga 20 fil o syt a l up— 1 Rev. M. Maophail. 212. PROCEEDINGS OF THE SOCIETY, APRIL 12, 1875.

whe reae nth l distanc less sewa tha miles—h0 n7 e .has applie falsda d ean exaggerated pla s scal vacanhi e orden i o th et filp o t ru tl space. inferencn A somf eo e importanc gatheree b y ema d from this error, viz., that many more places were names authoritiehi n di enus s ha tha -e nh merate tables e gives th her r ha n dstatioa ni e fo , e h t abouna t ever0 y1 miles, while elsewhere the average is 30 miles apart. orden I instituto t r compariso. ea n wit e reahth l feature s e coastoth f , the Ptolemaic distances must be reduced two-thirds. C. 1. Novantarumprom.—The Mull of Galloway; see note, B. 1. O. 2. Abravannits flu.—The Ptol. distance from Novant. prom, to Abra- vannus flu. is 63 miles S. 51 W., which reduced is 21 miles S. 39 E. The place intended is the Cree, the upper part of which is still called the Mannach e loweth e rived r th paran ;f ro t must have born e sameth e e positios timee showi th namon s a y ,t f b na Monny-gaffeeo n ; i.e., Mannach-gaffe (p. 20 "Symson's Galloway"). The "mouth of the Mannach" woul e "b d Abe r Mhannaich" (pron. Aber Vannich)e Th . distance fro Mule m th Gallowaf o l Wigtowo yt miles1 2 s ni . C. 3. lenae aastuarium.—The Ptol. distance from Abravann. flu. to lenae aest. is 32 miles S. 19 W., which reduced is 10 miles S. 71 E. aestuare e placFleee Th th th s f ei tyo intended, whic s abouhi mile6 t s from Wigtown presene Th . t e rivenamth Teutonics f i r o e n Norsei ; , " Fliot." . GDevae4 . flu. ost.—The Ptol. distance from lenae aest o Devt . a flu. W.5 mile3 4 i4 s . , S place swhic Th eh . E reduce 5 mile4 4 1 . s S si d 5 t e distancbu th Dee e s t i th e bu ;s e i De betwee e th e Flee d nth an t miles. . GNovii5 . flu. ost.—The Ptol. distance from Dev a mile2 flu3 . s .i sS 18 E., which reduced is 11 miles K 72 E. The place is the Nith, and the distance between the Dee and Nith may be taken as 23 miles. . ItunaeC6 . aest.—The Ptol. distance from Novius flu Ituna. .to e aest. mile5 iE.6 4 s . , S swhic he plac reduce Th mile5 e1 . s E si d 4 N8 . Edene th monte namew th , no whicf hmile0 s o di 1 s hi s fro e Nithmth . Fro Edee e directiom erroth e th e meridianth n th i r f o n n ceasese th ; Antonine Itinerary will check the position of the towns, and from thence thef eaturee coast th n Scotlan I f o .s e solth de corroboratioe b o t s ni e Anonymoufounth n di s Geographe Ravennaf o r , whose string namef so s PTOLEMAIE ONTH C GEOGRAPH 3 SCOTLANDF 21 YO .

are so many puzzles, the greater number of which are likely to remain without solution; he seems to be beyond the reach of Grimm's or any other law. The persistency in the same form, or almost the same form, in the local names in this quarter is to be remarked; the station next south of the Ede s n"i MO/D/CO//,/?^," whic representes i h modere th n di n "Morcambe" same Bath en yi letters .

EAST COAST OP SCOTLAND.

