. 
 
 
EXCAVATIONS AT CARNAC. 
 

EXCAVATIONS 
 
 AT 
 
 (BRITTANY) 
 
 A RECORD OF ARCHAEOLOGICAL RESEARCHES IN 
 
 THE BOSSENNO 
 AND THE MONT SAINT MICHEL 
 
 BY 
 
 JAMES MILN 
 
 EDINBURGH 
 
 DAVID DOUGLAS, No. 9 SOUTH CASTLE STREET 
 
 MDCCCLXXVII. 
 
Printed by K. & R. CLARK, Edinburgh. 
 
PREFACE. 
 
 IN the summer of 1873 I went on a tour in Brittany for the 
 purpose of seeing some of the antiquities for which it is so 
 famous, my chief object being to visit the celebrated align- 
 ments of Carnac. 
 
 Landing at St. Malo, I proceeded by the beautiful scenery 
 of the Ranee to Dinan, and thence on to Morlaix. There a 
 few days were pleasantly passed in making excursions to visit 
 the curious churches and calvaries of St. Thegonnec and St. 
 Guimiliau, and in examining the houses of the fifteenth, 
 sixteenth, and seventeenth centuries, which are still to be 
 seen in the Grande Rue, and in the Rue des Nobles at 
 Morlaix. These picturesque specimens of Breton architecture 
 usually present their gable ends to the street, each successive 
 storey projecting over the one below it, and they are often 
 adorned with quaint carving on the woodwork. 
 
 Proceeding onwards and making short stoppages at Brest, 
 Pleybe, Ouimper, Concarneau, Pontaven, and Quimperle, I 
 eventually arrived at Auray, where carriages can always be 
 obtained for Carnac. 
 
 Shortly after my arrival at Carnac, my attention was drawn 
 to the Mounds of the Bossenno by a French archaeologist, 
 
vi PREFACE. 
 
 who stated that he had discovered and intended to excavate 
 them. In the end of the summer of 1874 he renounced his 
 intention, and I then undertook the work, which I super- 
 intended and carried on at my own expense. On becoming 
 better acquainted with Carnac and its surrounding country, 
 I found that these Mounds were perfectly well known by 
 the name of "Caesar's Camp," that they had long attracted 
 the attention of antiquaries, and had been mentioned in 
 archaeological works, though no attempt had been made to 
 ascertain their true character by systematic excavations. 
 
 Foreseeing the advantages to be derived from possessing 
 the ground occupied by the Mounds of the Bossenno, I 
 endeavoured to purchase it, but difficulties too numerous 
 to mention compelled me to abandon the idea. Had I 
 succeeded, many difficulties as to obtaining permission to 
 excavate would have been obviated, and much expense in 
 carting away the excavated materials would have been saved, 
 as the stones could then have been utilised in constructing 
 a wall round the ruins, so as to preserve them from injury. 
 
 There still remains much to be done at the Bossenno to 
 complete the excavations described in the following chapters, 
 and I may perhaps continue these excavations if the necessary 
 permission can be obtained. 
 
 In compliance with the wishes of friends in France, the 
 description of the excavations of the Bossenno was first written 
 and printed in French. Every one knows the difficulties 
 and disadvantages of writing in a foreign language ; I 
 therefore had recourse to my friend the Count de M artel, 
 and to M. le Men, Arche"viste of Ouimper, to whom I am 
 
PREFACE. vii 
 
 indebted for revising the text of the French edition. I am 
 also indebted to M. Paul Gervais, Professor of the Museum 
 of Natural History, Paris, for classifying the bones, and to 
 M. H. Cohen, Attache of the Bibliotheque Nationale, for 
 classifying the coins found at the Bossenno. 
 
 The wood engravings in this work have been executed by 
 M. Emile Nicolay ; the pottery is mostly produced from 
 drawings by MM. Henri Du Cleuziou and Louis Cappe" ; 
 while the plans, landscapes, and statuettes are from my own 
 drawings. Although the woodcuts introduced as vignettes at 
 the end of the several chapters in the text of this volume have 
 no direct bearing on the excavations at the Bossenno, the 
 objects thus represented are generally remarkable for their 
 local or special characteristics. Amongst these I may notice 
 the Lech at Plouharnel, as an example of a peculiar class of 
 sepulchral monument, believed by many antiquaries to be 
 the intermediate link between the ancient unsculptured 
 menhir and the sculptured headstone of later times. 
 
 I have much pleasure in expressing my obligations to 
 the clergy and the authorities for their kind offices, to the 
 proprietors of the ground for their permission to excavate, 
 and to the worthy proprietress of the Hotel des Voyageurs 
 at Carnac for placing all the resources of that well-known 
 establishment at my disposal throughout my stay at Carnac. 
 
CONTENTS. 
 
 PREFACE ........... 
 
 I. THE BOSSENNO. 
 
 Description of Carnac Occupation of Armorica by the Romans Juxta- 
 position of Celtic monuments and Roman ruins Researches of the Poly- 
 mathic Society of Morbihan Unexplored Gallo-Roman remains The 
 Bossenno, a group of mounds, otherwise called " Csesar's Camp "- 
 Etymology and legends of the Bossenno . . . . 1-8 
 
 II. MOUND A. 
 
 Method of the excavations Discovery of a ruined dwelling Its plan answer- 
 ing to the villa urbana Description of the rooms Evidences of its 
 pillage and destruction by fire Traces of earlier dwellings on the same 
 site Description of the relics found Stone implements Superstitious 
 use of these Objects in bronze, iron, and glass Food refuse Kitchen- 
 midden of the dwelling Its pottery The pottery of the Dolmens 
 Recent hand-made pottery of the Hebrides Ornamentation of the 
 pottery of the Mound A Resemblance to Roman pottery found in 
 London Head of Venus in terra cotta .... 9-24 
 
 III. MOUND B. 
 
 Description of the excavation Discovery of a building 62 feet by 47 
 Description of its eleven rooms Curious hypocaust Objects found in 
 
CONTENTS. 
 
 % 
 
 it Mural decorations in fresco The cubicnla or sleeping apartments 
 of the mansion Traces of earlier buildings on the same site Celtic 
 brooch in bronze The kitchen of the mansion Concrete floors 
 Relics found in the different rooms Objects of iron, glass, and pottery 
 Roman coins Ornamentation of the pottery found Its persistence 
 in Brittany to the present day Modern Breton pottery Reconstruction 
 of the ancient vases found in the ruins. .... 2 5-5 
 
 IV.-MOUND C. 
 
 Its connection with the previous mound Discovery of the ruins of a Gallo- 
 Roman bath Description of its rooms, furnaces, and general arrange- 
 ments Cold and hot baths Ornamentation of the ceiling in polychrome, 
 and geometric patterns of shells inserted in the plaster Reconstruction 
 of the patterns Objects found : Coins, pottery, bones, beads, armlet, 
 whorls, sinkers, etc. . . . . . . . 51-74 
 
 V. MOUND D. 
 
 Excavation of the ruins of a small temple or Lararhtm Description of its 
 construction Objects found in it : Statuettes of Venus Genetrix and 
 Anadyomene, pottery, tooth of a bear, broken stone hammer, beads, 
 coins . 75-86 
 
 VI. MOUND E. 
 
 Discovery of the Ruins of the -villa agraria Description of its five rooms 
 Objects found in them : Pottery of Celtic type, iron implements, 
 bronze ornaments and engraved signet-ring, glass, potters' stamps 
 Statuette of an ox in bronze Excitement of the workmen on its dis- 
 covery Description of " The Pardon of St. Carnely " Nocturnal 
 processions of the bestial of Breton farms to the Fountain of St. 
 Carnely Fete of the Solstice Passage of the bestial across the 
 fires of St. John Bonfire on the Mont St. Michel . . 87-102 
 

 
 CONTENTS. xi 
 
 VII. MOUND F. 
 
 Description of the ruins excavated Concrete floors Discovery of a 
 Smith's forge Objects found in the different buildings : Polished 
 stone axe, objects in bronze, pottery, iron implements, etc., 
 coins 103-112 
 
 VIII. MOUND G. 
 
 'Evidences of conflagration Objects found in the ruins Modes of 
 sepulture Excavation of an alignment of small Menhirs Diggings 
 in a ruined Dolmen Tiles, Samian ware, and flint flake found 
 under a large Menhir Results of similar researches by French 
 archaeologists . . . . . . . . .113-122 
 
 IX. MOUND H. 
 
 Discovery of the wall of the enceinte Rude construction of building 
 Objects found : iron, glass, pottery, coins .... 123-128 
 
 X. SUMMARY AND CONCLUSIONS. 
 
 Evidence of the existence of a Gallo-Roman settlement at the Bossenno 
 Characteristics of the buildings The villa urbana : its baths ; its 
 Lararium The villa agraria : its dependencies The enceinte or 
 wall of defence Deductions from the relics as to the condition of 
 the people Their food Luxury of the table The ceramics of the 
 Bossenno Modes of Manufacture and ornamentation Intermixture 
 of objects characteristic of two civilisations, Roman and Gaulish 
 Celtic mythology The two oxen of Hu-Gadarn The Taurus 
 Trigaranus The coins of the Bossenno Period of the settlement 
 fixed by them The Pagan worship of the Bossenno Legend of 
 its destruction . . . . . . . .129-144 
 

 xii CONTENTS. 
 
 XL MONT ST. MICHEL. 
 
 Description of the tumulus of the Mont St. Michel Its excavation by M. 
 Rene Galles Its contents Discovery of structures at its base, pre- 
 viously unknown Their excavation Description of the ruins L, M, 
 N, and O ; and objects found in them These structures not Gallo- 
 Roman, but probably of the Middle Ages Resemblance of O to 
 the primitive Christian churches of Brittany The Chapel of St. 
 Agatha : Fresco in it of Venus rising from the sea ; described in the 
 twelfth century as " Ecclesia Sancti Veneris " Saint Vener Vener 
 as the Breton name of Venus Significance of the superposition of 
 structures of such different ages and character on the Mont St. 
 Michel Traces of three great periods, Celtic, Roman, and Christian, 
 in Brittany . . . . . . . . . 145-162 
 
LIST OF ILLUSTRATIONS. 
 
 VIEW FROM THE BOSSENNO 
 
 Frontispiece. 
 
 i. THE BOSSENNO. 
 
 PLATE I. Plan of the country round Carnac 
 PLATE II. View of the Alignments of Kermario . 
 PLATE III. General Plan of the Bossenno . 
 Croaz-ar-Guen . 
 
 PAGE 
 
 3 
 4 
 6 
 
 2. MOUND A. 
 
 A. PLATE I. Plan of the Dwelling under the Mound A 
 A. PLATE II. Figure i. Stone Burnisher 
 2. Flint implement 
 3. Stone Implement cut in facets 
 A. PLATE III. Figure i. Bronze Buckle 
 ., 2. Bronze Ring 
 3. Cup in red lustrous ware 
 4. Do. do. 
 
 5. Dish in brown earth 
 Fragment of Pottery found in London 
 A. PLATE IV. Chromo-lithograph of a Vase 
 Window at Bourgerel ...... 
 
 16 
 
 22 
 22 
 24 
 
XIV 
 
 LIST OF ILLUSTRATIONS. 
 
 3. MOUND B. 
 
 B. PLATE I. View of part of the ruins of the Mounds B and C 
 B.PLATE II. Plan of the Villa Urbana .... 
 
 Section of the Villa Urbana from north to south 
 B. PLATE III. Figure i. Quern, or Hand-mill, in granite . 
 
 2. Polished Stone Axe in diorite 
 
 3. Bronze Brooch . 
 
 4. Fragment of gray pottery 
 
 B. PLATE IV. Chromo-lithograph, Mural Painting, in fragments . 
 B. PLATE V. Chromo-lithograph, Mural Painting, do. 
 
 B. PLATE VI. Figure i. Neck of an Amphora . . . . 
 
 2. Antler of Deer . 
 
 3. Ornament in bronze . . . . 
 
 4. Bronze Ring, incised . 
 
 5. Painter's Palette . 
 
 6. Lagena, with two handles 
 
 7. Do. turbiniform, without handles 
 B. PLATE VII. Figure i. Terrine 
 
 2. Vase ....... 
 
 3. Fragment of red lustrous ware 
 
 4. Do. do. 
 
 5. Vase in brown earth . 
 B. PLATE VIII. Figure i. Vase 
 
 2. Bowl ....... 
 
 3. Twisted Handle . 
 
 ., 4. Fragment of yellow pottery . 
 
 5. Do. of red do. . 
 
 6. Do. of black do. 
 Pone of Malansac ......... 
 
 B. PLATE IX. Figure i. Fragment of red lustrous ware 
 
 ,, 2. Do. do. 
 
 3. Do. of a bowl 
 B. PLATE X. Figure i. Vase in yellow earth . 
 
 2. Fragment in black earth 
 
 3. Do. in gray earth 
 
 4. Do. do. 
 
 5. Do. do. 
 
 PAGE 
 26 
 
 28 
 
 32 
 32 
 
 37 
 
 40 
 
 43 
 
 45 
 
LIST OF ILLUSTRATIONS. 
 
 B. PLATE XI. Figure i. Fragment of Vase, with engraved letter^ 
 
 or sign .... 
 
 2. Do. with geometric figures . 
 
 3- Do- do. 
 
 4. Bone Bodkin .... 
 ,, 5. Vertebra of a Whale 
 6. Part of a Hunting-call in bone 
 7. Hunting-call, entire, in bone 
 
 B. PLATE XII. Figure i. Grooved Stone .... 
 2. Stone Mortar .... 
 3. Do. with the mark IXII 
 
 4. Rim of a Bowl with the same mark 
 
 B. PLATE XIII. Figure i. Vase in gray earth 
 
 2. Do. do. ... 
 
 ,, 3. Bowl in red lustrous ware 
 4. Fragment of a Vase in the same ware 
 ,, 5. Do. do. do. 
 
 B. PLATE XIV. Figure i. Fragment of red lustrous ware 
 
 2. Bowl ...... 
 
 ,, 3. Part of a Suet Lamp in yellow earth 
 4. Bowl ...... 
 
 Door at Bourgerel ....... 
 
 xv 
 
 PAGE 
 
 4 6 
 
 47 
 
 48 
 
 49 
 
 4. MOUND C. 
 
 C. PLATE I. Plan of the Baths discovered under the Mound C 
 Conduit Pipe in brick earth ....... 
 
 C. PLATE II. General View of the Baths 
 
 C. PLATE III. Figure i. Section following C H 
 2. Do. do. E F 
 
 3. Do. do. C D 
 
 4. Do. do. AB 
 
 C. PLATE IV. View of the Interior, Hypocaust, and Baths 
 C. PLATE V. Chromo-lithograph, Fragment of Plaster, decorated 
 C. PLATE VI. Do. do. do. 
 
 C. PLATE VII. Do. Reconstruction of a Window, 
 
 and Fragment of Plaster, decorated . 
 C. PLATE VIII. Chromo-lithograph, Fragment of a rosace in fresco, 
 
 and reconstruction . 
 
 53 
 53 
 54 
 
 56 
 
 53 
 62 
 62 
 
 62 
 64 
 
XVI 
 
 LIST OF ILLUSTRATIONS. 
 
 C. PLATE IX. Chromolithograph, Reconstruction of Border in 
 
 fresco ... 
 C. PLATE X. Chromo-lithograph, Reconstruction, part of Ceiling 
 
 in fresco 
 
 C. PLATE XL Reconstruction, design of Ceiling 
 C. PLATE XII. Furnaces beside the Baths . 
 Implement formed of deer's horn 
 
 C. PLATE XIII. Figure i. Iron Window-bar . 
 
 2. Bottom of a Vase with three legs . 
 
 3. Bronze Brooch of Celtic type 
 
 4. Part of a Drill . 
 
 5. Fragment of Pottery with geometric 
 designs ... . 
 
 ,, 6. Do. of a Vase . . . . 
 
 Whorl in terra cotta ......... 
 
 Tile found at Carhaix ........ 
 
 5. MOUND D. 
 
 D. PLATE I. Plan of the Building discovered under the Mound D 
 D. PLATE II. Figure i. Iron Compasses ..... 
 
 2. Fragment of Pottery, ornamented in relief 
 3. Reconstruction of a Bowl 
 4. Neck of a Vase ..... 
 5. Fragment of the Handle of a Vase 
 
 D. PLATE III. Figure i. Upper part of a Vase .... 
 2. Tooth of a Bear ..... 
 3. Necklace Bead ..... 
 4. Reconstruction of a Bowl 
 5. Do. of a flat Dish 
 
 6. Do. do. ... 
 
 Part of a Stone Hammer ..... . 
 
 D. PLATE IV. Statuettes of Venus Genetrix 
 
 D. PLATE V. Do. of Venus Anadyomene . 
 
 D. PLATE VI. Fragments of the same ...... 
 
 Lech at Plouharnel 
 
 64 
 
 64 
 64 
 66 
 70 
 
 70 
 
 73 
 
 80 
 
 81 
 
 81 
 82 
 82 
 82 
 86 
 
 6. MOUND E. 
 
 E. PLATE I. Plan of the Building discovered under the Mound E 89 
 
 Brick pierced with holes . . ..... 92 
 
LIST OF ILLUSTRATIONS. 
 
 E. PLATE II. Figure i. Granite Pestle 
 
 2. Reconstruction of a Glass Cup 
 
 ,, 3. Fragment of Celtic Pottery 
 
 4. Orifice of a Vase . 
 E. PLATE III. Figure i. Stone Amulet with bronze ring 
 
 2. Bronze Ornament 
 
 3. Iron Arrow-head 
 
 4. Iron Hook .... 
 
 5. Iron Knife .... 
 
 6. Bronze Signet-ring 
 
 Impression of the Signet-ring ..... 
 Mark of the Potter Cassigneti ..... 
 E. PLATE IV. Figure i. Vase of hemispheric form 
 
 2. Bowl of the same . 
 
 3.' Fragment of a Vase 
 
 4. Do do. 
 
 E. PLATE V. Figure i. Fragment of a Vase 
 
 2. Do. pierced with holes 
 
 3- 
 
 Do 
 
 Gallo-Roman 
 
 4- 
 
 Do. 
 
 do. . 
 
 5- 
 
 Do. 
 
 do. . 
 
 :amps . 
 
 . 
 
 
 E. PLATE VI. Chromo-lithograph. Statuette of an Ox in bronze 
 E. PLATE VII. View of the Fountain of St. Carnely 
 Cross on the Mont St. Michel 
 
 xvn 
 
 PAGE 
 
 93 
 
 95 
 
 96 
 96 
 
 96 
 
 97 
 
 97 
 
 98 
 
 100 
 
 7. MOUND F. 
 
 F. PLATE I. Plan of Buildings discovered under the Mound F 
 F. PLATE II. Figure i. Bronze Ring . 
 
 ,, 2. Incised Ornament in bronze 
 
 ,, 3. Fragment of a Bronze Plate pierced with 
 holes ...... 
 
 4. Fragment of a Vase in black earth . 
 
 5. Bottom of a Bowl in red lustrous ware 
 
 6. Vase reconstructed . 
 Bullock Shoe in iron .... 
 Cross at Coet-a-Tous ......... 
 
 10 5 
 
 107 
 
 109 
 in 
 
xviii LIST OF ILLUSTRATIONS. 
 
 8. MOUND G. 
 
 PAGE 
 
 Vase at Mayenne . . . . . . . . . 1 1 6 
 
 G. PLATE I. Section of a large Menhir . . . . . 119 
 
 Cross at Han hon 121 
 
 9. MOUND H. 
 
 H. PLATE I. Plan of the Buildings discovered under the Mound H 125 
 H. PLATE II. Figure i. Fragment of a Clasp in bronze 
 2. Stone Weight or Sinker . 
 
 ,, 3. Fragment of a Bronze Buckle 
 
 126 
 4. Reconstruction of a Vase 
 
 5. Do. do. 
 
 ., 6. Do. do. . . . j 
 
 Door at Kergouellard .... . 128 
 
 io. SUMMARY AND CONCLUSIONS. 
 Chalice cut on a tombstone ... . . . 143 
 
 ii. MONT ST. MICHEL. 
 
 PLATE I. View and Section of the Mont. St. Michel . . . 149 
 PLATE II. Plan of the Mont St. Michel and excavated buildings . 151 
 PLATE III. Figure i. Whetstone . . . . . . ^ 
 
 ,, 2. Flint Knife . . . . . I 
 
 ,, 3. Stone Celt, polished . . . . . } 152 
 
 4. Whorl in brown earth . . . I 
 
 5. Fragment of a Vase . . . . . j 
 
 PLATE IV. Figure i. Neck of a Vase ^> 
 
 2. Handle of the same Vase .... 
 
 3. Fragment of a Vase ..... 
 
 4. Do. do. 
 
 Ground-plan of Chapel of St Agatha . . . . . . 160 
 
 Cross on the Mont St. Michel . 162 
 
I. 
 
 THE BOSSENNO. 
 
THE BOSSENNO. 
 
 THE little town of Carnac, situated in the Department of the 
 Morbihan, about nine miles to the south of Auray, is visited 
 by numerous travellers from all parts of the world, attracted 
 thither by the celebrity of its alignments and other stone 
 monuments. 
 
 In following the road from Auray to Carnac, a striking 
 difference is observed in the appearance of the country. In 
 approaching the coast the land becomes less fertile ; on every 
 hand sterile tracts of moorland appear, on which stand the 
 monuments called Celtic, viz. menhirs,"" dolmens, t and tumuli, 
 all of which become more numerous within a mile of Carnac. 
 
 The traveller's first impressions, however, are not favour- 
 able ; the country seems flat and uninteresting, and his expect- 
 ations are not immediately realised. It is only after he has 
 remained some time at Carnac, and become more intimately 
 acquainted with the surrounding country, that he begins to 
 appreciate the number, extent, and variety of the Celtic monu- 
 ments and Gallo- Roman ruins with which it is studded, and 
 to realise that the time thus devoted to the examination of 
 these in detail has been amply repaid. 
 
 * Men-hir, from the Breton Men, a stone, and Mr, long literally a long 
 stone. 
 
 t Dol-men, from the Breton Dot, a table, and men, a stone literally a 
 stone table. 
 
4 THE BOSSENNO. 
 
 According to Caesar's Commentaries this country was in- 
 habited before the Roman Conquest by the Veneti, the most 
 powerful people of Armorica. 
 
 Its occupation by the Romans until the year 409, that is 
 for more than four centuries, must of necessity have left very 
 marked traces of the masters of the world. These are to be 
 seen in the numerous Gallo- Roman ruins which have been 
 explored of late years, and in the Roman tiles and pottery 
 which are so frequently met with in traversing the fields. 
 During all this lapse of time the country was inhabited (accord- 
 ing to the Roman authors) by a population who had in a large 
 measure adopted the manners and customs of the Romans. 
 
 It is especially on the sea-coast of Armorica that the juxta- 
 position of Celtic monuments and Gallo-Roman ruins attracts 
 the attention of the observer, and suggests questions of what 
 had led to this juxtaposition of monuments whose origin is so 
 different. This question is an interesting one. Without pre- 
 tending to solve it, I may be permitted to cite a few places 
 where this proximity may be easily verified. At Locmariaker, 
 Roman ruins are found alongside of the Celtic monuments ; 
 at Kerilio, in Erdeven, the same may be seen in the sandy 
 downs to the south of the village ; at Carnac, the Roman 
 ruins of the Bossenno are surrounded by Celtic monuments. 
 
 It is owing to explorations which have been made, during 
 the last thirty or forty years, by the different Archaeological 
 Societies in Brittany, and also by private individuals, that 
 numerous Gallo-Roman establishments have been discovered 
 in that province. This is a proof of the progress of archaeo- 
 logy in Brittany, for at that time several antiquaries held that 
 the Romans had never penetrated to the extreme point of 
 Armorica. 
 
 The Transactions of the Polymathic Society of the Mor- 
 
- 
 
THE BOSSENNO. 5 
 
 bihan, for 1857, contain reports on the exploration of the 
 following Gallo-Roman establishments : 
 
 At St. Christophe, in the Commune of Eloen description 
 of the explorations, and plan. At Lodo, in the Commune of 
 Arradon description, plan, drawings of objects found, and 
 account of Roman coins collected. At St. Symphorien, near 
 Vannes description and plan. 
 
 In the Department of Finistere there has also been dis- 
 covered, at Parc-en-Croix, near Ouimper, a Gallo-Roman 
 military post, of which there is a description and plan in 
 the third volume of the Transactions of the Archaeological 
 Society of Finistere. 
 
 At Poulker, near the embouchure of the Odet, on the left 
 bank of the river of Quimper, Gallo-Roman ruins were ex- 
 plored in 1866, for the Archaeological Society of Finistere, by 
 M. le Men, Archeviste of the Department. 
 
 M. Fornier has also published a description and plans of 
 his explorations at Haut-Beeherel, in the Commune of Cor- 
 seul (Cote du Nord). 
 
 Numerous other discoveries might easily be cited, but it 
 would be superfluous to do so, since Gallo-Roman remains 
 have been found almost everywhere in Brittany, thus proving 
 that Armorica was occupied as long and as fully as the other 
 parts of Gaul. 
 
 Amongst the number of unexplored Gallo-Roman remains 
 in the Department of the Morbihan several works on archaeo- 
 logy mention certain mounds * called "Caesar's Camp," situated 
 a mile to the east of Carnac, and near the village of Cru- 
 Carnac, their origin being indicated by the fragments of Roman 
 
 * Amongst others, the Repertoire Archeologique du Morbihan, published 
 by M. Rozenvveig. These mounds had also been pointed out to several 
 archaeologists by M. Charles de Kerenflech-Kevegne before 1860. 
 
6 THE BOSSENNO. 
 
 tiles strewn on their surface. Shortly after my arrival at 
 Carnac in 1873 these mounds were pointed out to me; and 
 the more I saw of them, the more I desired to explore them. 
 
 These mounds are situated in the following portions of the 
 section M, second subdivision of the Cadastre.* 
 
 No. 610. Er Vurten. No. 649. Bossenno. 
 
 611. Boceno. ,, 651. Bocenneu. 
 
 612. Boceno Hir. 652. Bocenneu. 
 
 647. Clos Gurunne. 
 
 These portions of land form part of an elevated plateau, 
 extending from them towards the sea. 
 
 The etymology of the word Bossenno appears to show 
 that it is derived from the Breton word bossen (a mound or 
 heap) in the plural boceno, bossenno, or bocenneu, according to 
 the different dialects. This etymology of the word is in 
 harmony with the character of the objects to which the name 
 is given, consisting of a cluster of mounds occupying an area 
 which may be estimated at 6578 square yards. 
 
 Commencing at 200 yards to the east of the village of 
 Cru-Carnac, this area extends eastward to the edge of the 
 valley now called Ker-ine-goh, but in the old titles of 1740, 
 Ker-guin-goh^ These old titles will be hereafter alluded to 
 as proving that the sea came up to Ker-guin-goh during the 
 last century. 
 
 A talus of unequal dimensions stretches along the north 
 
 * See the general plan of the Bossenno. 
 
 t This word signifies the old town of Guen, or the town of Old Guen. 
 A curious old cross, figured at the end of this chapter, is situated about 500 
 yards to the north-west of Ker-guin-goh. It is named Croaz-ar-Guen. The 
 Breton word guen or van, or ban in the Celtic, signifies white. Guened is 
 the town of Vannes, and Guenedi the Veneti. The most common names of 
 the inhabitants of the village of Cru-Carnac are Guen-ac (whites), and 
 Baelagu (priests). 
 
British Imp^Feet o 50 100 
 
 Joint j Xi 
 
 To face page 6. 
 
 GENERAL PLAN 
 
JS.jVori&u, jr. 
 
 DSSENNO. 
 
THE BOSSENNO. 7 
 
 side of this ground, and appears in some parts also on the south 
 side. When making several trenches across the northern 
 talus, I found the remains of a wall measuring 5 ft. in breadth 
 and 5 ft. in height, built with large undressed stones without 
 lime, similar in its masonry to the walls of the Middle Ages. 
 
 From the commencement of the work at the Bossenno, 
 this talus seemed to me to have been an enceinte, and the 
 discovery of the stone wall strengthened this impression. 
 This destination is also indicated by the difference of level 
 between the north and south sides of the talus : on the north 
 side the ground is fully three feet lower ; and this leads me to 
 suppose that on this side a moat had been cast, the earth from 
 which had served to heighten the enceinte. 
 
 The legends of a country are never to be overlooked, 
 especially when they bear upon the researches in which one 
 is engaged. 
 
 The group of mounds called the Bossenno have their 
 legend, and it has a certain analogy with certain facts which 
 the excavations there have brought to light. The tradition 
 of the country is that the Bossenno were inhabited by the red 
 monks (the Templars) who, having exasperated the country 
 by their crimes, drew down upon themselves a terrible 
 punishment, their neighbours having killed them and burnt 
 their habitations in one night. It will be afterwards seen 
 that in the exploration of the ruins of the Bossenno the 
 unmistakable proofs were everywhere found of their having 
 been destroyed by fire. 
 
