image/svg+xml
163
XIII/2/2022
INTERDISCIPLINARIA ARCHAEOLOGICA
NATURAL SCIENCES IN ARCHAEOLOGY
homepage: http://www.iansa.eu
Examination of Metal Finds from the 10
th
Century Cemetery of
Kiskunfélegyháza (Hungary)
Béla Török
1*
,
Alessandra Giumlia-Mair
2
1
Institute of Metallurgy, University of Miskolc, B/1, 3515 Miskolc-Egyetemváros, Hungary
2
AGM Archeoanalisi, Via della Costa 4, I-39012 Merano (BZ), Italy
1. Introduction – archaeological background
The excavation of the 10
th
century cemetery identifed on the
Kiskunfélegyháza-Terjék-tanya site was performed by the
Kiskun Museum and the Research Centre for Archaeology
of the Institute for Hungarian Studies within the framework
of a joint project in April 2020 (Gallina
et al.
, 2021). The
excavation showed that the cemetery had been heavily
disturbed by sand quarrying and metal detecting. When the
burial site was excavated, the connection between certain
graves and some scattered fnds could be reconstructed
only partially and it remained often hypothetical. The
archaeological excavation revealed that the cemetery
consists of a number of graves arranged along one line. The
site can be dated to the mid-10
th
century AD. Coins of Hugo
of Provence and Lotar II of Pavia were found in one of the
graves, giving thus the year 931 as the
terminus post quem
of
the burial (Gallina
et al.
, 2021). In the conquest period this
cemetery must have been the burial site of an elite Hungarian
community – or possibly of a family –because, besides
the silver jewellery, a gold hair band and harnesses were
recovered from the graves. As we know from excavations
on other sites dated to the period of the Hungarian conquest
(9
th
–10
th
centuries) that surround Kiskunfélegyháza, the
region was densely populated and was quite an important area
at that time (Tóth, 1974; Somogyvári, 1992; Balogh, 2003;
Varga, 2011). Important archaeological sites, cemeteries of
the conquering Hungarians, and a map of the excavation site
of Terjék-tanya can be seen in the Figure 1. The location of
grave 1 could not be precisely determined, but it could be
between grave 2 and grave 3. Most of the fnds belonging to
grave 1 were found with a metal detector.
2. Materials and methods
Some objects made of precious metals and copper-based alloys
have been analysed by energy dispersive X-ray fuorescence
(henceforth ED-XRF). An Oxford Instrument X-MET8000
portable ED-XRF spectrometer (50 kV, Rh anode, Silicon
Volume XIII ● Issue 2/2022 ● Pages 163–177
*Corresponding author. E-mail: bela.torok69@gmail.com
ARTICLE INFO
Article history:
Received: 22
nd
February 2022
Accepted: 9
th
September 2022
DOI: http://dx.doi.org/10.24916/iansa.2022.2.6
Key words:
metals
10
th
century
Hungary
ED-XRF
OM
SEM-EDS
ABSTRACT
This case study presents the results and conclusions of chemical and metallographic analyses carried
out on metal fnds (gilded silver mounts, jewelry made of silver- and copper-based alloys, and iron horse
fttings) found in nine graves excavated at the 10
th
century site of the Terjék-tanya at Kiskunfélegyháza
(Hungary). The examinations were performed with portable handheld X-ray fuorescence spectrometer
(ED-XRF), optical microscopy (OM) and scanning electron microscopy equipped with an energy
dispersive spectroscope (SEM-EDS). Beside the determination of the chemical composition of the
non-ferrous artefacts and the inclusions of the iron samples, the aim of the study was to detect traces
and characteristics of diferent manufacturing methods such as fre gilding, forging,
etc.
image/svg+xml
IANSA 2022 ● XIII/2 ● 163–177
Béla Török, Alessandra Giumlia-Mair: Examination of Metal Finds from the 10
th
Century Cemetery of Kiskunfélegyháza (Hungary)
164
Drift Detector) has been used. Three calibration methods
were employed for the measurements. In most cases we
used the Precious FP mode, developed for the analysis of
elements found in alloys and in particular precious metals
(Au, Ag). Further, we used the mode Alloy FP, developed for
the analysis of the most common elements found in alloys,
and the Alloy LE FP, which is similar to Alloy FP but also
includes light elements, such as, for example, Mg, Al and Si.
The concentration range for each element goes from 0% to
100% in all three cases. Fundamental parameter (FP) methods
use a complex mathematical analysis of X-ray fuorescence
to calculate the concentration of elements. For metals with
inherently unknown composition, such as historical fnds,
this method is highly suitable and recommended. In the case
of these metal objects only a non-destructive examination
was possible. The general aim of the analyses was that of
determining the chemical composition of the alloys and of
the possible coatings. The measurements were carried out
Figure 1.
A: 10
th
-century archaeological sites (marked with red dots) of the area surrounding Kiskunfélegyháza (Hungary) B: excavation map of the site
of Terjék-tanya.
image/svg+xml
IANSA 2022 ● XIII/2 ● 163–177
Béla Török, Alessandra Giumlia-Mair: Examination of Metal Finds from the 10
th
Century Cemetery of Kiskunfélegyháza (Hungary)
165
under a radiation shield (dome) whenever it was possible –
i.e.
, when the size of the object allowed it – or by hand using
a collimator and a camera pointed at specifc locations, here
marked with numbers in the fgures. In the latter case, when
the object was too small or narrow or when the measurement
was carried out at the edge of the object, the Fe level resulted
higher than usual. The Ti content might have been caused
by the paper placed under the object. In all cases, the tests
were performed with a measurement time of 30 seconds. The
analysis results are reported below in tabular form for each
object. The data are expressed as weight percent.
Regarding the XRF analysis of both the silver- and the
copper-based objects we have to keep in mind that this kind
of surface analysis can only be considered indicative and
semiquantitative at best, because the composition of the
surface is in most ancient objects very diferent from that
of the core metal. We know that in the period of this site
a range of unusual alloys were employed over the entire
European territory, and this was due to the recycling of
metals. In particular, precious metals were melted down
and re-used just as they were (without any refning) to
produce ornaments more pleasing to the owners. Thus, the
analytical data must be discussed in every single case and
all possibilities must be kept in mind, paying attention to the
instrumental error as well. Due to centuries of corrosion, tin
might have redeposited on the surface, so the average values
of this element inside the alloy may actually be lower. Iron
always appears as a contaminant, but its measured value
can be infuenced by external factors as well. Therefore, the
analyst must be very careful in the evaluation of the data.
Two samples from iron fnds were examined by optical and
scanning electron microscopy (OM and SEM-EDS). For the
examination the samples were cut and embedded in epoxy-
resin. The surface of the examined sections was mechanically
polished and etched with a 2% nital solution. A Zeiss Axio
Imager M1m microscope was employed for optical imaging.