As Ptolemy make ease sth t sid f Englaneo n north ru d thao df t an ,o t Scotland to run east, he naturally speaks of the east coast of Britain as having two sides. goinn I g dow ease nth t coas Scotlanf o t t dmusi rememberede b t that "Orcas cautes s beeha " n projecte a bearin y b d g from Cape Wrath (Nabeus flu.) far to the northward of the real position of Dunnet Head. It is necessary to start from this imaginary point to arrive at any argu- ment betwee nPtole naturth .d positionsean . D. 1. Orcas cautes.—Dunnet Head: see note, B. 11. D. 2. Virvedrum prom.—From Orcas cautes to Virvedrum prom, the Ptol. distance is 18 miles S. 34 W. (i.e. S. 56 E.) Having projected Orcas cautes to the N.N.E. instead of to the east, Ptolemy then turns the coast S.W. for 43 miles; Orcas, Virvedrum, and Verubium proms, being e placesamth en i drigh t line. Virvedrum prom bearw e no ,th s Norse nam Duncansbf eo y Head (for Dungalsbo mile2 e1 Hofdi) s i s d an , eas Dunnef o t t Head. D. 3. Verubium prom.—From Virvedrum prom, to Verubium prom., bearine Th e Ptol g) th froE. . . 3 m(i.e.distanc5 W . S mile7 5 3 2 . s ei sS Duncansby Head to Noss Head is south, and the distance is 14 miles. From this point Ptolemy turns the coast west (i.e. south); but the true directio s south-westi n e presenTh . t nam f Noso e s Hea s Norsedi ; Snos Hofdi, meaning a projecting rock. There are other two Noss Heads in. Shetland. . Ilaeflu.4 . D ost.—From Verubium prom Ilafluo t , Ptole th . . distance is 15 miles west (i.e. south). Although more generally called the Helms- dale River, its Gaelic name, the Illy (properly Amhainn Thuilaidh, i.e. 4 21 PROCEEDING SOCIETYE TH F SO , APRI , 1875L12 .

Floody Biver, Torrent Eiver stils i ) l known r EoberSi . t Gordon) 4 . (p 1 names "the Holmisdell or Vlly;" at p. 3 he speaks of "Strathvely." The distance from Noss Head to the Helmsdale Eiver is 30 miles, and moreove beyons i t i r "Bipe dth a alta." D. 5. Ripa alta.—There is not a more remarkable place on the east coas f Scotlano t d than Ochthe Huphele' "Are th , d Chaitaobh,d Or r o " of Caithness; it appears from the sea as a range of cliffs of enormous northware th o t s i dd e Helmsdal oth Or f e heightth t eBu Eiver. e th ; river is in fact the southern boundary of the Ord. The Ptol. distance from Ila flu Eipo t . a0 mile 3 alt s i sa west (i.e. south). Ther s greai e t confusion here,.whic explainee b y ma h d awa thin yi s manner firse Th t . station after Verub.-prom, should be Eipa alta, 30 miles; the Ila flu. must be deleted, for it is the same Ptolemaically (for the Ptol. -unit, is 5 miles) as Eipa alta; the second station is—the ancient name not recorded, t modern—thbu e Dornoch Firth miles5 1 , thirde th ; , Loxaflu. miles5 1 , . D. 6. Loxae flu. ost.—The Ptol. distance from Eipa alta is 15 miles west (i.e. south). The Keltic loc is evident in Ao£a, but the Dornoch and Cromarty Firths (Fiorflar e equallar ) y s probabli lochs t I . y Cro- marty precedinFirte se h: g note. D. 1. Varar aest. -^-The latitude of this place, 59° 00', given in the text is evidently wrong, and the various reading, 59° 40', must be accepted. "With this latitude the Ptol. distance from Loxa flu. is 30 miles west (i.e. mile5 1 t s bu fro.south) s i m t I Cromart. y Firt Invernesso ht l al n I . the large estuaries, Clyde, Murray, Forth, Thames, the positions are un- satisfactory, partly from the nature of the case (for who can define the positio larga f no e singla are y ab e point1?) while wher coase eth straighs ti t the positions are laudibly approximate. The name Varar, here preserved, is interesting: the Gaelic form of Murray is Marthaobh, which ia the genitive cas Mharthaibs ei h (pron. Var-ive); thu have sw e Moreb, Morref, Muref, Murray. D. 8. Tuaesis aest.—The Ptol. distance from Varar aest. to Tuaesis aest. is 43 miles S. 22 "W". (i.e. S. 68 E.) Tuaesis is certainly Spe, in whic initiae hth l letter e corruptedsar distance Th . e from Invernessy b , sea Speymouto t , abous hi miles0 4 t . JJ. 9. Celniiflu. ost— The Ptol. distance from Tuaesis flu. to Celnius 1 Genealog Earle th Sutherlandf f so yo . ON THE PTOLEMAIC GEOGRAPHY OF SCOTLAND. 215

