- ^Xi^ 
 
 «v;»^ 
 
 -fc>N 
 
 ^ V 
 
 
 21 ^^»U/^. 
 
Class. 
 Book. 
 
 F 
 
 im 
 
 
 / 
 
 ^ 
 
DWELLINGS OF THE SAGA-TIME 
 
 ICELAND, GREENLAND, AND VINELAND 
 
 CORNELIA HORSFORD 
 w 
 
 [Rkpuintbd from The National Geographic Magazine, Vol. IX, 
 No. 3, March, 1898] 
 
 WASHINGTON, D. C. 
 
 JUDI) A DETWEILER, PRINTERS 
 
 1898 
 
,«'t<ii: 
 
 ^^^01 
 
 ^ 
 
DWELLINGS OF THE SAGA-TIME IN ICELAND, 
 GREENLAND, AND VINELAND 
 
 BY 
 
 Cornelia Horsford 
 
 The Siiua-time began with the colonization of Iceland in 875 
 and lasted for about 150 years. During this time the oft- repeated 
 accounts of the discovery, colonization, and early history of Ice- 
 land, as well as that of all Scandinavia, acquired the form of 
 Sagas or narrations. Ari Thorgilsson, the historian, who was 
 born in Iceland in 1067 and died in 1148, was the first to write 
 down these events in chronological order. In each of the four 
 books attributed to this writer Greenland and Vineland are 
 briefly mentioned.* Other Sagas relate the adventures, trage- 
 dies, and family histories of the colonists, and among these are 
 the Sagas which tell about Greenland and Vineland.t 
 
 We know that Scandinavia has been a rich field for collecting 
 relics of the stone, bronze, and earl}^ iron ages, but no ruin of a 
 dwelling dating from the Saga-time has yet been identified in Den- 
 mark, Sweden, or Norwa}-. This may be due to the lack of dura- 
 bility in the way of building the houses and to the custom of 
 using over and over again in new buildings all the suitable 
 material from the old walls. 
 
 In 1888 a young Icelander named Valtyr Gudmundsson, who 
 was studying for the degree of Doctor of Philosophy at the Uni- 
 
 * Islendingabok, Landnamnbok, Kristni-Saga, and Konungab6k. 
 
 t Hauksbok, Eiriks Saga Raathi, and Flateyjarbok. Greenland and Vineland are also 
 briefly mentioned in the Fornmanna Sogur, Eyrbyggja Saga, and in tliree vellum manu- 
 scripts in the Arna-MagniBan Library .at Copenhagen. An account of these will be 
 found in the first chapter of " The Finding of Wineland the Good," by Arthur Middle- 
 ton Reeves, London, 1890, Henry Frovvde. 
 
74 DWELLINGS OF THE SAGA-TIME IN 
 
 versity of Copenhagen, chose for the subject of his thesis " Pri- 
 vate Dwellings in Iceland in tiie 8aga-time."* In preparing for 
 this he read every saga of his native literature, comparing each 
 description, sentence, and word relating to his subject, until in 
 imagination he had reconstructed every form of dwelling and 
 outhouse of the Saga-days. These buildings differed considera- 
 bly from the design given b}'^ Finsen in his edition of Gunnlaug's 
 Saga, printed in 1775, which was the accepted model until the 
 publication of Dr Gudmundsson's work. 
 
 In 1894 Lieutenant Daniel Brunn, of the Danish nav}^, was 
 sent b}'- the Danish government to make extended researches 
 among the Norse ruins in Greenland. These researches went 
 far toward confirming the results of Dr Gudmundsson's studies. 
 
 It was therefore with much gratification that Dr Gudniunds- 
 son (who was by that time professor of Old Norse literature and 
 history at the University of Copenhagen) accepted my commis- 
 sion to direct arclieological researches for me among the ruined 
 dwellings and other works of man in Iceland during the summer 
 season of 1895.t He took with him from Copenhagen another 
 Icelander named Thorsteinn Erlingsson, and to him the greater 
 part of the work is to be accredited, for Dr Gudraundsson was 
 in attendance at the Icelandic Parliament and could not be 
 present in the field himself. 
 