 These mounds have an evil reputation to this day. The 
 peasants tell you " those who pass by them late at night may 
 happen to see each mound illuminated, and sometimes to hear 
 an animated conversation, of which not a word can be under- 
 stood, for it is in Latin." 
 
THE BOSSENNO. 
 
 The exploration of the Bossenno was commenced in the 
 beginning of September 1874. The excavation of each 
 separate mound was undertaken successively. Each of the 
 mounds so excavated is distinguished in the general plan by 
 a particular letter. Up to this time eight mounds have been 
 explored, and these are distinguished in the following pages 
 by the letters A, B, C, D, E, F, G, H. 
 
 I shall confine myself principally to the description of the 
 constructions found under the respective mounds, with an 
 enumeration of the objects collected from them, reserving for 
 the concluding chapter a few remarks regarding the special 
 nature of these constructions, and the time when they had 
 been destroyed. 
 
 CROAZ-AR-GUEN. 
 
II. 
 
 EXCAVATIONS IN THE MOUND A. 
 
THE MOUND A. 
 
 THE excavations at the Bossenno were commenced in the 
 early part of September 1874, on the mound indicated in the 
 general plan by the letter A. It was one of the smallest of 
 several artificial eminences in the field 649 Bossenno, of which 
 it occupied the south end. In form it was a low circular 
 mound, having a height of about 3 feet, and a circumference 
 at the base of about 1 30 feet. 
 
 Before proceeding farther, it may be as well to indicate the 
 method which was followed in conducting the diggings, not 
 only of this mound, but of those successively explored. 
 
 The different objects brought to light by the labours of 
 the workmen such as bricks, tiles, pottery, metals, bones, etc. 
 were carefully laid aside. The fragments of pottery and tiles 
 were first of all submitted to a preliminary examination, with 
 the view of finding the potters' marks or names. At the end 
 of each day the more interesting objects were carried to the 
 Hotel des Voyageurs at Carnac, where they were submitted 
 to a more careful examination, previous to being classed and 
 ticketed, so that no errors might occur as to the precise 
 position in which they had been found. 
 
 The workmen commenced digging on the south side of the 
 mound, and were not long in turning up a quantity of debris, 
 
12 THE BOSSENNO. 
 
 consisting of fragments of Roman roofing tiles (Tcgula and 
 Imbrices], also sundry fragments of Gallo-Roman pottery in 
 red, gray, and black ware. After several hours' work we came 
 upon the outside of the north wall. 
 
 Here some unforeseen difficulties arose, which stopped the 
 work for several days. The lessee of the grazing of the field 
 set up a claim for indemnity which he had not previously 
 required, and which ultimately had to be submitted to. On 
 the other hand, the proprietor raised a still greater difficulty 
 by exacting that all the excavated stones and rubbish should 
 be carted away clear of his fields. Thanks, however, to the 
 kind intervention of Dr. de Gressy and of M. Broux, the 
 Mayor of Carnac, leave was obtained to deposit these on the 
 waste land below Kercado, about a mile distant from the 
 Mound A, and upon some parts of the old road which 
 traverses the Bossenno towards the Trinite-sur-Mer. As 
 soon as these arrangements were concluded, the diggings 
 were immediately resumed. 
 
 The outside of the wall previously mentioned was followed 
 by cutting a trench along it lengthways down to its founda- 
 tion. Several days were passed in carrying this trench round 
 the outside of the dwelling so long buried under the Mound 
 A. When finished it exposed a building of quadrilateral 
 form, each of its sides measuring nearly 34 feet, excepting that 
 on the south, which was somewhat less than the others. 
 This irregularity of construction is not uncommon in Gallo- 
 Roman buildings. 
 
 The walls measured 2 feet in breadth, and were con- 
 structed with small cubic stones, carefully laid, and bedded in 
 lime mortar. Some of the neighbouring peasants who came 
 to see the diggings were much astonished at the regularity of 
 the courses of small cubic stones, and remarked as well that 
 
Metres 
 
 .....;. " I 2 3 S 
 1 Bn t ish Imp! Feet ......... | 
 
 - fecit. 
 
 20 
 
 30 
 
 To face page 13. 
 
 A, PLATE I. PLAN OF THE BUILDING. 
 
EXCAVATIONS IN THE MOUND A. 13 
 
 the breadth of the walls was the same as in the modern 
 buildings in the environs of Carnac. 
 
 The following days were employed in clearing out the 
 interior of the building. The result of this was to expose a 
 dwelling of four rooms, of which I shall now give the descrip- 
 tion, designing each by the number it bears on the plan A, 
 plate I. 
 
 No. i. This apartment, by far the largest, measured 28 
 feet by 18, and occupied all the east side of the house. Two 
 pieces of masonry projected from the inner side of the north 
 wall, and marked the places which had been occupied by two 
 constructions, whose remains were now so dilapidated as to 
 afford no definite indication of their original destination. It 
 was, nevertheless, agreed by all who saw them, that they 
 might have been furnaces. 
 
 Nos. 2 and 3. The first of these rooms formed a square, 
 measuring 9 feet on each side. The second measured 9 feet 
 by 7 feet. 
 
 The flooring of the rooms i, 2, and 3, was formed of a 
 thick layer of concrete cement. 
 
 No. 4. This room was also square, and measured 9 feet 
 on each side. It had been paved with square tiles, but these 
 were found displaced and broken. Other fragments of tiles 
 were collected here, which were ornamented with diagonal 
 lines and concentric circular markings. Subsequent expe- 
 rience showed that these tiles had been sometimes used for 
 decorative purposes, while in other cases the hollow lines had 
 served to make the plaster adhere. Three paving tiles were 
 found which bore the impression of the feet of dogs, which 
 had passed over them when the clay was in the soft state. 
 
 The floors of these four rooms, which have just been 
 described, were covered with a layer of cinders, charcoal, 
 
U THE BOSSENNO. 
 
 and ashes, mingled with pieces of iron, scoriae, and even of 
 granite vitrified. The impression derived from the presence 
 of these objects, and from the consideration of other circum- 
 stances revealed during the progress of the diggings, was that 
 the building had been destroyed by fire, while the quantity of 
 broken pottery'" found within its precincts renders it probable 
 that the dwelling had been previously pillaged ; for it is pro- 
 verbial that the marauder destroys all that he is unable to 
 appropriate or does not care to carry away. 
 
 After the four rooms had been entirely cleared out, trenches 
 were dug across the floor of the room No. i from north to 
 south, and also all round the base of the walls of the building 
 on the inside. 
 
 At a depth of 1 2 inches under the floor the workmen came 
 upon a layer, 4 inches thick, composed of bones of ruminants, 
 shells of edible molluscs, ashes, cinders, charcoal, and fragments 
 of pottery. Another bed of the same nature was found 22 
 inches lower still. These two layers seemed to be the evidences 
 of two burnings anterior to that which had destroyed the 
 house buried under the Mound A. An apparent corrobora- 
 tion of this view is found in the fact that, as we got deeper 
 down in the trenches, the forms of the pottery became more 
 simple, and showed only plain edges. 
 
 The debris which was carted away from the inside of the 
 building, was composed principally of small cubic building 
 stones, fragments of flat roofing tiles ( Tegula) and of ridge 
 tiles (Imbrices]. From all this quantity of debris, however, 
 there were only extracted entire one flat roofing tile and six 
 ridge tiles. Subsequently, when exploring the other mounds 
 of the Bossenno, a considerable number of both Tegula and 
 
 * For detailed descriptions of the pottery, see pp. 18-24. 
 
To face page 15. 
 
 A, PLATE II 
 
EXCAVATIONS IN THE MOUND A. 15 
 
 Imbrices were found entire, so that I was enabled to recon- 
 struct a portion of the roof, which was deposited and may be 
 seen in the newly-formed museum in the Mairie of Carnac, 
 and where, it is to be hoped, that it will remain after the ruins 
 of the Bossenno have entirely disappeared. * 
 
 The following is a descriptive list of the most important 
 objects found in clearing out the inside of the buildings : 
 
 Several large pieces of querns or hand-mills in granite and 
 conglomerate or." pudding stone," reddened by the action of fire, 
 were found scattered on the floor of the room No. i. 
 
 Part of a broken implement in yellow sandstone, cut in 
 facets, polished by use. This appeared to have been a hone 
 or sharpening stone. (A, pi. II. fig. 3.) 
 
 Several fragments of implements in gray slate polished by 
 use, which appeared to have been used as rubbing stones or 
 burnishers. (A, pi. II. fig. i.) 
 
 Several worked flakes of black flint similar to the flint of 
 the English Channel as found near Dieppe. 
 
 An implement in gray flint unpolished, and measuring 
 4^ inches by 3 inches. (A, pi. II. fig. 2.) 
 
 Mr. Le Guennec, proprietor of the field in which the mound 
 A was situated, discovered, when removing the ruins of a 
 dolmen from one of his fields a few years ago, a stone axe also 
 in gray flint and of the same form as that here described. 
 Mr. Le Guennec placed his axe in the museum, in the Mairie 
 of Carnac. The dolmen in which it was found was situated 
 in a meadow field near the village of Kerfraval, distant about 
 half a mile to the south of the Bossenno. 
 
 * A few years ago there was discovered near Redon, in the grounds be- 
 longing to the Eudiste fathers of that town, a Gallo-Roman furnace, having 
 its roof constructed with Tegula and Imbrices. It is to be regretted that this 
 interesting monument was not better cared for. It fell in ruins in 1873. 
 
16 THE BOSSENNO. 
 
 It happened one day when I was absent during the dinner 
 hour of my workmen, that an English lady and her son came 
 to see the diggings. The latter amused himself in working 
 with a pick about that part of the construction in the room 
 No. i which resembled a chimney, where he discovered a 
 polished stone celt of a white colour, which he showed to his 
 mother : neither of them, however, was aware of its value, 
 and it was flung aside amongst the debris to be carted away. 
 It was not until the following day, when I happened to show 
 them the polished stone celts in the museum at Carnac, that 
 they informed me of their discovery, and regretted that they 
 had not known better. Exertions were made to recover the 
 lost axe, but without success. This discovery of a stone axe 
 in what appeared to be a chimney was all the more 
 interesting from its correlation with a custom still observed 
 at Carnac, that of building into the chimney of the dwelling- 
 house a stone celt which is supposed to preserve the house 
 from being struck by lightning. It is to be noted also that 
 the name of the stone axe or celt in the Breton language is 
 Mein-Gurun, that is to say, the Thunder Stone. 
 
 There were only a few objects in bronze found, viz. 
 
 A buckle of a square form. (A, pi. III. fig. i, natural 
 size.) 
 
 A piece of wire or rod measuring 4 inches in length, and 
 inch in breadth. 
 
 A ring of elliptic form, probably part of the buckle of a 
 waist-belt. (A, pi. III. fig. 2, natural size.) 
 
 The head of a nail or button i J inch in diameter, which 
 may have been mounted either on leather garments, or on 
 harness. The zingaris or gipsies of Hungary, and the 
 Wallachians, to this day, employ similar buttons to adorn their 
 
To face page 16. 
 
 A, PLATE III. 
 
EXCAVATIONS IN THE MOUND A. 17 
 
 leather vestments, as may be seen by referring to the studies 
 made by Valeric in Hungary and Wallachia. The Breton 
 peasants, in many communes, still wear a number of metal 
 buttons on their garments which are disposed more for 
 ornament than use. 
 
 Lumps of iron and scoriae were found in considerable 
 quantity in the room No. i, especially near the construction 
 which has been indicated as resembling a forge. Several other 
 pieces of iron, which appeared to have been parts of sword- 
 blades and spear-heads, were also found on the floor, but they 
 crumbled away after extraction and exposure to the air. The 
 only objects in iron obtained entire were a few nails 2 to 3 
 inches in length. These were extracted from the east wall, 
 where the lime mortar had preserved them. 
 
 Among the few fragments of glass found, some were 
 portions of cups in white glass, while others of a bluish-green 
 colour, ornamented with fluted mouldings, were evidently 
 parts of square-shaped bottles similar to the Gallo- Roman 
 bottles exhibited in the museums of Caen, Quimper, and 
 Vannes. These fragments were found in the rooms Nos. 2, 
 3, and 4. 
 
 There were also found in these rooms some fragments of 
 fine clear glass, very thin and pure, and a large piece of a bowl 
 in fine clear glass, ornamented with two circular lines engraved 
 not quite parallel. 
 
 A considerable quantity of bones and teeth of oxen of a 
 small size were collected here. One of the bones had clean 
 incisions on it, made by a sharp cutting implement; others 
 bore the marks of the teeth of carnivora which had gnawed 
 them. Several had been split lengthways, either to extract 
 the marrow, or as the first step in the process of fabricating 
 them into implements. 
 
 C 
 
18 THE BOSSENNO. 
 
 There were also found bones and teeth of the deer, goat, 
 sheep, and pig. One tooth was found which belonged either 
 to the dog or wolf. 
 
 In a locality so close to the sea, it was to be expected that 
 the remains of edible molluscs would not be wanting, and so 
 it proved, for there was an abundance of the shells of oysters, 
 clams, ormers, limpets, and mussels. The oyster-shells were 
 the most numerous. On comparing these with shells fresh 
 from the Bay of Quiberon, no difference could be observed 
 either in the enamel or in the dimensions. 
 
 Besides these, there was found in the trench cut on the 
 outside of the east wall a veritable kitchen midden, composed 
 of shells, bones, and fragments of pottery. This heap appeared 
 to have been the refuse which had been thrown out of the 
 dwelling. 
 
 The quantity of fragments of pottery collected during the 
 excavation was very large in proportion to the small size of 
 the dwelling, and thus denoted a certain amount of luxury. 
 The whole of the fragments when collected together were 
 sufficient to fill three baskets, which weighed 71 Ibs. net. 
 
 I attempted, but without success, to reconstruct complete 
 vases from these numerous fragments. It was with the 
 greatest difficulty that two or three fragments were found 
 which adapted themselves so as to be cemented together, and 
 it became a puzzle to ascertain what had become of all the 
 other fragments. 
 
 But the principal cause of astonishment was not so much 
 the abundance of these fragments as the extreme diversity of 
 the forms of the vases to which they belonged. This 
 diversity of form, when taken in connection with their 
 difference of texture and variety of ornament, indicates that 
 they must have belonged to very different epochs. 
 
EXCAVATIONS IN THE MOUND A. 19 
 
 A mere cursory inspection is in fact sufficient to show that 
 some of these vases had been fashioned with neither taste 
 nor art, and from rude materials collected from the surface of 
 the ground ; whilst others indicate, by the elegance of their 
 forms and by the careful preparation of their material, that 
 the makers had a considerable knowledge of the ceramic art. 
 
 A more careful examination of the fragments belonging to 
 the first of these two categories enables one to perceive, from 
 their irregularity of form and from their inequality of thick- 
 ness, that they had been made by hand without the aid of the 
 potter's wheel. The marks of the fingers are plainly visible 
 on many of these fragments, whilst others bore a sort of 
 ornament formed by the print of the potter's finger round the 
 rim, and, judging from the smallness and shape of these, one 
 would say that they had been made by women."" 
 
 Their paste is formed of earth, consisting of the debris of 
 primitive rocks, and badly worked, in which one can easily 
 distinguish, without the aid of a magnifying-glass, small frag- 
 ments of quartz, felspar, mica, etc. The fracture is earthy, and 
 the firing imperfect and unequal. They have in this respect 
 a remarkable resemblance to the funeral vases found in the 
 dolmens and kistvaens which are generally regarded as 
 belonging to a very remote period. 
 
 It may be as well to add that in pointing out this striking 
 similarity, I do not believe myself authorised to assign the 
 rude pottery of the Bossenno to the remote epoch of the 
 dolmens ; but it appears to me to be shown by this com- 
 parison that the resemblance which exists between the 
 
 * At Rieux and at Malansac, which are villages inhabited by potters, 
 situated in the Department of the Morbihan, it is the women who make all 
 the pottery for ordinary use. The men make only the large vases called 
 pones or charniers. 
 
20 THE BOSSENNO. 
 
 products of ages so far apart may become, in many cases, a 
 source of error, and that great caution must be exercised 
 when rude pottery is made the subject of a chronological 
 classification. 
 
 The following narrative is adduced in support of this 
 opinion : 
 
 I happened to be in the Hebrides during the summer of 
 1868. The weather was tempestuous, and I had found a 
 shelter from a thunderstorm in the cottage of an old woman, 
 originally from the island of Tiree, but then living in lona. 
 Two or three small vases that were placed on the table 
 attracted my attention. They were so like the ancient 
 pottery of the dolmens or kistvaens, that I asked where they 
 had been found. To my surprise the old woman replied, " I 
 did not find them, I made them myself." 
 
 " That must be impossible," I said, " for you have no 
 furnace to fire them." 
 
 " I do not need a furnace for that, and if you like I will 
 make you some." 
 
 " Do so by all means, I will be very glad to have two or 
 three of them." 
 
 The old woman went immediately out to her garden, and 
 brought back a quantity of argillaceous earth. She then 
 wetted the surface of the table, on which she moulded very 
 neatly, and in a short space of time, three small vases, which 
 bore the impress of her fingers round the outside of the neck, 
 and on the inside of the rim there was a border formed with 
 the points of her finger nails. She then said to me, 
 
 " I cannot proceed further just now, because the pots must 
 be dried before they are put on the fire. Come to-morrow 
 morning at ten o'clock, and you shall see that I do not require 
 a furnace to fire them." 
 
EXCAVATIONS IN THE MOUND A. 21 
 
 At ten o'clock next morning I returned to the cottage. 
 The only fireplace was a slab of stone in the centre of the 
 room, on which a few pieces of turf were burning. The old 
 woman placed the vases on the fire, poured milk into them, 
 and allowed them to remain until the milk had boiled for 
 some time. When this operation, which did not much exceed 
 half-an-hour, was concluded, the vases had acquired sufficient 
 hardness, and had all the appearance of the ancient pottery of 
 the dolmens. 
 
 Returning to the description of the pottery of the Bossenno, 
 several fragments of coarse gray pottery were found in the 
 Mound A, which bore a more elaborate decoration than those 
 described in the preceding pages. This decoration, which is 
 by some authors designated geometric, consisted of a series 
 of hollow black lines forming chevrons and lozenges. (See C, 
 pi. XIII. figs. 5 and 6.) Several very fine specimens of vases 
 of this kind, which had been found in excavations at Arradon 
 and Carhaix, are exhibited in the museums of Vannes and 
 Quimper. Several specimens similar to these are to be seen 
 in the museum of the manufactory at Sevres. These had 
 been found in Egypt, however, and they are figured and 
 described by Brogniart, 1 as of the time of the Ptolemies, 
 that is about the middle of the third century before the 
 Christian era. Can this similarity of pottery found in Egypt 
 and in Brittany be accounted for by the fact of African sol- 
 diers having been stationed in the Morbihan and in Finistere ? 
 The Notitia Dignitatum of the Empire gives the principal 
 military stations for the fourth century, and mentions 
 prefectus militum Maurorum Venetorum Venetis, and again 
 prefectus militum Maurorum Osismiacorum Osismiis. 
 
 f Traite des arts Cemmiques par Alex. Brogniart. Tome premier, page 
 503. Paris, Bechet Seune; 1854. 
 
22 THE BOSSENNO. 
 
 To the pottery described in the preceding pages there 
 are to be added some fragments of a small vase, very thin, 
 well fired, and having a fine hard sonorous paste. The outer 
 surface was covered' with a dark metallic-looking glaze, with 
 red ornaments in relief. The Museum of St. Germain-en- 
 Laye possesses several similar vases found in the Department 
 of the Lozere. Three 'fragments of vases of the same kind 
 are exhibited in the Guild Hall Museum, London, which 
 
 FRAGMENT OF POTTERY FOUND IN LONDON. 
 
 were found at a depth of 42 feet when the foundations of 
 the Royal Exchange were sunk. In my subsequent excava- 
 tions in the Mounds B and C, numerous fragments of this 
 curious class of pottery were discovered, which enabled me 
 to reconstruct the vase figured in the accompanying chromo- 
 lithograph, A, pi. IV. 
 
 The greater portion of the pottery found in the Mound A 
 was formed of common yellow, gray, and black earth, and 
 consisted of fragments of diotae, amphorae, and ordinary 
 domestic vessels. I succeeded in determining the forms of 
 two flat dishes, in common gray earth, which resemble those 
 
PL.l 
 
 To face page 22. 
 
 A, PLATE IV. 
 VASE FOUND IN MOUND C. (SEE p. 67.) 
 
 L . Cappe .Del 
 
 Imo. Leroy QcRichon Pans. 
 
EXCAVATIONS IN THE MOUND A. 23 
 
 still in daily use at Carnac, called in the Breton dialect of 
 Vannes, er Cass. (See A, pi. III. fig. 5.) 
 
 Besides these fragments in gray, yellow, and black earth, 
 the progress of the excavations brought to light some frag- 
 ments in common red earth, and others, not so numerous, in 
 the fine red earth, and with the lustrous glaze commonly called 
 Samian. From the latter fragments two small vases were 
 reconstructed, one 5^ and the other 3^ inches in diameter 
 (See A, pi. III., figs. 3 and 4.) Some of these latter fragments 
 appeared to me to belong to the class of pottery called " False 
 Samian." Birch describes this variety of pottery as covered 
 with a thin red glazing due to a solution of sulphate of iron 
 in which it had been plunged. On trying to wash one of 
 these fragments, so as to clear off the adhesive earth which 
 covered it, the red glazing dissolved in the water, to which it 
 imparted a bright red colour. This experience showed the 
 inexpediency of washing the fragments of pottery, even after 
 they had been first carefully dried. 
 
 This enumeration of the different specimens of pottery 
 found in the Mound A may now be concluded by mentioning 
 a number of rods or stalks in coarse red earth, which were 
 scattered about the floor of the room No. i. These measured 
 4 to 7 inches in length, and i to ij inches in diameter. It is 
 believed by many antiquaries that these rods were made use 
 of, in the places where pottery was made, to support the 
 apodal vases and pieces with round bottoms during the firing,"'' 
 and also for cleaning the hands of the potter, by rubbing, when 
 too much empasted. 
 
 Lamps and coins are generally found in the diggings made 
 
 * To this day, in the Breton potteries, unvarnished vases are supported, 
 whilst enamelled vases are isolated by little rods of refractory earth, having 
 different forms, called ferneites. 
 
24 
 
 THE BOSSENNO. 
 
 in Gallo-Roman establishments, but no traces of either of these 
 were found in the Mound A. 
 
 As soon as the diggings were completed the peasants com- 
 menced to clear away the ruins which had been exposed, as 
 they wished to put the ground under pasture. In demolishing 
 the east wall they discovered a small head of Venus in terra 
 cotta, having an ornamant round the forehead formed of stars 
 and small circles, with a point in their centre. As the neck 
 bore the evidence of having been freshly broken, I caused 
 a search to be made, in the hope of finding at least some 
 other portion of the statuette, but without success. 
 
 The season was too far advanced to think of attacking 
 another mound this year, and the diggings were therefore 
 adjourned until the spring of 1875. 
 
 WINDOW AT BOURGEREL. 
 
III. 
 
 EXCAVATIONS IN THE MOUND B. 
 
THE MOUND B. 
 
 THE results obtained from the excavation of the small mound, 
 undertaken in the month of September 1874, were so far 
 satisfactory as to lead me to look for still better results from 
 the larger mounds, and in this I was not disappointed. As 
 soon as the weather permitted, in the spring of 1875, the dig- 
 gings were recommenced ; and with a dozen of workmen, 
 most of whom had been rendered more skilful by the expe- 
 rience of the previous year, an attack was made successively 
 on the mounds marked in the plan of the Bossenno by the 
 letters B and C. Although the work was commenced in 
 reality on the Mound C, still it will render the description 
 more clear to the reader if the diggings in the Mound B are 
 first described, as the constructions found under the Mound 
 C were only a dependence of the habitation found under the 
 Mound B. 
 
 The Mound B measured from north to south 82 feet in 
 length, and had a breadth of 56 feet. Its average height was 
 about 7 feet. An excavation on its summit indicated that 
 some one had already been at work there. My workmen, 
 who belonged to the locality, stated that there was a tradition 
 to that effect in the country. 
 
 The example of my predecessors was not followed in our 
 commencing operations on the summit, but on the north side, 
 
28 THE BOSSENNO. 
 
 where we soon struck on a wall, at the base of which was a 
 square space 5 feet 1 1 inches of a side, paved with Roman 
 roofing tiles, or Tegulte. These tiles had their crotchet 
 edges turned downwards, and were solidly imbedded in 
 cement, so as to form a smooth piece of pavement.""" 
 
 The wall was followed all round the mound until the 
 exterior of the building was entirely exposed to view, after 
 which the clearing out of the interior was proceeded with. 
 
 When the latter operation was concluded, we had exposed 
 to view a building of rectangular form, measuring 62 feet in 
 length by 47 feet in breadth, and composed of eleven sub- 
 divisions viz., three large apartments and a passage, Nos. 
 2, 3, 4, and 10 of the ground-plan. These occupied the centre 
 of the building. On the east side there were the room No. 1 1, 
 and the corridor No. i. On the west side there was a suite 
 of small rooms, Nos. 5, 6, 7, 8, 9, probably cubicula. The 
 dilapidated state of the exterior wall on the west side pre- 
 vented our ascertaining if there had been any entrances on 
 this side. The principal entrance of the building is indicated 
 on the ground-plan. 
 
 The thickness of the exterior walls was 24 inches, and that 
 of the interior 20 inches. 
 
 The west wall was prolonged to the north to where it 
 joined on to the building found under the Mound C (the 
 baths). A court, which measured 27 feet in breadth, separated 
 the two buildings. 
 
 These indications being sufficient to give a general idea 
 of the form, character, and relations of the two buildings, we 
 now proceed to give a more detailed description of each of 
 the eleven subdivisions of the larger structure, all of which 
 
 * See the ground-plan B, pi. II. 
 
N?3 N? 2 N?10 
 
 B, PLATE II. PLAN AND SECTION OF THE BUILDING IN THE MOUND B. 
 
EXCAVATIONS IN THE MOUND B. 29 
 
 bore marks of having been destroyed by fire. Their relative 
 positions and dimensions will be more readily understood by 
 reference to the ground-plan of the Mound B. 
 
 No. i. A passage 42 feet in length, by 7 feet 6 inches in 
 breadth. The height of what remained of the walls was on 
 the south side 6 feet, on the east i foot 3 inches, on the west 
 6 feet 6 inches. 
 
 This passage had a doorway on the east side, which 
 appears to have been the principal entrance to the building ; 
 whilst on the north it opened on the court already mentioned 
 as situated between the buildings B and C.""~ The east wall 
 seemed to be continued until it joined on to the baths under 
 the Mound C ; but this could not be verified, as the tenant of 
 the grazing objected to the diggings being carried on in the 
 little bit of pasture between the Mounds B and C. 
 
 Two square buttresses measuring i foot 4 inches on each 
 side supported the west wall of the passage. One of these 
 buttresses was not bonded into the wall. This seemed to 
 indicate that it had been added to the building after its con- 
 struction. 
 
 An opening to the south of these buttresses led into the 
 interior of the building by a small staircase which ascended 
 from the passage to the room No. 2. This staircase was in a 
 very dilapidated condition, and fell to pieces during the 
 excavation. Underneath it were found the traces of a fur- 
 nace, with a very small aperture, which had served to heat 
 the hypocaust under the room No. 2. 
 
 The few courses of masonry that remained of the east wall 
 of the passage were in a very dilapidated condition. 
 
 * See the plan of the Bossenno. 
 
30 THE BOSSENNO. 
 
 No. 2. This room measured 21 feet in length by 16 in 
 breadth. The height of what remained of the walls was 
 
 ft. in. 
 
 On the north side 3 4 
 ,, south ,,28 
 east ,,25 
 ,, west 2 10 
 
 This room appeared to have been the principal chamber 
 of the mansion, and had two entrances ; the one from the 
 staircase above mentioned, the other from another passage, 
 No. 3. It was built upon a hypocaust of a very remarkable 
 construction. The ground-plan shows the direction of the 
 conduits which served to circulate the heated air under the 
 floor of the apartment. These were built with undressed 
 stones without mortar, and formed a sort of St. Andrew's 
 cross, in the axis of which was a pillar, of similar construction, 
 measuring 2 feet 8 inches in length by 1 1 inches in breadth. 
 They were covered with large slabs of granite, except at the 
 eastern extremity of the pillar, where a large brick tile, 
 measuring 3 feet square and 3 inches thick, and pierced with 
 a round hole at one of its corners, had been substituted. Two 
 layers of cement 6 inches thick formed the floor of the room, 
 and covered the hypocaust. 
 
 The walls, and also the covering slabs of the conduits, 
 were still black from smoke, whilst the interior was one-third 
 full of soot and charcoal. This circumstance leaves no doubt 
 as to the use of these conduits. They had evidently been 
 constructed for the purpose of heating the apartment, and not 
 to act as drains. In the Gallo-Roman villa discovered at 
 Mane-Bourgerel, in the Morbihan, a similar hypocaust was 
 
t./v. 
 