mile5 1 flu. is s. south (i.e. east). From Speymout o Culleht mile0 1 s nsi east. I suspect, however, that the Douveron is the station intended. D. 10. Taezolorum prom.—The Ptol. distance from Celnius flu. to the promonotor . (i.e. ) E e Taezol ProE. 6 th mile2 4 N2 4 f my . 4 o .s S i si Cullen to Kinnaird Head is 26 miles east. . Devaeflu.11 . D ost.—The Ptol. distance from Taezolorum promo t , Deva flu. is 31 miles west (i.e. south). From Kinnaird Head to the Dee in a straight line, is 36 miles, by sea it is about 40 miles. D. 12. Tava' aest.—The Ptol. distance from Deva flu. to Tava aest. is mile1 3 s west (i.e. south). Ther hers ei edisplacemena f nameso t e th ; place intended is the South Esk, which is 37 miles from Aberdeen. . Tinna13 . D flu. ost.—The Ptol. distance from Tava aest Tinno t . a mile1 3 flu s i s. west (i.e. south)y Ta . e Froth f mo r Montrosba e th o t e miles0 i2 s Dundeo t , miles9 place 2 e Th e. intende Taye th ,s d i whic h nam bees ha en transferre Soute th o hdt Esk; whil e importanth e t river s namTynit s e eha omitte d fro ordee mth whicn i r t i hshoul d have been written, and is introduced here. There is another Tyne in East Lothian, whic havy hma e helpe confuso dt geographere th e . . Boderia14 . D aest.—The Ptol. distance from Tinna flu o Boderi.t a e actuamile8 aest4 Th s l si .) W.distancK2 E. 7 8 (i.e.1 . e S from Tay Bar to May Isle is 18 miles; but from Dundee, by sea, to Inchkeith abous i miles2 4 t . Boderi writtes ai n Bogderi . (fioySepiai's,M n i a where the is is put for eio-fuo-cs); and Bodotria by Tacitus. It is probably the same word which afterwards appears as Fothren and Fothreve (for Fother thaobh 1n Welchi ) : , Werid d whic an s ; beeha h n abraded inte th o modern Forth. D. 15. Alauniflu. ost.—The Ptol. distance from Boderia aest. to Alaunus ) e TweedThiE. th 0 . s (i.e.i 3 s W ,. 0 S whicmile 0 6 3 . flus S s si hi . firse th t rive importancf o r e sout he Firt oth f f Forthho Tweee Th s .i d 30 miles to the southward of May Isle (the Maiona? of the Bav. Geo.), t wouli t bu d tak mile0 e2 s mor reaco t e o Inchkeitht h s beforA - . re e markede largth l e al estuarie, e unsatisfactoryar s e namTh e. Alaunus belong Alne th ,o st whos e geographical positio t giveno t alls na i : this conclusio supportenis relativthe dby e positio Bremenuimnof . D. 16. Vedrae flu. ost.—From Alaunus flu. to Vedra flu. the Ptol. distanc mile7 4 s i es west (i.e. south). e Vedrworth ds i a Tweed with 6 21 .PROCEEDING SOCIETYE TH F SO , APEI , 1875L12 .

the'i! aspirated; thus Thvedra. pron. Vedra e samTh e. grammatical accident occurs at the Thames, where " lamesa "is written for " Thamesa." iBut this statio factn i e Tyn , th ,nis e j which nam s beeha e n removeo t d the Tay. (See D. 13.) From this station the error in direction ceases; it is difficult to imagine how it could have originated. It is evident that it was at this point the mischief began, for although the direction is changed from north to east, the distances continue moderately correct, while the dislocation and error causefalse th ey dbearinb l throwal s gi n int e nortoth he Sol th sid f -o e way .Eirth. geographee Th r proceed describo st e coaseth t southward Cantiuo t s m promontorium Soute th w h Forelandno , : from thenc returne eh whao st t islande th f norto e believe, e th namind h e hen b differeno e st g th t peoples or tribes, and giving the geographical positions of the principal towns. Beginning wit N"ovantae hth Selgovaee th d ean firs e h , t proceeds eastward (i.e. northward) to the end of the island; he then returns to the Selgo- vaegoed an ,s southward, namin peoplee gth d townsan w n whai sno s i t England. SCOTTISH SEPTS AND TOWNS. So long as the geographer is describing the coast, naming the features thein i r natural order, therdifficulto n s ei identifyinn yi stationse gth t ,bu it is far otherwise when he professes to give the geographical positions of townse th f thesI . e positions were even approximately correct e siteth , s could be easily determined, but in the south of Scotland they are grossly inaccurate. The reason of this will be readily seen by comparing the two maps largee Somth f ero town Englandf*sucn si Londos ha d Yorknan , bear names almost identical, and hold the same relative importance that they did 1600 years ago; but in Scotland, with one exception, perhaps (Aberdeen), the centres of commerce bear names which are not to be recognised in those transmitted to us by the Alexandrian geographer. Borne help is afforded by the natural features- of the country; no towns lire wese give th f Scotland o tn i , e nortClydeth f ho , while they appear havo t e been placed where they were naturall havo yt e been expected, as centres .of fertile districts and upon the main roads or tracks which traverse countrye d th e RavTh .. Geographer gives long lists t unforbu , - ON THE PTOLEMAIC GEOGBAPHY OF SCOTLAND. 217