The instrument is equipped with a computer-controlled
stage featuring composite imaging for the examination
of the whole surface. The SEM-EDS examinations were
performed on a Hitachi S4300 CFE electron microscope,
equipped with a Bruker energy dispersive spectroscope.
The main objectives of the examination were to determine
the characteristics of the microstructure and to characterise
the nature of the hypothesised manufacturing process. The
production technology can be deduced from the grain size,
the fne or rough microstructure and the distribution of
carbon. The inclusions were also examined and evaluated.
3. Results and discussion
3.1 Owl-shaped mount (grave 1)
The various measurements carried out on diferent areas of
this small object have shown (Figure 2) that the material
employed was silver, alloyed with the so-called gunmetal –
i.e.
, with the quaternary alloy of copper, tin, lead and zinc that
was the most common alloy in use in this period. Gunmetal
was mainly the result of the mixing of scrap with some fresh
additions of suitable metals, when needed. We know that for
a long stretch of time after the 3
rd
century AD throughout
European territories only very little mining activity existed,
and the surviving metallurgical production was mainly
based on the recycling of scrap metal of all kinds. Only
a few larger mines were still being exploited, and there was
very little freshly-smelted metal in circulation. This situation
lasted almost unchanged for many centuries and only by
the 12
th
–13
th
century AD there was a sort of revival of deep
mining and extractive metallurgy, aided by the invention
of various hydraulic machines (see, for example, Le Gof,
1983, pp.222–223; Gille, 1993, pp.660–661). Because of this
situation in vast areas of Europe, the addition of gunmetal or
the addition of scrap to silver to render it harder and more
resistant to wear does not represent a surprise and was a very
common occurrence. After casting, the object had to be
properly fnished by cutting of pouring channels, vents and
casting fns and removing the casting skin.
After the polishing process, the surface had to be carefully
degreased, for example with strong vinegar or alum, as
preparation for the gilding. The analysis data indicate
clearly that the surface was amalgam gilded because of
the regular presence of mercury all over the object. The
amalgam gilding method is also called mercury gilding
or fre gilding and has been known in Europe since the
Hellenistic period, around the third century BC (Craddock,
1977, pp.109–110; Martinon Torres and Ladra, 2011; Perea
et al.
, 2008; Giumlia-Mair, 2020, pp.5–7) – and even earlier
in China,
i.e.
, in the 5
th
century BC (Jett and Chase, 2000). It
is even possible that it spread from China to the West; for the
moment, however, there exists no study that might confrm
this hypothesis (Giumlia-Mair, 2020, p.5). The method of
amalgam gilding involves mixing gold fragments, flings,
and broken wire pieces with mercury to form a greyish
amalgam by grinding it in a mortar. The excess mercury can
be removed from the amalgam by squeezing the mixture in
a leather- or cloth bag so that the liquid mercury can run out
of it. The ready amalgam (
i.e.
, Au
2
Hg with around 10–20%
Au) can be spread on the degreased surface of the object
to be gilded, for example with a hard brush. The object is
then slowly and prudently heated to a temperature that
must remain under the boiling point of mercury (356.73°C)
Figure 2.
Owl-shaped mount.
0 1 cm
image/svg+xml
IANSA 2022 ● XIII/2 ● 163–177
Béla Török, Alessandra Giumlia-Mair: Examination of Metal Finds from the 10
th
Century Cemetery of Kiskunfélegyháza (Hungary)
166
because too much heat would spoil the gilding. The ideal
temperature would be 280–300° C. When the grey amalgam
turns to yellow the mercury volatilises and leaves a thin,
but durable layer of gold on the surface. The gilding is frst
rather dull and porous and has to be burnished, but after the
polishing with special tools or even simple polished stones
like carnelian or hematite, the gilded area looks like shiny
solid gold. Some mercury always remains in the gold layer
and can be easily detected (Giumlia-Mair, 2020, p.7). A very
similar technique was described by Theophilus Presbyter
in his work
Schedula diversarum artium
(early 12
th
century
AD) (Theobald, 1933; Dodwell, 1961; Takács, 1986). As our
analysis data indicate, this gilding technique was employed
on several other objects from this site. The amount of gold
left on the back of this object was only around 2%, while
on the front 12–20% of gold was measured (Table 1).
Obviously, the quantity of mercury decreased proportionally
to the decreased quantity of gold, (as it did here on the more
damaged parts of the front). This is perhaps the reason why
mercury could not be detected with this method of analysis
in the gilding on the reverse of the owl mount, but it is
possible that originally it was present in the gold layer. As we
determined some mercury on the ornamented side, however,
and none on the reverse, where there was only 1 to 3% Au,
there is still the possibility that some gold was present in the
silver, which might have been recycled from gilded silver
objects. A further explanation for the presence of the gold
might be a “technical error” during the gilding process: the
artisan might inadvertently have touched the reverse.
3.2 Gilded silver rosette (grave 1)
The base metal of this decorative rosette (Figure 3) contains
a rather high amount of copper, with some tin, lead and zinc
(Table 2). Apparently also in this case the silver was alloyed
(and diluted) with some gunmetal. Even a relatively small
amount of copper can greatly increase the strength of silver
alloy, whilst even an equal-parts silver-copper alloy retains
the brilliant white colour of silver. In this case, it is worth
noting that the eutectic point of the Ag-Cu diagram is 28.1%
copper in silver,
i.e.
, the alloy theoretically has the best
castability with this composition.
Some areas of the mount (the petals of the fower-like
pattern) have been gilded by using the fre gilding process
described above, but the gilding has partially worn of and
is essentially the cause of the diferences in composition
determined on the various areas of the rosette. The analysis
results of the measurement carried out on the reverse of this
object shows again a low amount of gold, for which we can
hypothesise the same origin as in the previous object.
3.3. Five silver plaquettes (grave 1)
Both sides of the plaquettes (Figure 4) appear to be similar and
without any gilding. We call the sides from which the holes
were pierced the “front”, as they were visible when in use. All
Table 1.
Composition of the owl-shaped mount (wt%).
