THE EFFECT OF HUMIDIFICATION ON ARTIFICIALLY AGED TIN-WEIGHTED SILKS

In the late nineteenth and early twentieth centuries a variety of weighting agents were applied to silk yarns and fabrics to increase their weight in an attempt to make a greater profit on their sale. In addition, the weighting process imparted a more desirable drape to silk, making it more popular with consumers. Consequently, weighting became an almost ubiquitous practice during this time until sanctions on the amount of acceptable weighting were enacted in the 1930s due to concerns around the rapid deterioration of heavily weighted silks. However, many of these textiles have been accessioned by museums and other cultural institutions and are now in need of extensive conservation. Weighted silks, particularly those weighted with tin salts, present a unique problem in textile conservation in that the medium itself is compromised and subject to continued degradation. While treatments to physically stabilize or consolidate already damaged weighted silk are commonly used, research that has aimed to slow or reverse the effects of these metallic salts have been largely unsuccessful. However, hydration of unweighted historic silk has been shown to increase the tensile strength of treated samples. This research investigates the effect of moisture level on artificially aged tin weighted silk. This study applied two of the most common tin-weighting procedures and then artificially aged samples of weighted silk habutae. Those samples were then subjected to a series of humidification treatments and complete submersion in water with the aim of increasing the weakened fiber's tensile strength without affecting the drape of the fabric. Samples were evaluated for breaking strength and cantilever bending length (as a proxy for drape) before and after receiving treatments. ACKNOWLEDGMENTS I would like to thank my major professor Dr. Martin Bide for all his patience and guidance. This research would not have been possible without his wisdom. I would also like to thank Dr. Saheli Goswami and Dr. William Euler for their dedication and time serving on my thesis committee. Thank you to Kelly Kummers for being my tensile testing troubleshooting guru. Thank you to Rebecca Kelly for all her help, both with the conservation aspects of this research and for allowing me to rant when things went sideways. I would also like to thank my family and friends for their support and innumerable long distance phone calls. Thank you to my parents for acting as a sounding board and a source of comfort whenever the unexpected happened. Chelsea and Helen, thank you for always being my sisters despite the hundreds, and even thousands, of miles between us.

However, many of these textiles have been accessioned by museums and other cultural institutions and are now in need of extensive conservation. Weighted silks, particularly those weighted with tin salts, present a unique problem in textile conservation in that the medium itself is compromised and subject to continued degradation. While treatments to physically stabilize or consolidate already damaged weighted silk are commonly used, research that has aimed to slow or reverse the effects of these metallic salts have been largely unsuccessful. However, hydration of unweighted historic silk has been shown to increase the tensile strength of treated samples. This research investigates the effect of moisture level on artificially aged tin weighted silk.
This study applied two of the most common tin-weighting procedures and then artificially aged samples of weighted silk habutae. Those samples were then subjected to a series of humidification treatments and complete submersion in water with the aim of increasing the weakened fiber's tensile strength without affecting the drape of the fabric. Samples were evaluated for breaking strength and cantilever bending length (as a proxy for drape) before and after receiving treatments.