 To face fnge 31. 
 
 U, PLATE III. 
 
EXCAVATIONS IN THE MOUND B. 31 
 
 found, of which De Caumont gives the plan in his Ere gallo- 
 romaine* 
 
 On lifting the covering slabs at the corners of the apart- 
 ment, we were surprised to find on the ends of the conduits 
 the following objects, which appeared to have been placed 
 there intentionally : 
 
 1. At the north-east corner the broken fragments of a 
 quern or hand-mill in granite, 16 inches in diameter, and red- 
 dened by the action of fire. The granite of Carnac takes this 
 tint after being burnt. (B, pi. III. fig. i, bis, reduced.) 
 
 2. At the south-east angle a polished celt in diorite, 5^ 
 inches long, and beside it some fragments of blue glass, and 
 of the red lustrous ware called Samian. (B, pi. III. fig. 2.) 
 
 3. At the north-west corner fragments of common pottery 
 in gray and black earth. 
 
 4. At the south-west corner a small opaque bead of a 
 turquoise-blue colour. Several similar beads, which had 
 apparently formed parts of a necklace, were afterwards found 
 in the progress of the diggings. 
 
 These objects, it may be repeated, appeared to have been 
 placed intentionally in the places where they were discovered, 
 and possibly in conformity with some superstition which is 
 
 * " Villa de Mane-Bourgerel (Morbihan). Je dois a M. Jaquemet, 
 inge'nieur en chef des ponts et chaussees, le plan d'un autre villa de'couverte a 
 Mane-Bourgerel, mais on n'a reconnu qu'une partie des fondations ; il en reste 
 d'autres dans les pieces voisines, de sorte que je me borne a presenter (p. 392) 
 le plan qu'a bien voulu me communiquer M. Jaquemet, sans essayer d'indi- 
 quer la destination des appartements. La piece J etait remarquable par un 
 hypocauste d'une forme particuliere. Le plan montre la direction de conduits 
 qui faisaient circuler la chaleur sous le sol ; j'ai vu a Vieux un hypocauste 
 semblable dont les conduits etaient en partie creuse's dans la roche calcaire 
 au-dessous du pave, ce qui prouve que les moyens les plus simples etaient 
 quelquefois employe's." Ere gallo-romaine, 2 e edit., p. 390. 
 
32 THE BOSSENNO. 
 
 now unknown. In support of this statement, the reader is 
 referred to the description given in the account of the excava- 
 tions of the Mound A (page 16), of a very ancient custom still 
 existing in Brittany, viz. that of placing a polished stone celt 
 in the chimney to preserve the house against lightning. 
 
 As fewer roofing tiles were found when clearing out the 
 room No. 2 than in the other apartments, it was supposed 
 that it might have been the atrium of the dwelling, but after 
 the discovery of the hypocaust this hypothesis became less 
 admissible. 
 
 No. 3 was a passage communicating with the vestibule 
 No. 5, and with the apartments Nos. 2 and 4. It measured 
 22 feet 6 inches in length and 4 feet 6 inches in breadth. 
 
 A red band, two inches broad, had been painted along the 
 plaster which covered the walls at a height of 2 feet 5 inches 
 from the floor. Other remains of mural decoration were 
 found in the bottom of the passage, but in such a bad state of 
 preservation that the design could not be fully made out. 
 
 No. 4. One of the principal rooms of the house measured 
 21 feet in length and 16 feet in breadth. The entrance to it 
 was from the passage just described. 
 
 The average height of the walls was on the north side 
 3 feet; on the south 3 feet 10 inches; on the east 5 feet; 
 and on the west 4 feet. 
 
 The walls of this room had been covered with three coats 
 of plaster, each half-an-inch in thickness. Traces of mural 
 decoration in red, yellow, dark gray, and blue, were visible on 
 the second coat of plaster, and a series of small holes had been 
 picked in it, evidently for the purpose of making the third 
 coat of plaster adhere. This also bore traces of having been 
 
EXCAVATIONS IN THE MOUND B. 33 
 
 painted in similar colours, but of a different design. (See the 
 section of the Mound B, pi. II.) An attempt to find out 
 the complete design of the decoration of this chamber failed 
 from the want of sufficient data, and owing to the small 
 number and diminutive size of the painted fragments of plaster 
 which remained adhering to the walls. 
 
 No. 5. This little room, situated at one end of the 
 passage No. 3, appeared to have been a vestibule, and 
 measured 12 feet in length by 7 feet in breadth. It would 
 seem probable that there must have been an opening here for 
 communicating with the exterior, but this hypothesis could 
 not be verified, as only the foundations of the west wall 
 remained. 
 
 Nos. 6, 7, 8. These three little rooms appeared to have 
 been the sleeping apartments, cubicula, of the mansion. 
 
 The first of these, No. 6, measured 14 feet in length by 
 1 1 feet 6 inches in breadth, and communicated by a doorway 
 with No. 5. Although the north wall was in a very dilapi- 
 dated condition, one could still make out the remains of a 
 doorway which opened on the court, situated between the 
 Mound B and the Mound C (the baths). 
 
 The second bedroom, No. 7, measured 14 feet in length 
 by 1 1 feet 6 inches in breadth. 
 
 The third bedroom, No. 8, measured 9 feet in length, and 
 10 feet 6 inches in breadth, and communicated with the rooms 
 No. 7 on the north and No. 9 on the south. 
 
 Following the same method of working as in the Mound A, 
 trenches were cut under the floors of these bedrooms to a 
 depth of 4 feet. At that depth we found traces of an anterior 
 building and diverse objects, amongst which were a circular 
 
 D 
 
34 THE BOSSENNO. 
 
 bronze brooch of Celtic type with short pin (B, pi. III. fig. 3, 
 natural size) ; fragments of a cup in white glass, and of square 
 formed bottles in blue glass ; several iron objects very much 
 oxidised, which seemed to have been the fastenings of doors ; 
 also numerous fragments of pottery, in gray and black earth, 
 of simple forms, and with straight rims. 
 
 No. 9. This little room, measuring 8 feet 6 inches by 
 1 1 feet 6 inches, communicated with the rooms No. 8 on the 
 north and No. 10 on the east side. 
 
 The partition walls between these five rooms, Nos. 5, 
 6, 7, 8, 9, were very faintly indicated on the floor, and were 
 difficult to make out. 
 
 No. 10. This room measured 21 feet in length by 18 feet 
 in breadth. The mean height of the walls was on the north 
 4 feet; on the south 3 feet 10 inches; on the east 5 feet 5 
 inches ; and on the west side 5 feet 5 inches. 
 
 It communicated by a doorway with No. 9. The floor 
 was 1 2 inches below the level of the other apartments. 
 
 A curious construction was discovered in the south-east 
 corner, shaped like a beehive or baker's oven, and roughly 
 built of large stones of irregular shape. Between it and the 
 north wall of the apartment there was found a layer of large 
 round stones, cemented into the floor. I caused a hole to be 
 dug inside this beehive-like structure to a depth of 6 feet 
 below the level of the floor, but the excavation yielded nothing 
 save the neck of a bottle in blue glass. This curiously shaped 
 structure may possibly have been an oven or furnace, and this 
 conjecture is strengthened by the fact of an opening having 
 been discovered in the wall of No. 1 1 choked up with rubbish 
 which might have answered for the mouth of the furnace. 
 
EXCAVATIONS IN THE MOUND B. 35 
 
 The large quantity of bones, shells of edible molluscs, 
 and fragments of domestic pottery, found in the room No. 10, 
 leave but little doubt as to its destination. It had been the 
 kitchen. 
 
 Besides the bones and shells there were also found in this 
 room several round discs in brick earth, and some weights in 
 polished granite, which may probably have been used as 
 sinkers for nets. 
 
 At this date the sea does not come within a mile of the 
 Bosseno, but formerly it had come close to it. The proprietor, 
 M. le Guennec of Kerfraval, showed me one of his titles of 
 that holding, dated 1 740, which proves that in the last century 
 the sea came up to the lower fields of the Bossenno. This title 
 makes mention of a number of pieces of land, one of which is 
 therein described as 
 
 " Une piece de terre sous pature Ker-guin-go/i, au mette 
 (sic) de laquelle est une source d'eau, et une piece de terre 
 sous pature et jonc qui se couvre de mer aux grands eaux," 
 etc.'"" 
 
 No. ii. This room measured 18 feet in length by 7 feet 
 6 inches in breadth. 
 
 The average height of the walls was, on the north 5 feet 
 10 inches; on the east 3 feet; and on the west side 3 feet 
 8 inches. 
 
 This room seemed to have been made use of as an annexe 
 to the kitchen, and for attending to the furnace above men- 
 tioned. As no traces of even the foundations of a stone wall 
 
 * Description et declaration par mesurage et debornement, fait en 1'etude 
 de M. Henry, notaire royal registrateur, controle'e a Auray le 24 decembre 
 1740, a titre de domaine congeable. 
 
36 THE BOSSENNO. 
 
 could be found, it must be supposed that it had been closed at 
 the south end by a barrier or wooden partition. 
 
 This mansion, of which we conclude a detailed descrip- 
 tion by that of the room No. 1 1, had been very carefully built 
 of small cubic stones of equal sizes, the courses of which were 
 regulated, at regular intervals, by layers of tiles. The floors 
 of all the apartments were formed of thick layers of cement, 
 and the walls were covered with coats of plaster, excepting 
 the room No. 6, which had none. 
 
 The site had been fitly chosen, as well from its convenience 
 as for the pleasing view it commanded of the surrounding 
 country. The Mont Saint Michel had sheltered it against 
 the west winds which came off the Bay of Biscay to this day 
 the most violent in Brittany. Numerous springs of good 
 water still exist all round the Bossenno. The field numbered 
 665 in the map of the Bossenno bears the name Er goh fetan 
 tal pare piard (the old fountain near the stony field). Another 
 fountain (which is said never to fail) is situated in a field 
 farther to the east, and is named Fetan ker in goh (the 
 fountain of the old town). The means of communication 
 may possibly have been by the old road leading from Carnac 
 to the Trinite-sur-mer, which still traverses the Bossenno. 
 Tradition points it out as having been a road before the 
 Roman occupation. 
 
 The following enumeration contains the description of the 
 different objects found in each of the rooms of this dwelling: 
 
 Iron nails of different sizes, and numerous fragments of 
 Gallo-Roman pottery of many different kinds were found in 
 all the rooms. With the view of avoiding repetition, the more 
 noteworthy objects only will be indicated in the order of the 
 numbers of the apartments. 
 
To face / age 37. 
 
 B, PLATE VI. 
 
EXCAVATIONS IN THE MOUND B. 37 
 
 In the passage No. i were found many fragments of 
 curious pottery in fine paste, hard and sonorous, ornamented 
 in relief with diagonal bands and chains. From these frag- 
 ments several vases were reconstructed, similar to that figured 
 in the chromo-lithograph A, pi. IV. 
 
 Part of the neck of an amphora, 3^ inches in length, in 
 red earth, and having two handles, was also found here. (B, 
 pi. VI. fig. i.) 
 
 In the room No. 2, one of the antlers of a deer (Cervns 
 clap/ms), 2 feet 8 inches in length, was discovered near the door- 
 way communicating with the passage No. 3. (B, pi. VI. fig. 2.) 
 
 Near the antler was the head of a small brooch in bronze 
 (B, pi. VI. fig. 3, natural size), also an engraved ring in 
 bronze (B, pi. VI. fig. 4, natural size), and a small brass of 
 Claudius II., on which : 
 
 DIVO CLAVDIO. Head radiated and turned to the 
 right. 
 
 Reverse : CONSECRATIO. An eagle erect and turning 
 to the left. This coin had been struck after the death of the 
 Emperor Claudius. (Cohen, 49.) 
 
 Several pieces of iron very much oxidised were scattered 
 about the floor, and seemed to be parts of sword-blades. 
 
 Two objects in stone were also found here. The one was 
 a polisher in gray schist, similar to that found in the Mound 
 A. The other was a portion of a painter's palette in a fine 
 white stone, measuring 3 inches in length, by 3 inches in 
 breadth. (B, pi. VI. fig. 5.) 
 
 With the different fragments of pottery collected in this 
 room I succeeded in reconstructing 
 
 A lagena, diameter 7 inches, in yellow earth, having two 
 handles and a straight neck, with a bourrelet or collar in 
 relief. (B, pi. VI. fig. 6.) 
 
38 THE BOSSENNO. 
 
 A pear-shaped lagena, diameter 7.] inches, in yellow earth, 
 without handles, having two parallel bands engraved, the neck 
 wanting. (B, pi. VI. fig. 7.) 
 
 A terrine, diameter 10^ inches, in gray earth, with a 
 bourrelet or collar in relief on the outer rim, and having 
 a spout. (B, pi. VII. fig. i.) 
 
 A vase of ovoid shape, height 9^ inches, diameter 8J 
 inches, in coarse grayish brown earth, having a large mouth, 
 with a circular rim projecting outwards, and twelve horizontal 
 bands painted black superposed. (B, pi. VII. fig. 2.) 
 
 In the passage No. 3 the upper and lower stones of a 
 quern or hand-mill, in lava, 14 inches in diameter, were found 
 on the floor opposite the doorway of the room No. 43, and 
 also several fragments of red lustrous ware, two specimens of 
 which are figured, viz. 
 
 1. Fragment of a bowl with ornaments divided in com- 
 partments in relief. In one compartment a winged figure, in 
 the other a boar's head. (B, pi. VII. fig. 3, reduced one half.) 
 
 2. Fragment of the same kind ornamented with a circular 
 band of vertical lines, stamped hollow. (B, pi. VII. fig. 4, 
 reduced one half.) 
 
 In the room No. 4 : 
 
 Fragment of a deer's horn having cuts made by a sharp 
 instrument on its base. 
 
 A bodkin or stylus in polished bone, similar to one which 
 was afterwards found in the room No. 10. (B, pi. VIII. 
 
 % 4-) 
 
 A bead of a turquoise-blue colour, opaque, f of an inch in 
 
 diameter, probably part of a necklace. 
 
 A small brass of Tetricus the younger : 
 
 CAIVS PIVS. ESV. TETRICVS. CAES. (Caius 
 
To face page 38. 
 
 B, PLATE VII. 
 
EXCAVATIONS IN THE MOUND B. 39 
 
 Pius Esuvius Tetricus Caesar). His head radiated, turned to 
 the right, and having the paludamentum. 
 
 Reverse defaced. 
 
 A second brass of Gallienus : 
 
 GALLIENVS. AVG. (Gallienus Augustus); head turned 
 to the right. 
 
 Reverse: LIBERO. P. CONS. AVG. (Libero patri con- 
 servatori Augusti.) Panther passing to the left. On the 
 exergue B. (Cohen, 337.) 
 
 A small brass of Constantine I. 
 
 CONSTANS. NOB. CAES. (Constantius nobilis Cae- 
 sar.) The bust turned to the right, with the paludamentum. 
 
 Reverse: GLORIA EXERCITVS. Two soldiers hel- 
 meted, erect, facing each other, each holding a spear reversed 
 and leaning on a shield ; between them a staff surmounted by 
 a flag. (Cohen, 131.) 
 
 The following vases were reconstructed from the various 
 fragments of pottery found in this room : 
 
 A vase in brownish-gray earth, of ovoid form, having a 
 large mouth, the rim projecting outwards ; on the shoulder 
 three parallel lines coarsely engraved, and between two of 
 these a line of hollow points ; diameter 6f inches. (B, pi. 
 VII. fig. 5-) 
 
 A vase of the same form, in bluish-gray earth, very large 
 mouth, ornamented all round with hollow vertical lines in 
 juxtaposition, separated at the shoulder by three hollow 
 horizontal lines painted black ; paste fine and well fired ; 
 diameter 6^ inches. (B, pi. VIII. fig. i.) 
 
 A bowl in grayish-brown earth, having a narrow base ; 
 diameter 7 inches. This form of bowl is still in use in some 
 parts of Brittany. (B, pi. VIII. fig. 2.) 
 
 Three other fragments of pottery ought to be mentioned. 
 
40 THE BOSSENNO. 
 
 Part of the handle of a vase ; ornament formed by twisting. 
 (B.pl. VIII. fig. 3.) 
 
 Fragment of a vase decorated with undulating lines. 
 (B, pi. VI 1 1. fig. 4.) 
 
 Fragment, in red earth, of the neck of a vase having the 
 same ornament ; length 4^ inches. (B, pi. VIII. fig. 5.) 
 
 The persistence of an ornamentation similar to that on 
 the two fragments just described is very remarkable, and may 
 perhaps be sufficient excuse for the following digression : 
 
 Many nations have made use of these undulating lines as 
 the symbol of running water. 
 
 We find it in the eleventh sign of the Zodiac yz 
 Aquarius. 
 
 The ancient form of the Chinese character for water was 
 similar to the eleventh sign of the Zodiac, and the symbol 
 for distress consisted of the same sign with a bar across, 
 fs picturing the stoppage of the irrigation in the rice- 
 field. 
 
 It is also found in the symbols employed by the Egyptians 
 to represent water J^ll3 mai \\C2 mou. This symbol 
 
 J\ ] 1 1~~~*. _H\ 2T/"""* J 
 
 conveyed the idea of freshness or flowing. On the summit 
 of funeral monuments it seemed to symbolise the flowing 
 of time."" 
 
 The persistence of this ornamentation in the Morbihan 
 from the most ancient times up to the present is to be 
 remarked under very many conditions. It is found not only 
 on the drinking cup in the funeral pottery of the dolmens, 
 but also on the monoliths which form the supports of these 
 monuments. One is tempted to compare these with the 
 rendering of the Egyptian symbol just alluded to. 
 
 * See the Dictionnaire d 1 Archcologie Egyptienne, par Paul Pierret Paris, 
 Rollin and Feuardent, 1875. 
 
B, PLATE VIII 
 
EXCAVATIONS IN THE MOUND B. 41 
 
 These undulating lines are to be found on nearly all the 
 Breton pottery of the last two centuries. They are to this 
 day one of the most common forms of ornament used by the 
 potters of Malansac, who fabricate without a wheel, that is by 
 hand, by the method of procedure called colombin, large vases 
 called charniers or pones.'" 
 
 PONE OF MALANSAC. 
 
 The above engraving represents one of these pones which 
 I drew in the workshop of a potter at Malansac.t 
 
 Whilst I was drawing this pone I asked the potter if he 
 knew the meaning of these ornaments, and why he applied 
 them. He replied to me, " I know several sorts of ornaments 
 
 * Vases 3 feet 6 inches in height, used for washing purposes, on which 
 this design is applied in relief. For the manner in which these vases are 
 made, see the Traite des Arts Ceramiques, par Alex. Brongniart, Tome i. 
 page 391. 
 
 f This pone had another ornament, which is shown in the engraving, 
 placed above the undulating lines. This sign is identically the same with 
 one of the signs of the dolmens, classed by M. de Closmadeuc as the 
 jugiform sign. There is in this fact another proof to add to so many 
 others of the persistence of customs and traditions in Brittany. See also 
 Sculptures Lapidaires et Signes graves des Dolmens. Vannes, 1873. 
 
42 THE BOSSENNO. 
 
 which we apply to different vases because we have always 
 done so, but as to their signification I know nothing, and I 
 never heard it spoken of." 
 
 This ornamentation of undulating lines is also to be seen 
 engraved on the arches over the doors of houses* (see the 
 drawing, page 50), and on the tombstones in the cemeteries, 
 ancient and modern, in the same department. 
 
 We now return to the description of the objects found in 
 the rooms of the building in Mound B. 
 
 In the room No. 5 nothing of any importance was found. 
 
 In the room No. 6 : 
 
 A second brass of Victorinus : 
 
 IMP. C. VICTORINVS. P. F. AVG. (Imperator 
 Caesar Victorinus Pius Felix Augustus.) The bust radiated, 
 turned to the right, with the paludamentum. 
 
 Reverse: PROVIDENTIA. AVG. (providentia Augusti). 
 Figure of Providence erect to the left, holding a rod and horn 
 of plenty ; at the feet a globe. (Cohen, 57.) 
 
 Two small implements in yellow flint polished. The 
 cutting face of one of these is notched from having been used. 
 
 A blue bead like turquoise, J inch in diameter. 
 
 A flat dish in gray earth, 8 inches diameter, reconstructed 
 from fragments, and of the same description as the one 
 figured A, pi. III. fig. 5. 
 
 In the room No. 7 : 
 
 A small brass of Tetricus senior : 
 
 * I owe this drawing to the kindness of Mr. Huray, architect. The 
 curious curve of the arch, reproduced mathematically, merits attention. 
 
To/acefage 43. 
 
 B, PLATE IX. 
 
EXCAVATIONS IN THE MOUND B. 43 
 
 CAIVS. PIVS. ESVVIVS. TETRICVS. 
 
 Reverse defaced. 
 
 Several pieces of marble pavement, white streaked with 
 red. 
 
 A stone implement of a yellowish colour, polished by use, 
 2 inches in length, of the same kind as that figured A, pi. II. 
 
 % 3- 
 
 Several dressed mouldings, in a fine white calcareous stone 
 not found in the Morbihan. Several other dressed stones of 
 the same nature were here found, which seemed to have 
 formed parts of pillars, and which bore the traces of a thin 
 coating of cement. 
 
 The archway of a door, constructed of white stones, 
 cement, and bricks, placed alternately, lay upon the floor in 
 the middle of the room. These stones, the quoins for the 
 arch, bore the marks of the saw with which they had been cut 
 from the block. 
 
 Fragment of a bowl in fine black earth, slightly glazed, 
 having two letters complete, and a portion of a third skilfully 
 engraved after the firing. (B, pi. VIII. fig. 6, natural size.) 
 
 In the room No. 8 : 
 
 Fragment of a vase in fine red lustrous ware, ornamented 
 with zones of small rectangular figures in juxtaposition. 
 Within these was another form of decoration, consisting of 
 oblique lines and points in relief, obtained by pressure in 
 moulds. (B, pi. IX. fig. i, natural size.) 
 
 Fragment of the same nature, ornamented with zones of 
 concentric ovolos or festoonings, also obtained by moulding. 
 (B, pi. IX. fig. 2, natural size.) 
 
44 THE BOSSRNNO. 
 
 In the room No. 9 : 
 
 A granite pestle, 5^ inches in length, of the same kind as 
 that figured E, pi. II. fig. i. 
 
 A small hemisphere in bronze, one inch in diameter, 
 probably part of a button or ornament for harness. 
 
 Several boars' tusks. 
 
 A large quantity of bones of ruminants and shells of 
 edible molluscs, especially of oysters. On submitting the 
 latter to the inspection of Dr. du Gressy, who has contributed 
 so largely to the advancement of oyster-culture in the Depart- 
 ment of the Morbihan, he at once pronounced them to be of 
 the same kind as the oysters of Quiberon. 
 
 There were also found here a great many fragments of 
 pottery of different kinds. 
 
 Rounded discs of tile from 2 to 4 inches in diameter. 
 
 A bottle, reconstructed, in dark yellow earth, turbiniform, 
 with narrow neck and base, and furnished with a handle 
 inserted in the neck below a circular band in relief, orna- 
 mented with painted dark brown horizontal lines. Over all 
 the vase, from the base of the neck downwards, there are 
 horizontal dark brown bands in relief. Its height is 8 inches, 
 diameter 6 inches. (B, pi. X. fig. i.) 
 
 Numerous fragments of coarse pottery of a Celtic type 
 were collected in this little room. Of these the following 
 specimens represent the most noteworthy. 
 
 Fragment of the rim of a bowl in brownish-black earth, 
 badly fired, decorated with horizontal lines in relief; below 
 these a series of hollow grain markings obtained by stamping. 
 (B, pi. IX. fig. 3, natural size.) 
 
 Fragment of the same kind having a large band of 
 undulating lines finely cut. (B, pi. X. fig. 2, natural size.) 
 
B, PLATE X. 
 
EXCAVATIONS IN THE MOUND 13. 45 
 
 Fragment of the same kind decorated with circular lines 
 in relief, having hollow grain markings rudely cut. (B, pi. 
 X. fig. 5, natural size.) 
 
 Fragment of the same nature decorated with five hori- 
 zontal bands of hollow grain markings. The middle band is 
 composed of larger elongated hollows of an oval form 
 irregularly and rudely cut by hand. Below these bands an 
 indication of fern-leaf ornamentation similarly formed by 
 hand. (B, pi. X. fig. 4, natural size.) 
 
 Fragment of the same sort adorned with hollow horizontal 
 lines ; between two of these lines a large band of hollow grain 
 markings similar to those in the preceding fragments, but 
 arranged obliquely in three rows. (B, pi. X. fig. 3, natural 
 size.) 
 
 Fragment of the rim of a bowl in dark gray earth having 
 a letter or sign (?) engraved after the firing. (B, pi. XL fig. i, 
 natural size.) 
 
 Fragment of a vase decorated with geometric figures of 
 different kinds stamped hollow ; paste bluish gray, well fired. 
 (B, pi. XI. fig. 2, natural size.) 
 
 Fragment of the handle of the same vase. (B, pi. XI. fig. 
 3, natural size.) 
 
 In clearing out the room No. 10, a large quantity of bones 
 of ruminants and other animals, also numerous shells of edible 
 molluscs, were found, as in the room No. 9. 
 
 Besides these there was also found a small brass attributed 
 to Constantine the Great or his sons : 
 
 VRBS ROMA. Bust of Roma turned to the left, with 
 the imperial mantle and helmet, an aigrette on the helmet. 
 
 Reverse without legend. The wolf turned to the left, 
 
46 THE BOSSENNO. 
 
 suckling Romulus and Remus, and looking at them. Above 
 these two stars. 
 
 A bodkin 6 inches in length, in polished bone. (B, pi. XI. 
 
 % 4-) 
 
 The vertebra of a young whale 5 inches in diameter, 
 
 having the edges rudely cut, so as to make it serviceable for a 
 pedestal or support. (B, pi. XI. fig. 5.) 
 
 The half of a whistle or hunting call in bone, polished. (B, 
 pi. XI. fig. 6.) This call is similar to some of those found 
 in the forest of Brotonne * now exhibited in the Museum of 
 Caen in Normandy, where the following drawing was made. 
 (B, pi. XI. fig. 7.) 
 
 A round stone, polished, in nephrite, called "ceuf de 
 serpent," similar to those found in the Gallo-Roman cemetery 
 at Saint Jacques de Lisieux, and at Saint Symphorien, near 
 Vannes. 
 
 Two stones of ovoid form, in polished granite, having a 
 groove cut all round in their longer axis. The larger of 
 
 * " Les os chez les Remains dtaient employes, comme chez nous, a la 
 confection d'un grand nombre de petits ustensiles, tels que des cuilleres, des 
 epingles a cheveux, des poingons, des des, etc. On sait qu'on en fabriquait 
 aussi des flutes. Le nom latin tibia est du a cette origine." 
 
 No. 312. " Fragment de flute long de 10 centimetres, perce' de deux trous. 
 Tron9on perce' d'un seul trou circulaire a usage de sifflet. Ces deux objects 
 proviennent des fouilles de Jort en 1852." 
 
 No. 313. "Deux tron9ons semblables au precedent, dont la longueur 
 n'excede pas 25 millimetres, provenant des fouilles de la foret de Brotonne." 
 
 " Ces sifflets dont il est facile de tirer un son aigu, en bouchant les extre'- 
 mite's de la fosse me'dullaire avec le pouce et 1'index, e'taient vraisemblable- 
 ment a 1'usage des chasseurs ; cela parait d'autant plus incontestable qu'ils 
 ont e'te trouves avec des de'bris de cornes de cerf et de dents de sangliers." 
 Catalogue du Musee de la Soriete des Antiqnaircs dc Normandie, rc'dige' par M. 
 Gervais, p. 66. Caen, 1864, F. Le Blanc- Hardel. 
 
To face page 46. 
 
 B, PLATE XI. 
 
To face page 47. 
 
 B, PLATE XII. 
 
EXCAVATIONS IN THE MOUND B. 47 
 
 these measures 8 inches in length by 6 inches in breadth, 
 and weighs 20 Ibs. (B, pi. XII. fig. i.) 
 
 Several discs rudely formed from brick, 3 to 5 inches in 
 diameter. These, and the two preceding ovoid -shaped 
 stones may have probably been used by fishermen as sinkers 
 for their nets ; or they may have been loom weigh ts.*~ 
 
 A flat-shaped mortar in granite, measuring 10 inches in 
 length by 8 inches in breadth, the face polished by use ; had 
 probably been used for bruising grain or preparing food. 
 (B, pi. XII. fig. 2.) Similar stones are to this day used by 
 the natives of Bengal and other parts of India for preparing 
 their food. 
 
 Another stone, in every respect similar to the former, 
 except that on the face the mark IXII was engraved. 
 (B, pi. XII. fig. 3.) Near this stone was found the fragment 
 of a small bowl, having the same mark engraved on its rim 
 after firing. (B, pi. XII. fig. 4.) It looked as if the two 
 latter objects had been marked for some special service. 
 
