Metal Salt Weighting of Silk: Understanding Practices and Their Historical Context Through Textual Sources

Abstract

Treating silk with metal salts was a common practice starting in the second half of the 19th century until the early 20th century. It aimed to increase the weight and thickness of the fibres. However, the presence of metal salts is believed to accelerate and aggravate the deterioration of historical silk textiles, and weighted silks are nowadays considered one of the most pressing issues in textile conservation. This paper explores the history of the practice of metal salt weighting of silk: the materials and methods used, the reasons behind weighting, and how this practice developed as the product of a specific historical and economic context. A total of 147 primary textual sources (patents, dyers’ manuals, and books) were investigated and from these 136 weighting methods were collected and reviewed. The results highlighted tin salts as the most commonly mentioned metal salts for weighting silks of any colour. Iron compounds combined with tannins were the method of choice for dark silks, although also in combination with tin in half of the cases. The knowledge gained from this research will help further the study of the degradation pathways of historical silk fabrics, as representative reproductions of weighted silks will be produced based on the findings.

1. Introduction

Silk is a natural fibre produced by some insects; the majority of silk used in the textile industry is spun by the moth species Bombyx mori, the domesticated mulberry silkworm. Silk is one of the oldest fibres known to humankind; thanks to its unique qualities among natural fibres, such as lustre, elasticity, strength, and affinity for dyes, it has been a highly valued and popular high-end material for centuries [1].

Silk is made of two proteins: fibroin, which forms long continuous filaments, and sericin, which acts as a glue holding the fibroin strands together [2]. The sericin, also known as “silk gum”, is usually removed from commercial silk through a process known as “degumming”. This softens the fibre and increases the typical lustre of silk; however, it also reduces the weight of the fibre by around 20 to 25% [1,3].

The “weighting” or “loading” of silk is the practice of adding different substances, with the aim of increasing its weight and the volume of the fibres. While it is not clear exactly how or when the practice started, weighting was most likely developed by dyers as a way to compensate for the decrease in weight caused by degumming. As silk has historically been sold by weight, loading would have been a way to make up for the “lost value” compared to that of the initial raw material [4,5]. Written evidence of weighting can be found as early as the 17th century, when this was accomplished mostly by applying organic agents such as sugars, waxes, or protein glues. While effectively increasing the weight, these substances would eventually wash off [3,6]. In the mid-1900s, the practice was revolutionised by the introduction of metal salts as weighting agents, particularly tin chloride salts, which not only allowed reaching higher loading effects, but were also permanent, as the metal bound to the silk. Nowadays, when the term weighted silk is used, it usually refers specifically to the metal-salt-weighted silks produced between the second half of the 19th century and the start of the 20th century. Metal salt weighting started to decrease in popularity in the 20th century as the focus of research in the textile industry shifted, new synthetic fibres entered the market in the 1930s, and the demand for silk goods greatly declined [3,7]. Metal salts weighting practices developed and used from the mid-19th century to the early 20th century are the main focus of this paper, and for ease of understanding, when the term weighting is used throughout this paper, it refers to metal salts weighting unless otherwise stated [5,6].

Historical silks are found in many museum collections, not only in garments but also in accessories, such as shoes, fans, parasols, upholstery, banners, and flags. However, pieces produced at the time when silk weighting was a common practice have often been observed to be in significantly poor condition. The presence of metal salts in the fibres is believed to accelerate and aggravate the deterioration of historical silk textiles, and although this link is yet to be proven, weighted silks are nowadays considered as one of the most pressing issues in textile conservation [3,5,8].

In order to study the role metal salts may be playing in these degradation pathways, it is essential to first understand how weighted silks were produced in the past. This paper aims to illustrate metal weighting practices: how these were carried out, which materials were used, and how they were applied and also to understand weighting in the context of the silk production industry from the 1850s to the 1930s, when the practice was popular, exploring how and why weighting came into existence. This paper distinguishes itself from previously published studies on silk weighting for its use of patents as primary sources and the discussion of the practice in the wider silk market and the economic and historical context of the late 19th and early 20th centuries.

The Reasons Behind Weighting

Beyond the weighting method and its developments, it is important to consider the motivations that drove the widespread adoption of the practice in the first place.

The main reason for weighting was that it was economically advantageous, as one could sell the same amount of silk for a higher price due to it being heavier and thicker. The practice, in fact, increased not only the weight, but also the diameter of the silk yarn. This meant that to make a finished product, fewer threads per unit surface were used, reducing the amount of raw material needed [4,7].

However, economic value is not the only reason found in literature as to why the silk was weighted. Silk fabric that had been treated with metal salts created a characteristic rustling and swishing sound as the wearer moved. This was particularly in fashion and a desirable effect, so much so that a 1908 patent discussing a finishing treatment made of beeswax explains not only how to apply it, but also the subsequent steps to take so that the finishing would not disrupt the sound [9].

Another aspect often mentioned is the hand, as the silk becomes easier to work with. The silk is thicker, it has a greater body, and it is less slippery—all very desirable qualities for specific applications, such as the making of ties [4].