ACKNOWLEDGMENTS
I would like to thank my major professor Dr. Martin Bide for all his patience and guidance. This research would not have been possible without his wisdom. I would also like to thank Dr. Saheli Goswami and Dr. William Euler for their dedication and time serving on my thesis committee.
Thank you to Kelly Kummers for being my tensile testing troubleshooting guru.
Thank you to Rebecca Kelly for all her help, both with the conservation aspects of this research and for allowing me to rant when things went sideways.
I would also like to thank my family and friends for their support and innumerable long distance phone calls. Thank you to my parents for acting as a sounding board and a source of comfort whenever the unexpected happened. Chelsea and Helen, thank you for always being my sisters despite the hundreds, and even thousands, of miles between us. iv LIST OF TABLES  Silks produced in the late nineteenth and early twentieth centuries were frequently treated with additives in an attempt to increase their weight before sale. This practice was particularly profitable when silk goods were sold by the pound, rather than by the yard. In addition to its economic benefits, weighting silk imparted a series of desirable qualities, including altering the drape of the fabric.
Weighting evolved somewhat by chance from coloration techniques that used metallic salts. These metallic salts were soon used independently of any dyeing process to increase the weight of both yarns and fabrics. Tin salts became the most popular, although a variety of other metallic salts and compounds were also utilized with varying degrees of success.
Unfortunately, the metallic salts, as well as the process used to apply them, created weakened fabrics prone to degradation. Although economic sanctions and regulations imposed in the 1930s curtailed the production of weighted silks, many garments and other objects made from these fabrics have been accessioned into the collections of museums and other cultural institutions.
The fragility and often advanced state of decay of weighted silks is a well known problem in textile conservation. However, very little progress has been made in attempting to stop or reverse the continued degradation of these textiles. Although physical stabilization with stitching and adhesives or the application of consolidants 1 are common practice, attempts to counteract the weighting agents themselves have thus far been unsuccessful.
A study on unweighted silk have suggested that re-hydrating silk fibers restores some of the objects' tensile strength. 1 How such a treatment would effect weighted silks was proposed as an area for further research by the authors, but at the time of this writing no further study on this matter has been conducted.
This study focuses on recreating two of the most common tin weighting methods, pink and dynamite, and artificially aging these samples. After aging, these modern samples were exposed to an extremely lowered relative humidity, a moderately increased relative humidity, and complete submersion in water to determine if rehydrating the weighted silk had a significant impact on the tensile strength of the fabric. The stiffness of the samples was also evaluated before and after treatments to determine if the treatments had a negative impact on one of the more desirable qualities imparted by weighting.
A review of the literature associated with weighted silks, both their commercial production and their conservation, and the degradation of weighted silks is provided in Chapter 2. The recreation of weighting methods, artificial aging of modern samples, and the application of various treatments to these samples are provided in Chapter 3.
The results of this testing are reported in Chapter 4. The conclusion of this study is provided in Chapter 5. Fibroin crystallizes into β-pleated sheets that form layers of lamellar-like microfilaments when secreted. 7 During the crystallization process, the evaporation of water from the fibroin to the atmosphere allows for the formation of micropores. 8 These micropores average 1.1 nm in size 9 and allow for water to remain bound to the fiber after secretion. 10 The micropores also account for the hygroscopic nature of silk; silk is capable of absorbing up to 30 percent of the fibroin's weight without appearing wet. 11 The fibroin protein is composed of two main components: the H-chain (heavy chain) and the small L-chain (light chain), 12 which are linked by disulfide bridges. 13 The earliest attempt to sequence the amino acids of the fibroin protein was in 1902, 14 but it was not until 1957 that the main hexapeptide repeat in the H-chain of Ser-Gly- In contrast, sericin is composed primarily of serine, glycine, and aspartic acid. 20 The high number of amino acids with polar side chains in sericin contributes to the hydrophillic nature of the protein, allowing it to be dissolved in water. 21 Sericin also has a high water content (86%), giving it an amorphous, random coil structure 22 unlike the β-pleated sheets of fibroin.
For the bave to be unwound from the cocoon and reeled for textile production, the silk worm pupa must be killed to prevent it from emerging and damaging the filaments. This is frequently accomplished through steaming the cocoons, a process that also softens the water-soluble sericin enough for the silk to be reeled. 23 Although some 1,500 meters of fiber are present in each cocoon, only 500-800 meters of this are capable of being reeled. 24 The remaining broken, shorter filaments are frequently turned into spun silk, a somewhat less desirable product. 25 Silk fabric with the gum-like sericin still present is referred to as "raw silk." 26 However, the sericin dulls the fiber, giving it a yellowed color, as well as stiffens it. 27 This frequently prompts the removal of the sericin both for aesthetic reasons and to improve the feel. 28 The process of removing the gum is known as "boiling-off" or "degumming" and is accomplished in a hot soap solution (typically 95 o C, 5-7.5 g/L soap). 29 Although the process softens the silk and improves its luster, both desirable qualities for consumers, removing the gum also removes 20 to 30% of the weight of the silk. 30