turiately nothing but the slightest verbal'resemblance by which'their sites can be identified. In the south of Scotland any conclusion or compari son between the Ptolemaic and actual geography is untrustworthy by reason of the great distortion. E. 1. Novantae.—The country of the Novantae is described as being under—really, to the eastward of—the Chersonesus, now the Ehinns of Galloway generalls i d an , y modere supposesame th th s e a b n o dcountt y of Wigtown. . Lucopibia.—2 . E The nearest given plac Lucopibio t e s leni a a flu. whic hI hav e identified wit Fleete hfroth d man ; whic s madi e b t hi o et mile0 1 s south; when corrected mile3 , s dependenco eastN . e shoule db placed in this result. Lucopibia probably occurs under the form, of " Lucotion, Lucocion," in the Eav. Geo. Glen Luce is the only name in the district which bears any resemblance to it. Camden has proposed to identify it with Whithorn, but it is more likely that Wigtown, which is not a Keltic name, is the place intended. E. 3. Rhetigonium.—This is an error for Rherigonium, and it has to be e sam sinuse shore th th th f elookef o n eo s nameprobablo i r t dI fa . y commemorate Eyanm neighbourhooe Ca th , n di capitae whicf dth o s hlwa of this quarter before Stranrae builte nams Th wa rn tha i e. t collection of riddles left us by the Eav. Geo. is " Brigomono." E. 4. Selgovae.—The Selgovae are placed to the east of the Novantae, with the Damnonii to the north and west; the Otadeni to the east, and the Brigantes to the south. They appear to be the Segloes of the Eav. "e Solwath Geo, d an , y probabls "i modere yth nsame forth f em o word . E, 5. Carbantorigum.—The Carbantium of the Eav. Geo. may be some "cathir" nea Deee th r . E. 6. Uxellum.—The Uxella of the Eav. Geo. is some place on a height, apparently not far from the Nith. . Cor1 . da.— E Also e RavCordth f o .a Geo. ; site unknowne Th . latitudes and longitudes here are useless, having been made to suit a false theory. . Trimontium.—8 . E Also Trimontiu e Eavth f .mo Geo Ptole Th . . position is such that it might represent Birrenswork at Ecclefechan. Whe thougho n nd therha t I etha e Trimontiut i mightth e b t f mo Ptolemy, and did not notice if such a name could be applied to it, nor do 218 PROCEEDINGS OF THE SOCIETY, APKIL 12, 1875.