 ICELAND 
 
 The Icelandic Antiquarian Society has done some good work 
 in the field. They have identified and roughly measured the 
 ruins of many historical farms and of several hundred booths at 
 some of the old open-air law courts called " things." One or two 
 pagan temples have been dug out and carefully described, and 
 many burial mounds, which also belonged to the pagan days. 
 The ancient dwellings were situated on sloping ground, near 
 rivers or fjords. 
 
 From the early days this has been believed to be the ruin of 
 the house built by Erik the Red in the Hawk River valley soon 
 after his marriage with Thorhild, and here his eldest son Leif 
 was probably born. Erik lived in four difii'erent places in Ice- 
 
 * " Privatboligen paa Island i Saga-Tiden " af Valtyr Gudimindsson. Copenhagen, 1889. 
 Andr. Fred. Host & Sons, Forlag. 
 
 fThe report of this expedition will soon be published by the Viking Club of London 
 under the title of " Ruins of the Saga-Time." 
 
 JThe researches of this society are published yearly at Reykjavik, Iceland, in the 
 "Arb6k hins Islenzka FornleifafSlags." 
 
ICELAND, GREENLAND, AND VINE LAND 
 
 SUl'l'uSEll lilKTliri.ACE UK I.KIF Kltll^.SU^ 
 
 land before he finally settled in Greenland. The supposed ruins 
 of his liouses on Oxney and Sudrey can still be seen also * but 
 I do not know that any ruins have been identified at Drangar^ 
 The ruins of these dwellings, when undisturbed, are low, grass- 
 grown ridges and hollows often difficult to detect, except when 
 stones protrude through the turf. A dwelling usually consisted 
 
 
 at" -:r^- i 
 
 VERTICAL SECTION OF AN APARTMENT 
 
 of three apartments : a hall or principal room, in which there 
 was always a fireplace; a sitting-room for the women, and a 
 store-room or pantry .t These apartments were like small houses, 
 
 * " Finding of Winehind the Good," by A. M. Reeves, p. 105. 
 
 t"Fortidsniinder og Nutidslijem paa Island " of D.uiiel Krnnn. Copenliagcn, Ernst 
 Bojesen, p. 161. 
 
76 
 
 DWELLINGS OF THE SAGA-TIME IN 
 
 each with a separate roof. l)ut attached to eacli otlier, with pas- 
 sages tlirougii the tliick walls. Near b}' were usually oue or more 
 small outhouses. Tli^se dwellings were built on the surface of 
 the ground, which was probably levelled when necessary. The 
 floor was of firmly beaten earth. 
 
 The walls were one and a half meters thick and from one to 
 one and a half meters high. The inner side was l)uilt of unliewn 
 stones and the interstices were filled with earth. The outer side 
 was of alternate layers of turf and stones, and the space between 
 the two sides was filled in with earth kneaded hard. When these 
 walls fall, the stones necessarily slip down on either side, and the 
 bottom row with the space between remains almost intact, unless 
 
 a Pi/rli, S/fel a htyi, og I7fel a hreidd GolA^ i afhysimi ef steinlagl 
 alhusinu . a ^nnsi aska. ab- fannsl aska og hrossionnan 
 
 nUIN OF PAQAN TEMPLE AT THYRLI 
 
 Ai'b6k Hins Islenzka Fornleifafelags, 1880-1881 
 
 unnaturally disturbed. Often, however, the waits were built en- 
 tirely of layers of turf or with only disconnected rows of stones 
 at the base. 
 
 The drawing of the pagan temple at Thyrli shows the manner 
 of laying the inner and outer sides of a wall with the earth be- 
 tween the two. A large stone, of course, extends farther back 
 into this earth between than a small one does. 
 
 The inside measurement of a hall varied from 3 to 7 meters 
 in width and from 10 to 17 meters in length. The plan is of the 
 ruin of Erik the Red's house, shown above from a photograph. 
 A long narrow fire-place usually extended through the middle of 
 the room. This was either paved or surrounded with stones 
 standing on edge, and was about 3 meters long and from 60 to 80 
 centimeters broad. Besides the long fire which served to warm 
 and light the hall, there was a small cooking fire made in the 
 same way, about 1 meter square and raised a few centimeters 
 