Silk Weighting
Despite some early attempts, American sericulture failed spectacularly after the 1839 mulberry tree market collapse. 31 However, a combination of increased 23 Matthews, Textile Fibres, 1913, 127. 24 Robson, Silk, 418-420. 25 Matthews, Textile Fibres, 1913 , 1929, 54. 6 its raw weight. 41 Adding more weight to the silk than had been lost in the degumming process was known as weighting above par. 42 Weighting also increased the silk fabric's thickness 43 and even the diameter of the yarn itself. 44 This increased yarn thickness also decreased the number of threads required to make the same surface area as unweighted silk. 45 Weighting stiffened the fabric and gave it a more desirable drape, 46 increasing its scroop (the highly desirable rustling sound silk makes when rubbed against itself 47 ) and even increasing the colorfastness of black dyed silk. 48 Weighting not only augmented profits for merchants and manufacturers, it imparted desirable qualities for consumers as well.
At its most extreme, weighting could extend a single pound of raw silk to more than twenty-five ounces, 49 a practice that was considerably profitable when silk was sold by weight, rather than the more modern yardage 50 and contributed to the booming American silk manufacturing industry. The swelling that resulted from the weighting process allowed manufacturers to use a lesser yarn count to create the same surface area as unweighted silk, 51 further lowering manufacturing costs and thereby increasing profits. 52 Some silks were weighted to such an extreme that they had to be passed through a breaking machine to restore some of the handle lost by the stiffening of the fibers before being sold. 53  Inorganic weighting practices began with combinations of iron salts with organic tannins to produce weighted black-dyed silks. 60 A similar increase in weight with substances that would not effect the final color of the silk was applied commercially using various forms of iron, tin, chromium, sulfates and chlorides of sodium, magnesium, 61 bismuth, tungsten, lead, antimony, and barium, 62 as well as with organics like glucose and gelatin. 63 Aluminum has also been identified on extant weighted silks. 64 63 Matthews, Textile Fibres, 1913, 551. 64 Miller and Reagan, "Degradation in Weighted and Unweighted Historic Silks," 101. 8 tannins and tin and iron compounds with their corresponding dates of popularity, can be found in . 65 Some, including sugar and tannic acid, were used commercially in the 1820s, 66 but were quickly replaced by other additives. Most early weighting agents were abandoned by the twentieth century, 67 either due to their lack of economic viability as a result of the weighting agent's price or the damage they caused to fabrics. 68 However, iron and tannin compounds remained in use specifically for fabrics that would be dyed black, as these compounds imparted a dark color to the silk, regardless of later dyeing. 69 Eventually, the commercial practice evolved into one based chiefly on salts of tin.
Marsh suggests that chemical weighting became a commercially successful in the 1850s, 70 although Federico claims that the first successful experiments with synthetic rather than natural weighting only occurred in 1856, with commercial viability coming decades later in the 1870s. 71 Another anecdote claims the discovery of tin weighting as a viable method was completely by chance in a Germany dyehouse in the 1850s: a worker threw a skein with a tin mordant at another worker where it landed in a vat of tannin instead. When removed from the vat, this skein was determined to weigh far more than the other skeins without the tin mordant. 72 Regardless of precise origin, tin weighting soon became the most popular method, reaching its peak in the 1890s. 73 Tin salts had the advantage of imparting far more weight with far less chemical input and effort on the finisher's part. A single cycle of tin and sodium phosphate baths could achieve the same extent of weighting as fifteen similar passes using iron compounds. 74 Other metals like aluminum could not approach the amount of weight imparted by tin compounds. 75 In spite of the expense of tin, even before the outbreak of World War I and the limits on its use created by wartime rationing, it was still considerably less expensive than alternatives that imparted the same weight gain like chromium salts. 76 Tin was also preferred as it, unlike other weighting compounds like tannin, could be used on lighter-colored or even white fabrics. 77 The tin on the tin-weighting could also behave as a mordant when the silk was to be dyed. 78 However, if the weighting was uneven, the weighted silk could produce unlevel dyeings and the weighting even reduced the affinity of certain acid dyes for the fiber. 79 Tin also served as a base for additional weighting agents that would otherwise fail to adhere to the silk, 80 further increasing the amount of weight that could be imparted and lowering costs.
Technological advances improved the efficiency of the process and lowered the cost. New machines allowed for the recovery and reuse of tin left in the bath after weighting. 81 Centrifuging and pressure were applied to the silk after the bath to recover even more of the unfixed tin for reuse, 82 and baths of different weighting agents could be combined to reduce the amount of labor and time required to treat the silk. 83 Weighting was initially confined to processing yarns, rather than piece goods, 84 with warps frequently weighted more heavily than wefts. 85 Silk,198. 13 allowing the silk to dry between dips can affect the total weight gain. 111 Lack of precise control over the pH of the various baths in the process has also been blamed for the inconsistent application of weighting. 112

