Thilo Rehren
Doris Vanhove
Herman F. Mussche
Mary Oikonomakou
Litharge from Laurion. A medical and metallurgical commodity
from South Attika
In: L'antiquité classique, Tome 68, 1999. pp. 299-308.
Citer ce document / Cite this document :
Rehren Thilo, Vanhove Doris, Mussche Herman F., Oikonomakou Mary. Litharge from Laurion. A medical and metallurgical
commodity from South Attika. In: L'antiquité classique, Tome 68, 1999. pp. 299-308.
doi : 10.3406/antiq.1999.1348
http://www.persee.fr/web/revues/home/prescript/article/antiq_0770-2817_1999_num_68_1_1348
Litharge from Laurion
A medical and metallurgical commodity from South Attika
Introduction
During the first century AD, both Dioskourides of Anazarbos and Pliny the
Elder discuss in some detail the preparation of lithargyros 'silver stone1 and spuma
argenti 'scum of silver' for medical use of which Pliny in his Naturalis Historia (33,
1 10) says: «It is used to make an eye-wash and for women's skin to remove ugly
scares and spots and as a hair-wash. Its effect is to dry, to soften, to cool, to act as a
gentle purge, to fill up cavities caused by ulcers and to soften tumours... it also
removes erysipelas and likewise chilblains....» (transi. H. Rackman [Loeb]) Both
authors apparently rely on the same source of information. Nevertheless Pliny gives
some additional information concerning the metallurgical background of its
production1. Dioskourides in his De Materia Medica (V, 87) says about lithargyros
that «one is made from sand called molybditis which is roasted until it is totally
burned, another from silver, a third from lead. The Attic is excellent, the second is
from Spain, followed by those from Dikaiarchia and Sicily» and Pliny (o.e. 33, 106)
notes «The same mines also produce the mineral called spuma argenti (scum of
silver). Of this there are three kinds, with Greek names meaning respectively golden
(chrysitim), silvery (argyritim) and leaden (molybditim)... the Attic kind is the most
approved, next the Spanish. The golden scum is obtained from the actual vein, the
silvery from silver, and the leaden from smelting the actual lead...» (transi. H.
Rackham [Loeb]). Despite in both cases the name silver is mentioned, the material is
virtually free from silver. It is lead oxide, in English litharge, which forms as a
'Waste' during the last stage of silver smelting. As will be shown, it was not at all a
waste, but an important commodity in Antiquity. In silver smelting, silver-bearing
lead ore2 is first smelted to gain metallic lead which collects all the silver present,
while the gangue minerals form the slags, or skoriai. The silver content of the lead
was usually small, depending on the ore. For the Laurion in South Attika, it is
generally assumed to be nearer to one tenth of a percent than to one percent3. To gain
this silver, the metallic lead is re-melted and oxidised in an open furnace, called
1
For a detailed and interdisciplinary interpretation of these and other chapters on
silver metallurgy see Projektgruppe Plinius, Silberbergbau und -Verhüttung in der Antike,
Bochum, 1998.
2
Some silver ores, e.g. in Spain, have insufficient lead mineral. Then, a suitable
lead compound has to be added to collect the precious metal. See H.-G. Bachmann, The
archaeometallurgy of silver, in R. Francovich (ed.), Archeologia delle Attivitá Estrattiva
e Metallurgiche, Firenze, 1993, p. 563-592, for more details of ancient silver metallurgy.
3
C. CONOPHAGOS, Le Laurium antique, Athens, 1980, p: 279; H.-G. Bachmann,
ArchäometallurgiSche Untersuchungen zur antiken Silbergewinnung in Ldurion, in
Erznietall, 35 (1982), p. 246-251.
300
LITHARGE FROM LAURION
SKORIAI
Fig. la. Smelting of silver-bearing lead ore.
cupellation hearth, forming lead oxide or litharge and pure silver (Fig. 1). Therefore,
one expects to find substantial quantities of litharge at a silver-producing region, like
one finds slags at a metal smelter or a smithy. In the Laurion, however, there is very
little litharge in the material record.
