Glass sherds from the Uppåkra settlement

In document Fynden i centrum Keramik, glas och metall från Uppåkra Hårdh, Birgitta (Page 75-94)

Hitherto 44 glass sherds have been found at the settlement (Stjernquist 1994a, 1996; Larsson &

Hårdh 1998). Six of these are stray finds (Fig. 3) without any information about the precise find spot. Of the remainder 34 were collected during the systematic walking with metal detectors (Fig.

2). Their find places are illustrated by the distri-bution map. Most of these finds come from area 7, that is, the area to the south of the church.

This material is supplemented by a rim frag-ment found earlier <luring the digging for the mortuary (LUHM), and by 3 sherds found at two small excavations (Table I).

This rim fragment found during the digging for the mortuary belongs to the category of glass with cut and polished decoration (Fig. 7).

Of the remainder, 4 are fragments of the Snartemo type (Fig. 4), 2 have reticella decora-tion (Fig. 6), 2 have a decoradecora-tion of applied gold

foil (Figs. 12, 18), one isa part of a snicked trail (Fig. 5), and 3 sherds belong to palm cups or funne! beakers (Figs. 9, 10). One of these is roade of dark green glass and two are rim frag-ments of light green glass.

Of the sherds impossible to identify accu-rately, two are yellowish-green pieces of thin glass. One ofthem isa rim fragment. The colour is the same as that of the Snartemo type of glass but the material is much thinner. Another sherd is green (Fig. 17).

Some sherds are characterized by the colour and by the decoration with applied trails. One sherd has applied trail in turquoise blue (Fig. 3), 2 are blue with self-coloured trailing. There is also a loose thread of this colour. Several sherds are intensively bluish-green (Fig. 13). A couple of these have very thin trailing. Some sherds are pale blue (Fig. 14). Some odd sherds are also found.

The material will be dealt with according to types. The colour is defined only visually. At the end, the very small sherds named as frag-ments or splinters will be mentioned.

Glass with cut and polished decoration This type of decoration characterizes a sherd found in connection with the digging for the mortuary (Fig. 7). It has been discussed and depicted in publications of the earliest finds from Uppåkra (Stjernquist 1994a, 1996). It isa rim fragment of light yellowish-green glass and some small fragments of the same vessel. By the rim two cut but not polished furrows run horizontally around the neck. The glass mate-rial is iridescent. The decoration consists of cut

Table I. Types of sherds

cut and polished 1

snartemo 4

yellowish-green, thin 3

reticella 2

gold-foiled 2

snicked trail 1

palm cup or funne! beaker 3 fragments, blue or bluish-green 8

sherds with trailing 15

odd sherds 5

Fig. 2. Detector finds. Uppåkra. Approx. 1: 1.

and polished facets (oval). One edge of the sherd seems to have been polished after a break in old times. According to Eggers the sherd belongs to the types E 227-237, an identification which was followed by Lund Hansen. It means a dat-ing to phases C2-C3 (Hansen 1987). It is per-haps possible to determine it more accurately through an analysis of the details of the rim.

It is not included in Straume's detailed study of the cut and polished glasses (1987). It can, however, be characterized by the shaping of the rim, which is evident as a characteristic feature.

Under a rather broad edge there are two cut furrows which form a pronounced ridge. It is similar to the rims of the vessels from Dybeck and Nyrup (Pig. 19) (Straume 1987:86 and 60).

The broad edge indicates that it is a rather big vessel and not a small one like the Himlingpje type (Straume No. 80). The cut decoration de-cides the type. The direction of the sherd shows that it is not a very conical beaker but a vessel with a rather straight wall. This indicates that it is similar to the above-mentioned numbers which have long-oval decoration design. In this case the study of the details of the cutting of the rim is very important for the identification of the

rim. The sherd then belongs to Straumes type IV, Eketorp glass 4 and Eggers 233.

Glass vessels with cut decoration are known from several findplaces in Skåne. Besides the beaker Straume No. 60 from Dybeck there are two beakers with more rounded ovals: Borrby, Borrby par. (E228-230), Forestad moar, Rise-berga par. (E 230) and two sherds.

A sherd, a stray find from Stockholmsgården, Valleberga par. (LUHM 28945, Strömberg 1953) has been dealt with by Straume (1987). It is light green, has a thickness of 4 mm, and traces af two polished oval facets. It seems to belong ta Straume type I. It has previously been identi-fied as belonging to E 228-230 (Hansen 1987) (Lund Hansen has a wrong inventory number).