I find notice of such a configuration elsewhere. The name is a translation, and descriptive—the Three Hills. Uxellum is descriptive of Birrens- work, while Trimontium is so thoroughly descriptive of the Eildon Hills tha I thav e great reluctanc o believt e e place th tha t stno thee yar intended. . Damnonii.—9 . E The Dannon Eave th f e Selgova.o iGeo th d .ha o et e Otaden soutth e e east d seeth d th han o ,an o havt m t i e included ithe entire basins of the Forth, Clyde, and perhaps the Tay. To the northward they were bounde Veniconee th e Caledoniie y th dth b o d t san d an , west partly by the Gadeni. Three of their towns are placed to the south- Forte war th Clyded f hnorthwardde an o thred th an , o et . In the article " Damnii" (Smith's "Greek and Roman Geography "), the write s overlookerha erroe dth r mad Ptolemy e b e directioth e n th yi f o n meridian ; that for "east " must be read " north," and for " north " must be read " west," and so on. The form Damnii is not quoted in Wilberg's edition. A people of the same name, Dumnonii (Dalesmen'?), occupied Cornwall and Devon gavd an ,e their lattee namth o ret country. . Colonia.—10 . E Also Colanica. Colonica, Eav. Geo. Positio- nin determinate. K. 11. Vanduara.—Vandogara, a various reading, is almost identical with Vindogara sinus althougd an , h Ptolemy place t thirtsi ye mileth o t s eastward, I believe it to be the same, and consequently that it is now represented by the town of Ayr. . Coria.—12 . E Position indeterminate Eave Corie th Th f . ao Geo. s .i named with the towns near the wall of the lower Isthmus. E. 13. Alauna.—Alauna is named by the Eav. Geo., but no doubt it represents Alnwick e threTh e. town f Alaunao s , Victoria 'Orread an , , are place Ptolemy d b same th en y i righ t line. From Alaun Victorio t a a is 31 miles ; from Victoria to 'Orrea is 21 miles; total, 52 miles. The three towns of Stirling, Perth, and Forfar are in one right line nearly; from Stirlin Perto g5 milest 2 s hi ; from Pert o Forfa5 milest h2 s i r ; total miles0 5 ,f thiI .s theor corrects yi , Lindum Callander= . E. 14. Lindum.—Lindum is 13 miles K 39 E.; corrected N. 51 W. from Alauna. Callander is 12 miles from Stirling. E. 15. Victoria.—Victoriae: Eav. Geo. Near Perth: see note E. 13. PTOLEMAIE ONTH C GEOGRAPH 9 SCOTLANDF Y21 O . Here the Greek MSS. have "TheGadeni farther to the north" (i.e. west); but this is omitted from the early printed editions. Supposing the existence of the Gadeni sept, they appear to have occupied Cowal; i.e. the country between Loch Lomond and Looh Fyne. E. 16. Otadini.—The Otadini are placed to the south (i.e. to the east) Damnoniie oth f theid an , r town Curie Bremeniumsar d aan . They occupied what is now the county of Northumberland, and perhaps Roxburgh Berwickd an . E. 17. Curia.—This station when projected comes about 30 miles N.E of Bremenium e dependenco b N y . e placema b t n i n that o deca t bu ; guesse soms i t di e e Tweed placth n o e , between Tweedmoute th d an h Lyne: perhaps Berwick. After Curia editioe on , n (M.) names Alauna, long , lat23 . 58-40. That there was an Alauna in the country of the Otadini is almost certain, for Rivee th r Al neas ni Rochestero t r Alnwicd an , probabls ki s represenyit - tative ; but the given long, and lat. would place the town in Fife, which cannot be accepted. The Rav. Geo. has an Aluna near to Bremenium. E. 18. Bremenium.—Bremenium; Geo. Rav. and Ant. Itn. Wil- berg's edition Lati e hasth nn ,i text, latGreee th . 56'30n ki t vq bu , &S; i.e. 58'45, whic corrects hi Ptole Th .. data place t betweesi n Vedra flu. i.e. Tyne, and Alauna flu. i.e. Tweed; and 15 miles from the coast. Camden makes it clear that Bremenium is the present Rochester, in Northumber- land, which is 22 miles from the nearest shore, and halfway between the Tyn Tweedd ean . E. 19. Epidii.—The description of the Epidii must be read thus: Afte Damnonie th r i fro northe mth t morbu , e toward e westh s t (i.e.o t the north-wes Damnonii)e th f o t , almose nortth f o tht ho e promontory of Epidii, the Epidii. These are, of course, the inhabitants of Kintyre, of whom the Greek nam moss ei t likel yparaphrasea . E. 20; Cerones.—These are next towards the north of the Epidii. As no boundaries are given of the peoples in the west of Scotland they can only be distributed on the equal dividend principle, and in some measure according to the features of the coast. The Cerones may be supposed to have possessed Lorn and Appin, or from Loch Crinan to Loch Leven; and their name to he represented by Loch Creran. E.1\. Cerones.—These are next (north) of the Cerones, and may be 0 22 PROCEEDING SOCIETYE TH F SO , APRIL 12, 1875.