The heydays of the Laurion as a major silver producing region were during the
fifth and fourth century BC. Mining in the region started, however, already in the
early third millennium, and lasted until the sixth century AD4. In his monumental
monograph on the metallurgy of this region, C. Conophagos calculated the amount
of metallic lead smelted during the classical period to about 1.4 million tons,
producing 350,000 kgs of silver metal. These figures were based on the quantity of
H. MuSsche, Thorikos: A Mining Town in Ancient Attika, Gent, 1998, p. 10-16.
T. REHREN, D. VANHOVE, H. MUSSCHE, M. OIKONOMAKOU
LITARGE
301
D.V.
Fig. 1b (A). Cupellation by overflowing of the litharge.
LITHARGE
J>.V. '50
Fig. Ib (B). Cupellation by immersion of bars.
302
LITHARGE FROM LAURION
ancient slag heaps (about 1.5 million tons) and mining debris (well above 10 million
tons) as measured during the second half of the 19th century AD5. The modern
exploitation of the region started by processing these materials, which were easily
accessible at the surface, already being crushed to a convenient grain size well suitable
for modern techniques. Thus, we are faced with an enormous amount of silver being
produced in the Laurion during the classical period, and the amount of litharge
resulting from the cupellation should be in the same range as the weight of the lead
metal, well above one million tons6. This is in sharp contrast to any noticeable
quantity of litharge in the archaeological record of the Laurion. The few and isolated
finds of litharge cakes from the excavations of C. Conophagos and H. Mussche
weight in total not more than a few kilograms. The reworking, by a second smelting,
of the litharge to soft or industrial lead may be an explanation; on the other hand even
if the Greeks used a lot of industrial lead as in architecture, for anchors, fishing, loom
weights, ex-voto's, sheets for packing etc., it is unlikely that all the litharge produced
was immediately reworked to industrial lead. In this paper, we try to give several
possible explanations for the apparent lack of this litharge.
Litharge in the Laurion
Until today, there were only two types of litharge known from the Laurion,
namely first solid lumps or cakes, and then hollow sticks7. Both types of litharge are
well known from almost all lead-silver smelting sites in the Old World, and are easily
recognised. Why are they so scarce in the Laurion?
During a recent survey and study program, we were able to analyse several
sediment samples from several ore washeries in the vicinity of the Belgian
Archaeological School in Thorikos. These include two samples of the Ore deposit1
next to the washery no. 1 at Thorikos (samples LTO and LTU, Fig. 2), and several
samples from current excavations of the Ephoria Β at the slopes of the hill
immediately west to the Belgian Archaeological School (LTH2 from the northern
slope, property Mexa, Fig. 3).
The sediments from Thorikos were excavated as two distinct heaps, in two
rooms adjacent to the washery, containing fragments of mill stones and huge slabs
with worked surfaces from grinding material. The sediments look, at a first sight,
quite similar to the local soil. On a somewhat closer inspection, however, they are
easily distinguished by their significantly higher density (measured densities of these
sampels are 5 grams/ccm, while the local soil density is less than half that value).
Also, they have a different texture, sometimes like fine sand, and often show an
internal stratification of layers of different grain sizes and hues of brown to red.
Sample LTO is from the northern heap, while sample LTU represents the more
southern deposit. The total volume of these two deposits is estimated to close to one
5
However, we have to be critical with these number, because some of the modern
calculations are inconsistent.
6
While the silver is extracted from the primary lead, thus reducing its weight, the
amount of litharge formed is in excess of the metallic lead due to the addition of about one
fifth of its weight in oxygen.
7
C. Conophagos, o.e., p. 313-323.
T. REHREN, D. VANHOVE, H. MUSSCHE, M. OIKONOMAKOU
2
Fig, 1. Thorikos, washery no. 1.
303
3
4
5WI
LITHARGE FROM LAURION
304
hundred litres, equalling several hundred kilograms of material. From the stratigraphy
of the installation, it is certain that the sediments date to the fourth century BC, and
not to a later period of production8.