No.CalibrationLocationMgAlSiTiFeCuZnAsAgSnAuHgPb
1Alloy_LE_FPfront5.120.540.72 0.262.830.210.0874.810.9412.111.680.50
2Alloy_FPfront 0.403.040.190.0077.121.3214.242.190.71
3Precious_FPfront 0.373.230.23 79.420.0013.652.210.79
4Precious_FPback 0.333.620.34 92.050.002.270.001.39
5Alloy_FPback 0.353.320.320.0090.711.232.200.001.27
6Alloy_LE_FPback10.840.830.48 0.202.560.260.1181.720.801.420.000.75
7Alloy_FPfront (coll.) 0.900.492.360.170.0083.801.428.561.250.77
8Alloy_FPfront (coll.) 0.390.303.490.210.0072.121.2918.132.810.71
9Alloy_FPfront (coll.) 2.521.272.260.220.0084.141.636.270.770.80
10Precious_FPfront 0.212.910.21 79.280.0014.172.270.83
11Precious_FPback 0.333.280.32 92.620.002.090.001.25
12Precious_FPfront (coll.) 3.881.302.900.27 77.640.0011.421.750.85
13Precious_FPfront (coll.) 1.900.792.440.19 83.630.008.861.240.95
14Precious_FPfront 0.003.530.20 72.040.0020.103.340.69
15Precious_FPfront 0.271.770.16 88.730.007.200.970.81
Figure 3.
Gilded silver rosette.
0 1 cm
image/svg+xml
IANSA 2022 ● XIII/2 ● 163–177
Béla Török, Alessandra Giumlia-Mair: Examination of Metal Finds from the 10
th
Century Cemetery of Kiskunfélegyháza (Hungary)
167
measurements were carried out under the radiation shield. As
with the previous objects the plaquettes are made of silver,
alloyed with a high amount of a quaternary copper alloy, and
show the same type of composition on both sides (Table 3).
The low gold admixture determined by the measurements
seems to come from the recycling of gilded silver objects (or
perhaps, less probably, of a gilded quaternary alloy) for the
manufacture of this ornament. Plaquette D is more damaged
than the rest, and at frst sight it seems to contain more
silver than the other pieces; however, the slight diferences
in composition among this group of items is probably
simply due to the diferent degree of corrosion of the pieces.
Obviously, oxidation frst attacked the less noble metals in
the alloy, dissolving them and leaving the silver. There is no
trace of the gilding process on the surfaces of the plaquettes.
3.4 Gilded silver mount (found with a metal detector)
and silver studs (grave 1)
When we calculate the composition of the silver mount
(Figure 5) by subtracting the gold of the gilding layer, we
Table 2.
Composition of the gilded silver rosette (wt%).
No.CalibrationLocationMgAlSiTiFeCuZnAsAgSnAuHgPbBi
1Alloy_LE_FPfront (coll.)12.040.620.23 0.1531.160.490.0852.230.221.970.240.440.00
2Alloy_LE_FPfront (coll.)8.790.610.51 0.1716.960.210.0748.220.2520.163.570.210.00
3Alloy_LE_FPback10.570.370.24 0.0022.420.310.1463.910.240.990.000.600.11
5Alloy_LE_FPback14.140.370.19 0.0028.610.390.1254.370.210.870.070.570.00
6Alloy_LE_FPfront6.700.000.28 0.0022.000.230.1151.130.3314.314.140.190.00
7Alloy_FPfront 0.1431.780.430.1058.180.166.931.210.480.09
8Alloy_FPfront 0.1616.340.140.0050.740.3226.055.150.350.05
9Alloy_FPback 0.1330.490.440.1766.030.131.300.000.670.12
10Precious_FPfront (coll.) 0.0017.080.12 50.260.0026.565.210.360.00
11Precious_FPfront (coll.) 0.0029.780.33 57.730.009.771.740.650.00
12Precious_FPback 0.0033.060.50 64.140.001.250.000.950.12
13Precious_FPfront (coll.) 0.340.1829.930.36 57.980.008.931.640.650.00
14Precious_FPfront (coll.) 0.0016.570.15 52.360.0024.785.330.400.00
15Precious_FPfront (coll.) 0.0018.760.14 52.190.0023.125.040.420.00
16Precious_FPback 0.0032.480.44 64.770.001.260.000.940.11
17Precious_FPfront (coll.) 0.0039.150.62 56.910.002.200.280.760.10
18Precious_FPfront (coll.) 0.0015.090.14 50.170.0027.955.870.370.00
19Precious_FPfront (coll.) 0.0023.070.14 51.990.0019.834.180.450.00
Figure 4.
Silver plaquettes.
0 1 cm
image/svg+xml
IANSA 2022 ● XIII/2 ● 163–177
Béla Török, Alessandra Giumlia-Mair: Examination of Metal Finds from the 10
th
Century Cemetery of Kiskunfélegyháza (Hungary)
168
come to an alloy containing 24–26% of copper, which
is almost the eutectic composition that gives the best
castability. The small number of contaminants due to
recycling comprises 1–2% of lead, and, apparently, around
0.5% of bismuth (Table 4). This latter element would greatly
disturb any kind of hammering process, both in copper and
in silver. It is known that as little as 0.02 % of bismuth in
a copper-based alloy would cause cracking when the metal
is intensively worked (Giumlia-Mair, 1992). In a cast piece,
however, this element is not as damaging as in an alloy that
has to be hammered, and it might come from the addition
of unrefned copper to the melt. The noticeable values of
mercury that are without doubt correlated with the amount
of gold clearly indicate the use of fre gilding for this item as
well. The residue of this decorative process is well detectable
Table 3.
Composition of silver plaquettes (wt%).
No.CalibrationLocationMgAlSiFeCuZnAsAgSnAuPb
1 AAlloy_LE_FPfront13.010.360.090.0043.094.320.2136.980.470.640.76
2 AAlloy_LE_FPback15.550.000.090.0042.174.190.2236.490.400.000.74
3 AAlloy_FPfront 0.1548.215.160.2543.590.430.860.88
4 APrecious_FPfront 0.0050.024.48 43.390.000.811.29
5 CPrecious_FPfront 0.0037.572.37 57.300.001.131.63
6 CPrecious_FPback 0.1634.382.52 59.920.001.181.84
7 BPrecious_FPfront 0.0043.481.69 52.510.000.941.38
8 BPrecious_FPback 0.0050.461.74 45.760.000.801.24
9 DPrecious_FPfront 0.0033.002.47 61.740.001.111.68
10 DPrecious_FPback 0.0025.681.87 69.210.001.371.87
11 CPrecious_FPback 0.1649.541.78 46.290.000.871.35
12 CPrecious_FPfront 0.0045.352.95 49.190.000.961.56
13 APrecious_FPback 0.0049.324.32 44.190.000.821.35
14 APrecious_FPfront 0.1949.954.71 43.030.000.791.33
15 APrecious_FPfront 0.0048.654.58 44.570.000.851.35
16 EPrecious_FPback 0.0037.741.63 58.100.000.991.53
17 EPrecious_FPfront 0.0039.451.58 56.380.001.001.59
18 EPrecious_FPfront 0.1537.691.54 58.220.000.951.44
19 EPrecious_FPfront 0.0039.821.83 55.760.000.991.60
Figure 5.
Gilded silver mount and studs for boots.
Figure 6.
Pendant ornament.