Deterioration of Weighted Silk
Despite all the advantages of tin weighting, the inherent flaws it created were recognized as early as the turn of the twentieth century. The reduced tensile strength and generally increased rate of deterioration when compared to unweighted silk were well known. 113 , 114 Weighting of 140% above par resulted in a loss of five-sixths of a fiber's strength when compared to unweighted silk. 115 The brittle and fragile fibers that resulted from the weighting process could lead to the silk's complete destruction with a few months of its manufacture, 116 often before the reaching the consumer. 117 The brittleness of degraded weighted silk result in what is known as shattering, in which a complete loss of extensibility in the fibers lead to their breakage when moved. 118 This breakage becomes so extreme that a textile can disintegrate completely over a relatively short time, even without handling. Perplexingly, not all heavily weighted silk has the same fragility, with the level of weighting often failing to correspond to the extensiveness of observed damage. 119 The weighting process itself was frequently responsible for the damage seen. The hydrolysis of tin(IV) chloride releases hydrochloric acid that remains on the silk.  Arts 60, no. 3092 (1912): 402. 117Marsh, Textile Finishing, 301. 118Hacke, "Weighted Silk," 8. 119Matthews, Textile Fibres, 1924 Although the secondary bath of sodium carbonate or sodium phosphate, which was often alkaline, was used to neutralize the hydrochloric acid freed by the earlier hydrolysis, 120 silk dissolves completely in high enough concentrations like a 30ºTw solution of hydrochloric acid. 121 The hydrochloric acid formed during the weighting process was directly blamed for the damaged silk, even after "considerable" time had passed since the treatment. 122 A variety of other problems were occur with weighted silks. In addition to decreasing the tensile strength, weighting often resulted in red spots, making the goods unsellable, and subsequent extreme tendering of these spot as copper or iron in wash water and sodium chloride from sweat enabled catalytic oxidation of the weighted silk. 123 These spots could occur within months of production. 124 Heavily weighted silks were also purported to be highly combustible, 125 although given the use of phosphorus compounds in flame retardant applications, this id doubtful. Perspiration, particularly when combined with heat, was also known to increase the rate of deterioration in weighted silk and decrease its tensile strength. 126 Many of the defects seen in lightly colored weighted silks were only amplified in black weighted silks. Black silks were traditionally weighted to a greater degree than other silks, often through a combination of tannin and tin, increasing their degradation in comparison to less weighted silks. 127 Textile Fibres, 1924, 310. 16 collections. 137 This has led to doubts over whether the metallic compounds are actually responsible for the damage seen in weighted silks.
The use of poor quality, and therefore already predisposed to damage, silk in weighting is one possible cause of this discrepancy. 138 In addition, the harsh conditions surrounding the weighting process, together with the fluctuations in pH that accompany bleaching, might be responsible for the observed damage. 139 However, due to the white color of degummed silk, bleaching was only deemed necessary when a "snow-white" color on the finished product was desired, 140 making the bleaching process unlikely to account for much of the damage seen in extant weighted silk.

Decline of Weighted Silk
The concerns over the rapid deterioration of tin-weighted silk led to early twentieth century attempts to counteract or limit the treatment's unfortunate effects.
Recommendations for the composition of the tin baths that would impart the most weight without damaging the fiber were abundant even before the turn of the century. 141 Special care was taken to remove impurities present in water from the different baths in the process and remove any fatty acids from soaping the silk prior to weighting as they were purported to injure the fiber when combined with tin(IV)  Silk,191. 17 weighting process that were frequently blamed for the tendering. 143 Thiourea was used to lessen weighted silk's sensitivity to light, 144 while tannin was proposed to counteract the crystallization of tin-silicate colloids that supposedly resulted in the loss of tensile strength. 145 By the twentieth century an extensive number of patents were given for compounds that purportedly imparted some protective quality to weighted silk to prevent its deterioration. 146 5 % hydrofluosilicic acid and hydrofluoric acid were also proposed as a means to strip inorganic weighting agents from silks with extreme weighting above par altogether to make them less brittle. 147 Alternative processing methods to reduce the damage caused by the baths' conditions were explored as well. Foam weighting was developed as early as 1915 in an attempt to prevent damage to the silk by the alkaline nature of some of the baths by reducing the amount of time spent in damaging solutions 148 , 149 Social pressure and economic regulation eventually emerged. The so-called "pure dyes" of latter part of the nineteenth century professed to be silks that were not weighted above par. 150 197 , 198 Wet cleaning is seen as a somewhat extreme step that should be avoided if at all possible due to the potential for damage to a textile that can occur during treatment. 199 However, early twentieth century research found that, when wet, weighted silk is less brittle than when dry. 200 While complete submersion in water to rehydrate fibers is not often possible or advisable for compound objects like the historic costumes found in museum collections, humidification is. This research aims to investigate the effect of complete wetting and humidification on weighted silk with regards to the possible benefits to tensile strength of the weighted silk fibers and how such a treatment will affect the drape of the textile. Hydrating the fibers will hopefully restore some of the A second dipped samples (40 cm (warp) x 15 cm (weft)) was prepared using the dynamite weighting method described in . The alternating cycles of pink and sodium hydrogen phosphate used in pink weighting were conducted three times. After the final bath of sodium hydrogen phosphate and subsequent overnight drying, the samples was immersed in a bath of 250 mL of aluminum sulfate at 7.0° Bé (9.54% solution) that was weakly acidified with sulfuric acid to achieve a pH of 4±0.5 for one hour at 60 ±2 °C before being rinsed in deionized water for one minute with After padding, samples were stored in the dark, at 60 % RH and 21 ºC. Padding at 50 psi was adopted as the application process for the remainder of the experiment and is the methodology that the conclusions of this study are based on.

Weighting
To determine the weight gain of silk samples at each stage of the weighting process, a circular die was used to cut samples and the samples were then weighed.
These samples were then weighed and compared to one another. Larger samples were cut with a 2.7 inch diameter die, and ,when less material was available, a 0. the tray with masking tape to prevent movement from air circulation in the instrument.
Areas of the sample that were direct contact with the tape were discarded. Samples were also aged in an oven at 80 ºC 24, 72, 96, and 120 hours. Samples were suspended between two wooden blocks with finishing nails to allow air flow around the samples.
Areas of the samples that were in contact with the wood were discarded. 10 hours in the Q-Sun weatherometer was eventually chosen as the aging condition for the experiment.

Humidification and Wetting
A control group of aged samples of both pink-and dynamite-weighted silk was retained with no further treatment and stored at 21ºC and 60% RH. Aged samples were exposed to either complete submersion in a bath of deionized water or an altered humidity. Complete submersion was conducted at 21ºC for 30 minutes. Samples exposed to altered humidity were placed in the Datacolor Conditioner with controller RP1 for 90 minutes until equilibrium was reached where no further weight was gained or lost. A slightly elevated humidity of 70% RH was applied to one group to simulate the standard testing conditions of ASTM D1776 of 21±2ºC and 65± 5% RH. A final group was exposed to a lowered humidity of 30% RH before being stored in a silica desiccation chamber.

Stiffness
The stiffness in the weft direction of the unweighted habutae, unaged pink-and dynamite-weighted habutae, aged pink-and dynamite-weighted habutae, and aged and treated pink-and dynamite-weighted habutae was evaluated using a "Shirley" Stiffness Tester according to ASTM D1388. Samples that received a treatment of 30 complete submersion in water were evaluated both wet and after they had been allowed to recondition overnight. Pink-weighted samples that were completely submerged in water were blotted to 0.48±02 grams (approximately twice the dry weight) before being evaluated while dynamite-weighted samples were blotted to 0.50±0.02 grams (approximately twice the dry weight). Neither groups conditioned at 30% RH or 70% RH was allowed to return to storage conditions, but rather were tested immediately upon conditioning. Samples that were completely wetted were evaluated both while wet and after they had been allowed to return to storage conditions overnight. Pink-and dynamiteweighted samples that were completely wetted in water were blotted to 0.040±0.002 grams (approximately twice the dry weight) before being evaluated. Groups exposed 31 to altered humidities were not allowed to return to storage conditions, but rather were tested immediately after conditioning.

CHAPTER 4 FINDINGS
This chapter presents observations related to sample preparation. These finding helped to ensure that samples were properly weighted and aged before the main experiment was conducted. This section also reports how samples' stiffness and tensile strength were impacted by the various treatments.

Weighting Recipes
Sodium hydrogen phosphate (Table 1) and aluminum sulfate (Table 2) were simply weighed out and mixed, often with the addition of low heat (less than 60ºC).
Sodium trisilicate, however, required the addition of a small amount of sodium hydroxide (NaOH) and the application of heat to solubilize it in water (Table 3).   Table 3. Relationship Between Density, ºBaumé, and Concentration of Na 2 Si 3 0 7 Determination of the density of tin (IV) chloride required more care, given the exothermic dilution and the offgassing of HCl. A balance was set up under the fume hood and the entire bottle of tin (IV) chloride was weighed. A small amount of tin (IV) chloride was poured into a beaker of deionized water and the weight lost from the stock bottle recorded. The diluted solution was allowed to cool before transfer to a volumetric flask and the appropriate amount of deionized water was added. The density of these solutions was then calculated to determine how much tin (IV) chloride was necessary to achieve the required solution density. The results can be found in 34 Mass of Na 2 0 7 Si 3 Mass of NaOH (g)