The LTH2 samples originate from a huge, multi-phased washery complex.
Here, in two large ore washeries dating to the mid of the fifth-fourth century BC. the
fourth century BC we identified lead-rich sediments. They are of a distinct dark purple
colour, indicating less soil contamination than in Thorikos. In accordance with this,
the measured density of these samples is about 6 grams/ccm, significantly more than
the Thorikos material. The most massive occurrence was found at the lower washery
A, where it covers one corner of the installation, up to ten centimeters thick, and
completely filling in the water channels in that part. No detailed volume
measurements were done, but we estimate the amount here to at least 60 liters,
equalling about 360 kilograms of material. In the vicinity of the second washery Β
more uphill, some quantities of the same material were found, probably ten to twenty
liters. Minor amounts were also found within the washery installations proper.
According to our analyses, all the samples presented here consist mainly of
litharge, ground to a fine powder and mostly weathered, over 2.400 years, to lead
carbonate and lead hydro-carbonate. The amount of other components is about 20
percent by weight in the Thorikos material, and about 10 percent by weight in the
samples from the property Mexa (Table 1).
Table 1. Chemical composition of sediment samples from washery no. 1 at Thorikos (LTO and LTU) and
from the property Mexa (LTH2). The total lead content is given as PbCO3; microscopic investigation
showed that not all of the lead oxide is actually transformed into lead carbonate. This and the presence
of other elements not listed in the table accounts for most of the difference between the calculated sum
of oxides, and 100 percent.
LTOa
LTOb
LTUa
LTH2a
LTH2d
SiO2
8,47
10,80
5,80
3,33
4,93
A12O3
2,28
2,87
1,10
0,78
0,94
Fe2O3
2,93
3,20
1,51
0,62
0,90
MgO
0,60
0,61
0,60
0,52
0,56
CaCO3
12,25
11,21
6,05
5,73
7,35
P2O5
0,18
0,30
0,24
0,19
0,18
S
0,13
0,10
0,09
0,07
0,10
Cu
0,64
0,09
0,22
0,76
0,78
ZnO
3,19
4,20
0,76
0,24
0,48
PbCO3
62,42
68,45
89,52
82,07
83,35
As
0,24
0,25
0,30
0,05
0,05
Sb
0,61
0,58
0,71
0,27
0,27
Sum
93,94
102,66
106,90
94,65
99,88
Based on a comparision with various kinds of cupellation debris, and after a
detailed microscopic and microanalytical study of the samples using thin sections, Xray diffraction and bulk chemical analyses by ICP-OES, there is no doubt concerning
the current, weathered, and the former, original, composition. It is in particular clear
that no natural lead ore, galena or cerussite, is present here9. Obviously, huge
quantities of ground litharge are present at Thorikos and at the property Mexa.
8
For details of the situation of ore washery A2 and its surrounding buildings, see
Thorikos I, Rapports préliminaires 1963, Bruxelles, 1968, p. 97-104.
9
Although the basic chemical and mineralogical data of the geological cerussite and
the artificial carbonate materials are identical, there are unambigous criteria to distinguish
these two, namely the shape and texture of the particles present. Cerussite developed from
litharge still preserves the lathy, sometimes wavy shape of the primary litharge crystals,
and often still has an uncorroded core of red litharge. Geological cerussite, to contrast, has
either its Own euhedral crystal faces, or mimics the cubic shape of galena.
T. REHREN, D. VANHOVE, H. MUSSCHE, M. OIKONOMAKOU
mm'
LTH2
Fig. 3. Property Mexa.
_J
I
O
D.V.
305
306
LITHARGE FROM LAURION
Interpretation
This discovery rises some questions, and allows several interpretations. The
main question is, why is a deposit of litharge in the ore washery, and why in this
condition. Its condition, namely as a fine sand or powder, implies that it was crushed
and ground like the ore for which the washeries were originally built. This indicates
that the cupellation debris, i.e. the litharge, had some value, enough to justify
transport from the cupellation furnaces back to the washeries. Then, time and effort
were spent to grind and treat the litharge here. On the other hand, its present
occurrence in a relatively significant quantity could indicate that this materials value
was limited, making the business secondary.
What, then, was it used for? The treatment in the washeries implies that the
use either required a certain grain size and quality of the material, or that it aimed at
only a certain fraction of it, which was accessible by washing.
There are at least two uses of litharge in Antiquity we have evidence for. Long
known is the addition of fine-grained litharge to produce the famous waterproof and
extremely lasting lining of the water-bearing installations in the Laurion, especially
cisterns. The demand from this usage, however, was necessarily limited, since the
installations where of a lasting nature, and not even all or all parts of them where
really done in this way.
The other use we know of is the one mentioned above by Dioskourides and
Pliny the Elder, as a basic ingredient for medical preparations. The demand from this
for Attic lithargyros will have been significant, considering that it was famous for its
superior quality. More important, this demand was independent from the local mining
activity, but arose from the daily grieve and illness throughout the Greek, and later
Roman, world. When quantities of litharge for the preparations are given, they are
usually minai or litrai, indicating that it was used in substantial amounts. Although
Dioskourides is speaking of «sand», Pliny (o.e. 33, 108-109) says «In the case of the
silvery kind ... the instructions are to break it into pieces the size of a bean and boil it
in water in an earthenware pot with the addition of wheat and barley ... Afterwards
they grind it in mortars for six days, three times daily washing it, with cold ... and
hot water and adding salt ...» (tranls. H. Rackham [Loeb]). Thus, a good deal of the
litharge which went into the medical business must have been exported as it formed
in the cupellation process, i.e. in solid lumps, and not as a fine-grained, washed
powder. Although the local preparation of litharge ointment 'ready for use1 can not be
excluded, we have yet no evidence for this. Therefore, medical use accounts for the
limited occurrence of litharge in the Laurion, but most probably it is not responsible
for the litharge found in the washeries.
For this, however, we can offer two possible explanations, both immediately
related to the metallurgical activity. The first possibility is the reworking of
cupellation debris to gather small droplets of silver and silver-bearing lead which were
regularly trapped in cracks and fissures of the cupellation furnaces. Without going
into any detail, it may suffice to mention that the silver content of the litharge is in
the order of a few hundreds of a percent, which is still one tenth of the original silver
content of the ore. This is a typical value for litharge cakes, while the canes of pure
litharge have usually less than a tenth of that value. In contrast to the ore, where
T. REHREN, D. VANHOVE, H. MUSSCHE, M. OIKONOMAKOU
307
silver and lead are mineralogically related to each other and require high-temperature
treatment to separate them, in the litharge most of the silver is present as tiny
particles mechanically trapped in the matrix. Due to the silver's physical properties
(density, ductility, wetting properties against water) being different from the litharge,
it can be gained similar to the way how gold is panned or washed from sand, and the
ore from the sterile waste. Therefore, one may assume that the cupellation debris was
crushed and washed like ore, to gather the residual silver. This may have occurred
immediately after the cupellation, more precisely after each repair or relining of the
cupellation furnaces, as it is done even today in modern silver works. As well,
however, this may have been the occupation of a later generation. Strabon tells us
that in his time (1st century BC/AD) there was a smelting industry based on the
reworking of the debris (ekvolades and skoriai) of the ancient activity.
Another possible metallurgical use of the litharge lies in the mineralogical
composition of the ore mined in the Laurion. Initial mining, in the Early Bronze
Age, may have made use of the rich cementation zone near the surface of the deposit.
In classical times, however, it is assumed that predominantly galena, lead sulfide, was
mined and smelted10. Although the direct smelting of sulfidic lead ore is possible, it
would have had profited from previous roasting in order to reduce the sulfur content of
the ore. Roasting beds, though, were never found in the Laurion. A similar effect
would have been achieved by the addition of lead oxide, which easily reacts with lead
sulfide to form metallic lead and gaseous sulfur dioxide. In order to have the reactions
taking place, however, the grain size of the reaction partners have to be as similar as
possible to allow thorough mixing. Thus, crushing of the litharge cakes would make
sense. Furthermore, washing would be necessary to separate the litharge proper from
adhering soil and furnace components, which, during smelting, would otherwise
eagerly react with the lead oxide, suppressing the slower reaction of lead oxide with
lead sulfide. Furthermore, by adding the litharge to the smelting charge, the silver
trapped in it would be gained with the metallic lead, making the whole operation
more economical.