suppose havo dt e occupie e ancienth d t distric f Garrnorano t ; 'i.e., from Lochabe e Sounth o f Skyeformt re do Th lasthe'namef o o s. tw t e th f so peoples, Kfptavfsd Kpeovtsan , suggest r both sfor(o a s thae f i m)o on t Cruithne (pron. Croon-yer), 'the Gaelic name of the Picts. . Carnonacae.—22 . E Perhaps fro e Sounmth f Skydo Assynto t e d an , commemorated in Loch Carron. . Careni.—23 . E doubto Thisn , erron is a ,r Cateni rfo theid an , r land wasame sth Strathnavi s ea s largesit n i r t sense r Loro , d Eeay's Country. The whole of this part of Scotland is still known as Cataobh. Cait was e fivth e f divisono e on f Pictaviaso . . Cornavii,24 . E i.e., peopl e e horth th r point.—Th t o nea e Cornavii occupied modem Caithness, which is cut off by high hills on the south side. Catte nese th modef t whilf th i o so s i Bu t ei m geography GallthaoLhs i t i , , Lane o th rForeignersf do , wit e Highlandershth , fro s havinmit g been settled by the Northmen. In England, the " Carnavii" were seated on the lands between the Mersey and the Dee. E. 25. Caledonii.—Having named the peoples upon the coast, from Kintyre, northwards o Caithnesst , , Ptolemy describe e Caledonith s i as- dwelling between the Lemaannonius sinus and the Yarar aest. This, if interpreted strictly, means between Cowal and the Murray Firth; but it really include e mountainouth l sal d Speys an countr y , Ta e y th wes f o t and, as I believe, of Eoss-shire also. The name—which appears to mean the Woodlands, from Coille, a wood—is used with great latitude by the Roman authors, and is made to include sometimes all the inhabitants Clyde nort th Tayd f sometimed heo an ,an l thossal e e Clydenortth f ho Forthd an . In fact t becami , generaea poeticallw l no term d yan , includes all Scotland. It would take a long article to tell all that has been said of Caledonia t mucbu , whicf ho h applied much e arablmorth o et e parts of Scotland—such as fighting from carts, &c.—than to the Highlands. : Caledonius26 . E saltus.—The Caledonian Fores abovee sais b i t o dt , really to the west of, the Caledonii. This would imply that it was on the eastern side of Drumalban. But, in point of fact, the woods must hav emoor w fille lo glene sth d throughousan countrye th t , except where they had been artificially cleared. In many places, especially about Loch Laggan kine f thinth ,o d g describe e Caledoniath s a d n Fores stily lma t be seen. That these were vast woods in the Highland glens, seems ON THE PTOLEMAIC GEOGRAPHY OF SCOTLAND. 221