0 1 cm
0 1 cm
image/svg+xml
IANSA 2022 ● XIII/2 ● 163–177
Béla Török, Alessandra Giumlia-Mair: Examination of Metal Finds from the 10
th
Century Cemetery of Kiskunfélegyháza (Hungary)
169
even on the worn surfaces. In the deeper recesses of this
mount a still signifcant gold content was detected. The
analysis of the reverse side of the mount shows that the silver
alloy also contains around 1% gold, and the same amount
of lead, both elements probably coming from recycling and
from the use of unrefned copper.
The material of the two studs – possibly originally
employed as decoration for boots or perhaps for leather
straps belonging to harnesses, bridles or a saddle – is silver
with a slightly lower copper content than in the case of the
previously-discussed plaquettes.
3.5 Pendant (found with a metal detector)
The material of this piece (Figure 6) is essentially the same
as in the previous objects: an alloy of silver and a quaternary
alloy. The diference in composition (Table 5) of the various
parts of the object are due to the corrosion of the less noble
metals present in the alloy and to segregation phenomena.
Nevertheless, we should keep in mind the possibility that
these pieces with a rather high copper content might have
been also surface-treated, if the colour of the silver resulted
to be too pink for the taste of the customer. The practice of
oxidising the baser metals from the surface of debased silver
is very ancient. This was done by placing the debased silver
objects in a strong organic acid to remove the copper and
any other base metals from the surface. The object was then
burnished and looked like solid silver or, at least, like a better
quality and more “silvery” kind of metal. Early examples of
treated rings from a burial in the Nahal Qana Cave are dated
to the 5
th
millennium BC, and they were surface-enriched
(Shalev, 1993); several other early examples of this practice
exist (Giumlia-Mair, 2020, pp.8–9). The silver artefacts
belonging to this group might have been at least partly
surface-treated before the gilding was applied.
The copper content of the round part of this pendant is
higher than that of the hanging part, and some measurements
(
e.g.
, on the back of the round part) gave a copper content
that was higher than silver. This is not strange, when we
think that debasing silver was a very common practice in
Roman times after the 2
nd
century AD and even more so later,
in the Middle Ages. We know that in the third century the
Roman “silver” coinage only contained around 4% of silver
in copper (Cope, 1972; Giumlia-Mair, 2001, p.770; Zwicker
et al.
, 1993). The traces of fre gilding on the front of the
Table 4.
Composition of gilded silver mount and silver studs (wt%).
No.CalibrationLocationFeCuZnAgAuHgPbBi
1Precious_FP(coll.)0.1722.770.0065.157.822.451.070.57
2Precious_FP(coll.)0.1617.910.0060.8315.044.560.820.42
3Precious_FP(coll.)0.0018.630.0060.0914.804.910.860.44
4Precious_FP(coll.)0.0024.640.0064.307.122.500.940.49
5Precious_FPback0.0027.920.0069.540.940.001.030.56
6Precious_FPback0.0032.160.1065.170.980.001.030.56
APrecious_FP(coll)0.4721.780.1974.491.170.001.420.49
BPrecious_FP(coll)0.4923.430.1872.651.210.001.530.51
Table 5.
Composition of the pendant ornament (wt%).
No.CalibrationLocationTiFeCuZnAsAgSnAuHgPb
1Precious_FP(coll.)0.000.1841.930.23 45.320.009.172.260.91
2Precious_FP(coll.)0.370.2642.710.18 49.430.004.840.951.25
3Precious_FP(coll.)0.000.1642.960.24 46.440.007.471.790.94
4Precious_FP(coll.)0.000.1528.890.16 58.670.009.381.481.26
5Precious_FPround part back0.000.0058.490.25 38.940.000.950.001.37
6Precious_FPhanging part back0.000.1942.880.23 53.790.001.310.001.60
7Alloy_FPround part front0.000.1940.720.200.1244.820.729.892.260.66
8Alloy_FP(coll.)0.530.2742.230.160.1851.560.732.730.440.90
9Alloy_FP(coll.)6.302.2130.710.240.2049.740.877.451.240.86
11Precious_FPround part front0.000.1942.510.22 47.410.007.111.570.98
12Precious_FPround part back0.000.0058.140.24 39.370.000.920.001.33
13Precious_FPhanging part front0.000.1627.400.14 60.080.009.461.441.32
14Precious_FPhanging part back0.000.2846.430.26 50.230.001.270.001.52
15Precious_FP(coll.)0.000.2044.290.26 43.430.008.692.150.97
16Precious_FP(coll.)0.000.2040.950.19 51.390.005.021.011.25
image/svg+xml
IANSA 2022 ● XIII/2 ● 163–177
Béla Török, Alessandra Giumlia-Mair: Examination of Metal Finds from the 10
th
Century Cemetery of Kiskunfélegyháza (Hungary)
170
object are clearly identifable. Originally the entire surface
of the small circular area in the centre of the round part was
gilded.
3.6 Pendant ornament (grave 16)
This ornament (Figure 7) has the same shape and design of
the previous one. The composition (Table 6) of the material
is also very similar and this suggests that the two objects
belonged to the same set of ornaments and represent a pair.
In the case of this particular piece, the gold layer, applied by
fre gilding, seems to be thicker than on the other objects or
it is simply much less worn and better preserved. It is again
amalgam gilding, because when the measurements indicated
a decrease in the gold content on the surface, there was
a decrease of mercury as well.
3.7 Fragments of a silver ring (grave 11)
This object (Figure 8) is very fragile and thin; therefore, only
one fragment (marked with an X) belonging to it has been
analysed. The item has a rather homogeneous aspect, and
there was no reason to think that there might be diferences
in the composition of the various fragments. The piece was
examined under a radiation shield dome, as it exhibits a very
thin cross-section. This thin strap-like object is made of silver
with a gold content of less than 2% and with a little copper
and lead (Table 7). Lead is a typical contaminant of medieval
silverware because lead and silver are usually present in the
same ore. The amount of lead in these objects is relatively
high to be a natural impurity of the silver coming from the
ore; however, we have to note that medieval silver typically
has a lead content in the range between 0.05 to 1.0%, (Merkel,
2016, p.25). The composition of this ring might also suggest
that the intensive hammering and annealing that must have
been carried out in the manufacture of this piece caused the
Figure 7.
Pendant ornament.
Table 6.
Composition of the pendant ornament (wt%).
No.CalibrationLocationFeCuZnAgAuHgPb
1Precious_FPround part front0.1628.600.1647.1817.675.110.79
2Precious_FPround part back0.0059.050.2238.541.050.001.14
3Precious_FPhanging part front0.1942.900.1949.625.430.481.19
4Precious_FPhanging part back0.0040.500.1656.541.400.001.40
5Precious_FP(coll.)0.0031.950.1848.1214.464.100.93
6Precious_FP(coll.)0.0029.480.1555.9010.462.860.98
Table 7.