Another sherd of the same type was found in a rich grave-find from Kristineberg near Malmö (Stjernquist 1994b ). This too is of light green glass and 5 mm thick. It has parts of two cut ovals. Pour fibulae date the grave to the Late Roman Iran Age. A faceted sherd from Stora Köpinge should also be mentioned in this con-nection.

A faceted beaker comes from the neighbour-hood of Lund (LUHM 10926) (Pig. 15). It has


been described in detail by Montelius and dis-cussed by Eggers (type 237), Lund Hansen, Näsman (Eketorp glass 6) and Straume (type VII B 1). The type is a little later than the beakers with cut ovals. The proposed dating is C3, that is, Late Roman Iron Age or the transi-tion to the Migratransi-tion Period. It was found in 1882, reportedly in a grave} pit in the neigh-bourhood of Lund. It was not possible to iden-tify the find place. It is, however, probable that it comes from a very rich complex and it is possible, therefore, that it comes from a grave with some connection to Uppåkra. It is difficult to find any other complex in the region which is in the running.

The motif with tight ovals and the motif with facets are similar and can be found on the same vessel. The beakers with this decoration are spread to the south-east towards the Black Sea region, a probable place of manufacture.

The glass vessels with long ovals have another distribution and are said to have been manufac-tured in a western area (Rau 1972; Koch 1977;

Näsman 1984; Straume 1987). In 1990 Näsman discusses the distribution maps and the glass commerce as a whole in the light of his intimate knowledge of the glass material.

1\vo earlier glass beakers with cut decora-tion from the well-known find from Öremölla belong to phase B2 and type E 187. They will not be discussed here.

Snartemo glass and other cone beakers with trailing

Four of the glass sherds from the Uppåkra sett-lement can be defined as fragments of the Snartemo type of vessels (U 3965, U 4063, U 4080, U 4236) (Fig. 4). They are all character-ized by yellowish-green colour and with thick-ness of 2-3.5 mm. Oval and round bubbles indicate their place on the vessel body. Three of these sherds are decorated with applied self-coloured trailing. Two of them have horizontal spiral trailing below the rim (with 4 and 5 trails respectively with a thickness of 1.5-2 mm) (U 4236, U 4080). One sherd (U 4063) hasa 3 mm thick trail which belongs to the vertical loops of a vessel. The sherd U 3965 has thin polished lines grouped in a way giving impression of

applied trails.

It is not possible to define the shape of the vessels in detail. The horizontal trailing of the two sherds indicates, however, that they belong to the type with a broad zone of spiral trailing around the neck. The dissimilarity of the trail-ing indicates that the two sherds represent dif-ferent vessels. Sherd U 4080 has a trailing which is similar to that of the vessel from Broa (Fig.

20), while sherd U 4236 is similar to the vessels from Eketorp and Nygårds (cf. Stjemquist 1986).

The trail of sherd U 4063 is similar to the verti-cal loops of all the vessels mentioned. With the knowledge of the glasses of the Snartemo type it is possible to assume that the sherds discussed belong to footed beakers. This type of glass vessel can be dated to the Migration Period (Näsman 1984, 153).

Because of their colour and quality three other sherds can be discussed in connection with the sherds of the Snartemo type, but they belong with certainty to other types of glasses.

The green sherd U 4182 has the same fabric as the Snartemo type. Another sherd, U 4183, isa rim fragment of yellowish-green colour. The fabric, however, is thin and the straight rim is not suited for a beaker of the Snartemo type.

Finally, there is U 4016 which is yellowish-green but which has a rather thin fabric. It is a little bent and seems to be part of a vessel with a curved body (Fig. 17).

The starting point for an analysis of the glasses from the Migration Period is the two main types defined by Rygh and Montelius:

Rygh 338/Mqntelius 203 or Snartemo glass and Rygh337/Montelius 202 or Kempston glass.

The difference between them has been de-scribed in Hunter & Sanderson 1982. The great-est typological differences are the proportions, the thickness of the fabric and the presence or lack of a foot. We can agree with the descrip-tions of these two authors. They add differences in the composition of the glass which elucidate that different raw material was used, as shown by technical analysis. The two types have in addition different distributions, which speaks for different areas of manufacture. The areas of manufacture are a problem which has been dis-cussed by many scholars. It is still unsolved,

however. Hunter & Sanderson do not rule out manufäcture in Scandinavia. In the publication of the glass from Gårdlösa other finds of Snartemo glasses from Sweden were discussed.