ICELAND, GREENLAND, AND VINELAND 
 
 77 
 
 
 ^■iT'm 
 
 ^ '-^ 
 
 JL-1 
 
 >4 
 
 -1 
 
 TW^WW 
 
 PLAN OF THE 
 
 IF KlUR TKK RED IN IIAUKADAI.E 
 
 above the level of the floor. Other noii-esseutial forms of fire- 
 place I need not describe here. A separate apartment was often 
 formed by erecting a thin partition across a room, as is shown 
 in this plan b,y the dotted line. Pavements, but more often 
 thresholds made of one or more long stone slabs, were some- 
 times in the doorways and also in the passages through the thick 
 walls between the apartments. The outhouse shown at the 
 
 
 < - -A 
 
 i-,,.-,^/-.-'r.Ar'''-.|-r--'-,-'''"''.-^ ""''■;, 
 
 i<».;t.l..i.«7i.(^.. 
 
 
 
 RUIN OF SAMSSTADIR IN lUUORSARDALB 
 
78 DWELLINGS OF THE SAGA-TIME IN 
 
 left was about 18 meters from the door of the house, on the 
 steep mountain side. It was 4 meters square, built of turf only, 
 and partially underground. There was a large square platform 
 of stones in one corner which had served for a fire-place. 
 
 Narrow platforms of earth faced along the outer edge with 
 ui)right stones, on which the inhabitants both sat and slept, ex- 
 tended along one or both sides of the hall. In the large halls 
 these platforms were about 23 centimeters high and 11 meters 
 broad. Sometimes there was also a broader platform at one end 
 of the hall. Samsstadir is one of the farmsteads in the Thor's 
 River valley which was buried during an eruption of Mount 
 Hecla in the fourteenth century. This valley is called the 
 Pom[)eii of Iceland. The farm was probably abandoned about 
 1300. It shows the first change in the evolution toward thicker 
 walls. 
 
 With the exception of some spinning-stones, which were found 
 in the sitting-room of a house not shown here, no relics were 
 found during these researches. It is also an interesting fact that 
 no runic inscription belonging to the Saga-time or for two cen- 
 turies later has yet been found in Iceland. 
 
 The evolution which has taken place in house-building since 
 the Saga-time has been in the stead}' increase in the thickness 
 of the walls until their breadth is nearly doubled, a slight in- 
 crease in height, not admitting a second story under the roof, 
 and the addition of many apartments, so that from a distance 
 the many roofs of a farmstead look almost like a little village, 
 
 GREENLAND 
 
 Greenland was discovered and colonized by Erik Thorvaldsson 
 toward the end of the tenth century, and flora that time two 
 Norse colonies, called respectively the eastern and the western 
 settlements, prospered for about three hundred* years. The 
 ruins of these two settlements have been studied with more or 
 less care by the Danish government. In the eastern settlement 
 a hundred and fifty farms, with all their outbuildings, have 
 been surveyed and measured. A few dwelling-houses have been 
 thoroughly dug out and examined.* 
 
 *Beskrivelse af Ruiner i Julianehaabs Distrikt i Aaret 1880, af G. F. Holm. Meddel- 
 elser om Gronland, udgivne af Commissioneu for Ledelsen af de geologiske og geo- 
 graphiske Undersogelser i Gronland. Copenhagen, 1883, vol. vi. 
 
 Undersogelse af Griinlands Vestkyst fra 64° til 67° N. B. af J. A. I>. Jensen, 1884 o3 
 1885. Meddelelser om Gronland. Copenhagen, 1889, vol. viii. 
 
 Arkseologiske Undersogelser i Julianehaabs Distrikt af Daniel Brunn, 1895. Meddel- 
 elser om Gronland. Copenhagen, 189G, vol. xvi. 
 
ICELAND, GREENLAND, AND VINELAND 
 
 79 
 
 As in Iceland, these farmsteads were situated on the shores of 
 rivers and fjords. Although in the main they resemble those of 
 Iceland, one is impressed at once with certain strikinoj differ- 
 ences. Even the undisturbed ruins suggest narrower, straighter, 
 and stronger walls. 
 