provee manneth y b dwhicn i r h they have been spokeagesr d fo an ,f no by what still remains; but it must not be supposed that the land was covered by a continuous forest, such as is seen in Canada, for the hills would always be bare. , Caledoni s probablwa a y represente Fidaey db h (from Fiodh = forest), e fivth e f o division e e countrae on th Cruithns th f o sf o y r Pictso e , whic hobsolete w namno s t i eCaledoni bu e ;continue b r y ou ma a o t d own time in the name of the county of Ross; for although Ros generally mean promontorysa t alsi , o mean forestsa . E. 27. Decantce: are placed north of the Caledonii. This should probably be Ducantae, which would mean the people of the Black Promon- applicable b tory d an , whao t e calles i t Blace th d k Isle, which forme th s greater part of the county of Cromarty. Kent, the south-eastern promontory of England, was the domain of other " Cantce." . Lugi.28 . —TheseE came aftee Decautaeth re th d nex o an t ,t Cornabii, i.e., they inhabited Easter Boss and East Sutherland. The Or Caithnesf do naturaa s si l division between Caithnes Sutherlandd san . The nam f thio e s people seem includo t s e Keltith e y c "loc,ma d an " imply the people of the loch district. E. 29. Smertce.—This people was above, i.e., to the west of the Lugi, and must have inhabibed about Loch Shin. Such is the description of the distribution of the tribes or peoples in the north and west of Scotland in the second century, and to those who are acquainted with the country it will appear to have all the character of truth. That the coasts and glens were well peopled at an early period is proved by Mr Anderson's very interesting map of the north of Scot- land shows whicn ha o ,site f seventy-nine th no eh e Pictish towern si Caithness, and sixty in the modern county of Sutherland. E. 30. Vacomagi.—These were below, in fact, to the eastward of, the Caledonii namee Th f fou.o sf theio r r town givene ar sfrod an m, their positions it is plain that their country included Murray, Strathspey, Badenoch d Athole Vacomagan , Th . i were probably includee th n i d "Fidaeh f Pictavia"o . E. 31. Banatia, or Bannatia.—This town is placed 45 miles west of mile4 3 Orrea d s an 1 WN.,f Victoria 6 o . . This would indicate th e neighbourhoo f Loco d h Rannoc impressio y r Loco m h s i h t i Tay t n bu ; 2 22 PROCEEDING SOCIETYE TH F SO , APRIL 12, 1875. that it was upon the Garry, say Blair Athol; the probability being that upos e naturawa th n t i l highway from name nortth southeo d ht An . (aseveran i s l other instances) e equivalen, th Banatia s i t i f Banf o f ti , f (Boineft'e, Baineffe) havy ema , been transferre errorn i t i .o dt . Tamea.—32 . E The Bav. Geoa "Tagea, s .I thinha t bu "k that that belong Taye th .o t s This statio mile2 3 . fromade s nsi E b mN8 o t e1 . Banatia, and 55 miles S. 11 W. from Tuaasis. As Tusesis is at or near e distancth s moute e Speya roay th th b d e f dan o h , from lochabero t s Kuthven Badenochn i , s mori , eassumee b tha y miles0 n5 ma d t i ,tha t an r has been dropt out of the name, and that the station was at or near junctioe th Truie Trbmie th th f r mno o e wit Speye hth . E. 33. Alata Castra.—Tirepwrrav crrpaToTreSov is undoubtedly a translation, the original of which may have been Dun Sgiath = Wing Fort, or other Keltic equivalent descriptioe Th . n suits Burgh Head (Borg Ilofdi) so well, and the fact of the ancient castle being still there, leaves no doubt in my mind that it is Alata Castra, the Pinnatis of the Eav. Geo. Ptole Th . distanc f Alato e a Castrmile9 1 ) . (i.e.s i saE W. 9 K 5 IsT1 3 . from Tusesis d Burgan , h 5 mileHeastraigha 1 n s i sdi t line froe mth Spey. . Tuassis.—34 . E This statio s certainl nwa r nea o e mout th t ra y f o h thhavy e eSpeyma beed an t ,Fochabersn a . . Venicones.35 . E Vernicones. Veniconies.— Therea choic s i f o e names, but none of them resemble the " Vecturiones" of Amuiianus. Ptolemy state Veniconee th s e belowb o t s , meaning sout factn h(i , east) of the Vacomagi, and towards the west (i.e. south); in other words, they were south-west of the Vacomagi. As Victoria was in the country of the Damnonii, it may be inferred that the Venicones occupied Angus and Mearns, that is, the present counties of Forfar and Kincardine. Together, they were one—Circinn—of the five divisions of Pictavia, represente w anname dno th s ei Mearnsy db . xxii(P .. Chron. Pictd san Scots.) . 'Orrea36 . s placei E 5 mile1 d s north (i.e. west f Tinno ) a flu.n i , other words, 15 miles up the Tay. But I have shown reason (E. 13) for supposing that Forfar is the ancient Orrea. The initial F of Gaelic names becomes aspirated (Fh), and is silent if the definite article is placed before it. (See Joyce's admirable work on the Irish Names of Places.) ON THE PTOLEMAIC GEOGRAPHY OF SCOTLAND. 223

'Orrea may be the "Porreo classis" of the Kav. Geo. It was in the country of the Horestrii (" in fines Horestorum") that Agricola took up winter quarters (A.D. 84) after the battle " ad montem Grampium," which must have been in the territory of the Venicones. . Taezali.37 . E Taizaloi. Taxaloi.—The countr f thio y s peopls i e placed towards the east (i.e. north) of the Venicones. The name does not occur elsewhere. It evidently comprised the districts of Mar and Buchan e presenth , t Aberdeenshire. e identifiedTaezalb - y ex ma i y b , haustion, with "Ce" ( = Keith?), one of the five divisions of Pictavia. E. 38. Devana.'-—Devoni; Eav. Geo. Although Devana is placed 45 miles westward from the mouth of the Dee, I doubt not that Aberdeen is intended. Aberdeen writte Gaelin i n c woul Aber-dhemhane b d d an , pronounced Aberdeen and Aberyeen. Strictly, a distance of 45 miles from the coast would reach Braemar. . 39.—FroE m Devana, Ptolemy return southware th o Tynee st th f ,do and describes the position and towns of the Brigantes, and so on through England, ending with Exeter n Cornwalli , e theH n. returne th o t s north end of Britain, and describes the islands which are near to Albion, Orcae th t sa promontory. . Ocitis.—I40 . E have show probabilite nth tha) 2 . t y(A Ociti n a s i s erro r Scitis fo r d tha e naman th ,t e belong o Skyee Ptolt s Th .. data makes Ocitis to be 50 miles E". 53 E. from Orcas prom. Fair Isle is 74 miles IT. 46 E. from Dunnet Head. This island, which is well , havinknowme o gt n surveye t whei d vera n y young man, has, likl al e northere th n islands, loss Keltiit t c name. Fair corruptioa Isl s ei f o n Fridara yPeacefu= l Island translatioa ,e whicb name y th hma f eno given e Culdeeth y b s t i o(Ceile-de) . "When ther I etooe th k t painge o t s names of as many of the " gias" and other topographical features as I could. An intelligent native went with me, and I endeavoured to write down the words from his pronunciation; but although English was the •only language spoken, the vowels and consonants were so queer and strange that I never could be sure of representing by letters the sounds whic I hheard . Since the e inhabitantth n s have emigratedy m d an , attempt is likely to be the only record of the name system of this solitary isle. If so much difficulty occurs with one's own language, how much greater must it be with a strange one—Gaelic, for instance—so that we 224 PROCEEDINGS OF THE SOCIETY, APRIL 12,. 1875.