 Three cylindrical stones, in granite, with the ends rounded 
 
 * " Rien de moins demontre que la veritable destination de ces objets, 
 ordinairement en brique, quelquefois en pierre, de forme conique souvent 
 tronquee et perces dans la partie superieure pour donner passage a un anneau 
 ou a une corde. 
 
 " L'opinion la plus repandue leur attribue la destination de poids a peser. 
 Us ont pu, tout aussi bien, servir de contre-poids ; ceux en pierre que possede 
 le musee, et qui ont ete trouves a Vieux, n'etaient certainement que des poids 
 d'immersion pour les grands filets de peche connus sous le nom de seines 
 (sagena)." 
 
 " La pesanteur inegale de ces cones, rend fort proble'matique leur emploi 
 usuel comme poids a peser. II est rare, en effet, d'en rencontrer qui corre- 
 spondent au poids de la livre romaine, de 6,144 grains, et a ses multiples ou 
 divisions decimales." Catalogue dn musee de la Societe des Antiquaires de Nor- 
 mandie, p. 43, redige par M. Gervais, conservateur. Caen, F. Le Blanc-Hardel. 
 
48 THE BOSSENNO. 
 
 off, measuring from 3 to 5 inches in length, were probably the 
 rollers or pestles which had been used with the mortars above 
 described. 
 
 Several fragments of cups in fine white glass, with striae i 
 relief. 
 
 A vase in black earth (reconstructed), having the form of 
 a water goblet, the mouth widened out from the neck, in which 
 the handle is inserted. At the base of the neck a cord in 
 relief, and on the shoulder at the junction of the handle a 
 frieze of enrolled ornamentation engraved hollow, between 
 two circular lines ; height icf inches, breadth, 8f inches. 
 (B, pi. XIII. fig. i.) 
 
 Another vase (reconstruction) of the same form and 
 description save that the ornament on the shoulder is formed 
 by undulating lines, with points between the undulations. 
 (B, pi. XIII. fig. 2.) 
 
 A deep plate or vase (reconstruction) with a hollow base, 
 in the red lustrous ware called Samian ; diameter lof inches. 
 (B, pi. XI 1 1. fig. 3.) . 
 
 Fragment of the same sort ornamented with small diamond 
 shaped markings, obtained by moulding. (B, pi. XIII. fig. 5, 
 natural size.) 
 
 Fragment of a bowl in the same ware, having palm leaves 
 finely engraved. (B, pi. XIII. fig. 4.) 
 
 Fragment of the rim of a vase in the same ware, orna- 
 mented with undulating lines ; engraved free hand. This 
 fragment is pierced by a circular hole, which appears to have 
 been drilled after the firing. (B, pi. XIV. fig. i.) 
 
 Vases pierced in this way with holes have been frequently 
 found in diggings on Gallo- Roman sites. Different uses have 
 been assigned to these vases, viz. for making cheese, for 
 
To face page 48. 
 
 B, PLATE XIII. 
 
To face page 49. 
 
 B, PLATE XIV. 
 
EXCAVATIONS IN THE MOUND B. 49 
 
 brewing beer, and for cooking purposes. I limit myself merely 
 to the mention of these conjectures. 
 
 A bowl (reconstructed), in brown earth, of hemispherical 
 form, having a circular projection lapping over from the 
 middle ; diameter 7 inches. (B, pi. XIV. fig. 2.) 
 
 Fragment of a lamp for burning tallow or fat, in yellow 
 earth, having an exterior rim which forms a socket."'" (B, pi. 
 XIV. fig. 3.) 
 
 No traces of lamps for burning oil were found. This at 
 first surprised me, but I afterwards learned that they have 
 been rarely found in this district of the Morbihan. 
 
 In the room No. n, the only object of interest that was 
 gathered from amidst a large quantity of tiles and pottery, was 
 a bowl (reconstructed), in gray earth, 7 inches diameter, having 
 vertical sides and a conical base. (B, pi. XIV. fig. 4.) 
 
 After the digging of the Mound B had been completed, 
 and when the different objects found in it had been classed 
 and arranged, the curious mixture of Celtic and Roman 
 
 * " Mon lampion gaulois a etc trouve dans un antique champ de sepul- 
 tures ou on a enterre depuis 1'epoque gauloise (incineration, fosses ovoides), 
 pendant tout le moyen age et jusqu'a la Revolution ; il est, comme son con- 
 frere, en argile bien cuite, jaune-rose, modele au tour ; le godet porte des 
 traces de carbonisation, et en dessous une vasque formant bobeche et portant 
 ces larges poussees qui caracterisent les poteries gauloises dans nos contrees, 
 et qui se sont conservees sur les charniers et pones a lessive, jusqu'a nos jours. 
 
 " J'ai pu constater que toutes les sepultures gauloises par incineration, du 
 Poitou, renfermaient un lampion a suif. Ces lampions se distinguent des 
 lampes a huile en ce qu'ils ne portent pas de bees pour recevoir la meche : 
 cette distinction a ete adoptee par 1'habile et savant directeur du Muse'e de 
 Sevres, M. Riocreux." Essai sur des Poteries antiques de I'ouest de la France, 
 par F. Parenteau. Nantes, Henri Charpentier, 1865, p. 7. 
 
 E 
 
50 
 
 THE BOSSKNNO. 
 
 pottery presented a striking appearance, and represented a 
 remarkable variety of forms, especially in the shapes of the 
 rims. 
 
 Seven baskets full of fragments of pottery had been 
 collected in this mound which weighed 159 Ibs. net. 
 
 DOOR AT BOURGEREL. 
 (From a drawing by M. Huray.) 
 
IV. 
 EXCAVATIONS IN THE MOUND C. 
 

 a a rfbJB 
 
 > n aM_L7K 
 
 -P : :..', 
 
 Meti.-, 
 
 01 2 3 
 
 ',,,,. BritUh Imp 1 Ft w. 
 
 .... ~ f o 
 
 Jajncji' HfUn, fecit-. 
 
 30 
 
 . MoricAts xc. 
 
 To face page 53. 
 
 C, PLATE I. PLAN OF THE BATHS, MOUND C. 
 
THE MOUND C. 
 
 As has been already mentioned in the preceding chapter, the 
 west wall of the habitation discovered under the Mound B 
 was found to be prolonged northwards until it joined on to 
 another mound designated in the plan of the Bossenno by the 
 letter C. 
 
 The Mound C was of an oval form, and had a medium 
 height of 5 feet. Its greater axis, from east to west, measured 
 82 feet, and its lesser axis, from north to south, 46 feet. The 
 exterior coating of this mound consisted of stones, earth, frag- 
 ments of pottery, bricks, and tiles. We commenced work on 
 the north side. 
 
 The result of almost the first few blows of the pick was to 
 bring to light a quantity of the shells of edible molluscs, viz. 
 oysters, whelks, and mussels. Shortly afterwards a hollow 
 tile or conduit brick, 18 inches long, ornamented with five 
 grooves'" in the form of a St. Andrew's cross, was turned up. 
 
 The discovery of this conduit brick was very pleasing, for 
 
 * It was afterwards seen that these grooves served a double purpose 
 sometimes as ornaments, at others for the adhesion of the plaster or mortar. 
 
54 THE BOSSENNO. 
 
 it at once gave rise to the hope that we were about to find a 
 Gallo-Roman bathing establishment under the Mound C. 
 These expectations were fully justified by the result of the 
 diggings. This may be seen on referring to the adjoining 
 ground-plan of the ruins discovered under the Mound C. 
 
 On making a careful examination of this building it be- 
 comes evident that it contains, and in their proper order, all 
 the rooms which, according to Messrs, de Caumont and Rich, 
 were required for a Balneum, viz., ist, the Apodytcrium; 2d, 
 the Tepidarium; 3d, the Sudatorium; 4th, \hs!Caldarium; 5th, 
 the Frigidarium containing its Baptisterium. 
 
 The small size of the different rooms of this establishment, 
 the impossibility of giving separate baths for males and females, 
 as was always found in the public baths, makes me suppose 
 that here we have discovered a Balneum or private bath 
 forming a dependence of the villa found under the Mound B, 
 which has just been described in the preceding chapter. 
 
 What leads me to believe that any other conclusion is 
 untenable is that the west wall of the villa and the west wall of 
 the baths form one and the same wall, and are of identical con- 
 struction. (See the general plan of the Bossenno.) 
 
 Of the south wall, near the letter H of the ground-plan, 
 there remained little more than the foundations. What seemed 
 to have been a wide entrance could here be made out, not 
 sufficiently distinct to admit of its being laid down on the 
 ground-plan, but still leaving the impression that it was the 
 place by which access was had from the court into the 
 passage No. i. 
 
 As in the constructions found under the Mounds A and B, 
 so in all the parts of this building, the proofs of the fire which 
 had destroyed all these edifices were everywhere abundant. 
 
 In order to avoid repetition, I will henceforth give to each 
 
EXCAVATIONS IN THE MOUND C. 55 
 
 room the number which it bears on the plan, also the name 
 by which, according to its destination, it was designated by 
 the Romans. 
 
 The passage No. i measured 47 feet in length by 5 feet 
 6 inches in breadth, and gave access by two doorways to the 
 rooms Nos. 2 and 8. The floor was formed of cement, as 
 were all the floors of the different parts of this building 
 excepting the room No. 9. As has been already mentioned, 
 the traces of a large doorway, which had been the principal 
 entrance to the baths, were visible in the south wall near 
 the letter H, but were not sufficiently distinct to admit of 
 their being laid down on the plan. 
 
 The room No. 2 appears to me to have been the Apody- 
 terii^m* where the bathers undressed and deposited their 
 clothes. It forms a square measuring 13 feet 6 inches of a 
 side. The remains of the four walls had an average height of 
 3 feet 4 inches. On the north side are two doorways com- 
 municating with the Frigidarium, No. 7, and on the east side 
 a doorway giving access to the room No. 3. The regularity 
 of the courses and excellence of construction of these walls 
 excited the attention and admiration not only of my workmen 
 but also of the numerous visitors who came to see the 
 diggings. 
 
 Some portions of the walls had a coating of plaster, deco- 
 rated, at three feet from the floor, with a red band two inches 
 in breadth. 
 
 Underneath the cement which formed the floor was a 
 
 * Apodytcrinm An undressing-room ; especially a chamber in the baths, 
 where the visitors undressed and left their clothes whilst bathing. Dictionary 
 of Ro man and Greek Antiquities, by Anthony Rich. London, 1860. Longman. 
 
 E3 
 
56 THE BOSSENNO. 
 
 deep bed of small round stones, having a few pieces of char- 
 coal and pottery interspersed. This bed appeared to have 
 been prepared so as to form a medium of drainage for the 
 apartment. On digging down to a depth of six feet the 
 natural soil was not reached. 
 
 The room No. 3 measured 13 feet 6 inches in length by 
 8 feet 3 inches in breadth. 
 
 In the middle of this room we came upon a long bench 
 formed of cement, but in a very soft and dilapidated condition : 
 notwithstanding all the care that was taken for its preservation, 
 it fell to pieces on being exposed to the air. Several polished 
 slabs in a fine white calcareous stone were embedded on the 
 upper face, and five pieces of marble slabs, white veined with 
 red, scattered on the floor, were probably the remnants of 
 what had once been the covering of this bench. This room 
 appeared to me to have been the Elceothesium* and that it was 
 on the bench above described that the bather placed himself 
 in order to be rubbed, manipulated with strigils, and anointed. 
 This supposition was so far strengthened by our having after- 
 wards discovered, when clearing the south wall, numerous 
 fragments of vases, in a recess there, which may have been 
 the place for keeping the oils and ointments with which the 
 bather was anointed after he had been rubbed and manipu- 
 lated. 
 
 On the floor a few feeble traces remained of a thin parti- 
 tion, separating this room from the Tepidarium,\ No. 4. It is 
 
 * Elaothesium. The oiling-room in a set of baths, where the oils and 
 unguents were kept, and to which the bather retired to be rubbed and 
 anointed. Dictionary. A. Rich. 
 
 t Tepidarium. A chamber in a set of baths kept at a moderate degree 
 of temperature, in order to prepare the body for the great heat of the Suda- 
 
^ 
 sp^S 
 
 Si 
 
 
EXCAVATIONS IN THE MOUND C. 57 
 
 quite possible that the rooms No. 3 and 4 may have had 
 large openings in the partition between them, so as to be 
 considered as forming only one apartment. This was, be- 
 sides, the most usual arrangement in the private baths ; 
 according to Rich, the Elceothesium and the Tepidarium often 
 formed one and the same room. 
 
 The room No. 4 measured 13 feet in length, by 9 feet 
 6 inches in breadth. The floor, formed of a bed of a red- 
 coloured cement 9 inches in thickness, had been covered with 
 large polished slabs of bluish-gray schist. Several of these 
 remained imbedded in the cement, which also bore the im- 
 press of those that were wanting. 
 
 Under this bed of cement a hypocaust was discovered, 
 having square pillars 2 feet 2 inches in height, placed at a 
 distance of 16 inches from each other. These pillars were 
 formed of bricks, measuring 8 and 9 inches square, set in a 
 thick bed of mortar. 
 
 The hypocaust was choked up with soot, charcoal, cinders, 
 and earth, mingled with a few fragments of pottery. 
 
 Two steps, rudely cut in the rock, were discovered in the 
 south-west corner, near the Sudatorium ; the use of these is 
 not quite apparent. 
 
 In the north-west corner two vertical flues formed of 
 hollow bricks, similar to the one found at the commencement 
 of the diggings, and figured at page 53, served to convey the 
 heated air from the hypocaust. 
 
 There can be little doubt as to the destination of the 
 room No. 4 ; it was the Tepidarium, a room kept at a mode- 
 rate temperature. The communication from this room to No. 
 
 torium or vapour bath, and to break the sudden transition after it, before 
 returning into the open air. Dictionary. A. Rich. 
 
58 THE BOSSENNO. 
 
 5 was by a small stair, of which a few steps remained when 
 the diggings were executed. 
 
 The room No. 5, nearly square in form, measured 5 feet in 
 length by 4 feet 8 inches in breadth. Its destination was not 
 evident until it had been completely cleared out, when it 
 became apparent that it was the Sudatorium * or Sudatio con- 
 camerata. The height of the walls on the north, south, and 
 east side was 6 feet 8 inches, on the west side 7 feet 6 inches. 
 It was constructed over a hypocaust, and was supported by 
 five square pillars. The floor and the walls were faced with 
 polished slabs of blue slate. Two steps, superposed, had 
 been worked into the north wall ; these also were faced with 
 polished slate. 
 
 Double rows of flues, formed of hollow bricks, ascending 
 through the walls from the hypocaust below, served to circu- 
 late the heated air. These hollow bricks, as may be seen on 
 referring to the section C D, were supported by crotchet 
 tiles, tegulce, fastened by iron double-headed nails driven into 
 the mortar, and covered over all by a layer of cement and by 
 the facing of polished slate already mentioned. This arrange- 
 ment of flues was not broken into on the south wall, as I 
 always endeavoured, whilst disengaging these ruins, to pre- 
 serve them as much as possible intact. The system for circu- 
 lating the heated air was besides perfectly shown in the east 
 and west walls. 
 
 In the bottom of this small room, a lead pipe, passing 
 through the south wall at the level of the floor, had served to 
 run off the condensed water. 
 
 * Sudatorium. The sudatory or sweating-room in a set of baths, which 
 was heated by flues arranged under the flooring (suspensurd)^ and sometimes 
 also constructed in the walls of the chamber, when it was specially termed 
 Sudalio concamcrala. Dictionary. A. Rich. 
 
EXCAVATIONS IN THE MOUND C. 59 
 
 A granite stone pillar was let into slabs of slate forming 
 the floor, so as to make a pedestal or seat. Facing this pillar, 
 in the east wall, was a recess, having an opening 8 inches 
 wide, which may have served to regulate the temperature, or 
 to admit fresh air from the exterior at the will of the bather. 
 
 The room No. 6 measured 1 3 feet in length by 5 feet in 
 breadth, and, like the rooms Nos. 4 and 5, was also con- 
 structed over a hypocaust. 
 
 A small bath or reservoir for holding warm water, measur- 
 ing 3 feet 6 inches in length by 2 feet 9 inches in breadth, 
 was discovered on clearing out the north side of this room. 
 Another bath of the same kind, measuring 3 feet 10 inches 
 by 2 feet 6 inches, was found on the south side. 
 
 The bottoms of these baths were laid with polished slabs 
 of a fine white calcareous stone, which is not found in the 
 environs of Carnac. The sides were faced with polished slabs 
 of blue slate. 
 
 Round the sides of each bath six vertical flues, formed of 
 hollow bricks, served to convey the heated air from the hypo- 
 caust below, and to warm the water. 
 
 The situation and arrangement of this chamber indicate 
 its having been the Caldarium* or room into which the bather 
 passed from the Sudatorium to be washed with hot water 
 after the action of the hot-air bath had induced a sufficient 
 perspiration. 
 
 A reference to the ground-plan of the Mound C will show 
 that the flooring between the baths just described had fallen in 
 ruins, and thereby exposed the pillars of the hypocaust below. 
 According to Rich, the Labrum would have been situated 
 
 o 
 
 between these baths, but nothing was found there save some 
 * Caldarium. The thermal chamber in a set of baths. Dictionary. A. Rich. 
 
60 THE BOSSENNO. 
 
 fragments of a large vase and a few curious pieces of frescoes 
 decorated with shells, one of which is figured in the chromo- 
 lithograph C, pi. VI. 
 
 The general aspect and character of this room denoted that 
 great care had been bestowed on its construction, and it was 
 still in a good state of preservation when first cleared out. 
 Unfortunately, numbers of tourists and country people came 
 to see the ruins. They soon carried off the fine white stones 
 of the baths, stripped and overset the walls, and demolished 
 the pillars of the hypocaust. 
 
 The room No. 7 measured 15 feet in length by n feet 6 
 inches in breadth. It contained on the north side a cold bath 
 or Baptisterium* having reservoirs for cold water at each end. 
 
 The cold bath measured 10 feet 6 inches in length by 5 
 feet in breadth. Its depth below the level of the floor of the 
 room was 4 feet. A conduit passing through the east wall, 
 at the bottom of the bath, had served to run off the waste 
 water. Two steps, superposed, were, worked into the south 
 side ; the first at i foot 8 inches, the second at 3 feet 1 1 
 inches from the bottom, destined, probably, to assist the 
 bather in ascending and descending. The sides and bottom 
 were formed of a thick coat of cement. See the section 
 No. i,G H, Cpl. III. 
 
 The reservoirs or cisterns measured each 5 feet in length 
 by i foot 6 inches in breadth. The depth of the reservoir 
 situated at the west end of the bath was 4 feet 10 inches, 
 whilst that at the east end was only 3 feet 4 inches. This 
 difference is caused by the conduit or discharge pipe from 
 the bottom of the bath passing below the latter reservoir. 
 See the section No. 4, A B, C pi. III. 
 
 * Baptisterium. A cold plunging bath constructed in the cdla frigidaria. 
 Dictionary. A. Rich. 
 
EXCAVATIONS IN THE MOUND C. 61 
 
 The sides of the two reservoirs were faced with cement 
 only, whilst the bottoms were carefully laid with slate em- 
 bedded in cement. 
 
 On ascending from the bath to the room above, another 
 conduit is to be seen in the section No. i, G H, passing 
 through the east wall at the level of the floor, and serving to 
 run off the water used by the bather : two seats or benches 
 were worked into the walls one on the east, the other on the 
 west side. 
 
 The construction here leaves but little doubt that it was 
 the room called by the Romans the Frigidarium* 
 
 Two doorways on the south side, separated by a square 
 pillar, gave access to it from the Apodyterium. 
 
 When the clearing out of this room was completed, the 
 pillar was in a good state of preservation, faced with plaster, 
 on which a red band, i^ inch in breadth, was painted 16 
 inches from the floor. 
 
 In clearing out the debris which choked up this room, 
 numerous fragments of curious frescoes were discovered. 
 These appeared to me to have been the remains of what 
 had once formed the ceiling of the Frigidarium, and which 
 had fallen down at the time of its destruction. 
 
 The greater portion of these fragments were quite soft and 
 friable, so much so that many had to be sacrificed in order to 
 extricate a single specimen. 
 
 These fragments, as may be seen on referring to the 
 accompanying chromo-lithographs, were painted in geometric 
 designs red, blue, green, yellow, and white ; decorated with 
 shells embedded in the plaster. Many of these shells were 
 
 * Frigidarium. It was certainly distinct from the cold-water bath (fri- 
 gido lavatio], with which it was enumerated, but situated in an opposite angle 
 of the edifice and adjoining the oiling-room (Elceot/iesiuni). 
 
62 THE BOSSENNO. 
 
 broken, and had fallen off the plaster. In some instances the 
 place which they had occupied on the plaster bore no trace of 
 colour, which shows that they had been placed on the plaster 
 before it was painted. 
 
 The marks of the compasses used by the painter in working 
 out his design were still visible on some of the fragments. I 
 mention this on account of a remark made by many of the 
 numerous visitors who came to see the collection of objects 
 discovered in the course of the diggings. A pair of iron 
 compasses was found in the Mound D, which they declared 
 to be an instrument of comparatively modern date, and they 
 also stated that it was doubtful if compasses were known 
 in Brittany during the Roman occupation. On seeing 
 the marks on the fragments above alluded to they at once 
 acknowledged their error. * 
 
 The colours of these fragments were bright and clear when 
 found, but after a few days' exposure they became dull and 
 faint. 
 
 With the view of simplifying the description of these 
 frescoes and of the reconstruction of the design of the ceiling, 
 I shall first direct the attention of the reader to some of the 
 more important specimens, as shown in the three chromo- 
 lithographs which follow. 
 
 1. Fragment of plaster, the decoration formed of parallel 
 bands, containing geometric figures in polychrome, with shells 
 of different species inserted. (Chromo-lithograph C, pi. V., 
 natural size.) 
 
 2. Fragment of plaster, the decoration formed of un- 
 dulating bands in polychrome, having shells embedded. 
 (Chromo-lithograph C, pi. VI., natural size.) 
 
 * Circinus. A pair of compasses employed by carpenters, architects, 
 masons, and sculptors, for describing circles, etc. etc., similar to those still in 
 use. Dictionary. A. Rich. 
 
O-i 
 
 s 
 
 , ) 
 
 r 
 
O 
 

 
PL. a. 
 
 PL. 4 
 
 Cappe .Del. 
 
 Imp . Leroy.ft. Richon . Paris 
 
 To face page 62, ,?rf. C, PLATK VII. 
 
 PART OF WINDOW AND OK CEILING (REurcun). 
 
EXCAVATIONS IN THE MOUND C. 63 
 
 3. Fragment of plaster, the decoration in polychrome, 
 forming part of a border surrounding an octagonal design, 
 having shells embedded. (Chromo-lithograph C, pi. VII. fig. 
 4, half size.) 
 
 The different fragments discovered, of which these three 
 chromos afford fair specimens, enabled me, with the aid of 
 M. Henry du Cleuziou and M. Louis Cappe, to reproduce 
 the decoration of certain portions of the ceiling, which are 
 represented in the following chromos. 
 
 1. Part of the ceiling reconstructed, the decoration in 
 polychrome, composed of a rosace surrounded by an octagon. 
 The central part of the rosace is divided into eight compart- 
 ments of varied colours ; over the whole shells of the genus 
 turbo and tellina are embedded. (Chromo C, pi. VIII. fig. 9.) 
 
 2. One of the original fragments made use of for the 
 preceding reconstruction. (Chromo C, pi. VIII. fig. 10, reduced 
 one half.) 
 
 3. Part of the border of the ceiling reconstructed, the 
 decoration in polychrome, composed of a large yellow band, 
 having shells of the Cardium echinatum embedded, and of 
 other smaller bands in various colours divided into compart- 
 ments, with shells of the turbo and tellina embedded. (Chromo 
 C, pi. IX., reduced one fourth.) 
 
 4. Part of the ceiling restored, the decoration in poly- 
 chrome, containing a diamond-shaped figure divided into 
 compartments, in which are embedded shells of the Cardium 
 echinatum, turbo, and tellina. (Chromo C, pi. X.) 
 
 M. H. du Cleuziou, after studying the combinations that 
 could be obtained with these fragments, succeeded in making 
 a design of the entire ceiling. 
 
 M. L. Cappe, in following the same method but making use 
 
64 THE BOSSENNO. 
 
 of other fragments subsequently found, made the reconstruction 
 represented in C, pi. XL, which I believe to be more exact. 
 
 I regret that the size of this volume does not admit of his 
 design being given on a large scale in chromo. 
 
 Several fragments of frescoes similar to those just described 
 were discovered a few years ago by M. R. F. le Men, when 
 excavating the Gallo- Roman baths at Poul-Ker, near Ben- 
 Odet, below the town of Quimper. Specimens of these are 
 exhibited in the museum at Quimper. 
 
 The apartment No. 8 probably served as a waiting-room 
 for the bather until the bath was ready, or for repose after 
 taking it. The entrance was by the doorway in the south wall. 
 
 The average height of the walls was on the north side 
 2 feet 6 inches, on the south 4 feet, on the west 3 feet. On 
 the east side the foundation alone remained. 
 
 These remains of walls were in good preservation and had 
 a coating of plaster still preserving some traces of painting. 
 On the north wall, at the height of 10 inches from the floor, 
 and on the right hand, was a horizontal red band 6 inches in 
 breadth. In the middle of the wall this was replaced by a 
 yellow band 6 inches in breadth ; 8 inches higher there was 
 another red band of the same breadth. 
 
 On the east wall traces were found of green leaves painted 
 on a white ground, and of a red sort of fruit on a black ground. 
 (Chromo B, pi. V.)' 
 
 Specimens of plaster similar to those here described, and 
 found in different Roman and Gallo- Roman establishments, 
 are exhibited in the museums at London, Paris, etc. etc. It 
 would seem that the most common style of mural decoration 
 in the Gallo- Roman epoch consisted of bands in primitive 
 colours painted on a white ground. 
 
o 
 
 dd 
 
To face page d^ zd. C, PLATE IX. 
 
 RECONSTRUCTION, BORDER OF CEILING IN FRESCO. 
 
To face page (14,. j,/. C, PLATE X. 
 
 RECONSTRUCTION, PART OF CEILING IN FRESCO. 
 
C. PI. IV' 
 
 
 C, PLATE XI. 
 RECONSTRUCTION, DESIGN OF CEfLINO. 
 
 A'. Jforieii .re. 
 
EXCAVATIONS IN THE MOUND C. 65 
 
 The floor of this apartment was formed of a bed of fine 
 smooth cement, which gave a hollow sound on walking over 
 it. On digging down below it for some distance, nothing was 
 found but small stones laid for drainage in the same way as 
 below the Apodyterium. 
 
 The room No. 9 measured 13 feet in length by 12 feet in 
 breadth. The walls were constructed with irregularly-shaped 
 stones of different sizes, in inferior workmanship, and had no 
 coating of plaster. The floor was without cement. It seems 
 to me that this room had been intended for the use of the 
 slaves who attended on the baths, more especially for those 
 who attended to the heating. The furnace whence the heat 
 was distributed through the baths was placed in the middle of 
 the west wall. In the north-west corner there was a square 
 construction, which appeared to have been the chimney. 
 Several visitors suggested that the room No. 9 had been an 
 open court, but the quantity of roofing tiles found on clearing- 
 out the floor showed that it had been roofed in part, if not 
 entirely. 
 
 The construction No. 10 was an enigma to all who saw it. 
 It had no connection with the baths, although it was joined on 
 to them. It was situated at a depth of 2 feet 8 inches under 
 the pasture which exists between the Mounds B and C, 
 already alluded to as a court. Three or four courses of rude 
 masonry in irregularly shaped stones were all that remained 
 of its surrounding walls. In the middle were two vaulted 
 constructions which appeared to have been furnaces, measuring 
 3 feet in length by i foot 8 inches in width. One of these 
 was filled with charcoal, ashes, and earth, the other with lime 
 and pounded brick dust. 
 
66 THE BOSSENNO. 
 
 Various ideas as to the use of this construction suggested 
 themselves. 
 
 Seeing that it was below the level of the floors of either 
 the villa or the baths, might it not have been the place where 
 the cement, which would be required in large quantities, had 
 been prepared anterior to the building of these edifices, and 
 which had been covered over after their completion ? Or 
 might it have been used for firing the tiles or coarse pottery 
 required for the use of the establishment ? This, however, 
 could hardly have been agreeable to the inhabitants of such a 
 luxurious dwelling. Seeing that none of its masonry bears 
 any trace of Roman workmanship, it may have been con- 
 structed after the destruction of the Bossenno, for some 
 purpose now unknown ; but whatever its destination may 
 have been, the workmen gave it the name of the Furnaces, 
 which it retained. (C, pi. XII.) 
 