 WALLS OF A NORSE RUIN IN GREENLAND 
 
 Meddelelser otn Gronland, vol. xvi. Daniel Brunn 
 
 ■^ lldgravti ,.*. 
 
 t5^o3»J 
 
 PLAN OF A NORSE EUIN IN OREF,NLAND 
 
 Meddelelser om Gronland, vol. xvi. Daniel Brunn 
 
 The dwellings were usually long and narrow, consisting of 
 from three to eight rooms, and were surrounded by numerous 
 outhouses and staljles for cattle, sheep, and goats. Close to the 
 houses are found enormous midden heaps, often larger than the 
 
80 
 
 DWELLINGS OF THE SAGA-TIME IN 
 
 ruins of the liouses themselves. The walls were narrower than 
 the Icelandic walls, and, although they were built of layers of 
 turf and stone or sometimes of turf on a foundation of stone, the 
 middle space, filled in with earth, had almost disappeared, as 
 may be seen in the sketch. The long j^latforms of stone along 
 the walls, the pavements, thresholds, and scattered fireplaces 
 recall similar constructions in Iceland. 
 
 In 1261 Greenland became sul)ject to the Crown of Norway, 
 and to this influence the Danes attribute certain differences be- 
 
 V f f =f j-.f^'^ 
 
 * '«/ * I" ■ "^^^ 
 ■ tMoAAifiq ^r^ 
 
 t: X v ^ t::.^^^ 
 
 * ■'•' overhrott rS^:^' 
 '■r •■ ^^* 
 
 * tlUaro/vet l£*^*' , 
 
 4 - ■'55^ 
 
 ■^ .v . ¥ ^rrcr^ - 
 
 
 
 
 ¥ ■ * V -y 
 
 SUPPOSED SITE OF THE HOUSE OF ERIK THE KED IN GREENLAND 
 
 Meddelelser om Gronland, vol. xvi. Daniel Brunn 
 
 tween the customs of the Norsemen in Iceland and in Greenland, 
 which I need not describe here.* Perhaps the difference in archi- 
 tecture is due to the same cause. The ruin of the house found 
 on the supposed site of Brattahlid, the abode of Erik the Red, 
 looks as if it might have been remodeled several times since that 
 fearless Norseman first settled in the land. 
 
 * Meddelelser om Gronland, vol. xvi, p. 490, 
 
ICELAND, GREENLAND, AND VIN ELAND 
 
 81 
 
 Numerous relics have been found in these ruins — iron nails 
 and knives, pieces of stone vessels, si)inning stones, bone combs, 
 and stone pendants bored with holes and incised with rune-like 
 but illegible characters. These, like all the ruins in Greenland 
 which have been thoroughly dug out, are attributed by the Danes 
 to a period later than the Saga time. 
 
 VINELAND 
 
 The ruins, found where one had every reason to hope to find 
 traces of the houses built in Vineland by Leif Erikson and his 
 followers, did not differ in their essential features from those of 
 Iceland in the Saga-time. The situations were similar. The 
 walls were laid in the same way and were of the same thickness, 
 and the fireplaces were constructed as they were in the habit of 
 constructing them at home. 
 
 
 !^ 
 
 1 I I I L. 
 
 10 Meters 
 Sri'I'OSED NORSE KUIN IN MASSACHLSEITS 
 
 The walls of this house can be little more than suggested. 
 They were i)robabl \' built almost entirely of turf, and they looked 
 as if they might have been intentionally destroyed. I show it 
 for its fireplace. Three or four fireplaces were on the site, one 
 of them being tlie familiar Indian clam-bake, with its neatly 
 paved, saucer-shaped hearth piled with ashes and unopened clam 
 shells, for this temptingly prepared feast had never been eaten. 
 One of these fireplaces, however, was very different from the 
 others, and of the Icelandic type, with its surrounding upright 
 stones at the four corners and a mass of charcoal and stones in- 
 side. This house is one of those on the place pointed out in 
 Cambridge by my father, Eben Norton Horsford, as the site of 
 
82 
 
 DWELLINGS OF THE SAGA-TIME IN 
 
 the group of houses built by the party of Thorfinn Karlsefni in 
 * Vine! and. 
 
 The second house I show for the construction of tlie walls and 
 the little pavement, presumably at the door, which resembles that 
 in the temple at Thyrli shown before. The outer side of the wall 
 contained only one layer of stones, the inner, according to cus- 
 tom, containing more and larger stones, some of which had fallen 
 in. The oblong platform of small stones occupied the place of 
 
 2 JV|eters 
 
 surruSKD nurse uuin in Massachusetts 
 
 and resembled a fireplace, but showed no trace of such use, un- 
 less in the dark sticky earth between and under the stones, which 
 I have since been told may have been ashes absorbed in the soil. 
 This house, with the other ruins near it, are about ten or more 
 miles from the settlement at Cambridge, and so far from the river 
 that it must be attributed to later visitors from the North than 
 those told about in the Vineland Sagas. 
 