Based on whole vessels, Näsman has sorted the Snartemo glasses, the footed beakers, into several groups. He has added the groups bowls and low beakers with the same decoration of applied self-coloured trails (Näsman 1984, 1990). In addition a large amount of sherds which belong to the types and in some cases to the Kempston type is presented. The Kempston type has been thoroughly mapped and discussed by Vera Evison (1972).

As the sherd material from settlements indi-cates the shape of the glasses only in certain cases, the finds from the Gårdlösa settlement were dealt with only with the aid of technical analyses which were carried out to elucidate the affinity of the sherds and the assumed number of glass vessels ( cf. Stjernquist 1986). No analy-sis of the shape was performed but the assump-tion was that it was in general the Snartemo type. Similar glass from the Uppåkra settlement will be dealt with in the same manner. The type can be defined by the decoration and by the quality of the fäbric. The number of beakers can then be characterized intensively by technical analyses if it is possible to carry out such inves-tigations.

The sherd material from Helgö is large, to-gether about 1600 pieces (A. Lundström 1981;

Henricson 1990). A large number of pieces be-long to beakers with cut and polished decora-tion or yellowish-green beakers of the Snartemo type (Holmqvist 1964; A. Lundström 1981;

Näsman 1984, 1990). The bulk of the material, however, is sherds of other types of beakers from the Late Iron Age, for instance reticella glass, which is discussed below (Näsman 1986, 1990). A large body of sherds is difficult or impossible to define accurately (Henricson 1990).

In his work on the glass from Helgö Holm-qvist has also discussed the sherds of the Snartemo type. He declares, however, that the shape of the vessels cannot be reconstructed. In 1981 A. Lundström discusses the sherds of Snartemo type in somewhat greater detail. She

discerns about 10 glasses of that type ( 1981, Pig. 1). She considers that some sherds may be of Kempston type (cf. Hunter & Sanderson 1982). Also Näsman discusses the glasses ofthe Snartemo type from Helgö (1986). The glasses from Helgö, however, have not been analysed in detail and it is therefore difficult to give a sum-mary. One has to limit oneself to the conclusion that a part of the large material is fragments of this type of glass. A look through the glass found at Helgö has shown that it is sometimes difficult to distinguish the Snartemo glasses from other types with similar decoration of trails. The trails on the Snartemo type are as a rule lightly melted into the surfäce, but that is not general.

The increase in the number offinds belong-ing to the Snartemo beakers or their variants is mostly due to the fäet that the attention has been focused on the sherd material. It has been clear that sherds appear in great numbers on settle-ment sites. It may be mentioned that, according to Margrethe Watt's mapping, there is a very large body of material found on Bornholm. The glass of the Snartemo type is rather easy to identify because of the characteristic colour, thickness, and decoration. The problem is, how-ever, that many sherds lack the characteristic decoration. They may come from a part of the body which lacks the trails, and the identifica-tion is then difficult. In connection with the publication of the sherds from Gårdlösa the sherd material from find places on Gotland was stud-ied intensively. It tumed out that lots of sherds were difficult to define. It was not possible to draw conclusions concerning the shape of the vessel even if the colour and thickness were similar to the Snartemo type. It also happened that the sherds were partly melted.

The same difficulties concerning the inter-pretation of the sherd material are stressed by Näsman (1984). The difficulties are also due to the fäet that several types of vessels have fäbric which seems to be similar on visual inspection.

The only method to solve the problem is in that case technical analyses of the fäbric (Hunter &

Sanderson 1982). There are, however, problems - not least financial - with technical analyses on a large scale. In his solid work on glass Näsman was not able to accomplish any such analyses.






5 8

9 10

Fig. 3. Stray finds. ppåkra. 1: 1. Fig. 4. Sherds of the Snartemo type. Uppåkra. 1: 1. Fig. 5. Apart of a snicked trail. Uppåkra. 1: I. Fig. 6. Sherds with reticella decoration. Uppåkra. 1: 1. Fig. 7. Rim sherd of a beaker.

Uppåkra. 1:1. Fig. 8. Sherd with reticella decorntion found in Trelleborg. 1:1. Fig. 9. Two sherds of a palm cup. Uppåkra. 1: 1. Fig. 10. Dark green sherd probably of a beaker. pp· kra. 1: 1. Photo: Bengt Almgren.

The glass from Uppåkra has not been techni-cally investigated either but it may be possible later on to complement the research in that way.

Technical analyses of glass from the Iron Age have been discussed in various

connec-tions. Schalars in Poland and areas nearby in the east have done intensive research in this field (Dek6wna 1990). Other schalars who have used technical methods are Chambon & Arbman (1952), Straume (1979), and Koch (1977, 1989).

11 12 13

14 16


18 18A

Fig. 11. Sherd of a funne! beaker found in Trelleborg. 1:1. Fig. 12. Blue sherd from Uppåkra with traces of gold foil. 1:1. Fig. 13. Bluish-green sherds from Uppåkra. 1:1. Fig. 14. Pale blue sherds from Uppåkra. 1:1.

Fig. 15. A beaker from the neighbourhood ofLund. Height 12.3 cm. Fig. 16. Blue sherd from Kverrestad. 1:1.

Fig. 17. Yellowish-green sherds from Uppåkra. 1: 1. Fig. 18. Light green sherd from Uppåkra with traces of gold foil. 1: 1. Fig. 18A. Details of Fig. 18 with traces of gold foil visible. Photo: Bengt Almgren.


~ t, -

... :---

- - .


. ···:·--


. ..




· . ..


- - -- :: .. --- -\,- 1

.... :,



- ..


- - . . -

.. -;,:

21 22

~ r - - - - .

~ --·


-~ ~ , ----.

. /

Fig. 19. Rim sherd from Uppåkra compared with a beakcr from Dybeck. Rim herd 1:1, drawing by the author; beaker 1 :2, af ter Straume 1987. Fig. 20. Sherd U 4080 from Upp, kra placcd on a m del of a beakcr from Broa. 1 :2. Drawing by the author af ter Stjernquist 1986. Fig. 21. The blue sherd from K verrestad placed on a model of a claw beaker from Gotland. 1:2. Drawing by the author after Nerman 1969. Fig. 22. The snicked trail from Uppåkra on a mode! of a claw beaker (Vendel XII). 1 :2. Drawing by the author after Stolpe

&Arne 1912.

As mentioned, technical analyses were carried out for the publication of the sherd material from Gårdlösa with the intention of illustrating the affinity of the sherds and the number of vessels at the settlement. It is quite clear that fragmentary material calls for an increased tech-nical treatment. The techtech-nical analyses must concentrate upon distinct problems.

A problem in the work with technical analy-ses is the possibility to compare the results from different laboratories. It has turned out that they differ every now and then from each other, per-haps because of different methods. Koch has declared in her 1989 paper that Hunter & Sand-erson's analyses of material from Helgö and Spong Hill (1982) differ from her analyses of equivalent material from southern Germany while her analyses from there correspond with Straume's, carried out by Christie and Brenna in Oslo (1979) and with the analyses of the Gårdlösa finds, carried out by Simmingsköld and Linzander in Växjö (Stjernquist 1986).

Arbman's analyses are important not least be-cause of his investigation of a production site at Mascquenoise in Belgium.

Technical analyses ofthe glass from Uppåkra can concentrate on elucidating the manufacture process and the number of beakers. With an increased number of analyses the question of the manufacture area may also be elucidated.

This is very important since some scholars have left open the question whether the glass vessels were manufactured not only on the Continent and in England but also in Scandinavia.

Claw beakers, bag beakers, squat jars There are no certain finds of claw beakers from the Uppåkra settlement. Many sherds of bluish-green or bluish-green glass are, however, characterized by a thin fabric of the quality which is typical of the claw beakers. Also blue sherds of the same fabric are present. Sherds with very thin horizon-tal trails could come from claw beakers which have this type of trails on some parts of the vessel body. As the claw beaker is a very common form in the Migration and early Vendel Period (cf.

Figs. 16, 21) it is very possible that some of these sherds come from this type of vessel. There is, however, no certain evidence of that.


A blue sherd (U 3997) (Pig. 5) with applied snicked trail, about 1 cm long, may come from a claw beaker, a squat jar or possibly a bag beaker, which all have similar applied trails.

The bag beaker is not common in Scandina-via but there is a glass vessel of that kind in a find from Alands, Hogrän par., Gotland (Fre-mersdorf 1955). The beaker is dark blue. The snicked trail is coarse and broader than the trail on the sherd from Uppåkra. The bag beaker occurs in England and rarely on the Continent.

The dating is late 6th century or 7th century.

Several claw beakers have horizontal snicked trails on the body. The bluish-green claw beaker from Vals gärde grave 8, for instance, has around the neck a horizontal, rather coarse snicked trail (Arwidsson 1954). This grave, according to Arrhenius's dating, is from ca. 550-600 A.D.