 Having now finished the description of the various parts of 
 the building discovered under the Mound C, I shall proceed 
 with the enumeration of the different objects found in it, 
 commencing with those found on the outside of the walls. 
 
 On the north-east face of the mound, near the place where 
 the conduit tile mentioned at the commencement of this 
 chapter was found, a number of square tiles were collected, 
 ornamented with hollow lines, forming concentric circular 
 figures, ovals, and diamond-shaped patterns. Similar tiles 
 found in the Roman ruins at Carhaix and Benodet are 
 exhibited 'in the museum of Quimper, one of which is repre- 
 sented in the engraving at the end of this chapter (page 73). 
 
 At the foot of the north wall of the room No. 4, on the 
 outside of the building, was a large quantity of broken window- 
 
EXCAVATIONS IN THE MOUND C. 67 
 
 glass of a green colour, having one face polished, and 
 the other rough as if it had been cast on sand. On the 
 border of some of the fragments a line of red cement still 
 remained. Alongside of these was an iron bar furnished with 
 cross pieces in lead. (C, pi. XIII. fig. i.) These were the 
 remains of a window which had been broken at the time of 
 the destruction of the building. The chromo-lithograph C, 
 pi. VII. No. 2, represents the reconstruction of a portion of 
 this window. 
 
 At a distance of about five feet from the wall of the room 
 No. 4 there was found a small bronze coin of Constantine the 
 Great, in a good state of preservation : 
 
 CONSTANTINVS. AVG. Bust turned to the left, 
 crowned with laurel, having the imperial mantle, and holding 
 a sceptre surmounted by an eagle. 
 
 Reverse: BEATA. TRANQVILLITAS. An altar 
 surmounted by a globe, above which are three stars : on the 
 altar VOTIS XX. (Cohen, 193.) 
 
 Several fragments of red lustrous ware, and of an amphora 
 in yellow earth, were found at a short distance from the north 
 side of the wall of the room No. 9. All the earth turned up 
 here was passed through sieves, with the view of collecting all 
 the fragments of a vase similar to those already mentioned as 
 having been found in London. It required a good deal of 
 perseverance to select, adapt, and cement these different 
 fragments, so as to complete the reconstruction of this vase, 
 which is represented in the chromo-lithograph A, pi. IV. 
 
 There were also found here a small bronze ring like a 
 marriage-ring, the head of a bronze button or ornament of 
 hemispherical form similar to the one found in the Mound A, 
 several iron nails of different sizes, and a few fragments of 
 wall plaster painted red. 
 
68 THE BOSSENNO. 
 
 In digging round the exterior of the Sudatorium, a large 
 bronze coin of Lucilla Augusta was found on the east side, 
 near to the foot of the walls of No. 9 : 
 
 LVCILL^:. AVG. ANTONINI. AVG. F. (Lucillae 
 Augustae Antonini Augusti filiae). Bust turned to the right. 
 
 Reverse: VENVS. S. C. (Venus senatus consulto.) 
 Venus erect, turned to the left, holding an apple and a sceptre. 
 (Cohen, 79.) 
 
 Near the south-east corner of the Sudatorium we came 
 upon an antler of the common red deer (Cervus elaphus), and 
 several fragments of deers' horns broken and cut. 
 
 The teeth, ribs, and other bones of oxen, which had been 
 of a small race. 
 
 Bones and teeth of the sheep. 
 
 Jaw-bones complete with teeth, and tibia of the red deer 
 (Cervus elaphus). 
 
 Boars' tusks ; teeth and bones of the pig. 
 Bones of hares and rabbits. 
 
 Mingled with these bones were the debris of edible shell- 
 fish, viz. 
 
 The whelk . . (Buccinum undatum). 
 Oyster . . . (Ostrea edulis). 
 Limpet . . . (Patella vulgata). 
 Scallop or palmer shell (Pecten Jacobcus). 
 Ormer . . . (Haliotis tuber cularis). 
 Mussel . . . (Mytilus ediilis). 
 Snail . . . (Helix pomatica!) 
 
 Mr. Wright, in his work The Celt, Roman, and Saxon, 
 refers to the fact that the proximity of Roman sites in 
 England is almost always shown by the presence of immense 
 quantities of the shells of the oyster, cockle, and mussel. He 
 
EXCAVATIONS IN THE MOUND C. 69 
 
 also mentions the curious circumstance that a large species 
 of snail is often found still existing about Roman stations. 
 
 The manufactured articles found here were 
 
 Part of a tool in tempered steel. 
 
 Numerous fragments of Gallo-Roman pottery, consisting 
 of pieces of handles, bottoms, and rims of vases in red, gray, 
 and black earth. 
 
 In the interior of the building there were found in the 
 passage No. i several fragments of common gray pottery, 
 and also of a plate in red ware. 
 
 In the Apodyteriiim, No. 2, only a few fragments of very 
 coarse pottery. 
 
 In the El&othesium and Tepidarium, Nos. 3 and 4, several 
 flakes of black flint, a flat hone or sharpening-stone, and 
 numerous fragments of pottery of different kinds. Amongst 
 these were the bottom of a tripod bowl in the red ware, called 
 false Samian. In the whole course of the diggings only one 
 other tripod bowl was found, viz. in the Mound A. Some of 
 the fragments were in common gray earth, ornamented with a 
 series of circular horizontal bands painted black. 
 
 In the Sudatorium, No. 5, nothing was found save a 
 quantity of bricks, tiles, and earth. 
 
 In the Caldarium, No. 6, there were two bodkins formed 
 of deer's horn, each two inches in length. 
 
 Two fragments of a marble slab, blue with white veins. 
 
 Fragments of a large vase in gray earth, which enabled 
 me to reconstruct the upper half of a vase similar to that 
 figured B, pi. VII. fig. 2. 
 
70 THE BOSSENNO. 
 
 In the Frigidarium, No. 7, several burnt bones were found 
 in the cold bath, also a tool formed from a deer's horn, having 
 the point bevelled off, and several fragments of small vases in 
 fine black lustrous ware. 
 
 In the room No. 8, near the centre, a penannular brooch 
 of Celtic type in bronze, formed from a flat ring, having a 
 simple but not ungraceful ornament engraved on the upper 
 flat of the ring, and having a short pin. (C, pi. XIII. fig. 3, 
 natural size.) 
 
 A small armlet in plain thin bronze wire. 
 
 In the north-east corner, the neck of a small bottle in clear 
 glass. 
 
 A small opaque bead of a turquoise blue colour, half an 
 inch in diameter, probably part of a necklace. 
 
 Several iron nails and pieces of iron of different forms- 
 use unknown. 
 
 Piece of a steel drill, similar to the drills used by carpenters 
 at the present time. (C, pi. XIII. fig. 4, natural size.) 
 
 Fragment of a marble slab, white with red veins. 
 
 The different marbles found at the Bossenno were sub- 
 mitted to the inspection of experts in London and Paris, with 
 the view of ascertaining the locality from which they had 
 been brought. They declined to give a decided opinion, 
 seeing that they had marbles exactly similar, found both 
 in Italy and in Africa, of which they produced specimens. 
 This is an interesting question, for as there are no quarries 
 of marble in the Morbihan, and as marbles have been 
 

 To face page 70. 
 
EXCAVATIONS IN THE MOUND C. 71 
 
 frequently found in excavating the Roman ruins of that 
 department, it points to a foreign commerce of some extent 
 during the Roman occupation.* 
 
 Several whorls in terra-cotta, and one made out of a piece 
 of blue slate, were found scattered on the floor. These had 
 been placed on a wooden spindle and used for spinning thread. 
 To this day in Brittany, the women watching their flocks on 
 the moorlands may be seen spinning in this way. Whorls 
 exactly similar have been used till quite recently in some 
 parts of the Highlands of Scotland, and in Shetland. 
 
 Several of these found during the course of the diggings 
 at the Bossenno were not pierced in the centre, but had an 
 
 oblique perforation near the edge ; this precludes the idea of 
 their having been used for spinning purposes. Perhaps they 
 may have been used as weavers' weights, or sinkers for nets 
 or fishing-lines. Stone sinkers similar to these, though of 
 larger size, are still used by the Shetland fishermen. 
 
 * " II n'existe done du terrain de transition, dans le Morbihan, que 
 deux etages, le Cambrien et le Silurien, separes Tun de 1'autre par 1'eruption 
 des granits k petits elements. L'etage cambrien est forme de gres a 
 grains, fins, de schiste talqueux ou chloritiques et de mica schistes stauro- 
 tidiferes." 
 
 " L'etage silurien comprend le gneiss, les poudingues, les quartzites, les 
 schistes argileux, les grauwaches schisteux et les mica schistes macliferes." 
 Bulletin de la Soci'ete Polymathique du Morbihan, pour 1'anne'e 1860. 
 
72 THE BOSSENNO. 
 
 Several pieces of tiles ornamented in the same manner as 
 those found on the north-east side of the Mound C were 
 found in clearing out this room ; also several fragments of 
 pottery in common gray earth, in fine black earth having a 
 lustrous glazing, in red lustrous ware, and rims, bottoms, and 
 handles of vases of varied sorts. From all these fragments 
 only one vase could be reconstructed, a hollow dish in common 
 gray earth similar to that described in the Mound A. (A, pi. 
 III. fig5). 
 
 In the room No. 9, near the south-east corner, were a 
 quantity of rounded pieces of tiles, from 3 to 4 inches in 
 diameter, probably sinkers or pieces for playing at some 
 game. To this day, in Brittany, the game of patit is a favourite 
 amusement. The workmen, on finding the rounded pieces of 
 tile, exclaimed, " They played at pa!6t then, these ancients as 
 we do now." The earth and stones cleared out here were, as 
 in all the other rooms, mixed with numerous fragments of 
 bricks and roofing tiles, tegulte and imbrices. Near the base 
 of the north wall were a few small fragments of wall plaster, 
 coloured red and yellow, which had probably fallen from the 
 adjoining room at the time of the destruction of the building. 
 
 Near the north-east corner was a small bronze coin, 
 slightly degraded, of Constantius the Second. 
 
 D. N. CONSTANTIVS. P. F. AVG. (Dominus noster 
 Constantius pius, felix Augustus.) Bust to the right ; diadem 
 and paludamentum. 
 
 Reverse: FEL. TEMP. REPARATIO. (Felix tem- 
 porum reparatio.) Soldier in a fury, erect to the left, piercing 
 with his spear and spurning with his foot an enemy who, 
 holding a shield, has fallen below his horse in endeavouring 
 to hold on to the mane. A shield on the earth. (Cohen, 223.) 
 
EXCAVATIONS IN THE MOUND C. 73 
 
 Near to the east wall were several iron nails of different 
 sizes, also three implements in steel, resembling punches and 
 drills. 
 
 Numerous fragments of coarse pottery enabled me to 
 reconstruct a flat deep dish, already described as er cass, and 
 another, 8 inches in diameter, pear-shaped, in gray-brown 
 earth, with black bands. Several fragments of vases, in 
 gray earth, similar to those described in the Mound A, 
 having a decoration resembling Moorish pottery, were also 
 found in this room ; specimens of these are represented, C, 
 pi. XIII. figs. 5 and 6. 
 
 The fragments of pottery collected in the Mound C filled 
 five baskets, and weighed 143 Ibs. 
 
 TILE FOUND AT CARHAIX. 
 
V. 
 EXCAVATIONS IN THE MOUND D. 
 
THE MOUND D. 
 
 BEARING in mind that the Romans generally placed a temple, 
 dedicated to Venus or Mars, in the vicinity of their baths ; 
 so soon as the exploration of the Mound C was terminated, 
 I sought all round for traces of a building that might have 
 had such a destination. 
 
 There seemed to be some prospect of finding this (taking 
 into account the persistence of names and traditions in 
 Brittany), when I learned from the peasants that the Mounds 
 D and E were known in the country by the name of the 
 temple. This statement was confirmed by my workmen. 
 
 On placing poles some distance apart on the north wall of 
 the baths, it could be seen that the southern face of the 
 Mound D was in the prolongation of a line drawn through 
 these poles. On digging holes in this line, in the cultivated 
 land between the Mounds C and D, the foundations of the 
 wall which united the baths to the building buried under the 
 Mound D were discovered. 
 
 The excavation of the Mound D was commenced on the 
 1 4th July 1875, and abundant proofs were soon found that it 
 covered the ruins of a small temple or Lararium* 
 
 * Lararium. A sort of shrine, small chapel, or apartment, where the statues 
 of the Lares or guardian spirits of a household, as well as other sanctified or 
 deified personages, were placed and worshipped. Lamprid, Alex. Sev., 29 
 
78 THE BOSSENNO. 
 
 This little mound was of a roundish form, and almost flat : 
 it measured 44 feet in diameter and 2 feet in height. In 
 appearance it was so insignificant as to cause it to be over- 
 looked at first, when searching for indications of a temple, but 
 the results obtained from its exploration were of great interest. 
 
 Work was commenced on the south side, and in a short 
 space of time a wall was struck upon, right in the prolonga- 
 tion of the line indicated by the poles. We continued work- 
 ing round the Mound, following and clearing the outside of 
 the wall until the ruins of a square building, measuring 32 
 feet 6 inches of a side, about 2 feet in height, and having an 
 entrance from the south, were completely exposed to view. 
 The clearing out of the interior was then proceeded with, and 
 soon disclosed the remains of another square building, measur- 
 ing 14 feet of a side, about 2 feet in height, with walls 31 
 inches in breadth, and without any entrance. These two 
 constructions represented (see the ground-plan of the Mound 
 D) two squares, one within the other, having a free passage 
 8 feet wide between their four sides. On the floor of the 
 inner square, at a distance of 2 feet from the north wall a 
 slab of fine white calcareous stone, 2 feet 3 inches square, 3 
 inches in thickness, and having mouldings cut on three sides, 
 was found embedded in a layer of cement. It appeared to me 
 that this had served as the base of a pedestal on which had 
 been placed the statue of the divinity to whom this little 
 temple had been dedicated. 
 
 As the walls surrounding the inner square showed no 
 indication of an entrance, it may be supposed that the access 
 
 and 31. Such an arrangement, however, was probably peculiar to particular 
 individuals, or to great houses and persons of wealth, for the usual situation 
 for images of the Lares was over or beside the hearth (focus) in the great 
 hall or atrium of the house. 
 
To face page 78. 
 
 D, PLATE I. PLAN OF THE LARARIUM, MOUND D. 
 
EXCAVATIONS IN THE MOUND D. 79 
 
 had been by wooden steps, which have long since disappeared, 
 leaving no trace of their existence. 
 
 The masonry of this building, constructed in the small 
 cubic stones used by the Romans, was not by any means so 
 carefully finished as those found under the Mounds B and C. 
 It had a much older appearance. The floor was of a hard 
 concrete, formed from argillaceous earth and small water-worn 
 pebbles. 
 
 The following objects were found during the excavation 
 of this mound. On the outside of the building, near the 
 entrance, a pair of iron compasses, very much oxidised. (D, 
 pi. II. fig. i, natural size.) 
 
 Several fragments of wall plaster painted blood-red. 
 
 Several fragments of pottery, in fine gray paste, well fired, 
 ornamented with a red trellis-work in relief, on a dark 
 ground. (D. pi. II. fig. 2, natural size.) 
 
 On the outside of the west wall of the building several 
 fragments of pottery, which served to reconstruct the form of 
 a bowl, in reddish-brown paste, of rude workmanship, and 
 badly fired. (D, pi. II. fig. 3.) 
 
 Fragments of statuettes, in white terra-cotta, of the god- 
 dess of maternity, and of Venus Anadyomene. (D, pi. IV., 
 V., and VI.) 
 
 On the outside of the north wall of the building : 
 Fragments of coarse pottery, in gray and brownish-black 
 paste, badly fired. 
 
 On the outside of the east wall of the building : 
 Fragments of statuettes, in white terra-cotta, of the god- 
 dess of maternity, and of the Venus Anadyomene. 
 
80 THE BOSSENNO. 
 
 Fragments of a vase, in yellow paste, having a three-lobed 
 mouth. (D, pi. II. fig. 4.) 
 
 Several fragments of rude pottery, and bottoms and rims 
 of vases, in gray and brown paste, badly fired. 
 
 In clearing out the interior of the building there were 
 found in the passage or enceinte on the south side : 
 
 A small piece of a marble slab, red, with white veins. 
 
 Fragments of vases, in coarse yellowish-red paste. 
 
 Part of the handle of a vase, having vertical striae in 
 relief, separated by lines of little holes which perforate the 
 handle ; paste dark brown. (D, pi. II. fig. 5.) 
 
 The upper part of a vase with a wide mouth, having a 
 rude ornament formed on the neck by the prints of the potter's 
 fingers, in coarse paste, and badly fired. (D, pi. III. fig. i.) 
 
 Several fragments of the same kind, with which two similar 
 vases were reconstructed. The smallness of the prints of the 
 potter's fingers on all these vases leads one to suppose that 
 they had been made by women. In the potteries of Malansac 
 and Rieux in Brittany it is the women who make and fire all 
 the varieties of the coarser domestic pottery manufactured in 
 these establishments ; while the work of the men is limited to 
 preparing the paste and modelling the larger and heavier 
 vases, as has been already described. Brogniart, in his Traiti 
 dcs Arts Cer antiques, gives a very interesting description of 
 this manufacture. 
 
 In the passage on the east side : 
 
 Fragments of coarse pottery, and of statuettes in terra-cotta. 
 
 The tooth of a bear, hollowed out and perforated with a 
 hole at the half of its length, so as to form a sort of whistle 
 or hunting-call. (D, pi. III. fig. 2, natural size.) 
 
To face page 80. 
 
To face page 81. 
 
 PLATE III. 
 
EXCAVATIONS IN THE MOUND D. 
 
 81 
 
 Within the inner square : 
 
 The half of a stone axe or hammer, in polished granite, 
 having a hole perforated for the handle. 
 
 A necklace bead in black glass, opaque, ornamented with 
 circular bands in white enamel, one of which is undulating. 
 (D, pi. III. fig. 3, natural size.) 
 
 A few fragments of thin clear glass. 
 
 Fragments of pottery, from which the three following 
 vases were reconstructed : 
 
 A sort of bowl, 6 inches in diameter, 2\ inches deep, in 
 coarse brown paste, badly fired ; sides thick and nearly 
 vertical; rim with a bevelled edge. (D, pi. III. fig. 4.) 
 
 A flat dish, 9^ inches in diameter, i^ inch deep, in 
 coarse brown paste, badly fired, and with deep bevelled edges. 
 
 (D,pl. III. fig. 5.) 
 
 A flat dish, diameter io| inches, depth f inch, in fine blue 
 paste, well fired, of nearly the same form as the preceding 
 
82 THE BOSSENNO. 
 
 vase, but having the sides more spread out. (D, pi. III. 
 fig. 6.) 
 
 Numerous fragments, in white terra-cotta, of statuettes of 
 the Venus Anadyomene and of the Goddess of Maternity. 
 
 Many of the fragments of statuettes found inside the 
 building united exactly with those found on the outside, and 
 thus enabled me in a measure to complete the various 
 statuettes figured in the accompanying engravings. (D, pi. 
 IV. V. and VI., one fourth of the actual size.) 
 
 These statuettes are not exactly alike, and although rudely 
 modelled, are not entirely destitute of artistic feeling. The 
 Venus Anadyomene is represented standing on a hemi- 
 spherical base, the right hand twining the long tresses of 
 hair, whilst the left holds the drapery. The goddess of 
 maternity is represented by a female figure clothed in a 
 long tunic, seated in a sort of arm-chair formed of wicker 
 work. In some figures she is represented suckling one, in 
 others, two infants. 
 
 Similar statuettes, which have been found in many parts 
 of France, have been described by M. Mazard in his de- 
 scriptive study of the ceramics of the Museum of Gallo-Roman 
 Antiquities at St. Germain-en-Laye, as imitations of Greek 
 and Roman statues. Other writers speak of them as repre- 
 senting Latona, the protectress of mothers and nurses ; 
 Lucina, who presided at accouchements ; and also as repre- 
 senting Leda, Juno, and even Isis. De Caumont states that 
 sometimes these have been taken for statues of the blessed 
 Virgin, and that in some places their discovery has been 
 regarded as miraculous, and has given rise to pilgrimages. * 
 
 * Ere Gallo-romaine, 2de edition, p. 586. Par. M. A. de Caumont 
 Caen, 1870. F. le Blanc Hardel. 
 
EXCAVATIONS IN THE MOUND D. 83 
 
 The reunion here of so many statuettes representing the 
 same divinity leads me to suppose that the little temple in 
 which they were found had been dedicated to Venus 
 Genetrix. The pedestal, of which the base was found in 
 situ, supported, in all probability, a statue of this goddess, 
 and one may look upon the numerous small statuettes as 
 votive offerings which had been deposited, and of which 
 some had been broken and others carried off, and with these 
 the principal statue, at the time of the pillage and destruction 
 of the Bossenno. 
 
 M. Tudot finds a proof that the worship of Venus was 
 general among the Gauls, from the reproach addressed by 
 Saint Augustine to his contemporaries, that they filled a 
 lararium with divinities, amongst whom Venus always pre- 
 sided.'"" 
 
 The pottery found in the Mound D is of a different 
 character from that found in the Mounds A, B, and C. It is 
 nearly all of primitive forms, badly fired, in a coarse paste, 
 which had a greasy unctuous feeling to the touch. The 
 weight of the fragments collected was 2 7 Ibs. 
 
 The earth we removed from the interior of this building 
 was laid aside and afterwards passed through sieves. This 
 operation yielded thirteen Roman coins. This is a much 
 Greater number than were found in the Mounds B and C. 
 
 o 
 
 May we suppose of these, as of the small statuettes, that they 
 had been offerings made to the presiding divinity of the 
 
 * Collection de figurines in argile, oeuvres premieres de 1'art Gauloise, 
 avec les noms des ceramistes qui les ont executees, par Edmond Tudot, p. 
 270. Paris : Rollin et Feuardent. 
 
84 THE BOSSENNO. 
 
 temple ? These coins are all in bronze, and embrace a period 
 of 184 years, from the year A.D. 169, the date of the third 
 consulate of Marcus Aurelius, to A.D. 353, the epoch of the 
 death of Magnentius. The following is a detailed description 
 of these coins in chronological order : 
 Two of Marcus Aurelius, large brass. 
 
 1. M. ANTONINVS. AVG. TR. P. XXIII. (Marcus 
 Antoninus tribunitiae potestate 23.) The head laurelled, 
 turned to the right. 
 
 Reverse: SALVTI. AVG. COS. III. S. C. (Saluti 
 Augusti consul tertium senatus consulto.) Health, erect, and 
 turned to the left, holds a sceptre and sacrificial cup, in which 
 a serpent drinks. (922, J. C. 169.) (Cohen, 618.) 
 
 2. M. ANTONINVS. AVG. TR. P. XXVII. (Marcus 
 Antoninus Augustus tribunitiae potestate 27.) Bust laurelled 
 and turned to the right. 
 
 Reverse: RELIG. AVG.; on the exergue IMP. VI. 
 COS. III.; on the field S. C. (Religio Augusti. Imperator 
 sexto consul tertium, Senatus consulto.) Mercury turned to 
 the left, erect, on a pedestal, holding a caduceus. An exchange 
 in a temple, with four columns and arches. The four columns 
 on the pediment are represented, a turtle, a ram, a caduceus, 
 a cock, a plumed helmet, and a purse. Eckhel explains, 
 according to Diodorus Siculus, the reason why Mercury is 
 represented as the type of religion on the coins of Marcus 
 Aurelius. "In Egypt Mercury was the director of the wor- 
 ship of the gods and of sacrifices." (Cohen, 614; Liv. i., 
 chap, xvi.) 
 
 A large brass of Septimius Severus, very much defaced. 
 Four of Gallienus, small brass. 
 
EXCAVATIONS IN THE MOUND D. 85 
 
 1. GALLIENVS. AVG. (Gallienus Augustus.) Bust 
 radiated, turned to the right, with a cuirass. 
 
 Reverse: FORTVNA REDVX. (Fortuna redux.) 
 Fortune erect, turned to the left, holding a rudder and 
 cornucopia. 
 
 2, 3, 4. GALLIENVS. AVG. (Gallienus Augustus.) 
 Bust radiated, turned to the right, with cuirass. 
 
 Reverse: MARTI PACIFERO. (Marti pacifero.) 
 Mars helmeted, erect, turned to the left, holding an olive 
 branch, leaning on a shield ; his spear placed on his left arm, 
 in the field A. (Cohen, 354.) 
 
 A small brass of Tetricus the elder, much defaced. 
 
 Two of Tetricus the younger, small brass, both alike. 
 
 C. P. E. TETRICVS. CAES. (Caius Pius Esuvius 
 Tetricus Caesar.) Head radiated, turned to the right, with 
 the paludamentum. 
 
 Reverse: PIETAS. AVGVSTO. (Pietas Augustorum.) 
 Aspersor, simpulum, sacrificial vase turned to the left, sacri- 
 ficial knife and divining rod. (Cohen, 32.) 
 
 A small brass of Constans I. 
 
 D. N. CONSTANS. P. F. AVG. (Dominus noster 
 Constans Pius Felix Augustus.) Bust, with diadem, turned 
 to the right, with the paludamentum and cuirass. 
 
 Reverse: FEL. TEMP. REPARATIO. (Felix tem- 
 porum reparatio.) Constans erect, to the left, in military 
 costume, on a vessel going to the left, holding a globe sur- 
 . mounted by a phcenix and imperial standard ; to the right, on 
 the vessel, Victory, seated, holds the rudder and regards the 
 emperor. (Cohen, 114.) 
 
86 
 
 THE BOSSENNO. 
 
 A second brass of Magnentius. 
 
 IMP. CAE. MAGNENTIVS AVG. (ImperatorCaisar 
 Magnentius Augustus.) Bust nude, turned to the right, 
 with the paludamentum. 
 
 Reverse: PEL. TEMP. REPARATIO. (Eelix tcm- 
 porum reparatio.) Magnentius in military costume erect, 
 turned to the left, on a vessel going to the left, holding a 
 Victory and the imperial standard. To the right, on the 
 vessel, Victory seated holding the rudder. (Cohen, 35.) 
 
 LECH AT PLOUHARNEL. 
 
VI. 
 EXCAVATIONS IN THE MOUND E. 
 
To face page 89. 
 
 E, PLATE I. PLAN OF THE VILLA AGRARIA. 
 
THE MOUND E. 
 
 THE Mound E, of an oblong form, and covered with dwarf 
 trees and bushes, measured 66 feet from north to south, 33 
 feet from east to west, and had an average height of 6 feet. 
 It was situated to the west of the Mound B, and lay nearly 
 parallel to it. Corn-fields surrounded it on the north, south, 
 and west sides, and a road passed along the east side. The 
 exploration of this mound was commenced on the 23d July 
 
 1875- 
 
 The remains of the walls of a house were soon exposed, 
 built with the small cubic stones used by the Romans, bound 
 with lime mortar, and having at intervals a layer of large 
 bricks. It soon became evident that these ruins had been 
 encroached upon by the cultivation of the surrounding corn- 
 fields, and by the formation of the road, both of which had 
 destroyed the walls in several places. Afterwards, when the 
 mounds E and F had been excavated, and the buildings found 
 under them exposed to view, we could see that the walls of E, 
 if prolonged to F, would join exactly to the buildings found 
 under that mound. The proprietors of the corn-fields kindly 
 permitted a few holes to be dug after the harvest was over, 
 and in these, at a depth of 3 feet 6 inches, the continuation of 
 the foundations of the wall between E and F were found. 
 In the adjoining field, on the north side of the building, it was 
 
90 THE BOSSENNO. 
 
 found, in a similar way, that the foundations of the walls were 
 also prolonged to the north. 
 
 Permission could not be obtained to pursue these researches 
 in a regular way, and this prevents my being able to give 
 either the complete form of the building, or to indicate with 
 certainty the destination of the different apartments. The 
 accompanying ground-plan of the Mound E represents these 
 ruins in the state in which they were when newly excavated, 
 and the following description of the different rooms is given 
 in the order in which they were discovered. 
 
 In the room No. i the west and south walls alone remained, 
 and of these only four or five courses of masonry. There was 
 neither plaster on the walls nor cement on the floor, which, 
 however, was sufficiently hard and had been formed apparently 
 from beaten earth. The walls were in a bad state of preserva- 
 tion, as were all the other walls of this house. A large 
 quantity of Roman roofing-tiles (tegulce and imbrices], for 
 the most part broken in fragments, lay scattered on the floor. 
 There were also to be seen, as in all the buildings hitherto 
 explored, the unmistakable proofs of the fire which had 
 destroyed them. 
 
 The division No. 2 extended along the west side of the 
 building, and seemed, judging from a few faint traces on the 
 floor, to have been divided into a number of small rooms. 
 The west wall had only its two lower courses remaining, so 
 that it was impossible to ascertain whether there had been any 
 doors on this side. The east wall had some portions of the 
 plaster which had once covered it still adhering, and the fact 
 that it was not bonded at the corners of the room No. 4 
 indicated that that room had been constructed before the other 
 parts of the house. 
 