Composition of silver ring (wt%).
No.CalibrationLocationFeCuAgAuPbBi
1Precious_FPfront0.000.3497.481.870.180.00
2Alloy_FPfront0.200.3796.461.890.180.10
3Precious_FPback0.000.5597.541.780.130.00
Figure 8.
Fragments of a silver ring.
0 1 cm
0 1 cm
image/svg+xml
IANSA 2022 ● XIII/2 ● 163–177
Béla Török, Alessandra Giumlia-Mair: Examination of Metal Finds from the 10
th
Century Cemetery of Kiskunfélegyháza (Hungary)
171
loss of the baser elements that had been introduced into the
alloy by the addition of gunmetal.
3.8 Bracelet (grave 11)
In accordance with the wearing custom of bracelets in this
period, this silver bracelet with widened rounded ends was
found on the middle part of the forearm of the deceased.
However, the ends were turned downwards, so that the
ribbed decorations at the ends were hidden (Gallina
et al.
,
2021, p.357).
All measurements on the bracelet (Figure 9) were carried
out free hand (
i.e.
, with camera and collimator). This ornament
must have been worn for a long time and the contact with the
skin might have oxidised out of the surface the baser elements
originally present in the material. This is also suggested by
the rather consistent amount of gold determined all over the
piece, while copper, lead and zinc show more fuctuation
(Table 8). This indicates that, with all probability, the alloy
contained more base metals. The presence of relatively high
bismuth is puzzling and is possibly due to instrumental error.
This element renders copper and silver very fragile under
the hammer, and, as the bracelet is made by hammering the
presence of bismuth in the alloy is doubtful.
3.9 Torque (grave 11)
In grave 11, the torque, twisted from double-layered bent
silver wire, had been found closed with the clasp below
the cervical vertebrae. The appearance of silver necklaces
made of similar thin wires can be dated from the second
quarter of the 10
th
century (Révész, 1996, p.92), however,
it is necessary to distinguish them from the similar artefacts
made by folding a single wire into two and by folding and
twisting a single wire into three layers. The former are the
rarer and probably earlier versions of this type of artefact,
although their use may have been parallel.
The examination of a complete torque (Figure 10) with
a portable ED-XRF spectrometer is not an easy task because of
the awkward shape of the object that cannot ft under the dome.
Collimator and camera were used for all measurements; as
the torque has a thin cross section; however, the measurement
results were strongly infuenced by the paper under the
object as refected in the high Ti and Fe determined by the
system (Table 9). Nevertheless, the determined values can
be considered informative in terms of the proportions of the
components. The torque is made of silver with a relatively high
copper content, and the baser elements in the alloy are better
preserved than, for instance, in the thin ring discussed above.
Table 8.
Composition of the bracelet (wt%).
No.CalibrationFeCuZnAgAuPbBi
1Precious_FP0.0027.580.1069.911.001.410.00
2Precious_FP0.1812.010.1084.661.221.710.11
3Precious_FP0.1815.250.1381.281.261.780.11
4Precious_FP0.0019.310.0078.701.110.880.00
5Precious_FP0.2010.480.0086.251.211.740.12
Table 9.
Composition of the torques (wt%).
No.CalibrationTiFeCuZnAsAgSnAuPb
1Precious_FP13.415.4728.950.39 47.710.001.211.30
2Precious_FP16.817.4728.980.52 40.150.001.161.49
3Alloy_FP12.645.3931.270.510.2145.620.971.130.96
4Alloy_FP18.287.9624.340.600.2043.031.021.191.23
Figure 9.
Bracelet.
0 2 cm
0 2 cm
image/svg+xml
IANSA 2022 ● XIII/2 ● 163–177
Béla Török, Alessandra Giumlia-Mair: Examination of Metal Finds from the 10
th
Century Cemetery of Kiskunfélegyháza (Hungary)
172
3.10 Fragments of a copper-based torques (found with
a metal detector)
The fragments of a torques (Figure 11) are made of a brass
containing 18–24% zinc. It is worth noting that only low tin
and lead amounts have been detected in the alloy. Higher
Sn- and Pb-percentages (2–4%) are more common in brass
objects (Török
et al.
, 2016; Szentpétery and Török, 2022)
found at 10
th
–11
th
-century sites in Hungary; however, we also
have to remember that, rather often, tin tends to redeposit on
the surface due to the long period of time in the soil, during
which the corrosion acted on the alloy.
The reddish colour of unalloyed copper becomes paler
with an addition of up to 7% Zn, but it is still red. The alloy
becomes yellowish-red when the amount of zinc increases
to up to 14–15%. With a Zn content of 15–17%, the alloy
becomes clearly yellow. The colour of brass is most similar
to that of gold with a Zn content of 20–30%. The addition of
Figure 10.
Torques.
Table 10.
Composition of the fragments of bronze torques (wt%).
No.CalibrationLocationTiFeCuZnAsAgSnAuPb
1Alloy_FPA (coll.)1.930.7272.3124.330.130.000.000.000.00
2Alloy_FPC (coll.)2.741.0872.9121.960.220.130.170.000.00
3Precious_FPA (coll.)3.591.3071.7621.45 0.000.000.150.27
4Precious_FPC (coll.)3.401.2872.3620.99 0.000.170.150.31
5Precious_FPB (coll.)0.310.2976.0722.92 0.120.200.000.10
6Precious_FPA0.000.0075.9224.08 0.000.000.000.00
7Precious_FPA0.000.0075.9624.04 0.000.000.000.00
8Precious_FPA0.000.0076.3623.64 0.000.000.000.00
9Precious_FPB0.000.3080.4218.88 0.170.220.000.00
10Precious_FPB0.000.2781.4917.77 0.150.210.000.11
11Precious_FPC0.000.0076.2723.42 0.130.190.000.00
12Precious_FPC0.000.0075.8423.82 0.140.200.000.00
13Precious_FPB (coll.)1.870.8073.9822.08 0.070.210.000.22
Figure 11.
Fragments of bronze torques.
0 2 cm
0 1 cm
image/svg+xml
IANSA 2022 ● XIII/2 ● 163–177
Béla Török, Alessandra Giumlia-Mair: Examination of Metal Finds from the 10
th
Century Cemetery of Kiskunfélegyháza (Hungary)
173
zinc renders harder and more wear resistant the copper alloy.
Brass reaches its highest hardness when it contains 28.5%
zinc. This percentage increases the tensile strength and
castability, but it reduces the malleability. The most cold-
workable brass alloy is the one containing 15–20% zinc,
but it cannot be hot worked (Bokor, 1998). The presence of
other metals in the gunmetal alloy further lowers the melting
point, facilitating the casting, and improves the malleability
as well. The only exception is lead that renders the alloy
more fragile, when added in higher amounts (above 3–4%).
In this case, however, only traces of lead were determined in
the alloy (Table 10). The artisans that produced these torques
wished to obtain a drawable and twistable alloy, easy to shape
when cold and with a colour as similar to gold as possible.
In this case they employed the best possible composition.
This shows the skill and empirical knowledge they applied
in their professional routine.
3.11 Fragments of an iron stirrup and a bit (grave 1)
Two samples of fragments of diferent iron objects were
examined with optical microscopy and SEM-EDS. Sample A
comes from a stirrup, and sample B from a fragment of a bit
(horse mouthpiece) or possibly a buckle (Figure 12).
3.11.1 Sample A
The metallographic examination showed mostly a pearlitic
area with ferrite network that can be observed on the left side
of the composite image of the whole section of the stirrup
sample (Figure 13). At higher magnifcations, the ferrite
grains are very angular. The structure becomes more ferritic
toward the right side (Figure 14). Although toward the lower
and upper edges of the cross-section (
i.e.
, the surface of the
object), the amount of pearlite decreases and thus the carbon
content is reduced; this is not a folded material. The layered
structure could rather have developed in this way as a result of
the object being rotated during the initial forging operations
(compacting). The ferrite grains and pearlite colonies are
small, and the pearlite structure is so fne that it cannot be
easily observed by optical microscopy. This implies a faster
cooling than equilibrium and a relatively high carbon
content. The stirrup is made from a single material and by
simple hot forging (so called “bulk forging”).
Based on the SEM-EDS examination, it can be assumed
that the angularity of the ferrite is caused by the relatively
rapid cooling and the so-called Widmanstätten efect, which
is an oriented, needle-like growth of ferrite grains. The
small black parts in the ferritic area (Figure 14/D) are most
probably carbide grains.
It is worth mentioning that metallographic analyses
of stirrups from the Avar period of the Carpathian Basin
(6
th
–9
th
centuries AD), often showed a slightly higher
proportion of pearlitic areas and the use of folding as
a forging process (Török
et al.
, 2017; Török and Barkóczy,
2022). This technology, however, did not give a signifcant
diference in the quality of the stirrup as a fnal product.
Figures 15 and 16 show SEM images of the inclusions
present in sample A, and the spots marked with green
numbers indicate the points measured by EDS method (in
wt%). The C content values given as composition are only
indicative, as the carbon detection by EDS is highly uncertain.
Figure 12.
Fragments of an iron stirrup and a bit or buckle.
Figure 13.
Composite image of the whole
cross-section of the sample of the stirrup.
0 2 cm
image/svg+xml
IANSA 2022 ● XIII/2 ● 163–177
Béla Török, Alessandra Giumlia-Mair: Examination of Metal Finds from the 10
th
Century Cemetery of Kiskunfélegyháza (Hungary)
174
Figure 14.
OM images of the samples of
the stirrup.
Figure 15.
SEM image of an inclusion of the sample of the stirrup.
8: C:1.91, O:46.29, Mg:2.48, Al:28.35, Mn:3.27, Fe:20.25; 9: C:0.47,
O:43.20, Na:2.23, Al:14.42, Si:18.47, K:1.46, Ca:6.47, Mn:3.42, Fe:9.86.
Figure 16.
SEM image of an inclusion of the sample of the stirrup.
40: Fe:100 (ferrite); 41: C:1.53, O:24.68, Al:1.31, Si:0.44, Fe:72.04;
42: C:1.36, O:45.71, Mg:2.96, Al:29.16, Si:0.62, Mn:2.54, Fe:17.62;
43: C:2.08, O:42.33, Al:33.92, K:3.27, Ca:0.93, Fe:17.47.
The inclusions we investigated may have originally been
slag inclusions that had formed during smelting present in
the bloom. They have a metallurgical origin: the composition
indicates that they derived from a bog ore, charcoal ash
(characterised by higher K, Na and Mg contents), and furnace
linings (characterised by Al-silicate); no phosphorus was
detected in any of the cases. Manganese, which is specifc
for this ore, was measured in all types of inclusions in the
sample and in rather signifcant amounts.
The shape of the complex microstructured inclusions
is strongly distorted by the efect of intensive forging. In
Figure 16 an inclusion (42) is coated by a layer of almost
pure iron oxide (41): hammerscale probably was formed
secondarily during the shaping process. The relatively low
values of silicon content are surprising.
3.11.2 Sample B
In the cross-section of sample B, there are metallic areas that
are smaller than in the case of the stirrup. This sample is
mostly ferritic, with a signifcantly higher fraction of almost
pure ferrite, compared to the previous sample (Figure 17).
Only a small amount of pearlite can be observed on the left
side of the composite image (Figure 18/A). A large area of
the sample consists of large grains, which are a consequence
of the low carbon content. Tiny precipitates can be observed
in the ferrite grains as black spots, which are presumably to
be interpreted as tertiary iron-carbide (Figure 18/C).
This material is slightly softer than the previous one. The
object is made by simple “bulk forging” of a single material.
image/svg+xml
IANSA 2022 ● XIII/2 ● 163–177
Béla Török, Alessandra Giumlia-Mair: Examination of Metal Finds from the 10
th
Century Cemetery of Kiskunfélegyháza (Hungary)
175
No special forging, heat treatment or other special processes
were used.
Fewer inclusions were found in this sample than in the
case of the stirrup. The lens-shaped inclusions (Figure 19)
were smaller and generally of the same type. They were
arranged in a row in the soft, ferritic structure, according to
the direction of the hammering. Their composition suggests
that they are Fe-Al-silicate slag inclusions with high
phosphorus content, and a legacy from smelting.
4. Conclusions
Nine members of the local community were buried in the
cemetery of the Kiskunfélegyháza–Terjék-tanya, and the
graves were dug in a single row. The gilded silver artefacts we
examined adorned the clothes of the buried persons. Most of
the more impressive silver objects from the cemetery –
i.e.
,
the owl mount, the rosette, the decorated silver mount, the
studs and the plaquettes – come from grave 1. The exception
are the pendants, one of which was found outside of context
with the metal dectector, while the second comes from
grave 16. The workmanship of the various silver fttings,
the gilding and the fnishing of the objects all suggest that
this group of items belonged together. We should not forget
Figure 17.
Composite image of the whole
cross-section of sample B.
Figure 18.
OM images of sample B.
Figure 19.
SEM image of an inclusion of the sample B. 1: C:1.53, O:26.65,
Al:1.09, Si:11.92, P:3.38, Fe:55.42.
image/svg+xml
IANSA 2022 ● XIII/2 ● 163–177
Béla Török, Alessandra Giumlia-Mair: Examination of Metal Finds from the 10
th
Century Cemetery of Kiskunfélegyháza (Hungary)
176
that the graves had been disturbed and mixed up, when sand
extraction had been carried out in the area, or by people
looking for “treasures” with the metal detector. The two-
piece pendant ornaments were typical Hungarian costume
decorations in this period, however, in the cemeteries of the
common people, pieces made of bronze are usually found.
Particularly typical (Bálint, 1991, pp.123–126) is the two-
piece ornament with a rhombus-shaped pendant found in
grave 16. Five more examples similar to this piece have
been found on the site with the help of metal detectors. The
owl-shaped mounts are also not unknown in other 10
th
and
11
th
century cemeteries on the Hungarian Plain. A good
example is grave 72 in the cemetery of Homokmégy-Székes,
where the bird’s head is more stylised and made of brass
(Török
et al.
, 2016, pp.214 and 218).
The fve silver plaquettes found in grave 11 were most
likely decorations for clothes (Gallina
et al.
, 2021), but similar
pieces have also been observed on horse harnesses in burials
of the period (Varga, 2016). The large, but thin and fragile
silver ring, the bracelet and the silver torque that come from
grave 11 could well represent the funerary set of a member
of the same family or social circle. Their manufacture and
fnishing, however, is very diferent from that of the previous
group of silver pieces: these are rather simple, hammered and
very linear objects, without any decoration or gilding. The
impression is that they were made by a diferent pair of hands,
not by the same artisan that produced the gilded silver objects.
The low but regularly appearing tin and lead percentages in
the torque seem to suggest that this alloy also contains some
scrap metal. In the cases of the gilded silver mount and studs
(Figure 5) and silver bracelet (Figure 9), part of the bismuth
content might be the result of an incorrect deconvolution of
a lead energy peak (peak overlap). These items may contain
some bismuth, but probably not in such quantity. The copper-
based torque fragments have been recovered with the metal
detector and they consist of double rods twisted to form
a ring. A relatively high zinc content (more than 10%) could
also be detected in some cases of ED-XRF and SEM-EDS
examinations of brass fnds from other 10
th
–11
th
century
cemeteries and settlements of the Magyars (Török
et al.
, 2016;
Szentpétery and Török, 2022); however, the zinc content
values in Table 10 appear to be extremely high. This high zinc
content raises questions about the production of brass in this
period. Were there smithsonite and sphalerite mines that could
be exploited at the time? Would the artisans of the period be
able to actually produce fresh brass? A possible answer could
be that the raw material for the brass objects came mainly from
the south or the southeast. Brass ingots with a zinc content of
nearly 30% have been found at the 10
th
century metalwork
centre near Nadarevo in Bulgaria. The extremely low levels of
tin and lead suggest that this was fresh brass and not recycled
metal (Doncheva
et al.
, 2017). Regarding the metal art centres
of medieval Bulgaria, it is worth noting that silver objects with
a copper content of almost 50%, around 1% of tin, between 1%
and 1.5% of lead and a gold content between 1% and 4% have
been observed on several occasions (Doncheva
et al.
, 2013).
The results can be compared with the composition of the silver
objects we have studied, especially the silver plaquettes and
the reverse of the gilded objects.
Neither the stirrup nor the other fragments of horse
fttings difer from the general types of the period in shape,
material, or manufacturing technique. If the silver plaquettes
mentioned among the dress ornaments belonged to a horse
harness, they would then have been sewn onto a leather
component.
The XRF analysis of the fnds from the Kiskunfélegyháza
– Terjék-tanya site helped to illustrate the habits of the local
metal artisans of the 10
th
century, with the recycling of both
copper-based alloys and precious silver, still with some
gilding remains recognisable inside the alloy, which testify
to the re-melting of partially-gilded silver items. The traces
of mercury identifed in the partly worn gilded decoration
also demonstrate the use of the amalgam gilding technique.
The carefully forged stirrup and its metallographic structure
show the skill and expertise of the local blacksmiths in the
10
th
century AD.
Acknowledgement
We are grateful to Péter Barkóczy (Institute of Physical
Metallurgy, Metalforming and Nanotechnology at
the University of Miskolc) for the co-operation in the
metallographic analyses of the iron samples.
References
BÁLINT, C., 1991.
Sudungarn im 10. Jahrhundert.
Studia Archaeologica,
11.
Budapest: Akadémiai Kiadó.
BALOGH, C., 2003. 10. századi sírok Kiskunfélegyháza környékéről. In:
A Móra Ferenc Múzeum Évkönyve.
Studia Archaeologica, 9. Szeged:
Móra Ferenc Múzeum, pp. 283–296.
BOKOR, J., 1998. Sárgaréz. In:
A Pallas nagy lexikona XIV
. Budapest:
Arcanum FolioNET, p. 69.
COPE L.H., 1972. Silver surfaced ancient coins. In: E.T. Hall, D.M.
Metcalf, eds.
Methods of Chemical and Metallurgical Investigation of
Ancient Coinage
. London: Royal Numismatic Society 8, pp. 261–278.
CRADDOCK, P.T., 1977. The composition of copper alloys used by the
Greek, Etruscan and Roman civilisations.
Journal of Archaeological
Science
, 5, 103–124.
DODWELL, C.R., 1961.
The Various Arts. De Diversis Artibus.
Oxford:
Clarendon Press.
DОNCHEVA, S., PENEV, I., TSEKOVA, G., FURU, E., SZIKSZAI, Z.,
and UZONYI, I., 2017. Elemental Composition of Metal Artefacts from
the 10
th
c. Metal Art Centre near the Village of Zlatar, Preslav Region, NE
Bulgaria.
Archaeologia Bulgarica,
17(1), 71–85.
DОNCHEVA, S., PENEV, I., TSEKOVA, G., FURU, E., SZIKSZAI, Z.,
and UZONYI, I., 2017. Elemental Composition of Artefacts Found at
the 10
th
Century Decorative Metalwork Centre near Nadarevo, Region of
Targovishte, Bulgaria.
Archaeologia Bulgarica
, 21(1), 75–88.
GALLINA, Z., GULYÁS G., HORVÁTH, C., KNIPL, I., and MAKOLDI,
M, 2021. 10. századi sírok Kiskunfélegyháza-Terjék-tanya lelőhelyről.
In: L.T. Vizi, ed.
A Magyarságkutató Intézet évkönyve 2020.
Budapest:
Magyarságkutató Intézet, pp. 345–374.
GILLE, B., 1993. Macchine. In: C. Singer, E.J. Holmyard, A.R. Hall,
T.I. Williams, eds.
Storia della tecnologia, 2/2
. Torino: Bollati
Boringhieri, pp. 638–673.