 No relics have been found at either of these sites which I attri- 
 bute to the Northmen. I have, however, one stone implement. 
 
ICELAND, GREENLAND, AND VINELAND 83 
 
 which was found imbedded in the yellow sand and seemed to 
 have been lost before the advent of the Northmen, and presuma* 
 bly belonged to the savages they found here. 
 
 Probably the reader will contrast these different dwellings of 
 the Northmen with those of tlie native tribes of Ngrth America, 
 from the magnificent ruins of Copan to the long, narrow houses 
 of the Iroquois, and will detect the similarities and difi'erences 
 between these and the habitations of the Greenland Eskimos. 
 
 The Spanish, Dutch, French, and English explorers visited 
 and might have built houses on these shores, but in Europe no 
 houses of this type are found outside of Iceland, except in the 
 Faroes, and, although ruins of Norse dwellings are probably 
 awaiting detection in England, Scotland, Orkney, and Shetland, 
 they have not yet been brought to the notice of archeologists* 
 
 The earliest examples of architecture on our shores, as w'ell as 
 the present knowledge of the evolution of European architecture, 
 as far as I have been able to find out, show that the walls of the 
 inferior houses in post-Columbian times were unlike those of 
 Iceland. Our oldest French house is the Sillery manor house 
 near Quebec, built by the Jesuits in 1637. The walls of this 
 house are built of stone, and are three feet thick, laid in mortar 
 which is now nearly as hard as the stone itself. I have been 
 unable to find anything more primitive of French workmanship 
 here. I have found nothing in English work which is not famil- 
 iar to you all, although I have followed up several mistaken re- 
 ports. ' The Dutch buildings show an equally advanced though 
 difierent type of development, and also the Spanish. 
 
 I am glad to have an opportunity to express publicly my sin- 
 cere thanks and deep indebtedness to the American archeologists, 
 both here and in Canada, who have come most kindly to my 
 assistance and taught me in the field the knowledge they had 
 acquired by their own experience, without which I could not 
 have learned how to gather many facts, a few of which I have 
 here presented. 
 
 Mr Gerard Fowke : Seven weeks of field work in and near Cambridge. 
 Two weeks of field work in Ohio, Pennsylvania, West Virginia, and Mary- 
 land, 1894. Five weeks in Cambridge, 1896. 
 
 Dr Fr.^nz Boas: Two days in and near Cambridge, 1894. 
 
 Mr D.wm Boyle, Curator of the Canadian Institute at Toronto: One week 
 in and near Cambridge. One week in Ontario, Canada, 1894. One week 
 in Cambridge, 1896. 
 
 ♦ Since writing this I have been notitio.l tliat ancient Norse niins liave been foiuul in 
 the Hebrides. 
 
84 COMPLETION OF THE LA BOCA DOCK 
 
 Mr F. W. Nouris, Hon. Editor of the Viking Club, London: One week in 
 Cornwall, 1895. Three weeks in Scotland, Orkney, and Shetland, 1896. 
 Two weeks in England, 1897. 
 
 Dr PHIL. ValtvrGudmundsson, Professor of Old Norse History and Liter- 
 ature at the University of Copenhagen : Direction of explorations in Iceland 
 for four months, 1895. Five weeks in and near Cambridge, 1896. 
 
 Mr Thorsteinx Eulingsson, Iceland: Four months in Iceland, 1895. 
 
 Rev. Henry Otis Thayer, Maine Historical Society : Two weeks among 
 old English ruins in Maine, 1896. 
 
 Sir James Lumoine, Past President of the Royal Society of Canada: Di- 
 rection of researches near Quebec, 1896. 
 
 MrC. C. Willoughby, Peabody Museum, Cambridge : Two days on Cape 
 Cod, 1897. 
 
 Mr W J McGee: Advice, criticism, and encouragement, both in Wash- 
 ington and Cambridge for over four years. 
 
^ 
 
X ^H 
 
 LIBRARY OF CONGRESS 
 
 011 270 537 5 
 
 y 
 
 b£^-: