Etruscan Granulation

 Gerhard Nestler
Edilberto Fomigli

Etruscan Granulation
An Ancient Art of Goldsmithing

Brynmorgen press

ISBN  9781929565733
 
English language edition
© 2010 Brynmorgen Press
 
Originally published as Granulazione Etrusca
© 1994 nuova immagine editrice
Via Quirico, 13
1-53100 Siena, Italy
 
Translated by Paola D’Amato
Edited by Tim McCreight
 
We want to thank the Museums and Institutions that supported our work and the authors Hans Ulrich Tietz, Michela Coriolaro, Monica Guerrieri and especially Sylvia.
 
All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any storage and retrieval system except by a reviewer who wishes to quote brief passages in connection with a review written for inclusion in a magazine, newspaper, web posting, or broadcast.
 
Also available in a print volume from booksellers and at www.Brynmorgen.com.

Preface
Jewelrymaking was one of the first ways that humans used metal. One of the earliest methods has remained in use over the centuries and continues to exert its extraordinary charm today: that method is granulation.
      Granulation, a sophisticated technique in which tiny granules are attached to a gold base, was practiced by ancient cultures for millennia. In these masterpieces of accomplishment, the number of granules can reach hundreds of thousands.
      Ancient jewelry is now confined almost exclusively to museum cases and therefore cannot express its true beauty. Without the reflection of the sun in the natural movement of a wearer, without the skin, hair, and garments of a human being, these wonderful pieces have been transformed into stunning but static showpieces.
      On the positive side, archeologists and museum curators made this jewelry the object of study—describing it, classifying it, giving it a chronologic location. With the advent of microscopic observation and sophisticated analytical tools, we have entered a new phase in the appreciation of granulated jewelry. Only recently have we begun studying these archeological finds as handmade creations with their own history and technical genealogy. We have begun empirically to reconstruct the phases of workmanship, recreating the surroundings, the tools, and the working methods in an attempt to bring together theory and practice, the mind and the most ancient tool of man, his hands.
 

 1
Pendant in the shape of the head of Achelous, with hook loop, about 480 BC, Louvre Museum, Paris

Introduction
A treatise by Alessandro Castellani (1), a member of a famous family of Roman goldsmiths, was handed out to the visitors of the 1862 London World Exposition. Here is a brief excerpt:
      “>Kind gentlemen, thanks to your courteous benevolence, I was given the opportunity to explain in a few words the results of my research into the goldsmith’s art in the way it was practiced by ancient populations. [...] I will refer to the procedures used by the artists of those times. Unfortunately, these procedures have been lost, as well as many other mysteries of a civilization which was the mother of our own. [...] Among other things, the almost invisible granules that had such an important role as decorative elements in the work of ancient goldsmithing, presented us with almost insurmountable difficulties. We have made countless attempts [...] we studied the writings of Pliny, Theophilus and Benvenuto Cellini, without neglecting other sources [...].”(2)
      The art which Castellani refers to is granulation, the art of attaching tiny spheres onto a metallic substructure at minute tangent points in such a way that the point of contact is almost invisible to the human eye. The name comes from the Latin granum: grain or seed.
      The Castellanis were never able to discover this “secret of the ancients,” but their many imitations of Etruscan gold work were of such high quality that only today are they recognized as forgeries. Meanwhile, they became museum and collectors items, at times more expensive than the originals.
      The newly created “mystery” of the granulation technique in the 1800s began to stimulate the imagination of goldsmiths and others. It was a time of great discoveries and immense opportunities. English, French, German, and American diggers supplied the museums of their own countries with treasures of the ancients, often shamelessly plundering archaeological sites.
      Findings were not only made in Asia Minor and Greece, but everywhere. Treasures such as the Regolini-Galassi tomb in Cerveteri were discovered in people’s backyards. These discoveries focused the attention of Europe on one people: the Etruscans.
 

2
Large disc Fibula, from the Regolini Galassi Tomb, in Cerveteri, about 650 BC, 31.5 x 24 cm, Gregorian Etruscan Museum, Vatican. On the arc, of elliptical shape and with leaves motif, 55 geese are aligned. They are built in the round and using two half cylinders of foil 24 cm long, with a zigzag pattern. All are made with very fine granulation. The oval bracket shows five lions embossed in bas-relief. All the outlines are realized with linear granulation, barely perceptible to the naked eye.

Ex Oriente Lux, The Light from the East

3
Details of the disc-fibula.
The most surprising findings unearthed at that time were works of Etruscan goldsmithing. They were not only expressions of dignity and wealth for the living, but religious relics for the dead. Through refined granulation techniques, Etruscan artisans achieved a level of technical and artistic mastery unsurpassed by other cultures. Remarkable is the complexity of the decorative pattern, but also the extreme refinement of the techniques. They seem to touch the limits of human possibility. In the 7th and 6th centuries BC, Etruscan goldsmiths were able to cover jewelry with a very fine granulation of 0.12 mm. At this scale it is impossible to distinguish individual spheres with the naked eye; the result appears as an opaque surface resembling velvet. Some sources call this variation dust or powder granulation. On certain jewelry, tens of thousands of minute spheres of equal size were applied. An example of this is the large disc Fibula of Cerveteri, located in the Vatican (Gregorian Etruscan Museum). For this remarkable work, more than 120,000 minute spheres were prepared, individually applied to the gold surface and soldered in an almost imperceptible manner.
      The goldsmiths who were able to achieve such mastery perfected their abilities through knowledge of their predecessors. Those who think the Etruscans first developed this refined decorative technique will be surprised to learn that when the Etruscans were using it, granulation was already an ancient technique with a tradition of more than two thousand years.
      The most ancient objects with granular decoration that we know of are dated 2500 B.C. and were found in the royal tombs of Ur in Mesopotamia. It is from here that granulation spread, especially after the destruction of that city. It spread to Anatolia, to Syria, and to Troy (2100 BC). The “Treasure of Priam,” >discovered by Schliemann, attests, in the second period of Troy, to a high level of technical proficiency, with linear and triangular pattern granulation.

4
Next to the Tutankamon mummy, two daggers were placed for protection from the demons of the afterlife, one with a hardened gold blade, the other with a well preserved iron blade—a very rare metal for those times. The finely chiseled images on the case portray the victory of good against evil. Both handles are adorned with geometric, lozenge, and triangular granulation. Egyptian Museum, Cairo, JE 61584.
      Starting in the second millennium B.C., Egyptian goldsmithing of the Middle Kingdom still shows clear elements of a Mesopotamian style. The most ancient findings with granulation come from the princely tombs of Dahour (3). Of course, we shouldn’t forget the treasures of the New Kingdom, with the extraordinary belongings of Tutankhamun: daggers, rings, and bracelets carry decorative patterns with linear, triangular, and rhombic pattern granulation, in red gold and electrum (gold-silver alloy). Granulation with different gold alloys implied, for those times, a thorough knowledge of metallurgy.

5
Sophia Engastomenos, Schliemann’s wife, wearing the tiara of “Priam’s Treasure.
      Around 1600 BC, also in Palestinian Territory, in Cyprus, there were Sumerian and later, Palestinian influences. Here a remarkable granulated piece was created, a pendant with a triangular granulation of nine rows and more than 4000 granules.
      The spread of this art in the Western world has been called Ex Oriente Lux, The Light Coming from the East. As early as 2000 BC, Minoan goldsmiths used this refined decorative technique for their traditional naturalistic patterns, more inspired by Mesopotamia than by Troy. This was the first step toward Europe. Four more centuries went by before continental Greece acquired the “new” technique.
      With the disappearance of Myceane in the 12th century BC, this artistic form died out for a few centuries, only to come back to life at the time of the Greek colonizations of the 9th century BC.

6
The map shows the spread toward the West of the granulation technique through different ages: Starting with the Sumerians, in Southern Mesopotamia and continuing toward Troy, Egypt, Palestina, Crete and lastly the Greek inland (black arrows). Later, the technique was revived and new waves of dispersal occured, this time through the Phoenician traders and artisans, starting from Syria, Carthage (red arrow), and Greece (yellow arrow).
      While the Greeks were founding their first colonies in the West, Phoenician traders maintained business relations and settled in small communities in the Tyrrhenian area. This could explain the fact that the first Etruscan granulated jewelry was more influenced by Phoenician than by Greek art. It is probable that immigrant Phoenician goldsmiths introduced new techniques in Etruria, including filigree and granulation.
      An example of this can be found in the presence in Tarquinia of a leech-fibula, which represents one of the most ancient examples of Etruscan granulation. If the leech shape of Villanovan tradition, is typically Etruscan, the Palmette motif, which also adorns it, originally belongs to the Phoenician tradition (4). This new style begins to appear in the 8th century BC and characterizes the following century, during the so-called “Orientalizing” phase.
      In the Villanovan Age, golden objects were rare, being mostly limited to foils decorated with the punch technique (repoussé) along with the solar disk motif. The materials used for decorative objects were mostly bronze, amber, bone, and horn.
      For the Etruscan artisans who were able to make refined bronze microfusions, it was probably not too complicated to master the new technologies. Particularly, the art of granulation seems to have immediately attracted both purchasers and artisans. The tombs of the noble families, rich in granulated jewelry, demonstrate this.
      In the beginning we can only find a geometric ornamental typology, with a limited number of granules. In the Late-Orientalizing Phase though, we can find more mature shapes with a high number of granules adorning the objects with Greek fret designs and meanders, or outlining figurative motifs. The number of small spheres increases as the technique becomes more masterful. A leech-fibula from Tuscania in the British Museum (dated to around 630 BC) has approximately 25,000 granules, some as small as 0.12 millimeter in diameter, attached to in the finest geometric style. In the second half of the 7th century, after the introduction of the silhouette technique in Vetulonia, or in Vulci, much jewelry is decorated using dust granulation to thickly cover the background. Smaller spheres and a wider variety of motifs characterize this extremely sophisticated age of Etruscan goldsmithing.
      In the Archaic Age, the 6th century BC, the art of granulation loses a little of the importance it had in the previous century. At this time, spiral and twined threads of various shapes accompany the use of granules. They are elegantly placed to create fanciful compositions and complicated floral motifs. Dust granulation was now needed to create an opaque background to better emphasize the smooth and glowing areas of a gold leaf.
      During the Classic period in the 5th and 4th century BC, granulation not only represents rich ornamentation, but it is also used to cover joints and as a bond to strengthen soldered joints. In these cases, the granules are almost thrown in bulk, without the creation of predisposed figures. Typical of this type of granulation are the “Disk Earrings” (5) of the second half of the 4th century BC.

7
Types of Granulation:
a) linear granulation with single and double rows, set apart or ajar;
b) plane triangular granulation, starting from an initial triangle of three granules;
c) granulation covering a hemispheric foil (usually with the base enclosed in round wire);
d) linear granulation on spheric foil;
e) complex geometric leitmotif with granulation, wires, and hemisphere;
f) linear granulation with meanders;
g,h) granulation of the outlines and of the details of raised figures in bas-relief;
i) “silhouette” granulation, with covering of the background;
j) field granulation of the background that leaves some of the foil smooth.

      At times the surface is worked with stamping (puntinatura), made with a stump from the back. This is a sort of imitation of the original technique (6).
      The pick of this sophisticated form of art rapidly became obsolete: it became the expression of a widespread Hellenistic style.

8
Previous double page: detail from a disk earring from Cetona, in Chiusi, with dust granulation on hemispheres and curved foils; 6th century BC, in Antikensammlung, Berlin.

9
Comb-Fibula from Palestrina, about 650 BC, with linear contour of embossed figures in bas-relief.

10 
Detail. Museum of Villa Giulia, Rome.

The Mystery
During the Renaissance, the resurgence of interest in antiquity resulted in the formation of rich collections that were eventually opened to the public. Into the 18th and 19th centuries, museum goers admired in astonishment Etruscan jewelry covered with gold granules. It seemed impossible that the ancient artisans had achieved such perfection using their “simple” means.
      The idea that contemporary culture was no longer able to do what the ancients did was gaining popularity and was nowhere better exemplified than in Etruscan granulation. Goldmsiths of the 19th century were forced to admit that in their attempts to duplicate the work of the ancients, modern techniques were falling short.
      They had to find a solder to attach minute spheres of a tenth of a millimeter to the thin gold sheet without leaving visible traces. To achieve the clarity of the Etruscan work, the solder must not fill the spaces between the spheres, but should be so minimal that light could filter through the spaces between granules, giving them a suspended effect. The parts should be joined homogeneously to obtain an object that would remain free of corrosion for an entire millennium. Soldering a second time, the heat shouldn’t melt the joints, but, on the contrary, should make them even harder.
      One of the people who took on the challenge was the Roman goldsmith Alessandro Castellani, collector, antiquarian, and researcher. Recall his words about the problem of ancient techniques:
      “>[We have] lost procedures, [and] many other mysteries of a civilization which was the mother of our own.”
      How did it happen that a procedure so old, already in use for 4300 years, disappeared without leaving any trace and how was it entirely forgotten? The answer is trivial: Technology had changed. Starting in the 19th century, oil, gas and then electricity replaced the use of charcoal for soldering.
      The new energy sources were better and more practical but their advent “put to sleep” >the ancient procedure of colloidal soldering, a joining technique based on a chemical solder that requires the reducing action of charcoal.
      An aura of mystery surrounded the granulation technique far into the 20th century. For far too long, archeologists and art historians were only concerned with the iconographic aspects and didn’t pay much attention to the techniques used by the ancient goldsmiths. Professional forgers took advantage of this gap.

11
The radiography shows in lighter color the metallic parts, less transparent under X-ray, of the disk earring, in Creton.

 

12 & 13
Details photographed with the electron microscope. The granules can reach a thickness of 0.07 mm. 

13-14
End of a strap with figurative leitmotifs in the round and in bas-relief. Linear granulation emphasizes details or follows the outlines of the figures (7th century BC). Antikensamm, Berlin. 

15
Metallographic section (not chemically attached) of a granule and a fragment of foil, of an authentic fibula of Marsiliana d’Albegna, 7th century BC. The alloy is made of about 68% gold, 30.7% silver and 1.3% copper.

16
The diagram shows the copper concentration along the dotted line from A to C. Note that in the tangent point between sphere and foil, the copper content is about 15 times higher than the content in the core of the sphere.

      During the1980s, tools commonly used in the natural sciences were introduced in order to expose forgery in the field of ancient goldsmithing. It has now become common practice to use radiography, infrared, and electronic microscopy to probe into previously hidden corners in order to better understand the working methods and tools of the ancients.
      This new knowledge gave us valid means to uncover numerous forgeries that otherwise would not have been recognizable using only stylistic clues. It also gave us the means to reevaluate objects previously considered fakes (7).
      The first experiments made using the ancients’ “secret” soldering technique were realized in the 1920s by goldsmiths including Wilms, Treskow, and Littledale. Sixty more years went by before Parigini, Formigli, and Mello finally provided the scientific proof of the ancient colloidal soldering used on the Etruscan finds (8).
      This proof, obtained through a metallographic section of small authentic Etruscan granulation samples, brought the research into scientific territory. Laboratory research was able to prove that the percentage of copper in the tangent point between the granules and the surface was much higher than in other areas (9). There are still many problems that will need to be solved, but there never was a real secret. Numerous ancient literary sources would have provided us with detailed information if only we hadn’t ignored them and given in to romantic misrepresentation.

 17-18
Details of a spherical bead from an Etruscan necklace (6th century BC, Antikensammlung, Berlin). Granulation covers the background with floral figures, designed in the round. Looking at its actual size, it is almost impossible to distinguish single granules. The eye perceives granulation like a velvety surface, an effect we describe as fine-dust granulation. 

19
No other ancient or modern civilization produced granulated jewelry as refined as that of the Etruscans. The disk earring is an example of this. The piece belongs to the Antikensammlung of Berlin. It has a diameter of 4.6 cm. On the disk are attached knurled and round wires (some of which are only 0.15 mm thick), and minute gold spheres with a diameter of 0.07 mm. Despite these infinitesimal measures, each minute sphere is individually soldered to the foil.

Chrosocolla
The green gemstone we now call malachite was known in antiquity as chrysocolla. It is not particularly rare and was widely known in the ancient world. It is well known by any goldsmith or mineralogist as a mineral of emerald green color, at times with spots of blue. It is also known under the name of Eilat Stone, a name derived from the mines of Solomon at Eilath in what is now southern Israel. The ore was used to make copper. Chemically, it is basic copper carbonate (10) 
      Cu2(OH)2 CO3
The ancient name of this stone comes from the Greek chrysos (gold) and colla (glue).
      Glue of gold? Has somebody tried to explain the name of this mineral and its origins?
      If we go back about  5000 years. Malachite was a material already familiar to the Sumerians, who used it for coloring and as a cosmetic. This is confirmed by the funerary equipment of Queen Pu-Abi of Ur, by the citation in the epic poetry of Gilgamesh, and by entire facades of ancient temples, all of which were enriched with the blue Egyptian faience. This glaze, made from a combination of quartz powder, malachite, and sodium carbonate was used to adorn bricks used in Egyptian temples. From the Nanna temple to Ur, this new luxurious style spread all over Mesopotamia.
      We know that long before, colored inlay materials were produced by covering the plain mineral steatite with glassy vitreous paste. The ancients considered glass to be a kind of metal because it could be molded when hot and it hardened when it cooled. (Still today, the word metal in English technical language means glass paste) (11).

20
The so-called secret of Etruscan soldering art always had a name. The Codex Marcianus presents it to us with a symbology still valid today: the circle of the sun with one ray and, next to it, the writing that translates to “gold.” Gold that contains a large quantity of silver, “electro,” is connected to the symbol of the moon; note the symbol on the right. The material that keeps the noble metals together, the glue of gold, is represented with one circle and two rays on the bottom line. The name of the solder has never been a mystery.

21
Leech-fibula, from Tarquinia, with the most ancient granulation found in Etruria (8th century BC, Tarquinia Antiquarium).

22
Gold cup, from Palestrina (first half of the 7th century BC) with decorations in the shape of small sphinxes on the two handles. The figures are in the round, built with foil, and decorated using linear granulation (Villa Giulia Museum in Rome).

23
Cylinder of a comb-fibula from Marsiliana d’Albegna with very fine linear granulation. The granulation is used to draw meanders and zigzags (first half of the 7th century BC, Archeological Museum of Florence).

24
Detail of gold cup from Palestrina. 

25
Bracelet with embossed figures and linear and geometric granulation (circa 650 BC, Gregorian Etruscan Museum, Vatican).

      We can imagine that a goldsmith allowed this aqueous solution of malachite to come into contact with some bits of gold that were then heated on glowing coal. This person must have been very surprised to find out that the gold parts were “glued together” >in an almost invisible manner. This hypothesis could explain how colloidal soldering was invented. The secret that would soon win over all the ancient world had a name: chrysocolla.
      The art historian H.J. Wagner, who studied painting in Rome from 1910 to 1913, had the intuition that the ancient name chrysocolla had to have a technical origin (12). The writings of Pliny the Elder and other old sources confirmed it.
      “>With regard to medium stones, known as emeralds of Laconia, the biggest is located in Tyrus, where, in the temple of Heracles, there is a stele of large scale, which is not a fake emerald, as those that sometimes can be found [...]. Only a few of these can reach the size of a seal, but for the majority they are smaller and this is why they are used for soldering gold. Gold can be soldered using fake emerald or chrysocolla” (Theophrastus, De Lapidibus, VI, 11-24, 315-305 BC).
      Pliny the Elder (Natural History, pre-77 AD) describes the procedure of making verdigris, chemically identical to natural malachite, and also called chrysocolla by the ancients: copper foil was hung in barrels filled with vinegar. The green layer that gradually formed on the surface was then removed. If the preparation was not working well, then, according to Pliny, it was necessary to add “for two portions of vinegar, one portion of urine of a young boy.” Moreover, Pliny asserts that “even goldsmiths use chrysocolla, especially for soldering gold, and because of this use they named it so.” Chrysocolla is prepared by treating the verdigris from Cyprus with the urine of a young boy and Nitrum. The latter is also known as or similar to potassium nitrate (saltpeter), KNO3. This mixture can be used to solder gold alloys containing silver (13).
      In the Middle Ages, a dense cloud of mystery dimmed these exact definitions. Philosophical schools started debating the subject: (Turna Philosophorum, About the Nature of Chrysocolla, >XXX and XXXIV, 9th and 10th century AD), “[..] Bacse said: You spoke well, Diamede, but I don’t understand how you explained to posterity the behavior of chrysocolla. Envious people spoke many times about the glue of gold, making it obscure with all sorts of names [...] So I say to future researchers of this art that chrysocolla is a mixture that must be baked seven times and, when it is completed, it colors any body [...]  Get the real chrysocolla, which resembles rust copper, and mix it with calf’s urine [pseudonym of mercury], until the true nature of chrysocolla changes, since the true nature of chrysocolla is hidden in itself!” (14)

26
Detail of a band from the Treasure of Alisedo. Printed figures in a Phoenician-Punic style with granulated substrate (6th century BC, National Archeological Museum, Madrid)

Colloidal Soldering
Modern soldering procedures that use solder in the form of wire, sheet, and paste have a common defect—the quantity of solder cannot be reduced at will or beyond a certain limit. To say it another way, it is difficult to use very tiny amounts of solder. The most ancient method of metallic soldering is entirely different, and it is also the least known. The proper name for this technique is colloidal soldering.
This type of soldering combines all the useful advantages needed for the complicated granulation technique. Colloidal soldering differs from conventional soldering in that it uses a chemical solder dissolved in water; this means it can be diluted as much as desired, which in turn allows for an infinitesimal amount of soldering alloy.
Copper compounds, for example copper sulfate (Cu2SO4), are dissolved in a one-to-one ratio with gelatin in a little water and then placed in small quantities on the base sheet. Excess solution is removed with a bit of tissue paper or a brush. The minute spheres remain attached.
When the object is dry, it is placed in the fire. Our descriptions generally refer to embers of burning coal, which have a reducing action. Obviously, we can obtain the same result with a gas torch as long as we are careful to use a reducing flame. At first we see a black coloration between the granules that is created as the copper sulfate is reduced to copper oxide (CuO). This occurs around 100°C (212˚F). Around 600°C (1110°F ) the gelatin burns, releasing carbon that joins with oxides to vaporize as carbon dioxide (CO2) around 850°C (1560°F).

27
Clockwise(from topleft): 

Azurite 2(CuCO3 ·Cu(OH)2)
Malachite CuCO3 · Cu(OH)2
Cuprite Cu2O

Copper sulfate CuSO4

In this specific moment, an expert eye can see the surface shimmer like a glare that originates from a mirror, precisely in those places where the granules make contact with the sheet on which they rest. The goldsmith has only an instant to determine the correct moment to reduce the temperature. This is done by removing the flame or wetting the coal.

28
On the left: a modern earring with field granulation. Gold 900 and steel, the foil is 0.12 mm thick.

29
Above right: a gold disk with concentric waves in relief, linear and background granulation (5 cm dimeter), made by Mechtild Baumann (Colon, 1979, private collection).

The eutectic point in the copper/gold system falls at 889°C (1632°F) (15). At this point, the copper atoms at the points of contact begin to alloy with the gold, creating a tiny point of metal with a lower melting point than the metal around it. This becomes the fused joint. Besides avoiding the relatively large fillet, this differs from conventional modern soldering because if the metal thus joined is reheated, the connection only grows stronger. In conventional soldering, reheating risks having the joint fall apart. This benefit, a result of the fact that the process enables a small amount of gold-copper alloy to be formed specifically at the point of contact, allows repairing, improving, and working while constructing the object in stages.
This is why it is logical to suppose, for example, that objects like the comb fibulas were at first worked as flat sheets, then worked a second time after having been formed into a cylindrical shape.
With regard to the thickness of the sheet onto which the granules were attached, our research showed that particular skill was needed when the sheet was very thin. In our attempts using foils as thin as a tenth of a millimeter (as the Etruscans did), some small holes formed very close to the granules. This happened where the diffusion action spread through the depth of the sheet and reached the opposite surface. On the ancient objects that suffered this type of damage, we can see tiny patching or an accumulation of granules placed afterwards to cover the holes. The solution is that when using thin sheet, chemical soldering should be used in as limited an area as possible.
The Twentieth Century
In the 1920s, the ancient granulation technique was rediscovered, especially in Germany and England. In 1918, Marc Rosenberg published Dr. Stanger’s theory, which was not supported scientifically. The theory stated that an enrichment with gold-carbide could lower the melting point on the surface of granules:, something we now know to be a nonsense (16). The leading German goldsmiths of the time were Hans Jochen Wilms (Senior) from Munich, who used chemical colloidal soldering as early as 1920, and Elisabeth Treskow from Essen, who achieved fame in the history of modern goldsmithing. Both of them participated in the 1936 London Exhibit along with the major British experts, W.T. Blackband and H.A.P. Littledale. In 1934, Littledale patented the colloidal technique in England. This, along with his exposure through conferences and publications contributed to the dispersal of scientific information about granulation (17).
Like the Castellani family, Blackband devoted much energy experimenting with sophisticated modern techniques. It turned out that he was following an incorrect wrong route, he deserves mention here because he was one the few who strived to get closer to the solution through a variety of different techniques. He was a true researcher.
Contemporary goldsmiths who use granulation minimize risk by using tools and techniques that guarantee the best results possible. They use more pure alloys, and either a gas torch (which permits observation at close range), or an electric kiln with controls that help avoid the danger of melting the object.

The Technique
Despite many publications in the field of technical, archeological, and art-historical literature, many questions related to the ancient granulation technique still remain unresolved. Archeologists organize, restore, and preserve. Goldsmiths who brought this technique back to life started debates and discussions.
The granulation process is a complicated series of alternative procedures that are accomplished with various methods (18). Only a few sources give us information about the true ancient techniques. Modern archeometric investigations applied to authentic Etruscan material and experimental archeology allow us to see behind the scenes.
      The application of other methods during the twentieth century has brought the ancient technique back to life. An example is diffusion bonding in which joints can be made without the introduction of solder.

30
Possibly the ancients used small shells as containers for granules. By lightly rubbing a finger across the natural ridges of the shell, the granules split into rows according to their size. In this way the granules can be easily picked up with a wet brush.

31
Sketch of a crucible kiln model (used with charcoal), in which a temperature of circa 1200°C can be reached. In section, also the crucible for the production of granules is shown.

Production of Granules
The goldsmith Vannoccio di Biringuccio, of Siena, in his work De La Pyrotechnia, 1540, gave us the first description of the manufacturing of granules: “… melt the metal, place it in a wooden barrel [bigonzo: small barrel in inverted half-cone shape], or into a terra cotta vat, filled with clear and fresh water and other types of broken wood… stir little by little and lightly, and while the water is moving you will sprinkle [the molten metal] in (19).
      It is common practice today to create irregular balls of metal called shot through this same process. These small pebbles of metal are easier to measure and melt than other forms of metal, which is why they are used for casting and making ingots. We can imagine that this process was used in antiquity for the same reason, and that this was a simple, though crude way to make relatively small amounts of granules. In some late Etruscan work it is possible to observe irregular granules of this type, but in this case they could also be native gold.
      Benvenuto Cellini, the most famous of all the goldsmiths, describes in his marvelous Trattati of 1568 a procedure similar to the Biringuccio’s: “Also provide [metal] that quickly will change into granules. Get gold or silver that you want to change into granules and put it to melt, and when it is well melted, throw it in a little vase full of charcoal, and so you will make any sort of granules [...]”. The master Cellini was working at that time on a filigree piece and he tells us what he intends to do with its granules: “and after twisting the wire, according to your own will, little by little start placing it on the plate you want to work, and with a small soft brush with Draganti’s water >(a type of glue), one by one, wet the wires and the little and large  granules, so that, while you are working on the base of your work or on another part, the Dragani’s water keeps all the parts together.” (20)
      Any goldsmith knows that  when bits of metal are sufficiently heated they will melt and form little balls. To produce single granules of perfect shape and of the desired size, the ancient goldsmiths used a substrate of flat and smooth form, something that we would call a soldering block today. On this, using a blowpipe, he melted the gold particles, which draw naturally into minute spheres. By tilting the soldering pad, the minute spheres, still semi-molten, would roll down into a water bath where they solidified. Today this job is done with a gas torch, a tool that the ancients obviously didn’t know. They most probably used a blow-pipe. In use, one end of this thin tube is held in the mouth and the other passes over a heat source such as an oil lamp to focus a high temperature flame at a specific point.

32
To produce the granules, goldsmiths use small clippings of foil or wire. These are melted individually or in bunches until they form minute spheres. The ancients probably didn’t use a blow-pipe to melt fragments one by one. Instead, we can assume they developed a more efficient method to produce the great number of granules they needed.
      According to Jüngst, ancient goldsmiths only used a blow-pipe to blow on burning coal (21). In reality, the images of the antique iconography portray goldsmiths with blow-pipes and fire tongs, but without other instruments. It is conceivable that the tongs were needed to hold a piece of burning coil whose flame was diverted by the air of the blow-pipe.
      Much more practical is the method used to produce a large quantity of granules. This method is still practiced today and still uses pretty much the same tools that were employed in ancient times. This method uses charcoal powder and a ceramic crucible.
      The first step is to clip small pieces of gold and it appears that the shape of these pieces influences the production of the granules. When gold (or most other metals) starts to melt, the grains that were organized into a crystalline structure become mobile. At a certain point, the crystals lose their place in the metallic structure. The gold starts to deform and, if heated to suffi cient temperature, to liquefy. This process begins on the surface and in peripheral areas such as corners, where the atoms are less bound to the metallic structure. Liquid metal, because of its high surface tension, tends to take the shape of the smallest surface—in other words, it draws into a sphere.
      This process of transformation happens more easily when the initial form is similar to the final one. A short cylinder, for example, is closer to the shape of a sphere than a flat bit of sheet. When making granules, pieces of round wire are more appropriate than clippings of sheet.
      In his research on authentic Etruscan sites, the restorer Hans Ulrich Tietz (22) of Berlin, recently discovered small clippings of round wire that hadn’t melted and had escaped the attention of the ancient goldsmith. Tietz also calculated that their volume corresponded to that of the surrounding granules.
      We know that Etruscan goldsmiths used a fairly complicated method to produce gold wires. It follows that they wouldn’t have gone to the trouble of making wires as the first step in granule production if they felt they could get the same result by clipping sheet. In fact, thin sheet (or foil) was used in the initial phase of the production of the wires.
      The Etruscan goldsmiths started the process of making granules by cutting a thin strip from a fine gold foil. This strip was then twisted and rolled between two smooth plates to form a wire of round section. Wires produced in this way still show helical marks (23). Making a piece of wire ten centimeters long requires between five and ten minutes of work, depending on the artisan’s ability.

33
When heated to melting temperature, metal clippings of either foil or wire tend to draw up into a sphere, the smallest surface possible. If two or more fragments are in contact with each-other they will form a single sphere.

      These wires could be cut into small segments, individually or in little bunches. The diameter of the wires also determines the limit of the size of the granules. In the fine-dust granulation practised by the Etruscans, the granules average an astounding 0.12 mm in diameter. In this case it is better to use foils very thinly forged or the smallest fragments of native gold (i.e., gold dust) found in rivers.
      Wolters supposes that in the 7th century BC the initial material for making Etruscan fine-dust granulation were gold filings (24). Until now, however, filings were not found in authentic ancient granulation. Only on forgeries of the last century can we find filings, apparently used as soldering material (25).

34
Detail of a granulated sphere of an Etruscan necklace, photographed with electronic microscopy (Antikensammlung, Berlin, photo from Rathgen Forschungslabor). Note the cylindrical fragment: This is one of the initial wire clippings, which, during the production of granules, didn’t melt. The sphere above, initially, was bigger than the others, but it was worn down over the years as the necklace was worn.

35
One method used to produce granules is to throw molten metal into a container filled with water as it is stirred with a ladle. This process produces granules of various sizes and of irregular shapes. This type of granulated material was used during the Middle Ages and the Renaissance as decorative elements on filigree pieces.

36 & 37
Etruscans probably used round wire as the starting material for the production of granules. Wires were made by rolling and pressing thin strips of foil, a process revealed under close inspection by one or more helical grooves. Minute spheres of similar diameter were obtained by using clippings of equal segments of this type of wire. 

38
Etruscan goldsmiths were able to produce granules smaller than a tenth of a millimeter.

39
In ancient granulation we sometimes find irregular shapes caused by an incomplete fusion of the initial fragment or the barbell shape that occurs when two granules are close enough together that they join.

The Process of Melting Granules
In a ceramic crucible the size of a small drinking glass, spread a thin layer of charcoal dust, then sprinkle gold particles onto this, distributing them evenly across the charcoal. The gold pieces, which shouldn’t touch each other, are held between the index finger and the thumb and dropped on the charcoal as we would sprinkle salt in our kitchens. A new layer of charcoal is sifted over the gold and the process is repeated. Continuing in this way, the crucible is filled up. Experience taught us the importance of pressing each layer lightly and cautiously to remove hollows that allow the gold pieces to shift. When two or more pieces touch, they are likely to melt into larger granules. When filled, the crucible is covered and heated to about 1100°C (2000°F). In our research we used a charcoal fire but in modern studios this is typically done in a kiln.

40
Etruscans probably used the method of alternating layers of gold snips and charcoal dust to obtain a large quantity of granules at the same time. They had the right tools for this method: crucibles, coal, kilns, and sieves.

41
In the center of this photograph (taken with the electron microscope at the Research Institute of Rathgen, in Berlin), it is possible to see two granules almost fused together. Shapes of this type can sometimes be seen among the granules produced inside charcoal dust.

42
This cross section illustration shows how the crucible was prepared, with gold flakes sprinkled between layers of charcoal.

Let’s now observe what happens inside the crucible. The lid reduces the entry of oxygen into the vessel as the heat slowly penetrates the inside of the crucible. A light oxidation of the coal dust is not entirely negative, however, because it can lighten the compactness of the material around the minute spheres that are forming. At this stage is it critical that movement of the crucible be avoided because the granules are at risk of dropping down and fusing into the layer beneath. If this were to happen, bigger or irregular granules of undesirable shape would form.
Even when done correctly, a couple of “odd” granules can always be found, born from two contiguous granules or from a longer clipping of wire. If the crucible is left in the furnace long enough, these irregular shapes will eventually draw up into a sphere. This explains why a few large granules are almost always found when using the crucible method.
What happens during the following phase of work is always stressful. When the crucible is allowed to cool and the lid is removed, all that is visible is the charcoal dust. Are the granules round? Do they have the desired smooth surface? We can’t tell until the gold has been filtered from the charcoal.
Especially in case of fine granules, the metal contents of the crucible are separated from the charcoal dust using strainers in the same way that gold diggers pan for flakes of gold in a stream. It is advisable to rinse the gold/charcoal mixture several times with fresh water because the finer granules float on top of the dusty surface of the charcoal laden water. Etruscans probably used rainwater.
When the charcoal has been completely cleaned away, the granules are dried with heat or placed on an absorbent surface then sorted according to size.

Washing and Sorting the Granules
Selection of the Granules
There are several ways to sort granules and in the end each artisan will find the method best suited to his tastes. First, granules of irregular shape must be separated from those that are perfectly round. The best way to do this is by letting the granules sort themselves spontaneously. To accomplish this, lay all the granules onto a smooth board. This board is then slightly inclined, so that all the round spheres can roll into one corner, while all the others remain in place. The application of this method was easy for the ancients. They used a marble board and a piece of leather for collecting the spheres.
After this, the granules must be sorted according to their size. It is possible to sort manually with tweezers but this method quickly becomes impractical when large numbers of granules are required. A second method uses a stack of filters or sieves with holes of various size, ranging, for instance, from one millimeter to a tenth of a millimeter. A collection of granules is poured into the top and the stack is shaken until all the granules have found the level where they no longer fit through the holes. The smallest ones are collected on the bottom. We don’t know if the ancients used precise filtering systems. In many museum collections we can see Etruscan kitchen strainers that are very precise.

43
The content of the crucible after the melting process.

44
Experimental reconstruction: granules, of two different sizes—circa 0.1 mm and circa 0.4 mm in diameter—have been arranged on a hemispherical surface and adhered with the mixture of glue and copper salt before the chemical soldering. 

45
Today the simplest and most efficient method for sorting granules is by using a stack of strainers with holes of decreasing size. 

It is not proven, but it is conceivable that they also used a different selecting method. Instead of strainers, this method used a small assortment of bone or horn combs. The biggest granules could have been raked out of a pile of granules spread onto a smooth surface, the process then being repeated by combs of decreasing size.
Another method is even simpler than this last one, and it was the one we used in our experiments. Granules of mixed sizes are placed in the corner of a marble board. By setting another board on top of this pile, larger granules are held between the boards while smaller ones are free to roll away when the boards are tilted. The process is repeated several times, each time gathering the spheres that are the largest remaining in the mixture (see picture 46). This method has the advantage of obtaining bunches of granules of the same diameter. This can be useful for realizing a row of single minute spheres or other specific geometric leitmotifs.

46
It is also possible to select granules by using a flat board that makes contact and holds only the largest and same-size spheres as others are removed by tilting the board.

Actual size

47
Experimental reconstruction: this image of a griffin was made by the authors using techniques and materials available to the ancient Etruscans. The small inset shows the actual size of the piece.

48
Granules are arranged with a wet brush (size #00) and placed on the foil individually (or at the most two by two). Another view of this piece is given on pages 56–57.

Placing the Granules on the Base
It is not our intention here to give an artistic evaluation of the esthetic value of jewelry. However we must note that granulation has always been a personally expressive medium, like graphics and painting. The ancient granulated jewelry comes back to life in a new dimension when the creativity of the contemporary observer is able to evoke the personality of the artist of the past. Perhaps this is a privilege reserved only for those in the position to have direct contact with the original piece of jewelry. This contact is further enhanced by sophisticated optical tools that allow total immersion into “the world of the thousand golden spheres.”
Of course the ancient artisans who created these wonders had no microscopes or microprobes. Their good eyes, steady hands, and patient concentration made their incredible work possible. As miraculous as it seems, in reality the practical tools they used are strikingly simple.
The material used to temporarily attach the granules needed to be organic so it would carbonize during the heating process and burn out without leaving any residue. In other words, the glue served two purposes, first as adhesive and then, after having been carbonized, as reducing agent.
Mediterranean people were familiar with the gelatin that forms when cooking fish. This glue met the requirements of granulation. Artisans not living near a supply of fish were equally familiar with similar variations of vegetal and animal provenance. Cellini, in his Trattati >mentions an adhesive he calls the Water of Dragante. Our experiments were done by using the sap of a fig tree; even human saliva will work.
The glue material is mixed with water (typically the copper salt is also added). We discovered that the density of this mixture is critical to the success of the granulation. If the mixture is too dense, it prevents the granules from getting next to each other and it hardens too quickly. On the other hand, if there is insufficient gelatin, the granules will not be securely held in position.
An alternate method is to add the copper salt after the gluing agent is completely dry. In this case it is preferable to use a water-soluble salt such as copper sulfate, CuSO4. A sable brush (in a small size such as #00), is dampened in the solution and lightly touched to the gold. Capillary action draws the liquid solution between the granules, providing a balanced distribution of the copper salt (e.g., chrysocolla). When malachite is used as the soldering agent this method is not recommended. Malachite is pasty and because it is not water soluble, it behaves differently.

49
Field granulation (as compared to linear granulation) allows a goldsmith to spread many granules onto the base sheet at the same time. Granules are carried to the work with a wet brush. All granules are held together on the foil by a liquid film. It looks as if they are attracted by an invisible magnet.
We should note that there is a different quantity of copper in various salts. Even at the risk of having a few granules fall off, it is preferable to have a low level of copper concentration so that the points of fusion are almost invisible. If there is too much copper, the spheres can be absorbed by the substrate so much that they appear to sink into the base sheet. Too much copper can also cause holes to form. In the worst case scenario, lowering the quantity of copper may cause gaps between granules. These, of course, can later be filled by adding more granules.
Anyone who tries to spread granules onto a sheet will quickly become aware of certain physical phenomena. The minute spheres, still surrounded by liquid, show properties similar to those of soap bubbles. They can draw close to each other in more or less predictable ways and often tightly adhere to each other. When this happens, three spheres naturally pull together to form a stable triangle. This explains why this triangular pattern appears frequently in all the cultures that used granulation as a form of decoration.

50
Cleaning is an essential part of granulation. The gelatin is prepared in a small container and mixed with pulverized copper sulfate. To wet the granules, the brush needs to be cleaned and soaked in water, and it must be rinsed each time.
Using a brush that is barely wet and continuously cleaned in the glue-water solution, we found that it is possible to spread granules into free shapes, almost like painting. The glue, already distributed in drops on the foil, holds the minute spheres when the brush is raised.
When the goal is to draw lines with granules, a different method is used. With a thin brush, the glue is applied to the base surface for a length of only a few millimeters, then granules are gently coaxed into place as the solution dries. This process moves slowly, with only two or three granules being placed at a time. Especially in the case of repeated geometric configurations and straight and regular rows (as we can see on so much Etruscan jewelry), it is advisable to use granules of the same size. In many pieces of Etruscan granulation we see single rows of granules placed on top of relief figures, created with the repoussé >technique or printed on fine gold leaf. These figures often portray animals, imaginary creatures or floral leitmotifs. Details, such as tails, eyes, and manes can be designed ex-novo by the granules themselves. Granules can also underline parts to provide greater emphasis. In this case they follow the figure’s hollowed rows, outlines or folds.
On flat surfaces we are more likely to find geometric decorations such as circles, zigzags, and meanders. Preparatory drawings were not found in any of these cases, even when the configurations were very complicated. However, we cannot confirm that those patterns were not previously painted with a brush and chrysocolla (which is colored and doesn’t leave traces).
So far only in one case was an engraved line discovered, double in certain points and running horizontally along the cylinder of a comb-fibula (Villa Giulia Museum). In this case we conjecture that this is a baseline to assist in placing the granules in a zigzag pattern rather than an ornamental line for its own purposes. After this specific portion, the granules were freely placed, without a pre-established pattern (fig. 52) (26). Even if the drawing is very precise and extremely complex, especially in the areas with meanders, creating these patterns was well within the talents of the ancient goldsmiths.
Timing and concentration were different for the ancient goldsmith in comparison with today’s standards. Technically, working for weeks at the same fibula was irrelevant. Castellani suggested that this type of work should be done by the sensible hands of young women. Today, perhaps we are admiring the skills of Etruscan goldsmiths’ daughters…

51
The attempt of using granulation on a foil 0.1 mm thick, as Etruscans often did, proved that small holes can easily form, even when the quantity of chemical solder is reduced to the minimum. On this very fine half-moon shaped foil, a few typical Etruscan elements were experimentally soldered by using chrysocolla. On the lower left, a few tiny holes formed.

52 & 53
The cylinder of a comb-fibula from Palestrina (Villa Giulia Museum, Rome) has a granulated decoration with zigzag leitmotif in the center. In the enlargement, a horizontal mark is clearly visible. So far, this is the only example of “preventive” drawing. This line was needed for aligning the vertexes of the leitmotif.

Flux
The minute diffusion that is critical to granulation can be achieved only if an alloy of lower melting point is formed at the point where the granules touch the base surface and each other. As with conventional soldering, the temperature at which this process occurs is also the temperature at which copper oxides form — oxides that will prevent proper joining. The reducing action of a charcoal fire, at least in part, prevents this from happening but something more is needed. Even 5000 years ago, goldsmiths recognized the need for a chemical agent to mitigate oxidation, an agent we have come to know as flux.
As with the reducing atmosphere of a charcoal fire, flux is also limited in its ability to prevent the formation of oxides. That said, to this day all goldsmiths use flux when soldering, even when working in a reducing environment.
Why is a flux needed in the colloidal soldering of granulation?
• It must absorb the oxides forming on the surface to permit an immediate contact between the metallic copper of the chrysocolla and the gold.
• It must protect the area from absorbing oxygen.
• It must lower the surface tension of the liquid copper-gold film that is forming.
• In an ideal case, it must melt 50°C before the chemical solder, allowing the chemical solder to take its place when reaching the temperature of 889°C (1632° F).

For the second time we find an “historical alliance” between two chemicals. We already know them: malachite and soda, between the “glue of gold” and flux.
Soda was already known in Egypt in the third millennia BC, and was widely avalable. As mentioned previously, soda was widely used to produce faience, glass, and detergents.
Chemically, soda is sodium carbonate (Na2CO3), a powder that melts at 850°C and it reduces copper oxide in carbonate. The flame becomes yellow during the oxidation of sodium, which prevents a precise observation of the soldering process.
The chemical reaction is the following:

Na2CO3 + CuO CuCO3 + Na2O

The sodium carbonate combines with the copper oxide to create copper carbonate and sodium oxide. The copper carbonate, which is the chemical solder, is reduced to metallic copper at the point of contact, creating an alloy with a lower melting point than the parent metals.

54 & 55
When using a gas torch, soldering happens in the reducing area of the flame. A flux can improve the reducing action.

Herodotus, circa 450 BC, describes the use of soda and Pliny the Elder describes specifically the use of nitrum (crude soda ash) for the soldering of gold. Together with the urine of a young boy and chrysocolla, it is possible to obtain the famous “santerna”, which name probably has Etruscan origin. Benvenuto Cellini, 1500 years later, regarded this recipe highly.
With further experiments we should check to see if the wetting action of soda could make it possible to reduce the percentage of chemical flux even more. In principle, this wetting action could help in the absorption of the granules by the substrate. This is why it seems more likely that granules connected to the base with a small attaching neck were joined without solder. In fact, a higher surface tension keeps the granules “raised,” leaving them more isolated and distinct on the surface.

Soldering on Charcoal
We may ask ourselves if it makes sense to be concerned today with ancient studio methods, particularly when we have already answered the question of how granulation was achieved. It appears that we have discovered the “secret” of chrysocolla and we now understand the ancient technique. But, do we really know it?
Learning the chemical soldering process is like having a key in our hand. Now we must use this key to shed light into obscure areas. In other words, our answers lead to many further questions.
For instance, were other techniques employed in addition to the main process described above? (27)
Also, in which conditions and with which tools did the ancients solder? If we think about the complicated Etruscan granulation of the Orientalizing age, this becomes an important problem.
These questions will immediately intrigue artisans, who think in a practical way and deal with these specific problems every day in the studio. Among the academics and historians who studied granulation, however, almost no one was concerned with such matters.
Let’s listen again to the person who became our link between antiquity and modern era, Benvenuto Cellini. In his Trattati we read a precise description of soldering on charcoal, as it was done at least until the 17th century and perhaps later:

“[...] so he started soldering with heat, which is done in this way. Take a little verdigris [chrysocolla] from its virgin pig [ingot], because it cannot have been used for anything else, and from this, wanting to solder similar works, you must take as much as a home-grown walnut without the hull, and with this you stir together a sixth part of Ammon’s salt, and an equal [part of] borax [.] Each thing, being ground together, then becomes liquid, in a little glass bowl, with a little pure and very clean water and with the above mentioned compound of powdered verdigris. [The solution] becomes liquid, with a color [similar to those used] for painting [.] From this [solution] take some with a small twig, Caradosso layered it quite thick on the joints [...], and on top of the above mentioned verdigris, he put, with his borax container, a little borax, ground to a fine powder. Then, he lights the fire with fresh charcoal, never burned before, he puts his work on the fire, adjusting the coals, turning them toward where he wanted to solder, blowing on the embers to raise the heat. After doing so, he arranged his work on a few coals in shape of a small grid, making sure, when his work had become the same color as the fire, to start blowing dexterously with the bellows on the charcoal so that the flames folded on his work. If the wind were to be too strong, the flames would open up and go outside, bringing the danger that the work could get ruined and damaged. Therefore, he, with diligence, controlling himself, as soon as he started to see the first layer of the gold flashing and moving [decisive moment for the diffusion of the copper metal into the surface of the gold!] quickly, with a little bristle infused in a little bit of water, sprayed on his work. In this way, the work was extremely well joined without soldering [meaning, without using metallic solder]. Since, the first time, he had soldered his work with heat, this joining method is not called soldering, but it is a transformation of all the work into one piece, because the virtue of verdigris accompanied by Ammon’s salt and borax is so great, that only the skin [superficial film] of gold moves. With the joining, gold becomes hard and whole. After doing this, Caradosso puts his work in very strong white vinegar, adds a little bit of salt, and leaves his work in it for an entire night. The result is that in the morning, [the work] is whitened and clean of borax [...](28).

56
Experiments using chemical solder were made in ceramic kilns built with soil from Murlo, a region in northeastern Italy. In the princely residence of Poggio Civitate in Murlo there was a shop where small objects of precious material were made (besides ceramic and bronze work).

Cellini’s description deserves special attention. It contains almost all the details of the soldering methods, although, for a contemporary observer, there are still some mysteries.
Let’s start from the description of the hearth and charcoal.
Fireplaces are still used today for heating and cooking in some rural areas of Italy. However, the charcoal that today is available in stores for barbecuing wouldn’t be the right kind, since it is already fragmented into little pieces and cannot be used in layers, the way Cellini is teaching us.
The pieces of beech and holm oak, completely carbonized, must be placed concentrically, so they form an air draft around the gold piece that comes from outside towards the inside of the structure. However, in this way, the piece would be irregularly heated because the upward convection would cause a loss of heat.
This is why Cellini builds a “tent” of coal over the jewelry; not only able to reflect heat, but also to become, itself, a heat source. At the same time, a reducing atmosphere is created all around the object. We know that this reducing action is the premise of the chemical soldering process.
A second phenomenon, observed and masterfully described by Cellini, is the “flashing” of the surface. We know that this is caused by the reduction of copper salt to copper metal. Just for a moment, the surface becomes like a mirror, covered with a liquid, glittering film. This is the beginning moment of the surface or point-of-contact alloying process of copper with gold. If the temperature is not immediately lowered, the object could melt. The temperature is lowered, as Cellini says, by sprinkling water with a small brush.
We have experimentally duplicated this process, applying it to a typical modern piece of jewelry, with foil 0.3-0.4 mm thick. The observation of the “critical point” >was difficult during the experiment. The heat of the burning coals made a precise observation from up close impossible. Therefore, in order to observe the work we used a screen with a hole.
Other attempts with foils 0.15 mm thick, as used in Etruscan age, were not satisfactory. The small movements of the embers that happened naturally as the charcoal burned caused the work to slip and led to a partial loss of the granules. In the areas where we had put copper salt, the foil (base sheet) became thin and even developed some holes.
Perhaps certain particularly delicate objects should use different solutions and specific precautions. This could be the case for jewelry with the following characteristics:
1. Objects built with extremely fine wires and foils. Note that these require close observation because the moment of soldering must be instantly seen to immediately start the cooling process.
2. Objects that, for the long preparation required, cannot undergo a risky soldering.
3. Objects with circular granulation, because they would lose granules in the embers at the slightest jarring.

57
In this kiln with a short chimney it is possible to reach temperatures around 1200°C with intake of air on burning charcoal.

Soldering in a Coal Kiln
Not much information comes from antiquity, so we are are forced to advance a hypothesis without the benefit of direct evidence.
The historical problem of the granulation technique was recognized and well described by Wolters (29). However, the shape and the function of the ancient goldsmiths’ >kilns represent another topic, with its own complex problems.
Some fragments of Attic vases show a mythologic iconography that is useful to understand the Etruscan age (30). Here, for example, you can see Hephaistos, the supervising god of metallurgic activities, carrying the bellows used for poking the fire in the metalsmiths’ and smelters’ >ovens (built at eye level). Hephaistos also appears in literature, e.g. in the 18th canto of the Iliad:

“ ... and he went back to the bellows: he turned them on the fire, he invited them to work: and the bellows, all twenty of them, blew on the furnaces.”

Egyptian parietal figures, much more ancient, show images that can be interpreted as melting and soldering processes. Four people with bellows are blowing air on the burning embers of a “crucible” kiln. Such a kiln could very well serve our purposes: melting the granules buried in charcoal dust.
Using hardwood charcoal, it is possible to reach temperatures around 1500°C.
In the Antiquarium of Murlo, where we can see Etruscan findings from Poggio Civitate from the 7th and 6th centuries, there are numerous little kitchen ovens. Recent excavations also revealed the ruins of a shop. Among the fragments there were ovens similar in shape to those used in kitchens. Probably they were used for melting metal that was placed inside small crucibles.
The soil of Murlo (Sienna, Tuscany, Italy) that we used during our reconstructioned kilns contains all those components that made it suitable to be used by the Etruscans as primary material for the construction of their ovens.
The crucible kiln, which is open in the upper part, is not suitable for soldering granules. The heat, raising up from the kiln, would prevent a close observation and would make working with pliers very difficult. Moreover, the free access of air from the top would limit the formation of that reducing atmosphere needed for our operations. Also, in these conditions, an object decorated in the round could not be soldered homogeneously.
Some Egyptian parietal images, for example those of the Rekhmara’s tomb, in Thebes (circa 1475 BC), show a kiln that has the important advantage of being accessible from the front. The basin that holds the charcoal is placed on top of a support that could also be used as an ash bin. The fire is on top of a supporting shelf (which could be perforated) and is protected from the wind on three sides by walls. These barriers also serve as heat reflectors. The goldsmith, seated in front of the fire on a low, three-legged stool, pokes the coals with a pointed blow-pipe nozzle, probably made of ceramic material.
This method of working compares favorably to the one described by Cellini. The kiln could be used for different types of goldsmithing, from simple soldering to the annealing of embossed foils. The delicate types of soldering we have described require an auxiliary construction to avoid loss of heat upward and out the front. Such a modified kiln can be easily developed by building a kiln in a shape of a dome. This is a type shown in some Egyptian representations.
We reconstructed a few types of kiln and we were able to verify that the most important and frequent problem was the shifting of the charcoal as it burned.

58
Reconstruction of a crucible-kiln and of bellows with triangular openings, made by using ceramic fragments found in Poggio Civitate. The sheepskin bellows were recreated after the iconography of Attic vases. This equipment can also be used to make gold granules.
As we know, the crucible containing the gold fragments, which is placed among the layers of burning embers, requires an absolutely stable position. Therefore, we see that it cannot rest directly on the coals, but must instead have a steady and secure base.

59
There are no Greek or Etruscan images of professional kilns. We must go back 1000 years, to Thebes, in Egypt, to find them. This kiln, painted in the Rekhmana’s tomb, could have been multipurpose. The work is done on the front, using a blow-pipe nozzle. The “brazier” has protection against wind and it is probably perforated to facilitate ash elimination. Heat loss inside this kiln is high.

Covering the working surface with a “tent” of coal (see image on page 80) was an ideal solution. In fact, even the covering itself, after a short time, becomes incandescent on the inside, facilitating the formation of a heat dome around the object. In this way, a uniform heat radiation is guaranteed even on cylinders, spheres, and similar rounded objects, including those of complex shape, such as some Etruscan pendants and earrings. With this expedient, the regularity of the reducing atmosphere around the object is improved. Moreover, the goldsmith can benefit from a close observation because the heat is more or less contained.

60
A little dome-kiln is used for melting. Five Egyptian workers are pressing on the bellows while flames are coming out from the dome.

61
Experimental reconstruction of a dome kiln for granulation. We found that this kiln suffered little heat loss and allowed good visibility of the object.

Ideally the coal pieces that form the tent should incinerate very slowly. This is why it is important to use hardwood charcoal. Moreover, they must be prepared in square and smooth shapes and they must be arranged one on top of the other, like a house of cards. The tent can be prepared in relation to the shape and size of the object.
A ceramic muffle kiln has the same function as the coal tent. As with the charcoal structure, a muffle kiln provides good heat distribution. It is particularly useful for very large objects such as the long cylinders of certain Etruscan comb-fibulas (see image on page 19). In this case, the object is arranged inside the muffle kiln. The muffle can be put into and removed from the fire without ever touching the jewelry.

62 
In this tall kiln, it is possible to work on two projects at the same time. The crucible is placed inside the cylindrical base tomake the granules, while soldering can be done on the covering section.

During our experiments, we also realized that it is possible (empirically and in a simple way) to solve a common problem: How were the ancients able to verify that soldering had occurred? The temperature at which the copper/gold alloy is formed when using the chemical soldering technique is approximately 900°C. In this temperature range goldsmiths must be very careful not to melt their work. If we place a piece of foil, also treated with copper-salt next to the jewelry, it is possible to check very well the moment when it starts flashing (to use Cellini’s term). Also, in the old times they knew well the melting points of certain substances. For example, we can place a lump of soda next to the object knowing that soda starts to melt at 850°C (1562° >F). This is the right moment to get ready to sprinkle the object with a brush loaded with water.
The dome also has another advantage: the steam that is formed by spritizing the object with water increases the reducing action and, at the same time, prevents the access of oxygen. Therefore, the object can be removed from the kiln while it is still shiny, with no oxidation.

63
The use of a small ceramic muffle requires a higher initial quantity of heat, but it guarantees better heat distribution.

Ultimately, it is important to remember that both the ancient and the modern methods are two aspects of the same discipline. What we learn from the ancients shouldn’t be used to imitate their forms, but should be creatively reused. The Florentine master, in front of Pope Clement the VII, refusing to copy the ancient examples, said that it was better to seek a new path than to attempt to match Etruscan metalsmiths. Trying to compete with them, in fact, would only demonstrate that we are clumsy imitators (31).

64
These sketches show the design of a dual function kiln: The crucible for the preparation of granules, which is placed down below in the hottest area (circa 1100°C), and its cover, which is used as working space for soldering (A). Probably, small objects could be soldered directly in a sort of muffle, built with large pieces of charcoal (B). 

Glossary
Blowpipe: >A thin tube, often tapered, through which a goldsmith blows air to increase the intensity of heat when soldering, melting metal, or stoking the coals. Air is blown by mouth through the flame of a candle or a portable oil lamp to create another flame of higher temperature that can be directed toward an object.

Castellani: Family of Roman goldsmiths active in the 1800s. They imitated Etruscan jewelry and were collectors of Etruscan antiquities.

Chrysocolla: >This ancient term from the Greek chrysos (gold) and kolla (glue) was mainly used in antiquity to indicate the stone we know today as malachite, but it might also describe other copper mixtures such as azurite, cupric oxide, and copper sulfate. All of these can be used as chemical solders. The mineralogical term for chrysocolla is copper silicate hydrate.

Crucible: >Vessel made of fire-clay or graphite used to melt metals.

Fibula: >Ancient shape of pin, decorated and provided with spring, arch and bracket used to clutch sections of clothing together. Leech-fibulas are a particular type with a bigger arch in the shape of a leech.

Filigree: From the Latin filum (thread) and granum (wheat). It is a decorative technique in which gold and silver threads are soldered onto a substrate or to each other.

Granulation: >from the Latin granum (wheat). A decorative technique that consists of attaching minute spheres onto a surface or to one another.

Silhouette granulation: Granulation with figures filled with granules, without any outline made of thread.

Fine-powder granulation: Granulation using extremely small gold granules.

Hephaestus: >Greek God of metallurgists.

Archaic Age: 6th century BC; before the Classic Age.

Classic Age: 5th-4th century BC; before the Hellenistic Age.

Orientalizing Age: 7th century BC; after the Iron Age and before the Archaic Age.

Domed Oven: In the context of this book, this is a clay structure in the shape of a dome that is fired with charcoal.

Littledale, H.A.P.: A noted British goldsmith who made experiments with ancient soldering techniques. In 1934, in England, he patented the technique that uses copper salts to create point-of-contact alloying.

Malachite: see Chrysocolla.

Minoan: >Cretan culture of the Bronze Age (3000-1100 B.C.).

Murlo: >Small medieval town south of Sienna. Etruscan excavations started here in 1966. There is an Etruscan Museum with important archeological findings from the princely residence. Seminars about experimental archeology are held here.

Disk earrings: Circular earrings, often with concentric decoration.

Gold: From the Latin aurum. An elemental metal long used for jewelry. Pure gold has a melting point of 1063°C (1945°F).

Native gold: Nuggets or flakes of gold found in the sand of rivers.

Paglione: From the Latin palea. Small clippings of soldering alloy. These are sometimes known by the French word paillion, but in US English more commonly called chips.

Punches: Steel rods with a points of various shapes. They are struck with a hammer to make ornamental marks and textures on metal.

Colloidal soldering: Chemical soldering of gold or silver that uses a copper salt. The copper mixture is reduced to metallic copper that bonds with the surface layer of gold, forming an alloy with a lower melting point. This process is also called copper salt soldering and is a type of diffusion bonding.

Embossing: >A decorative technique in which sheets of metal are formed with punches worked on both the front and back to create figures, patterns, and ornaments. Semi-rigid materials such as pitch, wax, and lead are used as a supporting base.

Steatite: >A white or greenish variety of talc.

Crystalline structure: Microscopic structures formed by individual metal atoms. For gold, silver and copper this structure as a cubic shape. The similarity of their crystal structure is one of the reasons these metals join together easily.

Sumerians: >Inhabitants of the Southern part of ancient Mesopotamia.

Theophrastus >(372-287 A.C.): A Greek philosopher of Lesbos who was a follower of Aristotle. He wrote History of Plants, the best botanical treaty of Antiquity.

Treskow, E.: German goldsmith of Colon promoted ancient granulation in the 1920s.

Ur: >Sumerian city, founded around 4000 B.C.

Wilm, J. M.: German goldsmith of Munich. In 1920 he used granulation with colloidal soldering.