EXCAVATIONS IN THE MOUND E. 91 
 
 The room No. 3 measured 20 feet 9 inches in length by 
 8 feet 3 inches in breadth. Three courses of masonry alone 
 remained of the north wall. The south and west walls had an 
 average height of 3 feet 6 inches, and were in a better state of 
 preservation. A few pieces of plaster remained adhering to 
 the south wall. No trace of even the foundation of a wall 
 was found on the east side. The floor was formed of a 
 mixture of earth and small water-worn pebbles. In the middle 
 of this room the foundation of a wall in large roughly-dressed 
 stones traversed the apartment in a slightly slanting direction. 
 The top of these stones was about six inches above the level 
 of the floor. 
 
 The room No. 4 measured 19 feet in length by 17 feet in 
 breadth. The average height of what remained of the walls 
 was 2 feet 9 inches. It was larger and in a better state of 
 preservation than any of the other rooms of this building. 
 A doorway on the east side gave access from the exterior. The 
 walls had been covered with plaster, and on the floor, which 
 w r as similar to that of the room No. 3, was the continuation of 
 the foundation of the wall described in connection with that 
 room. This foundation was afterwards found to be continued in 
 the floors of the rooms Nos. 5 and 6. On the right-hand side 
 of this foundation, in the room No. 4, was a piece of masonry 
 of an oblong form, the stones of which were reddened by the 
 action of fire. It was covered with cinders, scoriae, and lumps 
 of iron, and seemed to be the remains of a furnace or forge. 
 
 The room No. 5 measured 20 feet 9 inches in length by 
 8 feet 3 inches in breadth. No trace of a wall was found 
 on the east side ; those remaining on the north, south, and 
 west side had an average height of 3 feet. A doorway in 
 
92 
 
 THE BOSSENNO. 
 
 the south wall gave access to the room No. 6. No trace of 
 plaster was found on the walls nor of cement on the floor. 
 
 The dimensions of the room No. 6 could not be ascer- 
 tained, as it extended under the adjoining field, where I had no 
 permission to continue the diggings. The floor was formed 
 of earth and small rounded water- worn pebbles. 
 
 The room No. 7 was under the same conditions as No. 6. 
 The floor was exceedingly hard. 
 
 The following objects were found in working round the 
 outside, and in clearing out the different rooms of this 
 building : 
 
 Two fragments of a curious brick, in grayish earth, pierced 
 with circular holes and having their exterior edges polished 
 by friction, were found on the east side of the building, at the 
 foot of the wall of the room No. 4. 
 
To face page 93. 
 
 K, PLATK II. 
 
EXCAVATIONS IN THE MOUND E. 93 
 
 This brick had formed one of a series which admitted of 
 being placed in and withdrawn from the window of an apart- 
 ment into which the round holes served to circulate light and 
 air. Two similar bricks were found by M. Parenteau in the 
 diggings at Pouzauges.* 
 
 Near these fragments of a brick were the following 
 objects : 
 
 A pestle or rubbing-stone in granite, length 5^ inches. 
 (E, pi. II. fig. i.) 
 
 Part of a small cup, 3^ inches in diameter, in iridescent 
 glass. (E, pi. II. fig. 2.) 
 
 Pieces of an iron sword-blade, very much oxidised. 
 
 Iron clamps and nails of different sizes. 
 
 In the room No. i : 
 
 Fragment of a thin bronze plate, 2 inches long, covered 
 with a fine patina, and having two rivets affixed. 
 
 Numerous discs formed from brick tiles, similar to those 
 found in the Mounds B and C, varying in diameter from i^ 
 inch to 3^ inches. 
 
 Several fragments of pottery of Celtic type, in coarse 
 
 * " Grande brique en terre cuite modelee a la main, a base large. Elle 
 est percee de grands trous disposes regulierement sur une ligne circulaire ; 
 ses deux faces sont couvertes d'ornements graves en creux, avec une pointe 
 seche, avant la cuisson, et faisait partie d'un tout compose de deux briques 
 semblables placees 1'une sur 1'autre, et dont les trous se raccordaient. Les 
 dessins offrent la plus grande analogic avec les feuilles de fougeres de notre 
 pays, et les guillochures des bracelets gaulois, que Ton rencontre souvent 
 dans nos contrees. Cette brique a ete ramassee par moi sous les ruines du 
 vicus de Bournigal, au pied du vieux chateau de Pouzauges (Vendee), parmi 
 des debris gaulois sans nombre : murs eonstruits en pierres scenes ; fragments 
 de poteries, epees, fers de fleches et poignard en fer." Essai sur des Poteries 
 antiques deFoucst de la France, par F. Parenteau. Nantes, 1865. 
 
94 THE BOSSENNO. 
 
 orange-coloured paste, decorated with close vertical striae 
 engraved by hand. (E, pi. II. fig. 3, natural size.) 
 
 Several fragments of large amphorae in a yellow paste. 
 
 In the room No. 2 : 
 
 A large quantity of the bones of ruminants, teeth of the 
 ox and pig, and shells of edible molluscs. 
 
 A bodkin in polished bone similar to those found in the 
 Mounds B and C ; length 9 inches. 
 
 The upper part of a vase, larger diameter 4^ inches, in a 
 grayish-brown paste, having a large trilobed orifice ; the handle, 
 with cords in relief, joins on to the edge of the rim. (E, pi. 
 II. fig. 4.) Vases of this form are to this day manufactured in 
 Brittany : the orifice is obtained by the compression of the 
 anterior part of the neck by the thumb and forefinger. 
 
 In the room No. 3 : 
 
 Numerous discs formed from tiles, diameter i-J- and 3^ 
 inches. 
 
 Several nails and pieces of rod iron, much oxidised. 
 
 Several fragments of cups and bottles in white and blue 
 glass. 
 
 In the room No. 4 : 
 
 A small pestle formed of granite. 
 
 Fragment of an ornament cut in rock crystal. 
 
 A rounded flint stone like a beach pebble, perforated by a 
 circular hole, in which a bronze ring is inserted. This curious 
 object may probably have been worn as an amulet. 
 
 A bronze object, use unknown, but possibly intended to 
 be worn as an ornament. This interesting piece of fine work- 
 manship is formed by flat bands forming an equilateral 
 
To face page 95. 
 
 E, PLATE III. 
 
EXCAVATIONS IN THE MOUND E. 95 
 
 triangle, and at the angles, circles and dragons' heads. The 
 base is formed by another band forming a cross with four 
 circles cut open. The whole object is engraved with an orna- 
 mentation in lines and points. (E, pi. III. fig. 2, natural size.) 
 
 A Roman coin, second brass, re-struck. Part of what 
 appears to have been a tin buckle. 
 
 A statuette in bronze was found in the south-east corner 
 of this room which somewhat resembles the small statuettes 
 of the bull Apis exhibited in the Egyptian rooms of the 
 British Museum, and also in the Museum of the Louvre. 
 The chromo-lithograph E, pi. VI., represents this curious 
 object, natural size. It will be further alluded to at the 
 conclusion of this chapter. 
 
 A quantity of iron rings, clamps, nails, and pieces of 
 rod iron, were scattered about near the construction which 
 resembled a forge or furnace. 
 
 An iron lance or arrow head, having a hollow stem for 
 receiving the shaft, much oxidised. (E, pi. III. fig. 3, natural 
 size.) 
 
 A large iron hook, apparently a fishing-hook, much oxi- 
 dised. (E, pi. III. fig. 4, natural size.) 
 
 An iron knife, much oxidised. (E, pi. III. fig. 5, natural 
 size.) 
 
 Several pieces of ironwork, which seemed to have been 
 door fastenings. 
 
 Neck of a bottle in bluish-coloured glass, and a few frag- 
 ments, some of fine clear glass, and others of a blue colour. 
 
 Several whorls rudely formed of a coarse red paste, having 
 in some cases the holes which perforated them situated near 
 the edge and obliquely inclined ; similar to those already 
 alluded to in the Mounds B arid C, as having been probably 
 used by fishermen. 
 
96 THE BOSSENNO. 
 
 Numerous fragments of rudely-formed pottery of Celtic 
 type in a grayish-brown micaceous paste. 
 
 In the room No. 5 : 
 
 Three pestles or rubbing-stones in granite from 5 to 6 
 inches in length. 
 
 A signet ring in bronze. (E, pi. III. fig. 6, natural size.) 
 This ring contains a setting in blue glass, on which is engraved 
 
 the accompanying figure, 
 
 Fragments of pottery in a brownish paste, from which 
 a hemispherically-shaped cup was constructed ; diameter 6 
 inches, height 3^ inches, having a collar on the exterior of 
 the rim, and another rim lapping over from the middle part 
 of the vase. (E, pi. IV. fig. i.) 
 
 A bowl, also reconstructed, 7 inches in diameter and 3 
 inches in height, in a reddish-yellow paste, having a hemi- 
 spherical base and vertical sides. On the middle part are 
 three collars in relief, the intervals between which are orna- 
 mented with diagonal striae cut hollow. (E, pi. IV. fig. 2.) 
 
 Fragment of a vase in red paste, having a vitreous glazing, 
 bearing in relief the figure of the sea-horse. (E, pi. V. fig. i, 
 natural size.) 
 
 >Vt- 
 v , r 
 
 Bottom of a vase in red lustrous ware, on which is the 
 potter's mark, represented in the accompanying engraving. 
 
E, PLATE IV. 
 
To face fage 97 . 
 
 E, PLATE V. 
 
EXCAVATIONS IN THE MOUND E. 97 
 
 Fragment of Celtic pottery in brownish-gray paste, orna- 
 mented with points and horizontal lines incised. (E, pi. IV. 
 fig. 3, natural size.) 
 
 In the room No. 6 : 
 
 Several fragments of vases in red lustrous ware of the 
 same kind as those figured in the engraving, B, pi. XIV. 
 fig. 2. 
 
 Fragment of a vase in gray paste, pierced with small 
 circular holes which had been made after the firing. (E, pi. 
 V. fig. 2, natural size.) 
 
 Fragment of Celtic pottery in brownish-gray paste, orna- 
 mented with a large band of juxtaposed undulating lines, 
 incised. (E, pi. IV. fig. 4, natural size.) 
 
 In the room No. 7 : 
 
 One half of the bottom of a vase in red lustrous ware, 
 bearing part of the potter's stamp. 
 
 Several fragments in red lustrous ware, ornamented in 
 relief with garlands and running animals. 
 
 Several other fragments, also of Gallo- Roman type, in a 
 fine yellow paste, having a black vitreous glazing, ornamented 
 with figures, etc., in relief, two of which are represented (E, 
 pi. V. figs. 3 and 4, natural size). 
 
 Several fragments of a vase of similar type to the two 
 preceding specimens were found at the foot of the wall on 
 the outside of the house, near the entrance of the room No. 4 ; 
 
 H 
 
98 THE BOSSENNO. 
 
 represented for comparison with the others (E, pi. V. fig. 5, 
 natural size). 
 
 The fragments of pottery collected in the Mound E 
 weighed 196 Ibs. net. 
 
 Having now finished the enumeration of the objects found 
 under the Mound E, I .may be permitted to revert to the 
 statuette of an ox in bronze (E, pi. VI.) found in the room 
 No. 4, and as having some connection therewith to narrate 
 some customs still observed at Carnac, which would appear 
 to have been handed down from a very remote epoch. 
 
 When the statuette was found in the corner of the room, 
 the workmen all crowded round it in a state of excitement, 
 and exclaimed, " Here is a discovery ; we have found the ox 
 of St. Carnely." 
 
 This requires an explanation. The patron Saint of the 
 parish of Carnac is Saint Cornelius, the protector of bestial, 
 who is supposed to have been pope during the third century. 
 
 In the Breton dialect of Carnac his name is pronounced 
 Carnely. 
 
 The great fete of the year at Carnac is the anniversary of 
 Saint Carnely, which takes place at the autumnal equinox. 
 A pilgrimage, or as it is named in Brittany, the pardon of St. 
 Carnely, is then held, and goes on for a week. Numerous 
 services are held in the church, and processions march from 
 the church with banners flying to the fountain of St. Carnely 
 (E, pi. VII.), and thence round the village. A fair is held in 
 a field adjoining the village for the sale of bestial ; the streets 
 are lined with booths, shows, merry-go-rounds, and the usual 
 accompaniments of a country fair. Numerous pilgrims resort 
 to this from all parts of Brittany, clad in the costumes of their 
 respective districts ; they come to implore the protection of 
 Saint Carnely for their bestial. 
 
W > 
 

 
 
 
 f 
 
 
EXCAVATIONS IN THE MOUND E. 99 
 
 On arriving at Carnac they at once enter the church and 
 kneel down before a gilded bust containing the relics of Saint 
 Carnely ; thence they proceed to the outside, where a numerous 
 band of mendicants assail them with cries and lamentations 
 for alms. These are granted sometimes on the conditions 
 that the recipient makes the round of the church on his knees. 
 The pilgrims then proceed slowly, solemnly, and in silence, 
 round the outside of the church. They kneel down before a 
 statue of Saint Carnely, supported by two oxen, placed above 
 the west door, and repeat inwardly a prayer ; thence they 
 proceed in the same solemn manner to the fountain of Saint 
 Carnely, which is situated about 200 yards to the west of the 
 church. Here numerous boys belonging to the village, for 
 the remuneration of a few sous, present them with bowls of 
 water freshly drawn from the fountain. The pilgrims kneel 
 down before the fountain, taste the water, wet their face and 
 hands, and then elevating their hands towards heaven, allow 
 the water to trickle down their arms. 
 
 When this ceremony is finished they return in the same 
 solemn manner to the church ; there they deposit their offer- 
 
 j r 
 
 ings in the box placed alongside of the gilded bust of the 
 Saint ; should the donation, however, happen to be of a 
 considerable amount, it is placed in the hands of the church- 
 warden, and the amount is named from the pulpit on the 
 following Sunday. 
 
 Another curious ceremony, also having reference to the 
 health of their bestial, is sometimes performed by the peasants 
 in the neighbourhood of Carnac. It is called the nocturnal 
 procession, and is performed generally near midnight. 
 
 So soon as the cure of sick animals has been effected, the 
 proprietors arrange for a nocturnal procession. 
 
 On the following night all the bestial of the homestead 
 
100 THE BOSSENNO. 
 
 are assembled, and the procession marches solemnly, and in 
 strict silence to Carnac. A halt is made before the north 
 door of the church, where the conductors kneel down and 
 repeat inwardly a prayer ; thence they proceed round the 
 church, halting at the west door, and then go on to the 
 fountain of St. Carnely. The conductors again kneel down 
 for a few minutes, then draw water from the fountain, and 
 pour a little over each of the animals. This ceremony over, 
 the procession returns to the north door of the church, and 
 thence always in strict silence to their homes. 
 
 Another curious custom always observed in Carnac and 
 the neighbouring communes at the summer solstice, and in 
 which also the bestial are made to perform a part, is the 
 lighting up of bonfires in the evening, after sunset. This is 
 called in the Breton of Carnac tan heol (the fire of the sun), 
 and also tan Sant Jan (the fire of Saint John). 
 
 The young people of Carnac go from house to house 
 during the day, receiving contributions of faggots of wood and 
 bundles of furze, which they transport to the summit of the 
 tumulus or cairn which tops the Mont St. Michel. Towards 
 sunset numerous spectators proceed there, nearly all of whom 
 assist in preparing the bonfire. 
 
 Just as the sun sets long clouds of smoke are seen rolling 
 along the ground from the neighbourhood of adjoining farms. 
 If you ask the cause of all this, you are told that these are 
 fires prepared expressly for the cattle, with green broom and 
 fern, so as to produce a dense smoke. In fact it is at this 
 time that the cattle are brought home, and before being 
 housed for the night are made to pass across these fires, 
 the doing of which is supposed to preserve them from all 
 maladies for a year to come. 
 
 But let us return to the preparation of the bonfire. It has 
 
EXCAVATIONS IN THE MOUND E. 
 
 101 
 
 increased in size rapidly, many willing hands lending assistance, 
 and is at length finished ; before lighting, however, we must 
 wait until the fires of the surrounding country have shone out. 
 These soon begin to shoot out from every direction, at 
 Quiberon, Plouharnel, Erdeven, Ploemel, Crach, Locmariaker 
 Arzon, St. Gildas de Rhuis, in the islands of Houat, Hedic, 
 and Belle Isle, everywhere, as far as the eye can reach, is one 
 immense illumination. 
 
 The proper moment for lighting the fire of the Mont St. 
 Michel has arrived ; when it is desired to honour a stranger, 
 he is asked to set fire to the huge pile of furze, which, blazing 
 up in an instant, lightens up the whole hill and neighbourhood, 
 amidst shouts of joy from the spectators. 
 
 CROSS ON THE MONT ST MICHEL. 
 
VII. 
 
 EXCAVATIONS IN THE MOUND F. 
 
/ '"'... **'." :V /.','.'.',' >: --;s/C'''' : 
 
 / '-/' ^v^V'^"^-. . 
 
 / '. ^. . V ,,,,/' *'<.-. (,-- . . ^v*. 
 
 "^ ' - ' ' ' ?vs '"^-J> ^s* T?. - . ^j* 
 
 ;^- 
 
 -r-^:: - :^::;;.: %^-A.^ ; ^- 'i^"- Mr 
 
 .T^r: 
 
 :::'.<- ^v ^SS- , "^ ^ ' - - t 
 
 -....:"'- -^^-V: ^r^"^^^.^^%-" 
 
 ''">\V' ; ' v; -^^ ^r- ::./;'' '"'^-^^ 
 
 '-^sl^iSiSSglS i: - ; 
 
 Mrtr* I 
 
 Bntuh Imp! rt 5 
 
 To face page 105. F, PLATE I. PLAN OF THE BUILDINGS IN THE MOUND F. 
 
THE MOUND F. 
 
 THE exploration of the Mound F was begun on the loth of 
 August 1875. It consisted of several irregular heaps of earth 
 covered with furze and brambles, bounded on the north by the 
 road already mentioned in the diggings of the Mound E, and 
 on the south by the ancient road leading from Carnac to the 
 Trinite sur Mer. The widening and repairing of these two 
 roads, and the extension of the cultivated fields on the east 
 side of the Mound, have so cut in upon the ruins which my 
 diggings have exposed to view, that very little is left to give 
 an idea of their original construction. 
 
 The ancient road just mentioned was widened and repaired 
 when the Duke of Nemours passed this way in 1842. Whilst 
 the work was being done, there were discovered, at the place 
 where it passes close to the Mound F, remains of walls built 
 in small cubic stones bound with mortar, Roman roofing-tiles, 
 and numerous fragments of pottery. 
 
 It was not until 1849 an d 1850 that the new road between 
 Carnac and the Trinite sur Mer was made. When this was 
 being executed, there were discovered, near where it passes 
 the Bossenno and about 300 yards to the north, several Roman 
 coins, one of which was in gold, having on one side the figure 
 of a horse, on the other the effigy of Caligula. Close to these 
 were found remains of constructions in small cubic stones, 
 
106 THE BOSSENNO. 
 
 Roman bricks and tiles, and numerous fragments of different 
 kinds of pottery. These discoveries gave rise to a good deal 
 of discussion at the time they were made, but nothing was 
 done to follow up such sure indications, and they were soon 
 forgotten. It was only after the diggings of the Mound F 
 were commenced that I first heard of, and subsequently 
 verified, the above statements. 
 
 The ground-plan of the Mound F shows the state of the 
 buildings when the diggings were completed. The walls of 
 these buildings were all constructed with small cubic stones 
 bonded with lime, with the exception of the wall between No. 
 2 and No. 4, which appeared to be a continuation of the 
 southern enceinte, and was built of large irregularly-shaped 
 stones bound with a mortar formed of mud or argillaceous 
 earth. 
 
 The building No. 6 was first excavated. There could be 
 little doubt as to its having been used as a blacksmith's work- 
 shop, for a construction measuring 4 feet in length, 2 feet 10 
 inches in breadth, and i foot 10 inches in height, which 
 appeared to have been the forge, was built in large dressed 
 stones against the south wall. The upper surface was wasted 
 and reddened in the part which had been occupied as the 
 fireplace, and was covered with lumps of iron, scoriae, 
 ashes, and pieces of charcoal. A bar of iron one inch in 
 breadth and half an inch in depth was cemented with lead 
 into the west side of this construction just above the lower 
 course of masonry. It projected for twelve inches, and ter- 
 minated in a bend extending for two inches at right angles. 
 Close to the forge was the water-trough. It was constructed 
 against the west wall with slabs of granite bound at the corners 
 with a rose-coloured cement. 
 
 The floors of all the buildings found under the Mound F 
 
To face page 107. 
 
 F, PLATE II. 
 
EXCAVATIONS IN THE MOUND F. 107 
 
 were formed of a sort of concrete, and, with the exception of 
 No. 5, were covered with fragments of roofing-tiles, charcoal, 
 ashes, and the debris which had fallen on them at the period 
 when they had been burnt. As no other feature worth 
 mentioning was observed in the buildings Nos. i, 2, 3, 4, 5, I 
 shall now proceed with the enumeration of the different objects 
 found in excavating the Mound F. 
 
 In the building No. i : 
 
 Several fragments of common pottery in gray and in black 
 paste. 
 
 A fine polished stone axe in chloro-melanite * was dis- 
 covered in the angle formed by the two walls. This fine 
 specimen was unfortunately much injured by the blow of the 
 workman's pick : what remains of it measures 5 inches in 
 length by 2\ inches in breadth. 
 
 In the building No. 2 : 
 
 A solid bronze ring perforated with a circular hole. (F, 
 pi. II. fig. i, natural size.) 
 
 Fragment of a small object in bronze, having an orna- 
 mentation of circles and points engraved. (F, pi. II. fig. 2, 
 
 natural size.) 
 
 Fragment of a very thin bronze plate perforated with small 
 
 circular holes. (F, pi. II. fig. 3, natural size.) 
 
 Several other smaller fragments of the same description, 
 
 * Chloro-melanite approaches to jadeite in its hard, dense, and fusible 
 properties. It is a dark green mineral, which at first sight appears black, 
 but when the thin edge of a fracture is exposed to strong light, is trans- 
 parent and green. It may therefore be considered as a variety of jadeite, in 
 which a certain proportion of alumina is replaced by oxide of iron. 
 Damour. 
 
108 THE BOSSENNO. 
 
 and another fragment plated in silver. A stem of iron with a 
 round bronze head ; this seemed to have formed part of the 
 handle of a dagger. 
 
 A small bronze coin of Claudius the Second, in a bad state 
 of preservation. 
 
 Two small bronze coins, illegible. 
 
 Fragments of a vase in fine black paste, having a vitre- 
 ous lustre, large mouth, a collar on the exterior of the rim ; 
 badly fired, very fragile. The glazing dissolves on being 
 washed. Length 6 inches. (F, pi. II. fig. 4.) 
 
 Handle of a vase in red paste, the ornamentation formed 
 by twisting, similar to that figured (B, pi. VIII. fig. 3). 
 
 Rims of terrines in a fine red paste. 
 
 The neck of a large amphora in yellow paste. 
 
 Part of the handle of a similar amphora. 
 
 Bottom of a vase in red lustrous ware termed Samian, 
 ornamented with a herring-bone pattern in relief, obtained 
 from moulds. 
 
 Several fragments of the same kind of pottery ornamented 
 with garlands, running animals, and the leaves of aquatic plants 
 in relief, obtained by moulding. 
 
 Several fragments which served to reconstruct the greater 
 part of a bowl, diameter 9 inches, height 3 inches, in a dull 
 red paste, rudely ornamented with irregular zones formed of 
 coarsely punctulated lines. (F, pi. II. fig. 5.) 
 
 Two small cups measuring 4-^ inches in diameter by 2^ 
 inches in height were also reconstructed from fragments of 
 red lustrous ware. 
 
 Four dishes measuring from 6 inches to 8 inches in 
 diameter, in a grayish-brown paste, were also reconstructed 
 from fragments. These had been common culinary vessels, 
 similar to that figured (A, pi. III. fig. 5.) 
 
EXCAVATIONS IN THE MOUND F. 109 
 
 In the building No. 3 : 
 
 Some coarse fragments of common pottery of Celtic type 
 in a gray paste badly fired : 
 
 In the building No. 4 : 
 
 An iron bullock shoe very much oxidised, having three 
 square-headed nails in each side. It is here figured, full size. 
 
 The bones of oxen found in the other Mounds have been 
 recognised by Professor Paul Gervais as those of a small 
 race. This shoe has also been that of a small animal. The 
 bovine race to this day in Brittany are, as is well known, 
 celebrated for their small and fine proportions. We have 
 then in this shoe and in the bones alluded to, an indication 
 that there has been no great change in the size of the Breton 
 oxen since the Roman occupation. 
 
110 THE BOSSENNO. 
 
 Several iron nails of different sizes, very much oxidised. 
 
 Numerous fragments of common Celtic and Gallo-Roman 
 pottery, consisting of bottoms, sides, and rims of culinary 
 vessels. 
 
 In the building No. 5 : 
 
 An amber bead nearly opaque, diameter ij inch, per- 
 forated with a circular hole. 
 
 Several fragments of common pottery in a micaceous 
 grayish-brown paste. 
 
 In the building No. 6 : 
 
 Numerous objects in iron were found scattered about the 
 floor. These were of different forms, round, flat, and concave, 
 but all so much oxidised that, notwithstanding every pre- 
 caution, they could not be extracted from the earth without 
 crumbling into small pieces. 
 
 Numerous nails and pieces of rod iron of different sizes, 
 very much oxidised. 
 
 Numerous fragments of coarse pottery of Celtic type, in a 
 greyish-brown paste, badly fired. 
 
 Several fragments of vases in a brown paste, ornamented 
 with black vitreous bands. 
 
 Several fragments of vases, in a yellow paste, ornamented 
 with undulating lines similar to those figured (B, pi. VIII. 
 figs. 4 and 5). 
 
 Several fragments served to reconstruct a vase 7^ inches 
 in diameter, in fine black paste, with a vitreous glazing similar 
 to that figured (B, pi. VIII. fig. 2). 
 
 The bottom of a large bowl in red lustrous ware, 
 ornamented with parallel zones of lozenge-shaped figures, 
 stamped with a herring-bone pattern. The length of this 
 fragment is 7 inches. (F, pi. II. fig. 6.) 
 
EXCAVATIONS IN THE MOUND F. 
 
 Several fragments of red lustrous ware, ornamented with 
 garlands, figures of running animals, and leaves of aquatic 
 plants. 
 
 Other fragments of red lustrous ware served to re- 
 construct a vase 7^ inches in diameter, having a rim slanting 
 over from the middle part, the same as that figured (B, pi. 
 XIV. fig. 2). 
 
 The fragments of pottery collected in the Mound F 
 weighed 99 Ibs. nett. 
 
 
 CROSS AT COET-A-TOUS. 
 
VIII. 
 EXCAVATIONS IN THE MOUND G. 
 
THE MOUND G. 
 
 A VERY small mound, topped by an old hawthorn tree, was 
 situated in the southern enceinte, at a distance of 28 yards 
 from the ancient road which traverses the Bossenno. It is 
 marked on the general plan of the Bossenno by the letter G. 
 
 In cutting- a wide trench from north to south across this 
 mound, we came upon numerous small cubic building stones, 
 large dressed stones, portions of overturned walls, fragments 
 of roofing-tiles (tegulce and imbrices), and the debris of a 
 building which had evidently been completely destroyed. 
 Here also, as in all the mounds hitherto explored, were the 
 evidences of a violent conflagration, viz. ashes, charcoal, and 
 granite stones vitrified. 
 
 The following objects were collected during the cutting of 
 the trench, viz. 
 
 Several pieces of window-glass of a greenish colour, 
 polished on one face and rough on the other, similar to those 
 found in the Mound C (the baths). 
 
 Fragments of a glass bottle of square form, thick, and of a 
 bluish colour. 
 
 A few fragments of fine pottery in a fine black paste, with 
 vitreous glazing, tender and badly fired. 
 
 Numerous fragments of common gray pottery. 
 
 Handle and part of the body of a vase in yellow paste, 
 
116 THE BOSSENNO. 
 
 well fired, decorated with diagonal bands, and between these 
 numerous circles, irregularly placed, obtained by stamping. 
 A vase measuring 6 inches in diameter and 3 inches in height, 
 identically similar to this fragment, was found in the excava- 
 tions of the Gallo- Roman constructions of Jublains and is 
 now exhibited in the museum at Mayenne. It is figured in 
 the accompanying engraving. 
 
 As no remains of any construction worth following could 
 be found, we filled up the trench and abandoned the Mound G. 
 
 Two other trenches were next cut across the enceinte at 
 different points between the Mounds F and G, which yielded 
 similar results. 
 
 Up to this time we had explored seven mounds, in doing 
 which the ruins of the dwellings of people of a former age had 
 been exposed to view, and numerous objects collected, all 
 tending to throw some light on their mode of life and occu- 
 pation. It had frequently occurred to me, not only when 
 examining these ruins, but also the indications of others which 
 extended far on every side, that the population must have 
 
EXCAVATIONS IN THE MOUND G. 117 
 
 been of considerable extent, and that some evidence as to 
 their mode of sepulture ought to be found in the vicinity. 
 