Amongst the claw beakers with this decoration a beaker of yellowish-green glass from Vendel XII may be mentioned (Fig. 22) as well as light green sherds with snicked trails of a glass from Husby Södra, Trosa-Vagnhärad par. (SHM 26423). These sherds belong to a claw beaker which has been reconstructed and which can be dated to the Vendel Period (Särlvik 1962). An-other claw beaker with the same decoration comes from Grötlingbo or Hablingbo, Gotland (SHM 10928:3, cf. Nerman 1969, Taf. 84). It is made of deep blue glass and has horizontal trails which are very thin. As the sherd from Uppåkra has a rather thin trail, it is possible that it comes from such a beaker or from a neck beaker, a squat jar. This type of vessel some-times has a decoration which is less strong. The decoration with applied snicked trails occurs on several roughly contemporaneous types of glass vessels.

It is not probable that the sherd from Uppåkra can be defined accurately but it shows at any rate that a glass of one of the types mentioned existed in Uppåkra.

Reticella glass

Two sherds with reticella decoration were found at the Uppåkra settlement (F 37, U 3998) (Fig. 6).

The first sherd is of light bluish-green glass.

The reticella cable consists of a self-coloured and a yellow trail twisted together. Three whorls


\ \


I l

" I

,,,/ /1 '

Fig. 23. The reticella sherd Fig. 8 from Trelleborg on a model of a red squatjar from Helgö (after Holmqvist 1964). 1:2. Drawing by the author.

are visible with 5-8 mm between. The thickness of the glass is 2 mm. It is bent along the cable which indicates that it comes from a bulging beaker. The fabric has round bubbles.

The other sherd is of light green glass with a thickness of 2-2.5 mm. Round bubbles are vis-ible. The reticella cable consists of three trails:

one self-coloured, one white and one yellow.

The preserved part of the cable is 1.8 cm. It is not possible to discern where the sherd was placed on the body of the vessel.

Reticella glass dated to the Late Iron Age is well-known from many settlements dealing with commerce, from wealthy farms, and from some grave-finds. This decoration technique has been discussed by Arwidsson in connection with the publication of the Valsgärde finds. In Valsgärde 6 one of the few whole vessels with this decora-tion technique was found. Two others come from Birka and from Hopperstad in Norway. A fourth isa palm cup in the Vatican Museum with Rome as an uncertain find place (cf. Arwidsson 1932, 1942a).

The sherd material is large on sites in Scan-dinavia as well as on sites in England and on the Continent. Rich Scandinavian sites are Birka, Helgö, Kaupang, Dankirke/Ribe, Århus, and Sorte Muld on Bornholm. There are sherds of this kind in Eketorp, at the Åhus settlement, at some other sites such as Paviken on Gotland, and in some grave-finds on Swedish territory. A sherd with reticella decoration is also found in

Trelleborg. Interesting reticella glasses are in-cluded in the glass finds from the farm of Borg in Norway. Hedeby and Dorestad are important sites among the finds on the Continent. This decoration technique is also abundantly repre-sented in England (Arbman 1937, 1940; Holm-qvist 1964; Hougen 1969; Thorvildsen 1972;

Bencard 1978; Henricsson 1990; A. Lundström 1971, 1981; P. Lundström 1975, 1981; Evison 1982, 1983, 1988a and b; Baumgartner &

Krueger 1988; Isings 1978, 1980; Näsman 1986, 1990; Miiller-Wille 1985; Henderson & Holand 1992; Steppuhn 1998b).

The sherd with the reticella decoration from Trelleborg (find No. 616, 1989) (Fig. 8) was found in a pit house with many objects. It was excavated by Bengt Jacobsson. An equal-armed brooch gives a dating to the 9th century. The sherd is very interesting because of the colour.

The glass is red and the reticella cable twisted in two colours, white anda very dark colour which looks black but which is probably very dark red.

The red ground colour of the sherd is striped with horizontal brownish-red bands which can be seen in magnification. It seems to reflect a mixture of the fabric. It shows that the reticella cable had a vertical position on the vessel. Red vessels are not so common but a red vessel with reticella cables is included in the Helgö material (Holmqvist 1964) (Fig. 23).

With the finds from Eketorp as a starting point, Näsman has dealt with the production

In document Fynden i centrum Keramik, glas och metall från Uppåkra Hårdh, Birgitta (Page 75-94)