Weighting also played a key role in colouring silk. This is not only limited to the iron-tannin weighting overlap with black dyeing, but is also true for other colours. Tin compounds, the most commonly used metal salt for weighting silk, also acted as mordants by binding dyes to the fibre, in particular natural red dyes [10]. Weighting was therefore advertised as a practice which allowed for more even and vibrant colouring, or for specific colour shades to be created [11]. When combined with practices such as discharge printing, weighting was also used as a way to create particular decorative patterns on the silk, as areas weighted with different metals or different amounts would turn different colours [12].

2. Research Methodology

Building on the previous work by Hacke [3], a variety of historical documents were reviewed, including dyers’ manuals and technical journals, but with a particular focus on patents mentioning the weighting or loading of natural silk. To fully contextualise this practice, not only weighting recipes, but also publications on related treatments and specialised machinery were reviewed.

Previous studies were used as a starting point to identify suitable sources that were then expanded upon.

The main primary source used for the research was published patents available through online databases, both international and country-specific, including Google Patents, Espacenet, Patent Public Search USPTO, and the Privilegiensammlung österreichisches Patentamt. The most used secondary sources were articles published in technical journals, both contemporary to the patents and published later, as these not only offered an overview of how the practice developed and the state of the art, but also offered lists of published patents in their bibliographies. Technical journals were accessed through a series of online and physical libraries, including Archive.org, BnF Gallica, Hathi Trust, the libraries of the city and the University of Antwerp, the Rubenshuis Library, and the British Library. The types of sources used in the literature research are illustrated in

Table 1. Literature sources used and their use.

	Description	Information	Example
Primary	Original patents	Exact recipes and procedures; commentary on state of the art	[13]
	Dyers’ manuals and recipe books	Exact recipes and procedures; commentary on state of the art	[14]
Secondary	Technical journals contemporary to patents	Commentary on state of the art; refers to many patents at once; gives extra information compared to patents	[6]
	Later technical publications	Hindsight commentary on effects of weighting	[4,7]

.

Several search terms were used in various combinations across the databases. Searches were conducted in English, German, Italian, and French, giving results not only originally written in these languages but also, in the case of patents in the Google Patents database, as automatic translations into English from other languages. These specific languages were selected as the author (C.V.) is versed in them. Furthermore, as silk weighting was primarily a European and North American practice, it is fair to assume that a majority of the literature referring to it would be available in these European languages. A list of all the search terms used can be found in

Table 2. Terms used for patent searches in English and the corresponding terms used in German, Italian, and French.

English Term	Corresponding Terms
	German	Italian	French
Silk, natural silk	Seide, Naturseide	Seta, seta naturale	Soie, soie naturelle
Weighting, loading, weighted, loaded	Beschweren, Beschwerung	Carica, caricatura, appesantire, pesata, caricata	Charger, la charge, chargée
Metal, tin, metal salts	Metall, Zinn	Metallo, stagno, sale metallico, sali di metallo	Metal, étain, sel métallique
Method, treatment, procedure, process	Verfahrung, Verfahren	Metodo, trattamento, procedimento, procedura	Procédé, traitement
Machine, apparatus	Vorrichtung, Apparat	Macchina, apparato, dispositivo	Dispositif, machine, appareil

.

Databases were searched filtering for the time period 1845 to 1945, to ensure that all potential patents published in the period of interest could be found. For technical journals and books, later publications were also included in the research to ensure that retrospective discussions of weighting practices were also taken into account.

All the patents were assessed based on the country of both filing and the author’s residence; if they were filed in more than one country, the dates of application and publication of the patent and the presence of any follow-ups or modifications published at a later date were assessed. The weighting methods and recipes were also analysed so as to categorise them based on the materials used.

To study the development of the weighting practice through time, the application dates or, when indicated in the patent text, the dates specified as “patented since” were used. This choice was made to reflect the time closest to the actual invention, as the patents could sometimes be officially published up to a decade after being filed. This was the case, for example, with German patents, particularly in the period between the two world wars, because of an accumulated backlog of applications [15,16,17]. The actual publishing date was only used to date the patent when no further information was available.

When considering geographical distribution, two elements were considered: the country in which a patent was filed, and the country of residence of the inventors. This choice was made because the two provide different information. The countries of residence better reflect where the research and development were happening, while the countries of publication of patents give us insights into the wider geographical spread of where the inventions were potentially in use. When the same invention was patented in different countries, “repeat” patents arose. These inventions were considered as one in our study and defined by the earliest date. For example, Egon Elöd’s Method for silk loading was patented three times: in Germany, where the company was based, but also in the USA and France, and published in 1923, 1924, and 1928, respectively [18,19,20]. All three patents mention that the invention had been patented in Germany since the 14th of November 1921. Based on the criteria explained above, this method was counted as one, developed in Germany in 1921, but published in three iterations.