need not wonder at the distortion and confusion to be found in the ancient geography of Britain. E. 41. Dumna ins.—The Ptol. position of Dumna is .43 miles N. 69 Wf Orcao . s s usuai prom t identifo I t l . y Dumna with Stroma (Straurns-ey = Currents-isle), an insignificant . island almost always surrounded by a raging tidal sea, about 8 miles -to the eastward of Dunnet Head. (See E. 42.) Dumnam is .placed by Pliny with Scandinavia. . Orcades42 . E inss Dorcades= , Eav. Geo.; Orkn-eyar, Norse; Insi Orce, Gaeli Whale-islandsc= . I estimated thousanthao tw t d wern i e sight at one time between Fair Isle and Sauda. The long flat beaches in many of the Orkneys are very favourable for driving them ashore. islande Th frequentle sar y nameclassie th y dcb authors t thebu , y evidently knew little about theme middlTh .f Orcadeo e s insult s placei e 4 9 d miles W5 N2 . from Orcas middle prom.e Orkneyth th t f o ebu ; s i s less than 30 miles from Dunnet Head. I propose to bring about a better agreement with, fact in this manner. I assume that the latitude and longitud Dumnf eo a really belong Orkneyse th o st vice d an , versa. Kow, 43 miles from Dunnet Head will reach to Sanda, which is but 13 miles from middle beine Orkneysth th n gf i o e ; whil mile4 e9 s will just reach to Foula (Fowl-island), a solitary island to the west of Shetland. If Dumna is not some place in Scandinavia, and that the word is derived from the Keltic dun, it would be appropriate to Foula, which is a magni- ficent natural fortification. Both Fair Isle and Foula are visible from the Orkneys in clear weather. . Thule.—43 . E From "tuath," Gaeli northc= . Tacitus tells that Thule was seen by the Eoman fleet (A.D. 84). Thule is described by Ptolemy as 73 miles long and 35 miles broad, and the middle as 80 miles from the middle of the Orkneys. Shetland abous i mile0 6 t s long anmile0 3 d s broad I have. . shown reasor fo n supposing bhat the names of Dumna and Orcades have been inverted; to obtai n agreemena n t with fact t musi , t als e assumedb o , that aftee th r mistake was made, the reported distance of Thule from the Orcades was laif fro e dof falsm th e position othen i ; r words, that Thul lais f dof wa e from Foula instead of from the Orkneys. This would suit the circum- stances r frofo , m o groupcentrtw abouts e centri o st e th f miles0 o e.8 . PTOLEMAIE ONTH C GEOGRAPH SCOTLANDF YO . 225 Thert anotheye s ei r correctede errob o t r Ptole th ; . data place lengte sth h of Shetland easwestd an t , while Shetlan nearls di y nort southd han . There is a difficulty about the " septera Hsemoda, contra Germaniam vecte" (Pom. Mela) Septe,= m Acmodae, (Pliny) I presum t bu ,hav e ew e here another Keltic name for Shetland = Hialtland, i.e. the Land of Hialt notea , d pirate. This closes the long examination of the geographical positions in and around North Britain, and the conclusion is arrived at, that whatever the> number of errors, there is collected a large amount of reality and fact hardly to have been expected at so early a period' as the second century. Proceedings of the Societ Antiqmri.ef yo f Scotlandso . VOL.XI 'PLATM V.

O C£ANITS D rfECALED ONIUS

pVir\'edrum>pront-

OCEAN IT S MR NA G E C ITS I

EUROEAE TABU1.A PRIMA. ALBIONIS INSULAE BRITANNICAE SITUS.

Zlrmm, Copti,y • I' If- -T. Utcma-s JSS