GIUMLIA-MAIR, A., 1992. The composition of copper-based small fnds
from a west Phoenician settlement site and from nimrud compared with
image/svg+xml
IANSA 2022 ● XIII/2 ● 163–177
Béla Török, Alessandra Giumlia-Mair: Examination of Metal Finds from the 10
th
Century Cemetery of Kiskunfélegyháza (Hungary)
177
that of contemporary mediterranean small fnds.
Archaeometry
, 34(1),
107–119.
GIUMLIA-MAIR, A., 2001. Colouring Treatments on Ancient Copper
Alloys.
Revue de Métallurgie, Cahiers d’informations Techniques
, 98(9),
767–776.
GIUMLIA-MAIR, A., 2020, Plating and Surface Treatments on Ancient
metalwork.
Advances in Archaeomaterials
, 1, 1–26.
JETT, P., and CHASE, T.W., 2000. The gilding of metals in China. In:
T. Drayman-Weisser, ed.
Gilded Metals. History, technology and
conservation
. London: Archetype Publications, pp. 145–155.
LE GOFF, J., 1983.
La civiltà dell’Occidente medievale.
Torino: Einaudi.
MARTINÓN-TORRES, M., and LADRA, L., 2011. Orígenes del dorado por
amalgama: aportaciones desde la orfebrería protohistórica del noroeste de
la Península Ibérica.
Trabajos de Prehistoria
, 68(1), 187–198.
MERKEL, S.W., 2016.
Silver and the silver economy at Hedeby
. Der
Anschnitt, Beiheft, 33. Bochum: Deutsches Bergbau-Museum Bochum.
PEREA, A., MONTERO, I., GUTIERREZ, P.C., and CLIMENT-FONT, A.,
2008. Origen y trayectoria de una técnica esquiva: el dorado sobre metal.
Trabajos de Prehistoria
, 65(2), 117–130.
RÉVÉSZ, L., 1996. A karosi honfoglalás kori temetők. Régészeti adatok
a Felső-Tisza vidék X. századi történetéhez. In: L. Kovács, and L. Révész,
eds.
Magyarország honfoglalás kori és kora Árpád-kori sírleletei.
Miskolc: Herman Ottó Múzeum, Magyar Nemzeti Múzeum, p. 92.
SHALEV, S., 1993. The earliest gold artifacts in the Southern Levant.
Reconstruction of the manufacturing process. In: C. Éluère, ed.
Outils et
et ateliers d’orfèvres des temps anciens
, Antiquites nationales mémoire,
2. Musée d´archéologie nationale: Saint Germain en Laye, pp. 9–12.
SOMOGYVÁRI, Á., 1992. Bronzkori és Árpád-kori temetkezések
Kiskunfélegyháza-Pákán.
Bács-Kiskun Megyei Múzeumok Közleményei
(Kecskemét)
, 1990, 13–26.
SZENTPÉTERI, J., and TÖRÖK, B., 2022. A solti Tételhegy szórvány
leleteinek skandináv és balti kapcsolatai (régészeti és archeometriai
adatok). In: Á. Somogyvári, ed.
„Élőknek öröksége.” Tanulmányok
V. Székely György emlékére
. Kecskemét: Katona József Múzeum,
pp. 141–176.
TAKÁCS, V., 1986.
Theophilus Presbyter: A különféle művességekről
.
Budapest: Műszaki Könyvkiadó.
THEOBALD, W., 1933.
Technik des Kunsthandwerks im 10. Jahrhundert
des T. Schedula Diversarum Artium
. Berlin: VDI.
TÖRÖK, B., BARKÓCZY, P., KOVÁCS, Á., 2016. Homokmégy-
Székes 10–11. századi temetőjéből származó, jellemző fémleletek
komplex műszeres archeometriai vizsgálata. In: Z. Gallina, and
S. Varga, eds.
A Duna-Tisza közének honfoglalás és kora Árpád-kori
temetői, sír- és kincsleletei I. A kalocsai Sárköz a 10–11. században
.
Budapest, Szeged: Szegedi Tudományegyetem Régészeti Tanszéke,
MTA Bölcsészettudományi Kutatóközpont Régészeti Intézete, Magyar
Nemzeti Múzeum, Viski Károly Múzeum, Kalocsa, pp. 208–238.
TÖRÖK, B., BARKÓCZY, P., KOVÁCS, Á., KRISTÁLY, F., BÁNHIDI,
O., KREITER, A., SKRIBA, P., PETŐ, Á., GYULAI, F., SALÁTA,
D., 2017. Avar kori vasművesség az interdiszciplináris kutatások
tükrében – Zamárdi-Kútvölgy és Kaposvár-Fészerlak leleteinek
archeometriai vizsgálatai. Ironworking of the Avar period in the scope
of interdisciplinary research – Archaeometrical investigation of the fnds
from Zamárdi-Kútvölgy and Kaposvár-Fészerlak sites.
Archeometriai
Műhely
, 14(3), pp. 185–202.
TÖRÖK, B., BARKÓCZY, P., 2022. Hortobágy-Árkus lelőhelyről
származó vastárgy töredékek metallográfai vizsgálata. Metallographic
analysis of fragments of iron artefacts found at the Hortobágy-Árkus site.
In: G. Szenthe, E. Gáll, eds.
Hortobágy-Árkus kora középkori temetője.
Egy elit csoport hagyatéka a 8–10. századi Észak-Tiszántúlról. The Early
Medieval cemetery at Hortobágy-Árkus. A 8th–10th century elite group
from the North Eastern Carpathian Basin
. Budapest: Hungarian National
Museum, 376–391.
TÓTH, H.E., 1974. Honfoglaláskori sír Kiskunfélegyházán.
Archaeológiai
Értesítő
,
101, 112–126.
VARGA, S., 2011. 10–11. századi temetők Bács-Kiskun megye területén.
In: S. Bíró, and P. Tomka, eds.
„Hadak útján”. A népvándorlás kor
kutatóinak XIX. konferenciája
. Győr, Moson, Sopron: Megyei Múzeumok
Igazgatósága, pp. 215–229.
VARGA, S., 2016. Lószerszámok. In: Z. Gallina, and S. Varga, eds.
A Duna-
Tisza közének honfoglalás és kora Árpád-kori temetői, sír- és kincsleletei
I. A kalocsai Sárköz a 10–11. században
. Budapest, Szeged: Szegedi
Tudományegyetem Régészeti Tanszéke, MTA Bölcsészettudományi
Kutatóközpont Régészeti Intézete, Magyar Nemzeti Múzeum, Viski
Károly Múzeum, Kalocsa, p. 324.
ZWICKER, U., ODDY, A., LA NIECE, S., 1993. Roman Techniques for
Manufacturing Silver-plated Coins. In: S. La Niece, and P.T. Craddock,
eds.
Metal Plating and Patination. Cultural, Technical and Historical
Developments
. London: Elsevier, pp. 223–246.
image/svg+xml