Recipes involving iron and tannins posed the problem that most ingredients play multiple roles, not only adding weight, but also contributing to the creation of specific colour shades. These recipes are closely linked to the practice of dyeing with dark natural colours, particularly blacks and browns [10,14]. Some parameters needed to be defined for when a dyeing recipe could also be classified as a deliberate weighting method. These criteria were defined for this research as listed below, and in order to be considered a weighting method, a dyeing recipe had to fulfil at least one of them.

• The recipe specifically uses the word “heavy” in its description;
• An ingredient known to increase weight is applied three or more times;
• Three or more different ingredients known to increase weight are used;
• An ingredient with weighting as its main role is used.

The choice of three repeats or three ingredients was made because two instances could be a coincidence, while three or more were thought to suggest a deliberate intention for choosing specific ingredients known to increase the weight.

Limitations of the Research Methodology

There are a few limitations to the methodology used in this research: some are dependent on the nature of the weighting practice itself; others are specifically due to using patents as primary sources.

Particularly in its early days, the practice of weighting did not have a positive reputation and was seen by many as a way for dyers to cheat their customers. Because of this, weighting was often surrounded by secrecy, meaning that the methods are not well or accurately documented or have been post-dated, as an invention was sometimes patented years after it had started being used in the industry [4,6,21]. This is the case for example for the first mention of weighting using tin in combination with sodium phosphate and sodium silicate: the first patent for this method was issued in Germany in 1894, and in the USA, the invention was not patented until 1897. However, there is evidence that the practice was already in use by the Weidmann Silk Dyeing Company in New Jersey as early as 1892 [6].

The second limitation is for black silk weighting in particular: the metal salts and other ingredients used often played more than one role. Rather than just increasing the weight and swelling the fibres, they could also act as mordants, or contribute to giving the finished piece a specific colour shade. It was necessary for selection criteria to be chosen to distinguish between dyeing recipes and dyeing recipes that were also purposefully designed to increase the weight of the fabric. These criteria, however, are still arbitrary and not universally recognised. Another researcher might select different criteria which give rise to slightly different conclusions regarding silk weighting practices. An objective way to assess if a dyeing recipe is also a weighting method would be to create reconstructions of all the recipes identified and measure any weight increase occurring, but this would be very time-consuming and goes beyond the scope of this current research.

Another important aspect to consider is that not all new patented methodologies were used for large industrial-scale production. Some were used only in a limited capacity and some not at all. This means that while the study of patents is a good way to track the development and evolution of weighting practices, it is not perfect, and to gain a real understanding, it should be complemented with studies of other sources and where possible, museum objects. In the case of weighted silk, within the framework of the Safesilk project, a survey of museum collections in Belgium, the Netherlands, and Slovenia is currently ongoing and aims to assess to what extent the insights from patents and manuals are reflected in the preserved artefacts [22,23,24].

The last of the limitations to discuss is the overlap of the historical period studied with the development of patent offices in different countries. Many patent offices in their modern form were created in the nineteenth century, meaning that the earlier weighting inventions are often not found in patents simply because such intellectual property protection systems did not yet exist. The Austrian patent office, for example, was only founded in 1899. Before this, a system of so-called “privileges” existed, but this is not the case for all Western countries [25]. Switzerland first introduced a federal patent law in 1888, but this excluded any inventions from the chemical industry, which had strongly opposed the patent bill. It was not until 1907 that this law was modified [26]. The UK also made several reforms and updates to its patenting system, first creating a unified patent office system for the whole UK in 1852, which was then revised in 1907, 1919, 1931, and 1949 [27]. In Germany, as the German Empire collapsed with the end of World War I, the Kaiserliches Patentamts (Imperial Patent Office) came to an end and was later substituted by the Reichspatentamt (Patent Office of the German Reich), which, less than two decades later in 1938, also absorbed the Austrian patent office in its entirety, following the invasion of the country [15,25].

All of these changes cause issues in tracking down some of the patents mentioned in contemporary texts. Original documents from different years are often split between different archives with different access rules, or renumbering of the patents to follow a new system could have occurred. For example, in Walter M. Scott’s article “The Weighting of Silk”, published in 1931 in the American Dyestuff Reporter, a series of patents is listed with inconsistent identification numbers, some of which seem not to exist when searched for nowadays. However, these can sometimes still be located based on other identifiers, such as author names and publication dates, if mentioned [6].

3. Results and Discussion

Using the research methodology described, 147 primary textual sources (full list in Supplementary Information S1) were collected and analysed, of which 12 were dyers’ manuals and books, and 135 were patents. From the books and manuals, 53 silk weighting recipes were identified. The 135 patents were divided as follows: 83 distinct weighting methods; 24 repeat patents; 16 patents for protective and finishing treatments; 9 patents describing machinery and optimisations of the weighting process; and 3 which mentioned weighted silk, but were found to not be relevant to this research. Combining the weighting methods found in patents and the weighting recipes from books, a total of 136 distinct weighting methods were identified.

Sources published as early as 1846 were investigated. The earlier sources are dyers’ manuals, while the oldest patent found dates to 1882 [28,29,30]. According to Walter M. Scott’s 1931 article, the first patent discussing weighting using metal salts was published in the UK in 1855 and it suggested the use of zinc and barium chlorides as the loading matter [6]. While the existence of this early patent is known, it was not possible to locate the original document, so this patent was not included in the study of the primary sources.

Weighting in black and dark-coloured silks was already common in the 19th century; for example, the practice was being discussed in the 1873 Annual International Exhibition report in The Journal of the Society of Arts. This report pointed out how some of the black ribbons displayed had been “artificially made to cover four times the space of the natural thread”, meaning the silk threads shown were four times thicker than normal, and that the weighting practice was going to “leave competitors no choice but to discover the secret or be out of market” [31]. Reports such as this one highlight both how effective and economically advantageous weighting practices could be by pointing out how a manufacturer was doomed to go out of business if it did not adopt the same loading practices as its competitors.

The moment that led to a revolution in the practice and the start of silk weighting as an industrial process was the discovery of the high affinity of silk for tin salts, specifically tin chlorides, in the 1870s. What made tin weighting uniquely different from any of the earlier methods is that, not only did it increase the weight significantly, but it was colourless, so for the first time, silks of any shade, even white, could be advantageously treated [3,4].

With the exception of the period during the First World War, from the 1890s to the 1920s, new inventions related to silk weighting were constantly being patented; the 1930s started seeing a decline in weighting innovations until they completely stopped in the 1940s (Figure 1). It is important to note that textual sources and patents in particular provide information on when most of the research was conducted. While it can be assumed that there is a connection between the peak of production and use of these methods, this is not guaranteed, as was explored in Section Limitations of the Research Methodology. In this case, the assumption is supported by the number of published patents following the same trend as the silk market and experiencing a significant drop after the 1920s (Figure 1). Silk consumption in the 1930s and 1940s dropped to just a third of the 1920s level, due to the combined effect of economic crisis, World War II, and the advent of synthetic fibres. With the invention of the first artificial substitutes for silk, such as nylon, which was first commercialised in 1938, the focus of research in the textile industry also shifted away from natural silk [7].

Based on the countries of residence of inventors, the main centres of research and development of new innovations were concentrated in Germany, Switzerland, and, particularly in the 1920s, the US (Figure 1). The use of the practice was spread much wider, as is reflected by the countries in which the inventions were patented, including not only the US and most European countries, but also Canada and Australia.

The results highlighted how in the decades of its popularity, weighting was a geographically widespread practice in the Western world, with new technologies being patented, and, it is assumed, also used in industrial sites all over Europe and North America.

3.1. Categorisation of Weighting Methods

In order to better understand the variety of weighting techniques, the following classification system was developed based on the materials employed. The weighting recipes can be divided into different categories as illustrated in Figure 2.

The first category is recipes that include tin salts, either alone or in combination with other metals. The majority of recipes in this category fall into the definition of a tin salt bath, usually a tin tetrachloride solution, followed by a treatment with a phosphate. The development within this category is chronological: earlier recipes would stop after the phosphate baths, while later inventions also include additional silicate and alum baths. Tin salts were also at times used in combination with other metal salts. These methods were often attempting to achieve the same weight gain using cheaper starting materials or, as in the case of tin salts added to iron weighting recipes, aimed at reaching an even higher weight increase.

Other than tin, the other most commonly found metal in weighting recipes is salts of iron, most often used in combination with various types of tannins. But these methods could only be used for dark-coloured silks, most often black, as iron stains the silk fibres.

As weighting increased in popularity, various ways to optimise the process were being researched, giving rise to a variety of patents describing different application methods and machinery which reduced waste, improved efficiency, and reduced costs. These technological improvements are discussed later, together with the protective and finishing treatments associated with weighted silk, in Section 3.1.3 and Section 3.1.4.

3.1.1. Tin Weighting

Tin salts were by far the most commonly used metal salts employed in silk weighting, with 62% of the methods collected including tin, either alone or in combination with other metals. As mentioned earlier, tin reached this level of popularity because it not only increased the weight and thickness of the fibres significantly, but it was also colourless, meaning that silk could be pre-weighted with tin when still in skeins, regardless of the colour in which it was to be dyed.

The first loading method to employ tin salts was patented by Joseph Rubelin in 1882. However, the patent text did not include the words “weighting” or “loading”. Instead, it referred to the method as a treatment preparatory to dyeing and weaving that “gave a greater body to the silk”, as well as allowing for more even dyeing and an increased brilliancy of colour. The patent claimed the invention of a process subjecting silk to a solution of tin tetrachloride, followed by “fixing” the metal deposit with “a solution of a caustic alkali or carbonate of the alkalies, and repeating these operations until the proper effect is obtained” [30]. The specific choice of words may give insight into how weighting developed from dyeing practices, but it may also reflect the fact that at the time the general opinion was unfavourable to weighting practices, as is clear from the contemporary literature. In his 1931 article, Scott discusses how weighting had been at its beginning and for a long time considered an “extremely unethical practice” [6].

In 1893, about a decade later, Neuhaus published what was to become the most common weighting combination: the silk is steeped in a solution of tin tetrachloride for an hour, washed in water, and then put into a “fixing” solution of sodium diphosphate, through which the tin phosphate weighting complex is formed. The silk was then subjected to a final treatment in warm sodium silicate solution. The process could be repeated 1 to 6 times depending on the desired results (Figure 3). According to the patent, repeating these steps five times would give a weight increase of 100–120%, a substantial increase compared to the weight gain of only 20% that could be achieved with five passes of the older tin-plus-alkali method [32].

Variations of this method make up 18% of all the collected weighting methods. These recipes all follow the same format, with the most important innovation being the addition of an aluminium sulphate bath before the silicate treatment. First introduced in 1896 by Biermann and Puller, it was said to double the weight increase, so that the same loading levels could be achieved using half of the passes necessary using just the tin, phosphate, and silicate baths [33]. Variations within this category concern details such as the specific concentration and temperature of the solutions, the timing of each step, and the way the solutions are applied. Most commonly, the silk was submerged in open vat baths, but other ways of application, such as the spreading of a paste, were also experimented with. Publications focusing on optimising the practice and developing new machinery are discussed in more detail in Section 3.1.4 [3,4,7].

In the 1910s, several attempts were made at finding possible cheaper or more effective alternatives to the tin–phosphate-silicate method, often referred to as the “Neuhaus method”: salts of antimony, zinc, cerium, zirconium, and rare earth elements were all suggested as substitutes for tin, but they were not as effective or had unwanted side effects. Silk treated with cerium chloride solutions turned a yellow-brown colour very quickly [34,35,36,37,38,39,40]. Some of these adverse effects could be avoided if these metals’ salts were used as an addition to tin, rather than completely substituting it: the yellowing could be avoided if a quantity of tin chloride was added to the cerium chloride baths [39,41]. The plot in Figure 4 illustrates and summarises the patents that discuss other metal ions to be used as substitutes for, or in addition to, tin. These are divided into three categories, in which the different metal salts are suggested as tin substitutes, in addition to the tin solution bath, or in addition to tin weighting as an extra step.

In the category of recipes involving the addition of other metal salts to tin as a successive step, there are several examples of methods describing how the silk was first weighted with tin and only afterwards other metals were added. This could be conducted with salts of cerium, but the most commonly mentioned metal salts to be used this way are those of zinc and lead [42]. These were introduced through an additional treatment after the silicate step as solutions of their acetates, and resulted in a substantial weight increase. The two metals could be used alternatively or in combination. Using both zinc and lead compounds was particularly important if the silk was going to be used for a printed fabric made by discharge printing. In this technique, the fabric was dyed and then the colour was removed from specific areas using a stripping agent to create a “negative” pattern. When subject to discharge printing, lead-weighted silk turned grey, rather than the desired white, but this greying did not occur if zinc was also added [12,43].

Figure 4. Number of recorded methods which use salts of Sb [34,35], Zn [12,17,36,37,44,45,46,47,48,49,50,51,52], Ce [38,39,40,42,53], La and Di [54,55], Zr [20,39,41,56,57,58,59], Cr [50,60,61,62,63] and Pb [12,17,43,52,64,65,66] as weighting materials, either in place of or in addition to tin salts (* Didymium (Di), first discovered in 1839, was believed to be an element at the time. It was correctly identified as a mixture of praseodymium and neodymium for the first time in 1885 [67]).

Figure 4. Number of recorded methods which use salts of Sb [34,35], Zn [12,17,36,37,44,45,46,47,48,49,50,51,52], Ce [38,39,40,42,53], La and Di [54,55], Zr [20,39,41,56,57,58,59], Cr [50,60,61,62,63] and Pb [12,17,43,52,64,65,66] as weighting materials, either in place of or in addition to tin salts (* Didymium (Di), first discovered in 1839, was believed to be an element at the time. It was correctly identified as a mixture of praseodymium and neodymium for the first time in 1885 [67]).

Zirconium was also proposed as a potential substitute for tin several times, but there is reason to believe the proposed methods might not actually have been used much in the industry. This evidence is given by two patents by E. Stern published in 1921. The first patent describes the use of zirconium mixed with tin to weight silk, and recommends not using zirconium alone as it would yellow the fibres in a similar way to cerium [39]. A month after the filing of this patent, the author applied for a correction to the patent in which the zirconium is instead substituted by zinc [40]. These sources may suggest that while zirconium was experimented with, the methods may not have been as effective as hoped and were probably not used extensively. This hypothesis is further supported by another patent by the same author, which suggests zirconium might not have such a high affinity for silk. In this instance, Ernst mentions that any salt of tin or zinc could be used for the weighting of silk, while zirconium is absorbed by the fibres only if in the zirconium sulphate tetrahydrate form [56]_.

A potential variation in the tin weighting methods is the omission of the washing step between the tin chloride and the phosphate baths. A series of patents were granted between 1927 and the early 1930s to the Swiss scientist René Clavel, all focusing on how the washing step could be skipped if the silk was pre-treated with an acid before tin weighting. First suggesting a non-specified acid treatment, he later selected hydrochloric acid as the best choice for several reasons. The excess hydrochloric acid was more easily rinsed out; it was cheaper than other acids and the necessity of working with closed vats due to the toxic vapours caused by other acids such as formic acid could be avoided [68,69,70,71].

3.1.2. Iron Weighting

Following tin, iron is the second most common metal used in silk weighting. It is used in 59 of the collected methods, i.e., 43% of the total, and in 92% of these, iron is used in combination with one or more tannin compounds.

Tannin weighting is one of the earliest examples of silk loading practices, as it overlaps with methods of dyeing black with natural dyes: iron in combination with various tannins had been commonly used as black and brown colouring matter, both for inks and textile dyes for centuries [3]. One of the earlier records of dyers being aware of the advantageous additional weight gained by silk when treated with iron compounds in combination with specific tannins is found in Macquer’s L’art de la teinture en soie, first published in 1763, where a method involving repeated gall baths to obtain heavy blacks (called noir pesants to distinguish them from the light black noir légers) is described [72].

Interestingly, the added weight originates from the tannins rather than the iron compounds. However, not all tannins add weight, which is why not all dark dyeing methods can also be classified as loading methods. The tannin varieties most used for their specific ability to swell and increase the weight of the silk fibres, rather than only contributing to the colour, were cutch (Acacia catechu, Senegalia catechu), gambier or white cutch (Uncaria gambir), gallnut (Quercus infectoria), and chestnut (Castanea sativa) [4,10,72]. Tannins alone cannot bind to the silk fibres and require the presence of a metal ion. In the weighting recipes collected, this is most commonly an iron compound. These iron compounds, most commonly some form of iron sulphate, can be found either alone or, in 47% of the cases, in combination with tin. The other two recurring metals in iron-tannin weighting are copper and lead, present in 16% and 5%, respectively, of the iron-tannin methods. While weighting with tin was colourless, both iron and tannins stain the silk fibres. For this reason, tin can be found added to many iron weighting recipes, as a way to further increase the weight gain, since it would not alter the already dark shade. However, the opposite is not true. Adding iron or tannins to tin weighting would negate its main advantage of not affecting the natural silk colour.

Iron weighting methods generally consist of a bath in an iron solution, most commonly of iron (III) sulphate, followed by a bath with a tannin with weighting properties. These baths are repeated several times, and this can be conducted in two ways: either the silk is first passed repeatedly through the iron solution and only after it is introduced into the tannin bath, or the two baths are alternated and repeated several times, until the desired weight is reached [14,28]. The weighting methods that use both iron and tin compounds can be divided into two sub-categories: those that consist of applying a tin weighting method followed by treatment with iron and tannins, and those that introduce tin chloride mixed into the tannin bath [4,14].

Other than iron sulphate, the other iron compound often found in these recipes is potassium ferrocyanide, or Prussian blue. This was introduced as it contributed to the overall weight gain, but also for colouring purposes. Creating a blue base helped to produce a final “cold” black shade, rather than a black more akin to dark brown [14].

The combination of iron compounds and specific tannins, with or without the inclusion of tin compounds, was extremely effective at increasing the silk weight. Reportedly, some of these weighting methods regularly achieved weight gains of 200% and could even be pushed up to 400%, much higher percentages than were possible with tin weighting alone [14].

3.1.3. Protective and Finishing Treatments

Several patents mentioning the possible negative effects of weighting on the finished fabric and how to avoid these were published as early as 1905, indicating that at the time there was already an awareness of the weighted fabrics being more susceptible to damage [73]. A total of 16 patents were found that described different treatments ranging from lustring, which aimed to give silk its typical shine back as it could dull during the weighting process, to the addition of substances specifically aimed at maintaining the elasticity of silk and preventing it from becoming brittle.

The addition of diluted thiocyanate or thiocyanide solutions during weighting was suggested to improve mechanical properties and protect weighted silks from oxidation. Thiourea, its salts, and derivatives are also mentioned as widely used during the weighting process for similar reasons [74,75,76,77]. For example, a patent by E. Sisley filed in 1906 describes the use of thiourea, its salts, or derivatives to improve tensile strength and elasticity: weighted silks are immersed in a thiocyanide solution, to which citric acid is added for acidity. The patent also mentions this treatment prevents the formation of “rusty stains” that can appear due to a reaction with sweat or chloride compounds. The author justifies the protective effect by explaining that thiocyanides are more easily oxidised than fibroin in the presence of inorganic salts [76].

A couple of patents suggest the use of aldehydes by adding a solution of 1-5% during the process of weighting or dyeing, and claim this is more effective as the aldehydes do not alter the lustre and hand of the fabric [78,79].

Amines, amides, and proteins were suggested as a more advanced alternative method to preserve the feel and lustre of silk: hydrazine, hydroxylamine, hexamethylenetetramine, casein, albumin, animal glue, gelatine, agar-agar, or even raw silk cocoons dissolved into the weighting bath are mentioned [73,80,81,82]. Polysaccharide carrageenan is also suggested in one instance as having a protective effect [80].

Another group of protective methods involves inorganic sulphur-containing soluble salts, such as sodium hyposulfite and thiosulphates, used during dyeing or weighting or as finishing, or added into the stannic chloride bath [77,83]. The author of this last patent, Ewald Herzog, claims that his method would give more reproducible results as, in addition to a reducing effect, he argued that a deposition of fine sulphur particles would take place. This sulphur would react with the tin in the silk resulting in tin sulphides that, according to him, would not induce degradation of the fibres and would have a protective effect [83].

3.1.4. Innovations in Industrial Process and Machinery

At the beginning, silk weighting was carried out much like traditional dyeing: in large vat baths with the skeins suspended on wooden sticks being submerged and then taken out by hand. As the practice evolved, so did the technology associated with it, and while still practised in smaller dye houses, weighting in the 20th century had become an industrial-scale, semi-automated process.

Several patents were found that focused specifically on innovations in the machinery and industrial process of weighting. These new developments aimed at making weighting more efficient in various ways, for example, limiting the waste of ingredients, improving the uniformity of application, or allowing weighting of woven fabrics rather than the silk skeins [84,85,86].

A 1932 patent entitled Improvements relating to the loading of natural silk, for example, included not only ways to regulate and control the pH of the weighting baths, but also an extensive discussion of the acid-resistant materials the vats, pumps, and other parts of the industrial apparatus should be made of to ensure efficient and uniform weighting [84].

New inventions were also filed relating to innovations in the way the weighting materials were applied to the silk; for instance, instead of as a solution in a bath, tin chloride salts were applied to the fibres in the form of foams or mousses [86,87]. These methods aimed to give a more uniform, quicker application, and to use fewer raw materials, minimising the waste associated with bath solutions.

Weighting in the piece, meaning weighting of woven cloth rather than silk skeins, was not commonly conducted until the 1920s [6]. While some earlier patents already mentioned the possibility of weighting fabrics, the results were often uneven and not satisfactory [6,32]. Weighting in the piece became possible on a large commercial scale thanks to the invention of new machinery designed for this specific purpose. Swiss scientist René Clavel played a key role in the development of piece weighting. In 1927, he patented the invention of a new process and device for the weighting of silk fabrics, which later became known as the Clavel and Lindemeyer process [88,89]. The new process and machinery made weighting faster and less labour-intensive, and reduced the volume of the baths needed [6]. The weighting of woven fabrics gained popularity, as it was convenient and cost-effective. However, it did not fully replace the weighting of skeins. The two practices coexisted and both were still in use in the late 1930s, as detailed in a 1937 article in the American Dyestuff Reporter [90].

Another aspect that was experimented with was recovery methods. These are tin-specific processes that aim at precipitating any excess tin lost by the silk in the phosphate bath. This was conducted by adding silicate of soda to the used phosphate bath, boiling it, and filtering out any precipitate formed [91]. The purified phosphate bath could then be used again, instead of being thrown away. Innovations of this kind were of particular interest as they minimised the amount of raw materials used, making weighting even more convenient and cost-effective.

The fact that evidence of innovation, not only in weighting materials but also in machinery, can be found is an indication of how important and profitable the business of silk loading was at the time, as it was deemed worthy of investment and research.

3.2. Weighting as a Result of Its Historical Context

Weighting was an extremely popular and widespread practice in Western Europe and North America, but to fully understand this practice, it is essential to appreciate the historical context of silk production and trade in the 19th and 20th centuries. Metal weighting was the product of a very specific and not very long time period because of the concurrence of circumstances which made it possible for it to develop the way it did (Figure 5).

From the 1850s, the global silk market was highly affected by a series of external events that led to significant increases in the price of raw silk, pushing the industry to look for innovations in the field. Three major events explain this price increase: the pebrine epidemic in the Mediterranean area; the Taiping revolt in China; and the end of Japan’s period of seclusion [7,92,93]. Pebrine is a disease of the silkworm which causes the larvae to die before pupating, and therefore before reeling the silk cocoon. It is both contagious and hereditary: a worm infected during the raising period may at times survive into adulthood, but will lay infected eggs, and if a healthy batch of eggs hatches, one still has to hope the worms do not contract the disease later. Pebrine was first noticed in France in the late 1840s, reached Italy in 1850, and continued spreading around the Mediterranean basin in the following decades, reaching as far east as Persia. Due to pebrine, raw silk production between 1853 and 1865 decreased by 80% in France and by 50% in Italy [7,94,95]. The Taiping revolts started in China in 1851, but did not have major effects on the silk market until 1860, when hostilities reached the silk-producing area of Shanghai. The total Chinese export of silk in the period 1863 to 1866 was halved [7,92]. As Chinese exports dropped, Japanese silk entered the world market. The prices of silk in Japan doubled as their value was aligned with international prices, boosting production. It can be argued that Japanese exports at this time saved the European silk industry, as healthy silkworm eggs were being imported into the Mediterranean area to avoid pebrine, with Japan becoming the sole supplier of eggs by the end of the 1860s [7,21,93]. In 1870, Louis Pasteur recognised the silkworm disease as being caused by the presence of dark corpuscles, and developed a method to assess the health of egg batches and prevent the spread of the disease by isolating the moths. The European sericulture industry started recovering, but the market had changed: the epidemic had made silk harvests unpredictable and prices dependent on imports from Asia, making silk more expensive. This period of high prices played a key role in the push for innovation and technical progress in the silk industry [7,31,94,96].

As the European silk industry started recovering from the events of the 1850s and 1860s, another important factor developed: the ‘democratisation’ of silk. At this time no other fibre could yet compete with silk in the luxury fabrics market: rayon, which was often marketed as “artificial silk” and commercialised in 1896, mainly competed with cotton rather than silk, and nylon, the first true substitute for silk, especially for the production of stockings, did not enter the market until 1940. The demand for silk started to consistently grow, as for the first time it was not being purchased only by the richest, but it was becoming a mass consumption item for the growing middle class. The extent of this growth was such that global consumption of silk increased more than fivefold from the triennium 1875–1877 to the triennium 1927–1929 [7]. The increase in demand, newly developed technologies such as steam-reeling, and the advent of the US silk industry all contributed to a drastic change in the nature of silk goods too: while still at times being used for luxury pieces, silk was being used for the mass production of average quality, “cheap” pieces [7,97,98].

The weighting of silk was economically convenient: it increased the diameter of silk yarn meaning fewer threads per unit surface were needed. It is no coincidence that its popularisation happened as the industry was hit by a major crisis in production, which saw a significant increase in the price of raw silk, and that it remained in fashion as the demand for silk goods increased in the following 50 years. As the average quality of the mass-produced silk pieces in the early 20th century decreased, the earlier frowned-upon practice of loading the silk became more and more accepted [4,6,7,21]. This is evident also from the terminology used in patents: in earlier ones the inventions were simply called “treatments” and never explicitly mentioned the advantageous increase in weight, but used expressions such as giving “a greater body to the silk”; patents from the 20th century all had titles along the lines of “process for the weighting of silk” [30,71].

The growth in the silk market, which saw the quantity of silk traded on a global scale increase 20 times between the 1820s and the 1920s, was abruptly interrupted from the 1930s due to the combined effect of economic crises, World War II, and the invention of good artificial substitutes for silk. With the end of this golden era for the silk market, weighting in the form it had been most popular, also saw its end. While probably still practised to some extent throughout the 1930s and 1940s, its use gradually decreased until it went completely out of fashion, being substituted from the 1950s by the newly invented polymer treatments, which are in some form still practised today [7].

The practice of metal salt silk weighting was an extremely popular, yet short-lived one, and it was a product of the economic and historical context of the late 19th and early 20th centuries. While the context is much more complex, and many more factors were at play than can be discussed here, this practice would not have developed the way it did had it not been for the concurrence of a series of specific historic and economic factors.

4. Conclusions

In order to study the role metal salts may be playing in the degradation pathways of historical weighted silks, it is essential to understand how these were produced. This research analysed 136 different weighting methods, 83 from patents and 53 from dyers’ manuals and books, as well as several patents describing machinery and other treatments related to weighted silks.

Results highlighted how tin salts were the centre of the research associated with the weighting of silk from the 1880s. They were applied to all types and colours of silk, with some variety in application methods. In particular, weighting in the form first suggested by Neuhaus, which consisted of subsequent treatments with tin tetrachloride, sodium phosphate, and sodium silicate, was the most influential development in research in the weighting industry. For silk in dark shades, iron in combination with natural tannins remained the most popular choice of weighting method, but from the 1880s onwards, with the addition of tin salts in half of the identified methods.

Written sources discussing the negative effects of weighting on the finished fabric were published as early as the 1900s, indicating how contemporary practitioners were already aware of the potentially harmful effects of the practice. At the same time, many “protective treatments” and ways to avoid such negative consequences were patented, hinting at the fact that through time the practice was perfected and optimised and potentially less injurious to the silk fibres.

Silk weighting was discussed as the product of a very specific historic and economic context, which not only allowed the practice to develop, but encouraged it. While it was widespread, the practice was in fact not long-lived. Within two decades, it reached its peak in popularity and saw the start of its decline in research, and most likely also in use, as the demand for natural silk goods dropped.

The knowledge gained from this study will be linked and compared to what is observed in historical silk objects by comparing the results with those obtained from an ongoing survey using portable X-ray fluorescence spectroscopy in several European museum collections. Future steps in the research also include the reproduction of weighted silk fabrics according to methods that are representative of the most commonly published procedures: tin followed by phosphate; tin followed by phosphate and silicate; and iron tannins. The goal of these reproductions is their chemical characterisation and study of their differences in behaviour under artificial ageing in order to understand the degradation pathways of different weighted silks and consequently contribute to the development of targeted restoration treatments.