Bibliography
1540 V. Di Biringuccio, De la Pyrotechnia Libri X, Venezia.
1568 B. Cellini, Due Trattati: Uno intorno alle principali Arti della Oreficeria…, Firenze.
1862 A. Castellani, Antique Jewellery and its Revival, London.
1915 C. Densmore Curtis, "Ancient Granulated Jewellery of the VIIth Century and earlier," Memoirs of the American Academy in Rome, vol. I, pp. 63-85.
1918 M. Rosenberg, Geschichte der Goldschmiedekunst auf technischer Grundalage II, Granulation, Francoforte s.M.
1936 H.A.P. Littledale, A New Process of Hard-Soldering and its Possible Connection with the Methods Used by the Ancient Greeks and Etruscans, Lecture to the Worshipful of Goldsmiths 24th Feb. 1936, University Microfilms International, 1981.
1951 E. Frey, "Das Prinzip der Technik des Granulierens," Deutsche Goldschmiedezeitung, >8, pp. 207.
1951 J. Schneider, "Zur Granulation," Deutsche Goldschmiedezeitung 12, Stoccarda, pp. 318-319.
1952 G. Piccardi, "Sulla oreficeria granulata," Studi Etruschi 22, pp. 199-202.
1952 F. Chlebecek, "Beitrag zur Technik der Granulation," Studi Etruschi >22, pp. 203-205.
1954 H. Maryon, H. Plenderleith, "Fine Metalwork," A History of Technology II, London.
1954 G. Piccardi, S. Bordi, "Sull'oreficeria granulata etrusca,” Studi Etruschi 24, pp. 353-363.
1955 G. Becatti, Oreficerie antiche dalle minoiche all barbariche, Rome.
1956 E. Tresko, Kunst und Leben der Etrusker, Mostra di Colonia, 1956, pp. 46 ss.
1956 E. Coche de La Ferté, Les bijoux antiques, >Paris.
1961 R.A. Higgins, Greek and Roman Jewellery, London, pp. 19-23.
1966 H. Hoffmann, P.F. Davidson, Greek Gold. Jewelry from the Age of Alexander, Magonza.
1968 M. Trokay, Les origins du Décor à granulation dans l'orfévrerie egyptienne, Chron. d'Egypte, pp. 43, 85, 271.
1969 Diebeners, Handbuch des Goldschmieds, vol. III, Stoccarda.
1971 C. Aldred, The Jewels of the Pharaohs, Londra.
1973 A. Thouvenin, "La soudure dans la construction des oeuveres d'orfévrerie antique et ancienne," Revue Archéologique de l'est et du centre-est >91, XXIV, PP. 11-69.
1973 E. Brepohl, Theory and Practice of Goldsmithing, Maine.
1974 D.L. Caroll, "A Classification for Granulation in Ancient Metalwork," American Journal of Archaeology 78.
1980 H. Jüngst, Zu den Anfängen des Hartlötens, Symposium: Historische Technologie der Edelmetalle, Ludwigsburg.
1980 W.A. Oddy, N.D. Meeks, J.M. Ogden, The Scientific Examination of a Phoenician Earring, Symposium: Historische Technologie der Edelmetalle, Ludwigsburg.
1980 J. Wolters, Literary Sources on Granulation, Symposium: Historische Technologie der Edelmetalle, Ludwigsburg.
1982 J. Ogden, Jewellery of the Ancient World, London.
1982 P. Parrini, E. Formigli, E. Mello, "Etruscan Granulation: Analysis of Orientalizing Jewelry from Marsiliana d'Albegna," American Journal of Archaeology 86, pp. 118-121.
1982 V. Fell, "Ancient Fluxes for Soldering and Brazing," Masca Journal 2, p. 3.
1983 D.L. Caroll, "On Granulation in Ancient Metalwork," American Journal of Archaeology 87.
1983 J. Wolters, Die Granulation, Geschichte and Technik einer alten Goldschmiedikunst, Callwey, Monaco.
1983 E. Formigli, Appendice tecnica, L'oro degli Etruschi, Novara, pp. 321-33.
1985 E. Formigli, Tecniche dell'oreficeria etrusca e romana, Florence.
1987 R. Echt, W.R. Thiele, "Etruskischer Goldschmuck mit gelöteter und gesinterter Granulation," Archäologisches Korrespondenzblatt >17.
1989 A.R. Duval, C. Eluere, L.P. Hurtel, M. Menu, The Use of Scanning Electron Microscope in the Study of Gold Granulation, 25th Symposium in Athens: Archaeometry, Amsterdam.
1990 C. Eluere, Les secrets de l'or angique, Parigi, pp. 200-207.
1990 G. Nicolini, Techniques des ors antiques. La bijouterie ibérique du VII au IV siécle.
1990 R. Baines, The Significance of Double Row Granulation from Palestrina, Symposium: Decorative Techniques in Jewellery, Londra.
1990 A. Parea, Orefebreria prerromana, Casa del Monte.
1991 A.C. Robert-Hauglustaine, Le soudage de l'or: étude a partir des textes antiques et médiévaux, Symposium: Outils et ateliers d'orfevres, Saint Germain en Laye.
1992 E. Formigli, "Indagini archeometriche sull'autenticità della fibula prenestine," Mitteilungen des deutschen archäologischen Instituts 99, pp. 330-343.
1993 G. Platz, H.U. Tietz, "Etruskische Skarabäen-Colliers," Jahrbuch der Berliner Museen >35.
1993 E. Formigli, W.D. Heilmeyer, "Einige Fälschungen antiken Goldschmucks im 19. Jahrhundert," Archäologischer Anzeiger, pp. 299-332.
1994 G. Nestler, Die Granulation. Antike Goldschmiedetechnik im Experiment, Bild der Wissenschaft, DVA Stocarda.

Footnotes
1 A. Castelani, Antique Jewellery and its Revival, London, 1862.

2 J. Wolters, Die Granulation, Geschichte und Technik einer alten Goldschmiedekunst, Callwey, Monaco 1983, p. 283.

3 C. Aldred, The Jewels of the Pharaohs, Thames and Hudson, London 1971.

4 F.W. Von Hase, Zum östlichen Einfluß auf die Goldarbeiten des späteren 8. und 7. Jahrhundert v.C. in Mittelitalien, >Proceedings of the Xth International Congress of Classical Archaeology, Ankara 1978.

5 E. Formigli :Fälschungen etruskischer Goldschmiedearbeiten," Die Etrusker und Europa.

6 M. Cristofani, M. Martelli, L'oro degli Etruschi, Novara 1983, cat. 277.

7 E. Formigli, Tecniche dell'oreficeria etrusca e romana, originali e falsificazioni, >Firenze 1985; H. Hoffmann, V. Von Clear, Antiker Gold-und Silberschmuck, Magonza 1968; J. Ogden, Jewellery of the Ancient World, London 1982.

8 P. Parrini, E. Formigli, E. Mello, "Etruscan Granulation: Analysis of Orientalizing Jewelry from Marsiliana d'Albegna," American Journal of Archaeology 86, (1982), pp. 118-121.

9 J. Wolters, cit., p. 57.

10 W. Schuhmann, BLV- Bestimmungsbuch der Edelsteine, Monaco 1978.

11 H. Jüngst, Zu den Anfängen des Hartlötens, Symposium, Historische Technologie der Edelmetalle, Ludwigsburg 1980.

12 Diebeners, Handbuch des Goldschmieds, >1977, vol. III, p. 129.

13 J. Wolters, cit., p. 277.

14 J. Wolters, cit., p. 278.

15 E. Brepohl, Theory and Practice of Goldsmithing, Maine 1973.

16 M. Rosenberg, Geschichte der Goldschmiedekunst auf technischer Grundlage, Abt. Granulation, Frankfurt s.M. 1918.

17 H.A.P. Littledale, Improvements in hard soldering mixtures and hard soldering processes, >Brit. Pat. 415181 (23.8.1934).

18. J. Wolters, cit., p. 34.

19 Vannaccio di Biringuccio, De la Pyrotechnia, Venice 1540, l. IV, c. 68r.

20 B. Cellini, Due Trattati: uno intorno alle otto principali arti della oreficeria, l'altro in materia dell'arte della scultura…, Florence 1568.

21 H. Jüngst, cit.

22 G. Platz, H.U. Tietz, "Etruskische Skarabäen-Colliers," Jahrbuch der Berliner Museen >35, 1993.

23 E. Formigli, Sulla tecnica di costruzione dei fili d'oro nell'oreficeria etrusca, Symposium: Outils et ateliers d'orfevres, Saint Germain-en-Laye 1993.

24 J. Wolters, cit., p. 294.

25 E. Formigli, W.D. Heilmeyer, "Einige Fälschungen antiken Goldschmucks im 19. Jahrhundert," Archäologischer Anzeiger, 1993, pp. 229-332.

26 R. Baines, Technical Decisions in the Gold Cylinders from Praeneste, Symposium: Outils et ateliers d'orfevres, Saint Germain-en-Laye, 1993.

27 R. Echt, W.R. Thiele, "Etruskische Goldschmuck mit gelöteter und gesinterter Granulation, " Archäologisches Korrespondenzblatt >17, 1987.

28 B. Cellini, Due Trattati, cit., I, c. 18v e 19r.

29 J. Wolters, cit., pp. 34 ss., pp. 50 ss.

30 G. Weisgerber, C. Roden," Grieschische Metallhandwerker und ihr Gebläse," Der Anschnitt 1, 1986.

31 B. Cellini, Vita di Benvenuto Cellini scritta da lui medesimo (1558-1566), Florence 1728.