 The researches made of late years by the Polymathic 
 Society of the Morbihan have proved that the dolmens of that 
 country had been places of sepulture, in some of which cre- 
 mation, in others inhumation, had been used. The age of 
 these monuments, however, is still a mystery, although recent 
 discoveries show that some of them had been used as places 
 of sepulture in the third century. 
 
 The discovery made of late years of cinerary vases, ashes 
 and charcoal, under a few of the menhirs, tends to show that 
 these had been funeral monuments, whilst it does not prevent 
 others from having a different signification. 
 
 Caesar, who was not only a great captain* but an able writer, 
 mentions that the funerals in Gaul were magnificent for that 
 country, that the body of the defunct was burnt with every- 
 thing that he had most cherished, even his animals, and 
 formerly his slaves and freedmen.f The Commentaries, 
 however, preceded the date of the ruins of the Bossenno by 
 four centuries, judging from the evidence of the coins found 
 there. Time enough had elapsed for many changes to have 
 taken place in the manners and customs of a people. 
 
 The following researches were made with a view of 
 obtaining some information on the mode of sepulture which 
 had been practised by the inhabitants of the Bossenno. 
 
 * Tacitus, who wrote of Caesar Siimmus auctorum divtts Jtilius, differs from 
 him, nevertheless, in many particulars, particularly so as to the funeral rites. 
 The discrepancy may perhaps be accounted for by changes having occurred 
 during the time which separates these authors. Tacitus wrote about the end 
 of the first or beginning of the second century, fully a hundred years after the 
 time when Caesar wrote the Commentaries. 
 
 t Caesar's Commentaries, 6 lib. 
 
118 THE BOSSENNO. 
 
 On the west side of the Mound G was the commencement 
 of a row of very small menhirs which extended to the west- 
 ward from where the traces of the enceinte stop. At first 
 sight they appeared to have been large stones which had been 
 cleared away from the neighbouring fields and placed on the 
 low earthen wall which separates the fields 937 Er Prat 
 (the meadow), and 938 Narlier. But how came it that 
 the greater portion were sunk in the earth so as to stand in 
 an upright position ? No peasant farmer would have done 
 this, when the laying them lengthways would have answered 
 his purpose so much better. 
 
 A trench was cut along the south side of these menhirs : 
 from this we worked round and underneath those still left 
 standing. We found that they had been carefully built round 
 with stones, so as to maintain them in an erect position. In 
 digging round these menhirs we collected several pieces of 
 charcoal, fragments of coarse pottery of a Celtic type in gray 
 and in black paste, several pieces of roofing-tiles (tegulce], and 
 from right underneath the menhirs fragments of vases in red 
 lustrous ware called Samian. 
 
 Seeing that nothing more could be done here, we redressed 
 the menhirs which we had overturned, filled up our trenches, 
 and proceeded to the north side of the Bossenno. 
 
 The remains of a ruined dolmen were situated on an 
 eminence in the field 639 Er Roch (the rock), about 150 yards 
 to the north of the Mound C (the baths). We commenced 
 excavations both on the inside and on the outside of the ruins, 
 and soon found a few pieces of charcoal and common gray 
 pottery. These, however, were discarded as of no value, 
 when I learned that the fire of Saint John was made here at 
 midsummer ; and as this dolmen had evidently been pre- 
 viously ransacked, we renounced farther work here and 
 

 Mtr~ O o.So 
 
 British Imp 1 Feet i a" 
 
 3 k 5 6 1 8 
 
 To face page 119. G, PLATE I. LARGE MENHIR NEAR THE BOSSENNO. 
 
EXCAVATIONS IN THE MOUND G. 119 
 
 proceeded westward to examine a large menhir in the 
 vicinity. 
 
 This menhir measured 17 feet in length, of which 13 
 were above ground. It stood in the middle of a field (see 
 the general plan of the Bossenno, No. 619 Er Menhir the 
 menhir), 200 yards west from the ruined dolmen, which it is 
 supposed by some antiquaries to indicate. 
 
 We commenced digging on two sides of the menhir in the 
 field No. 619, and soon found that it was built round with 
 dry stones. These, however, had failed to keep it upright, 
 for it was so much off the perpendicular that I was afraid to 
 continue the diggings all round it. Underneath the menhir 
 was a bed of rich black earth 1 2 inches in depth ; below this 
 was the primitive rock. Amongst the stones which served to 
 prop up the menhir we found the bottom of a vase, in the 
 fine red ware usually called Samian,""' the glazing of which 
 was destroyed ; a rough piece of unworked marble, red, veined 
 
 * A few years ago, M. the Abbe du Marc'hallach, Vicar-General to the 
 Bishop of Quimper, executed some researches at the foot of a large menhir, 
 forming part of an alignment of menhirs situated a few miles from that city, 
 in the parish of Ploumelin. Underneath this enormous stone he discovered 
 a Roman coin (middle brass) and several fragments of Roman roofing-tiles 
 (tcgulce). 
 
 In 1871, M. the Abbe Collet, vicaire of Ploemel, had some diggings 
 made in the tumulus of Mane Ploerig, situated to the north of Han Hon, in 
 the parish of Carnac. He discovered in these diggings two iron lance-heads, 
 an iron ring, two flint arrow-heads, a piece of clear glass, several fragments 
 of roofing-tiles (tegulce), a large quantity of fragments of pottery of different 
 kinds, including a large portion of a fine vase in red lustrous ware (called 
 Samian), ornamented with flowers and garlands in relief. He also found 
 fragments of Roman roofing-tiles in the tumulus of Mane' Botgade, situated 
 to the west of Ploemel ; and in the tumulus of Mane Bihuic, situated to the 
 east of Ploemel. I am indebted for these facts to M. the Abbe du Marc'- 
 hallach, and to M. the Abbe' Collet. 
 
120 THE BOSSENNO. 
 
 with white ; a worked flake of dark flint, and several frag- 
 ments of roofing-tiles (fegulas). 
 
 I confine myself for the present to giving the above state- 
 ments, which I have verified de visu, without attempting to 
 draw any deduction therefrom. It would require a number 
 of analogous facts, constituted by careful observers, to enable 
 one to hazard a conjecture on this association of objects and 
 monuments of such different origin. 
 
 In the neighbourhood of Carnac, where there are so many 
 dolmens, a sharp-pointed menhir is generally situated either 
 on the top of the tumulus surmounting the dolmen or in its 
 immediate vicinity. The menhirs at the end of the great 
 alignments are also of a pointed form, and are supposed to 
 mark the rising and setting of the sun, in some at the solstices, 
 and in others at the equinoxes, to an observer situated within 
 the alignments. After having spent some time in making draw- 
 ings of the different alignments, one becomes convinced that 
 the forms of the menhirs are comprised in a few varieties, 
 which are continually recurring, viz. pointed or needle-shaped 
 stones, square blocks, long slabs, with their upper edge form- 
 ing an obtuse angle, pear-shaped monoliths set on their 
 narrow ends. 
 
 In the previous portion of this chapter the Gallo- Roman 
 antiquities of Jublains have been alluded to ; and as these 
 highly-interesting remains are not so generally known as they 
 ought to be, perhaps the following notice may serve to draw 
 the attention of archaeologists towards them, or induce the 
 tourist to visit what will amply repay the slight deviation he 
 will have to make from the ordinary railway route between 
 Paris and Rennes. 
 
 The village of Jublains has a population of 1846 inhabit- 
 ants, and is situated between Mayenne and Evron, in the 
 
EXCAVATIONS IN THE MOUND G. 121 
 
 department of Mayenne. At the time when Csesar's legions 
 invaded the country Jublains was a place of some importance 
 Naodunum, the capital of the Aulerces Diablintes. Under 
 the Roman civilisation the Gaulish town became a Roman 
 one of some splendour, having its castra stativa, temples, 
 theatre, forum, aqueducts, baths, and having roads communi- 
 cating with Tours, Angers, the valley of Mayenne, Rennes, 
 and Corseul. This Gallo- Roman town was destroyed about 
 the beginning of the fifth century, and, like the Bossenno, 
 was forgotten until the excavations of the last few years 
 brought its remains to light. 
 
 CROSS AT HAN HON. 
 
IX. 
 
I 
 
 ca 
 
THE MOUND H. 
 
 ON the wall which has already been alluded to in the intro- 
 duction as the northern enceinte there was, at a distance of 
 230 feet west from the Mound D (the temple), another mound, 
 marked in the plan of the Bossenno by the letter H. 
 
 This Mound was of an elliptical form, measured 83 feet 
 in length, 20 feet in breadth, and 4 feet 6 inches in height. 
 The height seemed to be partly due to the stones and rubbish 
 gathered on the adjoining fields having been thrown on it. 
 
 On the i /th August 1876 we commenced the excavation 
 of this mound by opening up the east side, and soon exposed 
 the large dry stone wall of the enceinte. A few feet to the 
 eastward it stopped short, with its end squared off; then an 
 interruption for 1 1 feet, when it again recommenced. This 
 break in the wall was observed on the ground by the differ- 
 ence of level before the commencement of the diggings. It 
 appeared to me to have been one of the entrances or ports to 
 the houses situated within the enceinte, which could be easily 
 barricaded. (See the plan of the Mound H, pi. I.) 
 
 Leaving the entrance we again commenced on the east end 
 of the mound, working to the westward along the wall of the 
 enceinte. We soon came upon a wall 2 feet 4 inches in breadth, 
 built in small cubic stones bound with lime mortar, running 
 parallel to the enceinte on the south side, at a distance from it 
 
126 THE BOSSENNO. 
 
 of 2 feet. At the foot of the eastern end of this wall was a 
 remnant of pavement formed of square tiles. This wall 
 continued for 8 feet ; at that distance it was traversed at a 
 right angle by another wall of the same description, but of 
 which the south portion had been destroyed by the cultivation 
 of the adjoining field. 
 
 Working round the transverse wall, we came upon another 
 of the same description running parallel to, but situated 2 
 feet farther from the enceinte, so as to form a second narrow 
 passage. We followed this wall to the west, and came on its 
 south side upon another, also of the same description, and 
 forming another narrow passage 3 feet in width. A doorway 
 was found in this last wall, near its western extremity, where 
 the cultivation of the adjoining field had destroyed it. An 
 examination of the plan of the Mound H will convey a better 
 idea of the position of these walls than can be given by any 
 description. 
 
 In excavating this mound we found that it was composed 
 for the most part of large unshapely stones. A little below 
 the surface, on the top of the mound, we came upon the 
 upper half of a stone axe in polished diorite. 
 
 The following objects were found lower down, on the floor 
 of the passages. 
 
 A bronze coin, illegible. 
 
 Part of a bronze buckle. (H, pi. II. fig. 3, natural size.) 
 
 Portion of a bronze clasp inlaid with silver. (H, pi. II. 
 fig. i, natural size.) 
 
 Several iron nails, much oxidised. 
 
 A small ingot of lead. 
 
 Two fragments of cups in fine thin glass, the rim turned 
 slightly outwards. 
 
EA/ 
 
 To face fnge 126. 
 
 H, PLATE II. 
 
EXCAVATIONS IN THE MOUND H. 127 
 
 Fragment of a cup in fine thin glass, with straight rim, 
 and having ribbed cannelures in the middle. 
 
 Fragment of clear glass with incised ornaments. 
 
 A granite stone, measuring 4 inches in length, and weigh- 
 ing 2 Ibs. 2 oz. The form reminds one of the plummet used 
 by carpenters and masons. At first sight it seemed to be a 
 weight, but the weight of the Roman pound (6144 grains) 
 does not bear out that idea. It may possibly have served as 
 a sinker for large fishing nets (sagence}. (H, pi. II. fig. 2.) 
 
 A sharpening stone or hone, 4 inches in length, much 
 abraded by use. 
 
 An unfinished spindle whorl, 2 inches in diameter, in a 
 paste of coarse brick earth, nearly half pierced through on 
 each face. 
 
 Several fragments of red lustrous ware, without ornament, 
 from which a vase, 5 inches in diameter, was reconstructed. 
 It is figured in H, pi. II. fig. 5. 
 
 Several fragments of a vase, in a fine black paste, with a 
 vitreous glazing, and having straight rim slightly turned out- 
 wards ; diameter 6^ inches. (H, pi. II. fig. 4.) 
 
 Several fragments which gave the form of a dish, in a 
 grayish paste, the sides spreading slightly outwards, and the 
 rim bevelled off to the outside ; diameter 8 inches. (H, pi. 
 II. fig. 6.) 
 
 Several fragments of red lustrous ware, some of which 
 were plain, others ornamented with leaves of aquatic plants 
 and figures of the Equus marinus on the middle part, with 
 flutings superposed ; the same as the fragments found in the 
 mounds previously described. 
 
 Fragments of large vessels having a spout on the rim, in 
 a yellow paste. 
 
 Handles of amphorae, in a yellow paste. 
 
128 
 
 THE BOSSENNO. 
 
 Neck and handle of a bottle, in yellow paste. 
 
 Fragments of different vases, in fine black paste, with 
 vitreous glazing ; plain. 
 
 Other fragments in black paste, ornamented with zones 
 formed of points incised. 
 
 Several bottoms and other fragments of vases, in gray 
 paste and red paste ; plain. Other fragments in gray paste, 
 ornamented with vertical and diagonal striae. 
 
 The weight of the pottery collected in this mound 
 amounted to 28 Ibs. nett. 
 
 DOOR AT KERGOUELLARD. 
 
X. 
 SUMMARY AND CONCLUSIONS. 
 
 K 
 
SUMMARY AND CONCLUSIONS. 
 
 HAVING described in the previous chapters the results of the 
 excavations in the Mounds of the Bossenno, I now proceed 
 to draw conclusions from them in regard to the age and 
 character of the buildings, the condition and modes of life of 
 their occupants, and the probable date of the final destruction 
 of the settlement. 
 
 It will be sufficiently obvious, from the details already 
 given, that in these buildings we have the remains of a Gallo- 
 Roman establishment. The evidence of this is found in the 
 construction of the walls with small cubic stones regulated at 
 intervals with layers of long bricks, in the roofing-tiles, in the 
 Gallo-Roman pottery, and in the Roman coins. 
 
 The excavation of the Mound A disclosed a small 
 dwelling divided into four apartments. In these were found 
 objects of bronze, iron, and glass, pottery, and food refuse, 
 all of which indicated a high degree of civilisation and 
 prosperity. 
 
 The removal of the Mound B disclosed a house of eleven 
 apartments, carefully constructed ; its walls plastered and 
 decorated. In its rooms were pottery of remarkable fineness, 
 objects of bronze and iron, fine glass ware, Roman coins, and 
 food refuse, all of which indicated a greater amount of wealth 
 
132 SUMMARY AND CONCLUSIONS. 
 
 and luxury than was implied in the contents of the 
 Mound A. 
 
 Under the Mound C there was disclosed a building, 
 forming a dependance of that in the Mound B, carefully 
 constructed, the walls plastered, and richly decorated with 
 painting and curious shell frescoes. It also yielded articles of 
 bronze and iron, glass, pottery, Roman coins, and food refuse. 
 
 In the Mound D there appeared another dependance of 
 the Mound B, a small square construction surrounding an- 
 other of the same form, in the floor of which was a slab, 
 which seemed to be the base of an altar. Its contents differed 
 widely from those in the other buildings, not only in the 
 absence of food refuse, but in the presence of numerous 
 statuettes of the Venus Anadyomene and the Venus Gene- 
 trix, Roman coins, and peculiar pottery. 
 
 The Mound E contained a building, of which only seven 
 compartments were determined, less carefully built than those 
 in the Mound B. It was distinguished from the others by 
 the absence of decoration on the walls, and the presence of a 
 statuette and other objects in bronze, articles of iron, pottery, 
 glass, and food refuse. 
 
 Under the Mound F was a construction consisting of 
 several buildings, part of which appeared to be a continuation 
 of those under the Mound E, but including a separate one, in 
 which were a forge, a stone trough, scoriae, and remains of 
 iron implements and pottery. 
 
 The Mounds G and H yielded traces of buildings, with 
 objects in bronze and iron, and pottery. 
 
 The following quotations from authors who are regarded 
 as authorities on the Gallo-Roman epoch may assist us in 
 deducing conclusions as to the destination of some of these 
 buildings. 
 
. SUMMARY AND CONCLUSIONS. 133 
 
 De Caumont speaks of certain villce as being so large as 
 to be more like camps than country-houses, and states * " that 
 the finest of the Roman country houses had only one storey ; 
 in other respects they did not differ from the houses in the 
 cities. The greater part of the villa were not only residences 
 for pleasure, but also included what was requisite for the pur- 
 suit of agriculture, whether as houses for the cultivators of the 
 soil, for the bestial, or for storing the harvest." 
 
 "Columella points out three divisions in the country 
 house which had its farm attached, and the greater part of the 
 Gallo- Roman vilUs would come under that category. These 
 three divisions were the villa urbana, or dwelling of the pro- 
 prietor ; the villa agraria, or dwelling for the labourers and 
 the bestial required for the farm ; and the villa fructuaria, 
 where the grain and other products were stored." 
 
 Applying these descriptions to the results of my excava- 
 tions, it appears to me that the dwelling found under the 
 Mound B may be considered as filling the conditions of a 
 villa urbana. The careful way in which it had been built, its 
 decoration, the fineness of its pottery, and the luxury implied 
 by the whole facts, lead to this inference. Its communication 
 with the Mound C (the baths) and with the Mound D (the 
 Lararium) may also be considered as lending additional 
 support to this view. 
 
 Comparatively few constructions have been discovered 
 similar to that of the Mound D, and of some of these the 
 destination is doubtful ; hence cle Caumont classes them as 
 " Temples incertaines," and recommends the further study of 
 such edifices to all who explore Gallo- Roman remains. But 
 I should be altogether at a loss to attribute to this building 
 any other destination than that of a Lararium. 
 
 * Ere gallo-romaine, page 376. 
 
134 SUMMARY AND CONCLUSIONS. 
 
 The building under the Mound E may probably be 
 regarded as the villa agraria. Its construction, as has been 
 already stated, is inferior to that of the villa urbana, and there 
 is neither decoration nor other evidence of luxury apparent in 
 it. Some of the rooms, from their having plaster on the walls 
 and concrete on the floors, may be presumed to have been 
 inhabited by the labourers ; others, destitute of these, would 
 seem to have been used as stables and storehouses. 
 
 Remembering the persistence of names in Brittany, it 
 may be observed that the field which adjoins this building on 
 the west is named Pare er poul (the field of the waterpool). 
 No pool of water exists there now, but it may have retained 
 the name from its having been the compluvium for watering 
 the bestial. 
 
 In the Mound F, the only portion of the building to which 
 a destination could be assigned with certainty was the work- 
 shop of the smith. The fireplace, with cinders and scoriae, 
 the trough alongside of it, and the quantity of iron scattered 
 about, left no doubt as to its object and use. 
 
 The ruins found under the Mounds G and H were so 
 much dilapidated that no special destination could be certainly 
 assigned to them. Seeing that the latter are situated so close 
 to the entrance in the enceinte, it is possible that they may be 
 the remains of a military post. 
 
 De Caumont, in speaking of enceintes, says* "When the 
 incursions of barbarians became frequent, I will ask if it was 
 not necessary to have not only camps for the troops, but also 
 places of refuge where the inhabitants of the country most 
 exposed to being plundered could unite and place themselves 
 under cover in the time of danger, and if this system of 
 defence was not extended even to particular habitations. My 
 
 \ 
 Ere gallo-romaine, page 613. 
 
SUMMARY AND CONCLUSIONS. 135 
 
 opinion is that many of the enceintes which are left to us were 
 the enclosures of Gallo-Roman habitations, which from their 
 importance were more exposed to pillage, and which could 
 even serve as a place of refuge to the inhabitants of the sur- 
 rounding country. Thus there were, even from the fourth 
 century, small towns and habitations fortified as in the Middle 
 Ages." 
 
 It may be supposed that in the troublous times and dis- 
 ordered state of the country during the latter part of the 
 Roman occupation some system of mural defence would have 
 been then erected by the inhabitants of the Bossenno if they 
 had not done so previously ; but the most of these have been 
 destroyed by the tillage of the land. On referring to the 
 general plan the reader will see what is still left there which 
 may be regarded as an enceinte. 
 
 From the exposed situation of the Bossenno, the inhabit- 
 ants were exposed to attack, not only from enemies by land, 
 but by sea also ; for, as has been already shown, the sea came 
 close up to the Bossenno, and besides a landing in its vicinity 
 could have been easily effected anywhere in the Bay of 
 Quiberon. 
 
 Besides the ruins which my diggings have exposed, there 
 must certainly have been numerous other constructions, both 
 within and without the enceintes, possibly built with less care, 
 and which have long ago disappeared."" The clearing of the 
 land for tillage has left no trace of these save in the names of 
 the fields, and in the fragments of tiles and pottery strewn 
 over their surfaces. Thus, on referring to the general plan, 
 
 * The proprietors of the Bossenno adopted this view, and stated that they 
 had not cleared away the mounds which I have excavated because the 
 expense of doing so would have amounted to more than the value of the 
 land. 
 
136 SUMMARY AND CONCLUSIONS. 
 
 there are two fields, Nos. 658 and 659, situated to the east of 
 the Mound B, which bear the significant names of Helle-duen 
 (the cinders) ; farther on are No. 727, Ker-in-goh (the old 
 town), No. 665, Er goh fetan tal pare piard (the old fountain 
 near the field of stones) ; and many others might be cited, not 
 only inside, but outside the enceintes. 
 
 About the year 1 790 the mother of the present proprietor 
 of the Hotel des Voyageurs at Carnac dwelt in the village of 
 Bomer, close to the Bossenno. She often spoke to her 
 daughter of having seen hollow bricks and a sort of conduit 
 which had been then discovered in working the fields of the 
 Bossenno. This discovery created much astonishment in the 
 country at the time, but no one attempted to follow it up. 
 
 The existence of the remains of buildings to the south 
 of the Bossenno has been already alluded to in the descrip- 
 tion of the Mound F. This shows that the town had 
 extended at least so far in that direction. 
 
 In an east and west direction it had extended from the 
 field Ker-in-goh to the foot of the Mont St. Michel, the 
 importance of which as a commanding point would not have 
 been overlooked by the Romans. This eminence (which will 
 be fully described in the following chapter) overlooks all the 
 surrounding country and the sea. It may be thus inferred 
 that it would be utilised as a post for observation. In fact 
 the levelling to which the top of the tumulus has been sub- 
 jected has not altogether obliterated the traces of ancient 
 works ; and the occurrence of Roman tiles and pottery also 
 serves to bear out this inference. 
 
 Other Gallo- Roman ruins have been discovered about 
 200 yards to the south of this tumulus in the field named 
 Grah ba dou (heaps of debris). The following objects, which 
 have been found in this field of late years, are now deposited 
 
SUMMARY AND CONCLUSIONS. 137 
 
 in the Mairie of Carnac viz., fragment of a statuette of the 
 Venus Anadyomene ; fragments of vases in fine red lustrous 
 ware, and in grey and in black ware ; a bronze axe, a waist 
 buckle, and a coin of Alexander Severus. 
 
 All the fields on the south side of the Mont St. Michel 
 are strewn with fragments of Gallo- Roman pottery, and the 
 small cubic stones used in building by the Romans may be 
 seen in the walls which form their boundaries. 
 
 The quantities of bones of ruminants and other animals, 
 and numerous shells of edible molluscs everywhere found in 
 these excavations, save in the Mounds D and H, supply the 
 evidence of what had been the staple food of the inhabitants. 
 Hunting and fishing would thus seem to have been largely 
 practised, if we judge from the frequent recurrence of bones 
 of the deer, hare, and rabbit, and from the remains of their 
 fishing gear. Hunting scenes and animals running form a 
 frequent ornamentation on the Gallo- Roman pottery found 
 here, as they generally do on that found elsewhere. 
 
 The fact of so many bricks having been found bearing the 
 prints of the feet of dogs which had passed over them in the 
 soft state, shows that many of these animals had been kept, 
 and the difference in the size of the prints also shows that 
 they had been made by dogs of all sizes from the small pet 
 or lap-dog to the great watch-dog and hound. 
 
 The luxury of the table at Rome during the third century 
 has been frequently described. Oysters were brought there 
 from the Dardanelles, the Adriatic, and Armorica."" At 
 present ostriculture forms a prominent industry in the 
 Morbihan. Large sums have been expended on the oyster- 
 
 * Sunt et Aremorici qui lauclant ostrea ponti. AUSON. Ep. xiii. v. 35. 
 
138 SUMMARY AND CONCLUSIONS. 
 
 beds thus cultivated, which give employment to numerous 
 hands, and their produce is regularly sold in Paris as huitres 
 Armoricaines. 
 
 Horace makes mention of snails as a stimulant to be taken 
 with wine." 5 '" Pliny the younger alludes to them as having 
 been served at table, although not so much esteemed as the 
 oyster.f 
 
 On looking over the ceramics of the Bossenno we see 
 that while some of the pottery is of a rude Celtic type, the 
 greater portion is Gallo- Roman pottery, nearly all of which, I 
 should be inclined to hold, had been made in Brittany. The 
 Gallo- Roman potter had attained a high degree of excellence 
 by adopting the Roman mode of working, while still preserv- 
 ing the ancient forms and the ornamentation characteristic 
 of his country, viz. undulating lines and concentric circles. 
 The leaves of aquatic plants and the figure of the sea- 
 horse occur frequently among the fragments found. The 
 hippocampus is still frequently caught by the fishermen in the 
 Bay of Quiberon. 
 
 The Roman law obliged potters to put their trade mark 
 or stamp on their products, but this does not seem to have 
 been enforced in Gaul. Although all the bricks, tiles, and 
 fragments of these found at the Bossenno were carefully 
 examined for such stamps or names, not one was found. The 
 only potters' marks discovered were on the two fragments of 
 vases in the Mound E. 
 
 * Tostis marcentem squillis recreabis et Afra 
 
 Potorem cochleae. HORACE, Lib. ii. Satire iv. 
 
 t Heu tu, promittis ad coenam, nee venis ! paratae erant cochleae ternae. 
 . . . At tu apud nescio quern ostrea. . . . maluisti. (Ep. i. xv.) 
 
SUMMARY AND CONCLUSIONS. 139 
 
 Great care was bestowed by the Romans on the manufac- 
 ture of bricks and tiles. Vitruvius states that they sometimes 
 remained two years drying in the shade, and that the firing 
 occupied from fifteen to twenty days, with an extreme heat at 
 the end. The bricks and tiles found at the Bossenno were in 
 as good a state of preservation as if they had been freshly 
 made. Roman tiles are found in the neighbourhood of 
 Guerande, and are still occasionally used by the paludiers to 
 close the water channels in the saltfields, in preference to 
 modern tiles, which only last a few years. 
 
 The red lustrous ware called Samian is frequently alluded 
 to by Roman authors, and has been found wherever the 
 Roman dominions extended. It is supposed to have been 
 made from B.C. 150 down to the fifth century of our era, and 
 it has been proved, by the discovery of the potteries, to have 
 been manufactured in several parts of France, amongst others 
 on the right bank of the Allier. The red ferruginous earth 
 necessary for its fabrication has been found in several parts of 
 Brittany where there are potteries still. Brongniart mentions 
 the magnificent glazing of this ware as the only one known to 
 the ancients, and one which cannot now be produced. He 
 describes it as being principally composed of a silicate, 
 rendered fusible by the introduction of an alkali, such as potash 
 or soda, and attributes the colour to a metallic oxide which 
 had been either introduced into the composition or derived 
 from the primitive paste."" 
 
 The same authority states t "that three modes of orna- 
 mentation were followed by the Roman potters : \st. Mould- 
 ing from single or from independent patterns representing 
 the same or different subjects ; 2d, The employment of 
 
 * Traite des Arts Ceramigues, tome t, p. 545. f Ibid. p. 424. 
 
140 SUMMARY AND CONCLUSIONS. 
 
 roulettes for forming circular zones on the circumference of 
 the vase, isolated ornaments being made by impressions from 
 a seal or stamp ; 3^, The forming of an ornament in relief on 
 the vase, with slip flowing from a pipette," a method pecu- 
 liarly characteristic of Roman potteries. 
 
 Numerous examples of the first and second modes of 
 ornamentation were found in all the mounds. 
 
 It is somewhat significant that specimens of the third 
 mode of ornamentation were only found in the Mounds 
 A, B, C, and D. The chromo-lithograph A, Plate IV., 
 represents a vase which had been ornamented by this curious 
 process. 
 
 In examining the objects in bronze, we find the two 
 civilisations, Gaulish and Roman, represented, as we have 
 just seen in the pottery; for whilst the two brooches (B, pi. III. 
 fig. 3 ; C, pi. XIII. fig. 3), and also most of the other bronze 
 ornaments, are of Celtic type, the coins are all Roman. 
 