Whatever is the right answer, the washing of litharge was already done during
the fourth century AD, and continued until much later times. Its use may have been
the same throughout this time, but it is as well possible that it changed, or
developed, over the various exploitation periods. At the present, we can not imagine a
way to unambigously determine the reason for this activity; neither in general, nor in
the individual case. Far too few technical details of ancient smelting practices at the
Laurion are known.
What would have happened with the litharge after washing? In the first case,
the finally exhausted litharge would have been dumped together with the other waste
from the washeries, more or less sterile mineral sand which became known in later
times as plynites. We can only speculate about the exact mineralogical composition
of these plynites, now that they are all gone. One single analyses is known with
separate figures for lead oxide (17 wt%) and lead sulfide (0.9 wt%). The total lead
content of this sample is too high for the typical plynites, of less then ten weight
percent lead, and too low for a typical litharge sediment, of at least 60 wt% lead.
10
H.-G. Bachmann, o.e., p. 249-250.
308
LITHARGE FROM LAURION
Also, the lead oxide measured may have been either geological lead carbonate mineral,
cerussite, or primary galena now weathered to lead carbonate. In any case, crushed
litharge as a part of the plynites would have gone unnoticed in the modern production.
In the second case, the litharge would mostly have disappeared already in
Antiquity. With little economic value in its own, its use would have been dependent
on the availability of fresh, primary, ore. A shortage in supply of galena could easily
result in the abandoning of a washery specialised in the reworking of litharge, leaving
behind large quantities of now worthless material. This seems to be the situation in
the property Mexa, where the litharge sediments even contain fragments of roof tiles
from the obviously ruined workshop.
Here it is necessary to point out that not all the washeries we have visited so
far contain litharge sediments. In many more cases, we identified by visual inspection
plynites, i.e., remains of the original ore. These, however, need sampling and
subsequent analytical work before any interpretations can be drawn. Clearly, the
working of litharge is a significant aspect of the operation of the washeries, not
limited to the Thorikos area. But not only more systematic sampling of sediments is
necessary, but also, more precise dates for the erection, and the ceasing of operation,
of the installations is needed before we can discuss these questions in more detail. In
view of their solid construction, and the good condition in which most of the
washeries were found, it may be safely assumed that they could have been operative
for a century, or re-opened several centuries later for a second period of use. The
sediments we find in these washeries now, of course, always represent the very final
stages of their operation only.
Conclusion
The apparent absence of significant quantities of litharge in the Laurion can
now be discussed in view of fresh evidence. Firstly, we now know with certainty that
significant quantities of litharge were processed on an industrial scale similar to the
washing of ore, using the very same installations and, hence, techniques. Further
more,it was utilised for non-metallurgical purposes, like the preparation of specia
lisedplasters and pharmaca, for which the Attican quality was famous throughout the
Roman Empire.
From the large-scale and persisting processing of litharge within the miningmetallurgical complex of the Laurion, secondly, we can conclude that it was either fed
into the ancient furnaces, as an additive to enhance the smelting of the primary ore, or
that it was dumped together with the regular plynites once the mechanically trapped
silver metal was extracted from it. In the first case, the litharge would have gone
already in Antiquity, in the latter, unnoticed, with the smelting of the plynites and
ekvolades during the years 1865 to 1910 AD.
Which one ever of the two possibilities is the correct one, the current findings
add an interesting and significant aspect to our understanding of the organisation, and
complexity, of the ancient metallurgical operations.
Thilo REHREN (Bochum), Doris VANHOVE (Gent), Herman MUSSCHE (Gent), Mary
OIKONOMAKOU (Ephoria Β Athens)