 The curious statuette of an ox (E, pi. VI.), resembling the 
 bronze statuettes found in Egypt which have no disc between 
 the horns, recalls the fact, as the pottery did, that the Romans 
 had Moorish soldiers stationed at Vannes. Are we to attri- 
 bute the statuette of the Mound E to the Egyptian mytho- 
 logy, to the Celtic mythology, or to the Mithraic worship ? 
 The worship of Mithras was introduced about the time of the 
 emperors at Rome, and thence spread over all parts of the 
 empire. The figure of an ox or bull frequently occurs on 
 Gaulish coins. In the Celtic mythology the two oxen of 
 Hu-Gadarn, a divinity of the ancient Bretons, are fabled to 
 have drawn out of the waters of the deluge, by means of 
 strong chains, the Avank, a monstrous crocodile which had 
 caused the submersion of the universe. This is alluded to in 
 
SUMMARY AND CONCLUSIONS. 141 
 
 one of the popular legendary ballads of Brittany, known by 
 the grotesque title of "The Frogs' Vespers,""" characterised 
 by Villemarque as one of the most singular, and perhaps the 
 most ancient of Breton Ballads. It consists of a dialogue 
 between a Druid and a child, in which a summary of the 
 Druidical mythology is given and recapitulated in a series of 
 twelve questions and answers. The second of these is as 
 
 follows : 
 
 AR BUGEL : 
 
 Kan d'in euz a zaou ranu 
 Ken a oufenn breman. 
 
 ANN DROUIZ : 
 Daou ejenn dioc'h eur gibi ; 
 O sachat, o souheti. 
 Edrec'hit aun estoni ! 
 
 The Gaul of the Bossenno, in adopting the religion of his 
 conquerors, had not entirely abandoned his former deities. 
 The Gallo- Roman altar found under the cathedral of Notre 
 Dame at Paris has on two of its sides the Celtic divinity 
 Hesus and the Taurus trigaranus ; on the other two sides 
 the Roman divinities Jupiter and Vulcan. 
 
 It has been said of coins, that one of the most important 
 uses to which they can be applied, when the circumstance of 
 their discovery on a particular site is beyond question, is in 
 approximating dates. The coins found at the Bossenno, 
 extending over the long space of one hundred and eighty-four 
 y ears v iz. from the third consulate of Marcus Aurelius to 
 the year three hundred and fifty-three, the epoch of the death 
 of Magnentius thus bring us down to the middle of the 
 fourth century. 
 
 * Barzas Breiz, par le Vicomte H. de la Villemarque. Paris, 1867. 
 
142 SUMMARY AND CONCLUSIONS. 
 
 Historians are in accord as to the deplorable state of Gaul 
 in the latter part of the fourth century, when its towns and 
 villages were burned and pillaged, and its rural districts 
 ravaged by invasions of barbarians both by sea and land. 
 The population of the towns occupied themselves in con- 
 structing enceintes and walls of defence, not formerly needed, 
 but now become of the first importance as an absolute neces- 
 sity of existence. 
 
 While the country was weighed down by these misfor- 
 tunes, the government, taking no steps to remedy them, only 
 thought of collecting more taxes. Then, in the beginning of 
 the fifth century, came the general rising, which took place in 
 Armorica, against the Roman government, and it appears to 
 me that it is to this epoch that the violent destruction of the 
 Bossenno may be assigned. 
 
 In the beginning of the fifth century Christianity had 
 made considerable progress in Armorica, but no trace or 
 indication of it was found at the Bossenno. On the other 
 hand, we have seen unequivocal evidences of paganism in the 
 statuettes of Venus found in the Mounds A and D, and in the 
 statuette of the ox found in the Mound E. 
 
 It may, perhaps, be surmised that, for some misdeeds, the 
 pagan inhabitants of the Bossenno had been driven away, 
 and that their habitations had been burned by their Christian 
 neighbours, and, in so far, the legend of the Bossenno would 
 be an echo of a forgotten event. That portion of the legend, 
 however, which states that they were slain, is not borne out 
 by the results of the excavations, for, among all the numerous 
 bones collected, not a single fragment of a human bone occurred. 
 This fact seems to indicate that the inhabitants had not been 
 surprised and slain by those who destroyed their dwellings, 
 but that they had found time to gain a refuge elsewhere. 
 
SUMMARY AND CONCLUSIONS. 
 
 143 
 
 In concluding these remarks on the excavations of the 
 Bossenno, I may, perhaps, be permitted again to draw atten- 
 tion to the persistence with which legendary incidents and 
 forms of ornamentation are transmitted in Brittany from 
 generation to generation. A striking example of the latter is 
 given in the accompanying engraving. 
 
 CHALICE ENGRAVED ON A TOMBSTONE. 
 
XL 
 THE MONT ST. MICHEL, 
 
THE MONT ST. MICHEL 
 
 IN the preceding chapter I have already alluded to the 
 vestiges of Gallo- Roman constructions found on the flank of 
 the Mont St. Michel. This eminence, from its commanding 
 position, must have been an important dependence of the 
 Bossenno, if only as a post for observation. 
 
 During the summer months of 1876 we could not work at 
 the Bossenno until the crops were cleared off the fields, so the 
 months of June and July were passed in making researches at 
 the foot of the tumulus surmounting the Mont St. Michel, 
 which yielded results different from those found at the 
 Bossenno. 
 
 The Mont St. Michel is situated at the distance of half a 
 mile to the east of the village of Carnac ; from its commanding 
 position and peculiar form it stands out as a striking feature 
 in the landscape, as seen from any point in the surrounding 
 country. 
 
 It is the first resort of the tourists who arrive at Carnac, 
 and they cannot make a better commencement in acquiring a 
 knowledge of the country than by giving a short time to the 
 examination of this hill, and of the fine panoramic view which 
 is best seen from the summit of the tumulus. 
 
 Facing the south the view extends over the sea studded 
 with islands and broken by promontories. On your left hand 
 
148 THE MONT ST. MICHEL. 
 
 you see Locmariaker, by some writers supposed to have been 
 Dariorigum ; farther on, in the same direction, the tumuli of 
 Arzon and Tumiac, excavated by the Polymathic Society, may 
 be made out with a telescope ; and still farther on the point of 
 St. Gildas de Rhuis brings up the memory of Abeilard. 
 Facing you are the islands of Houat and Hedic ; to the right- 
 hand Belle Isle, the largest of the islands of the gulf of the 
 Morbihan ; the promontory of Quiberon, the Fort Penthievre, 
 and isthmus leading to it, associated with the disastrous 
 landing of the emigres in 1795. Turning to the right you 
 see the isle of Groix, the spires of Plouharnel, Erdeven, and 
 Etel. 
 
 Facing the north, the spires of Ploemel, Auray, Saint 
 Anne, Arradon, and Qrach, are to be seen in the distance : 
 nearer at hand, and flanked by numerous dolmens, are the 
 three strange alignments of menhirs, or groups of gigantic 
 standing stones, viz. M6nec (the place of stones, or in another 
 rendering the place of remembrance) ; Kermario (the place 
 of the dead); and Ker-lescan (the place of burning.) Under 
 your feet almost, and to the right, are the Gallo- Roman ruins 
 of the Bossenno, and behind these the finely-wooded hill and 
 celebrated dolmen of Ker-Cado. 
 
 The base of the tumulus which surmounts the Mont St. 
 Michel forms an oval, measuring 400 feet in its major axis, 
 and 200 feet in its minor axis. The present height of the 
 tumulus above the crest of the hill is 33 feet, but it may be 
 supposed that it had been much higher in times previous to 
 the different levellings to which it has been subjected : very 
 possibly some of these may have occurred during the Roman 
 occupation, and certainly in later times, for the preparation of 
 the site of the chapel dedicated to St. Michael, which now 
 occupies part of the summit. Near the chapel stands a 
 
; 
 
 io 
 
 I 
 
 "i 
 
 IVv 
 
THE MONT ST. MICHEL. 149 
 
 curious cross, rudely sculptured in granite."' At the west end 
 of the summit are the ruins of a semaphore which was worked 
 in the beginning of this century. 
 
 This tumulus is similar to the mounds of stones called in 
 the Celtic Cam, in the lowland dialects cairn. Hence, it 
 seems to me, is the origin of the name of the neighbouring 
 village Carnac (the place of the earn), and this idea is 
 strengthened by the name of the village at the foot of the hill, 
 Cru-Carnact (the rocky hill of the earn). 
 
 M. Rene Galles, in his report on the diggings of Mont 
 St. Michel in September 1862, estimates that a mass of earth 
 and stones equal to 40,000 cubic yards had been transported 
 to the summit of this hill, the highest in the district, for the 
 construction of the tumulus. I shall here quote a passage 
 from the same report, which alludes to the construction of the 
 tumulus, a view and section of which is shown in pi. I. 
 
 " With the view of facilitating our preparatory researches 
 as much as possible, we marked on the platform of the tumulus 
 the line of its major axis. A series of pits were next sunk 
 from this line, which passed through a layer of stones for 
 2 feet 8 inches, and then came upon a bed of fine clay dried, 
 and probably beaten hard. A similar envelope of fine clay 
 has been found in the tombs which follow the line of the 
 coast. This for a long time was regarded as composed of 
 ashes. It had evidently been formed with the view of having 
 a compact and solid covering over the mass of dry stones, so 
 as to prevent the infiltration of rain-water. We determined 
 by trials that along the summit the depth of this bed of fine 
 
 * See the figure at the end of this Chapter. 
 
 f The inhabitants call this village indifferently either Cru-Carnac or 
 Clou-Carnac. In the ancient titles it is written Cru-Carnac. 
 
150 THE MONT ST. MICHEL. 
 
 clay was 5 feet 4 inches, and that below it a cairn of dry stones 
 extended all along the tumulus. 
 
 " Having determined the situation of a point a little to the 
 east of the centre of the tumulus, I commenced to sink a pit, 
 having an orifice, measuring 3 feet 4 inches by 6 feet 8 inches, 
 and which, according to our calculations, ought to pass very 
 close to the chamber in the interior of the monument, the 
 proximity of which would likely be indicated by some change 
 in the aggregation of the heaped-up stones. Near this place 
 we expected from our levels to find the summit of the hill which 
 supports the tumulus, is prolonged under it, and where we 
 expected to find that the central chamber had been constructed. 
 This difficult work was performed by skilful workmen, who 
 had been placed at our disposal by the kindness of the director 
 of the tin-mines of La Ville-d'Er, for it was found necessary 
 to take energetic measures to protect ourselves against the 
 falling in of the heaped-up stones which rolled about us like 
 lead shot in a sack."'' 
 
 M. Rene Galles then goes on to state that at a depth of 
 26 feet 3 inches they came upon the chamber, and found 
 therein eleven stone axes in jade, two large stone axes in a 
 coarser material, twenty-six small stone axes in fibrolite, nine 
 pendants, and one hundred and one beads ; some of which 
 were in jasper, others in turquoise ; two flakes of flint, and a 
 number of little beads formed of a sort of ivory. These 
 objects are now exhibited in the Museum of the Polymathic 
 Society of the Morbihan at Vannes. 
 
 One evening in the month of April 1875, when passing 
 
 * Fouilles du Mont St. Michel en Carnac, faitcs en Scptembre, 1862, par 
 Rene Galles, etc. Vannes, 3 e edition. 
 

 I &i*tul* M)"*. 'J3ti ' ', . '.vWJ^- 5 ^ 
 
 
THE MONT ST. MICHEL. 151 
 
 along the south side of the tumulus, I remarked for the first 
 time under the effect of a peculiar light some inequalities in 
 the surface of the ground which surprised me, as I had so 
 often passed there before without observing them. In poking 
 amongst the grass with my walking-stick I struck upon some 
 stones which appeared to be arranged in a line, and turned up 
 some fragments of roofing-tiles (tegulce) and Gallo- Roman 
 pottery. A short inspection of the dry stone wall forming the 
 demarcation of the adjoining fields showed that it contained 
 numerous dressed stones. These indications made me resolve 
 to prosecute some researches here so soon as time permitted 
 and the necessary permission could be obtained. 
 
 We commenced these researches in the month of June 
 1876 ; the results were the discovery of four curious construc- 
 tions running in an east and west direction, which are severally 
 noted in the accompanying plan of the Mont. St. Michel by 
 the letters L, M, N, and O. 
 
 We commenced by opening up the ground near the stones 
 in line above mentioned, and soon found remains of dwellings. 
 
 The construction L, on being laid bare, was found to be 
 of a rectangular form, measuring 54 feet in length by 1 7 feet 
 in breadth. What remained of its walls had an average height 
 of 2 feet, and i foot 8 inches in breadth. 
 
 The walls of this building, and also those of the construc- 
 tions M, N, and O, were all rudely built of undressed stones, 
 bound with a mortar of argillaceous earth, having the exterior 
 but not the interior corners rounded off, and the courses of 
 masonry, instead of being level, followed the slope of the hill. 
 The extraction of building stone from a quarry cutting into 
 the building L had destroyed the greater portion of it. 
 
 The following objects were found during the excavation 
 of the construction L : 
 
152 THE MONT ST. MICHEL. 
 
 Several fragments of roofing-tiles (teguUe) near the surface 
 of the soil. 
 
 Several pieces of grey schist, measuring i foot by 10 inches 
 and half-an-inch thick, probably parts of roofing-slates. 
 
 Near the north-east angle a sufficient number of fragments 
 were found to determine the greater part of the form of a vase 
 rudely moulded by hand, and still bearing the marks of the 
 potter's fingers all over it. It had a flat bottom and upright 
 sides, the rim turned out at a right angle, the paste reddish- 
 brown, coarse, with scales of talc visible, and badly fired. 
 The diameter across the exterior of the rim is 2 inches, and 
 across the interior 1 1 inches. A thick coating of soot on 
 the surface indicates that it had been used for cooking. In 
 one of the fragments of the bottom a band of iron was em- 
 bedded, which had doubtless served to support it over the 
 fireplace. 
 
 Several fragments of pottery in coarse yellow paste. 
 
 A few fragments of vases in a yellow paste, covered with 
 a green vitreous glazing. 
 
 Four flakes of flint. 
 
 A small round bead in brick earth. 
 
 A four-sided whetstone, having a groove cut round the 
 head, probably for suspension. (Mont St. Michel, pi. III. 
 fig. i, natural size.) 
 
 A whorl formed from a fragment of a vase in red paste, 
 the perforation oblique and eccentric. 
 
 In clearing away the ground between the constructions M 
 and N, the traces of fire, viz. ashes and charcoal, were found, 
 and these were more strongly marked at the north-east corner 
 of M, where we cut into the tumulus. It appears to me that 
 these traces of fire had no connection with the buildings M 
 
To face page 152. 
 
 MONT ST. MICHEL, PLATE III. 
 
THE MONT ST. MICHEL. 153 
 
 and N, but may be attributed to the time of the construction 
 of the Tumulus. 
 
 The following objects were found here : 
 A flint knife. (PI. III. fig. 2, natural size.) 
 A piece of iron measuring 6 inches in length, 4 inches in 
 breadth and J-inch thick, much oxidised ; use unknown. 
 
 Fragment of a curiously shaped vase, in a pale yellow 
 paste, well fired, sonorous, covered with a light green vitreous 
 glazing, and ornamented with a line of lozenge-shaped 
 markings. (PI. IV. fig. 3, natural size.) Pottery of this 
 description is generally believed to be not earlier than the 
 twelfth century. 
 
 In the centre of the large apartment No. i the fireplace 
 was indicated by some stones embedded in the floor : these, 
 reddened by the action of fire, were covered with a small 
 quantity of charcoal and ashes, whilst around them were 
 several flint flakes and shells of the limpet (Patella vulgata], 
 
 At the foot of the north wall we found : 
 
 A small fragment of a polished stone axe in diorite. 
 
 A piece of blue slate measuring 4 inches in length, i^ inch 
 in breadth, and \ inch in thickness ; rounded at one end and 
 having polished edges. 
 
 A piece of vitrified scoriae. 
 
 A small stone hammer of hemispheric form in grayish 
 flint, 2 inches in diameter. 
 
 A spindle whorl in a brownish paste mixed with small 
 grains of quartz, well fired, hollow on one face. (PI. III. fig. 
 4, natural size.) 
 
 Fragment of the tibia of a ruminant. 
 
 Several fragments of pottery in a coarse brown paste, 
 
154 THE MONT ST. MICHEL. 
 
 covered with soot on the exterior, similar to the vase described 
 in the building L. 
 
 Several fragments of very coarse pottery, some being in a 
 yellow and others in a brown paste, badly fired. 
 
 Near the north-east corner, on the floor : 
 A nucleus or core of silex, from which flakes had been 
 struck off. 
 
 Near the foot of the south wall : 
 
 A small fragment of fine thin glass. 
 
 A round piece of thin blue slate, having the edge cut 
 smooth, 3 inches in diameter. 
 
 Several pieces of gray slate, perforated with holes J 
 inch in diameter, and very irregularly formed ; probably part 
 of the roof of the building. 
 
 In the room No. 2, on the floor at the north-west corner : 
 Four flint flakes. 
 
 A small round piece of slate, smoothed, f inch in 
 diameter. 
 
 Several pieces of slate partly calcined. 
 
 On the floor of the doorway opening to the north : 
 
 A flake of flint and a fragment of a glass cup. 
 
 The building N, also of rectangular form, measures 50 feet 
 in length by 22 feet in breadth. It is composed of three 
 apartments of unequal dimensions, having no communication 
 with each other, each apartment having a doorway giving 
 access from the south. The walls and masonry are of the 
 same description as those of the buildings L and M, and the 
 floor is formed by a bed of argillaceous earth. 
 
 The room No. i measured 20 feet in length by 15 feet in 
 breadth. 
 
THE MONT ST. MICHEL. 155 
 
 No. 2 measured 15 feet 8 inches by 10 feet. 
 
 No. 3 15 8 10 3 inches. 
 
 The following objects were found here : 
 
 In the room No. i, on the floor, a flake of grayish flint. 
 
 Near the south-west corner : 
 
 A whorl, two inches in diameter, formed out of a fragment 
 of a vase, in a yellowish red paste. 
 
 A small ball, in brick earth, f inch in diameter. 
 
 A piece of red coarse pottery cut into a disc, i inch in 
 diameter. 
 
 Near the south-east corner : 
 
 Fragments of vases of different forms, in yellow and grey 
 pastes. 
 
 At the foot of the east wall : 
 
 Several fragments of a vase in a yellow paste, hard and 
 sonorous, covered with a light green vitreous glazing, and 
 ornamented with circular fluted bands. 
 
 Near the north wall : 
 
 Part of the tibia of a ruminant. 
 
 In the room No. 2, on the floor : 
 
 Three cores or nuclei in silex. 
 
 Fragments of a pestle in granite, 2| inches in diameter. 
 
 Two round stones in polished nephrite. 
 
 On the outside of the doorway of this room : 
 
 Several fragments of vases similar to that found in the 
 
 o 
 
 building L. 
 
156 THE MONT ST. MICHEL. 
 
 Fragment of the rim of a plate of octagonal shape, having 
 a rude ornament at each angle formed from the print of the 
 potter's thumb. 
 
 In the room No. 3, on the floor : 
 
 One of a set of discs cut in granite, 2^ inches in diameter 
 and f inch thick. 
 
 An unfinished whorl formed from a piece of coarse red 
 pottery, half perforated. 
 
 On the outside of the north wall, near the foot : 
 
 A few fragments of ridge tiles of a pale yellow paste, 
 
 having an ornament on the upper part of diamond-shaped 
 
 figures in relief. 
 
 The building O forms a parallelogram terminating in a 
 hemicycle at the east end, measuring 51 feet in length by 22 
 feet in breadth. The walls are of unequal thickness, and in 
 the north-east portion are merely blocked into the tumulus, and 
 consequently there they have no exterior face. This also 
 proves that the construction of this building was posterior to 
 that of the tumulus. The average height and style of masonry 
 are the same as in the preceding buildings. 
 
 The building O is divided into two rooms, the smaller of 
 which, No. i, measures 16 feet in length by 12 feet in breadth, 
 and communicated with the room No. 2 by a doorway 5 feet 
 6 inches in width. 
 
 The room No. 2, measuring 36 feet 10 inches in length by 
 1 6 feet in breadth, has a doorway 5 feet 6 inches wide, giving 
 access from the outside on the south side of the building. 
 Opposite this doorway, and embedded in the floor at the foot 
 of the north wall, is a dressed slab of granite measuring 3 feet 
 in length by i foot 6 inches in breadth. 
 
THE MONT ST. MICHEL. 157 
 
 When commencing the excavation of this building a 
 musket-ball was found near the surface of the ground, possibly 
 a souvenir of 1815. At that time both the village of Carnac 
 and the Mont St. Michel were held by the " Blues," when a 
 strong party of " Chouans " arrived, and after a sharp skirmish 
 gained possession of both these places. 
 
 The following objects were found during the excavation of 
 this building. 
 
 In the room No. i, on the floor : 
 
 Two pieces of flint. 
 
 Two discs made of brick tiles, 2 inches in diameter. 
 
 Half of a spindle whorl in brick paste, i inch in diameter. 
 
 Part of a horse-shoe, iron. 
 
 Two iron nails. 
 
 Several fragments of ridge tiles, the same as those found 
 in the building N, and others having a line of oval-shaped 
 ornaments on the top in relief. 
 
 Two teeth of the pig. 
 
 Knee-cap bone, and part of the thigh-bone of a ruminant. 
 
 In the room No. 2, on the floor : 
 
 Several slates ^ inch in thickness, perhaps part of the 
 roof of the building. 
 
 Two discs formed of brick tile ij inch in diameter. 
 
 Small round piece of slate f inch in diameter, having the 
 figure of a cross incised. 
 
 Fragment of a glass cup thin and clear. 
 
 A considerable quantity of fragments of coarse brownish- 
 red earthenware. 
 
 Fragment of a plate of an octagonal shape, in a gray- 
 brown paste, having an ornament at the angles formed by the 
 print of the potter's thumb. 
 
158 THE MONT ST. MICHEL. 
 
 Fragment of a vase in a yellow paste, having a band of 
 oval-shaped ornaments in relief, formed by the print of the 
 potter's fingers. (PI. IV. fig. 4, natural size.) 
 
 Several fragments of a vase in a yellow-red paste, having 
 thin sides and the rim turned inwards. One of the fragments 
 is riveted with brass wire. 
 
 The neck, handle, and other fragments of a vase in a 
 yellow paste, of the form of a water bottle, pierced with a 
 number of small holes. (PI. IV. figs, i and 2, natural size.) 
 Similar pottery was found in the Mound D. 
 
 Two fragments of a plate in yellow earth covered with a 
 green vitreous glazing, and having a collar round the rim. 
 
 Bottom of a small vase 2! inches in diameter, in a red 
 
 > 
 paste, and covered in the interior with a vitreous glazing of a 
 
 light green colour. 
 
 A large quantity of fragments of coarse pottery in yellow 
 and in gray paste. 
 
 Near the north-west corner were the traces of a fireplace, 
 ashes, charcoal, and shells of the limpet (Patella vulgata). 
 
 At the foot of the north wall : 
 
 Several fragments of slate having an ornament formed by 
 
 incised lines rudely cut in squares and lozenges. 
 
 mr 
 In the trench cut outside the semicircular termination, at 
 
 the depth of 3 feet 4 inches : 
 
 A polished stone axe in white flint, which seemed to have 
 undergone the action of fire. (PI. III. fig. 3, natural size.) 
 
 In taking a general survey of these ruins, and the objects 
 found in them, there arises a difficulty in the want of sufficient 
 data, which prevents any exact conclusion as to their date. 
 The evidence of the ruins is only of a negative character ; no 
 tradition exists about them, no document mentions them, and 
 
To face page 158. 
 
 MONT ST. MICHEL, PLATE IV. 
 
THE MONT ST. MICHEL. 159 
 
 so far as I can ascertain they were entirely unknown in the 
 country before the excavations just described brought them 
 to light. 
 
 At the Bossenno we had the remains of houses carefully 
 and well built in regular courses of small cubic stones, the 
 walls richly ornamented, and floors covered with cement. 
 Here the constructions are in rough irregular masonry, the 
 walls are destitute of plaster, and the floors of cement. They 
 are evidently not Gallo- Roman. But the fragments of Roman 
 tiles and pottery found on the tumulus, and the small cubic 
 building stones in the wall of the adjacent field, all indicate 
 that there must have been a Roman or Gallo-Roman con- 
 struction hereabouts, probably at a period anterior to that of 
 the ruins. 
 
 The rude masonry, the rounding off at the corners, the 
 hearth in the middle of the floor, all of these forcibly remind 
 one of the ancient form of rudely built dwellings still con- 
 structed in the Hebrides. 
 
 In cutting below the floors of the constructions M, N, and 
 O, we came upon the bed of fine clay which has been already 
 mentioned as covering the tumulus, then the wall of the build- 
 ing was in part blocked into the tumulus. These facts are 
 sufficient to show that the buildings were erected subsequently 
 to the construction of the tumulus. 
 
 The objects found in and about these buildings belong to 
 epochs so very different, that inferences drawn from them are 
 necessarily conflicting. It appears to me that there is no 
 connection between these buildings and the stone implements 
 found outside their walls, which may be classed with the 
 date of the sepulchral rites before the elevation of the 
 tumulus. 
 
 Some of the pottery collected in these buildings may be 
 
160 
 
 THE MONT ST. MICHEL. 
 
 classed as of Celtic type, but alongside of these fragments were 
 some belonging to the Middle Ages, and others which cannot 
 
 be dated before the seventeenth century. 
 
 
 
 The ruins may thus perhaps be the remains of one of the 
 primitive monastic establishments which were founded by the 
 
 i 
 
 emigrants from Great Britain, on their arrival in Armorica. 
 Under that hypothesisHhe building O would have been the 
 church of the convent ; its form and size are similar to the 
 primitive Christian churches in Brittany. On the other hand, 
 these buildings, after having been abandoned by the brother- 
 hood, may have been occupied up to a comparatively modern 
 time by men of different generations who have left the traces 
 of their occupation behind them. 
 
 De Caumont states in his Architecture Religieuse that the 
 primitive Christian churches were copied from the Roman 
 basilicas, which served the double purpose of tribunals and 
 commercial exchanges. 
 
 One of the oldest constructions which remains in Brittany 
 is the chapel of St. Agatha, in the department of He et 
 
 NORTH 
 
 GROUND-PLAN OF CHAPEL OF ST. AGATHA. 
 
 Vilaine. It is situated close to the parish church of Langon, 
 about a mile from the railway station of Fougeray Langon, 
 on the line between Rennes and Redon. The general form 
 
THE MONT ST. MICHEL. 161 
 
 of this chapel is that of the ancient basilicas. The nave 
 measures 26 feet 3 inches in length by 12 feet in breadth. 
 The apse or hemicycle at the east end measures 6 feet 7 
 inches by 1 2 feet. The walls are of Roman work built with 
 small cubic stones, averaging 3^ inches of a side, and having 
 the courses separated by a layer of lime mortar f of an inch 
 in thickness. These courses are regulated at intervals by 
 double rows of long bricks. The entrance, which is from the 
 south, is covered by a Roman arch, which seems, however, to 
 be of more recent construction than the body of the church. 
 The windows are very small, and, from their destroying the 
 uniformity of the masonry, appear to have been inserted after 
 the erection of the building. These are four in number, one 
 in the apse, two in the north wall, and one in the front. A 
 few years ago, on removing from the vault of the apse a thick 
 coat of plaster covered with the rude painting of the Middle 
 Ages, there was discovered the most curious fresco-painting 
 which the Romans have left in Brittany. It represents 
 Venus rising from a blue sea, surrounded with fishes, among 
 which is a dolphin playing. 
 
 This chapel is mentioned in a charter of the twelfth 
 century as Ecclesia Sancti Veneris. The patron of Pluvigner 
 near Auray is Saint Vener, or Vigner. The chapel of St. 
 Agatha must have been erected originally as a Gallo- Roman 
 temple of Venus, and afterwards dedicated to a saint bearing 
 a similar name. The Breton name for Venus is Vener. 
 Gregoire de Rostrennen gives it as Wener, Vener, Guener* 
 
 This superposition of constructions of so different a nature 
 on the Mont St. Michel is the very history of Brittany, where 
 at every point you meet with traces of three great periods 
 
 * Dictionnaire Fran^ois-Celtiqne sen Francois- Breton, par le P. F. Gregoire 
 de Rostrennen. Rennes, 1732. 
 
 M 
 
162 
 
 THE MONT ST. MICHEL. 
 
 confused together in the minds of the people viz. the period 
 of the stone monuments, usually called Celtic ; the period 
 of the Roman domination ; and the Christian period. The 
 earn or tumulus was erected on the hill ; then came Roman 
 constructions ; and, lastly, the Christian Church and cross 
 which now crown the summit of the tumulus. 
 
 
 CROSS ON THE MONT ST. MICHEL. 
 

 
GETTY CENTER LIBRARY