Density (g/mL)
Percent Solution°Baume   Table 4. The results were less precise than with other solutions: air movement caused variability of approximately 0.10g in the balance reading. The J.A. King Sample Cutter is a standard device (ASTM D3776) to determine fabric weight, and was used when sufficient material was available. The large size of these samples and the need to take multiple specimens required that a more efficient means be used.
A set of gasket cutting dies was purchased to perform the same measurements on a smaller scale. A 0.5 inch diameter hollow punch die was purchased. This die (intended for rubber) was not sharp enough to cut through fabric. It was sharpened by hand using 320 and 220 grit sandpaper and mineral oil as a lubricant. The average weights of the 0.5 inch diameter circles at various stages in the weighting process can be found in Table 5  Dipped samples had the greatest weight gain, with padded samples gaining less.
Within the dipped samples, those that received a fourth pass of tin-phosphate gained the most weight. However, this extra pass was abandoned as it made samples so stiff they became difficult to cut to the proper size for tensile testing. The pressure with 204 Matthews, Textile Fibres, 1924, 311. 36  which the samples were padded also affected the weight gain, with samples padded at a lower pressure gaining less weight overall than those padded at a higher pressure.
Surprisingly, the average weight gain differed between the initial batch at 50 psi and the final batch of samples padded at the same pressure used in the experiment.
Patents for the continuous application of the weighting process existed by 1928. 205 Adapting this technique to a pad-batch methodology would decrease the amount of labor expended agitating the various baths as well as reduce the variation in weight gain across a given piece by ensuring that the contents of the baths are evenly distributed. This application would also reduce the amount of bath lost in each pass, as the excess is easily recovered from the troughs. In this work, rather than run the samples in a continuous process as described in the 1928 patent, a semi-continuous method was used due to space and equipment limitations. The semi-continuous method also allowed the amount of time dipped samples were in solution to be mirrored in the padded samples. However, neither the pink-weighted nor dynamiteweighted samples prepared in this manner achieved even remotely the same weight gain as seen in the dipped samples. In addition, a white precipitate in the sodium phosphate bath after samples were padded suggesed that the unfixed stannic chloride was reacting in the padding tray rather than on the fabric. Padding was initially abandoned as a potential application method for this experiment as a result.
However, a considerable amount of strength variation was seen among the pinkand dynamite-weighted samples produced by dipping (Table 6). This variation was deemed too large to provide reliable data to asses the potential effect of any humidification treatment. Even though the padded samples did not achieve the same weight gain as the dipped samples, they did not display this same strength variability (Table 6). Therefore, padding was revisited as a potential application method. Initial tests used silk twill (Test Fabrics No. 611) padded with water and assessed for liquid pick up using the 2.7 inch diameter die. The pressure between the rollers was varied with each padding, ranging from 55 to 10 psi. The percent weight gain of the 2.7 inch diameter circles can be found in Table 7. The initial weighting tests described earlier had been conducted using 50 psi. A pressure of 25 psi gave a higher pickup and should thus allow more of the weighting solutions to remain on the fabric and correspondingly increase the weight pick up in an initial test run with a single pass of tin-phosphate.

Percent Weight Gain
samples (both with three tin-phosphate passes) was completed at 25 psi, the total weight gained in both the pink-and dynamite-weighted samples was lower than that gained by the earlier samples padded at 50 psi (Table 5). Despite the lowered weight gain, the samples padded at 25 psi had similar cantilever results as the higher psi counterparts ( Table 8). As a result, the initial 50 psi of the earlier samples was used for the main experiment. High humidity, drastic pH environments, and long term UV exposure have been rejected as accurate mimics for natural aging as they fail to replicate the same analytical markers seen in analytical aging on unweighted silks. 206 Therefore, heating in air was initially chosen as it has been shown to achieve a similar reduction of chain length and decrease in tyrosine present seen in naturally aged silk. 207 80ºC was chosen as the temperature for aging based on a previous study. 208 However, preliminary testing indicated that heat aging would take far too long to be a feasible method. After 120 hours only a small reduction (approximately 20%) in tensile strength was seen. In 206 Vilaplana et. al., "Analytical Markers," 1447. 207Vilaplana et. al., "Analytical Markers," 1442 Table 9 and the light preliminary tests in Table 10.  (p= 0.56) between this group and the control is unsurprising, given how little the treatment differed from storage conditions. All other treatments lowered the bending length of the aged samples. Water acts as a plasticizer, with humidity treatments used to increase historic textiles' flexibility, 210 and moisture can remain in the fabric, maintaining some of the newly imparted flexibility even after a humidity treatment. 211 This accounts for the decreased stiffness in the samples exposed to water as well as those wetted (p=7.13x10 -7 ) and later dried (p=4.11x10 -3 ). The decreased stiffness of the fabric while at a lower RH (p=4.15x10 -7) is surprising, as silk is known to become more rigid with desiccation. 212  (Table 13). Again, the lack of statistically significant difference (p=0.55) between this group and the control is unsurprising, given how little the treatment differed from storage conditions. All other treatments lowered the bending length of the aged samples. Similar to the pink-weighted silk, the decrease in stiffness of the wetted (p=8.44x10 -16 ) and wetted then dried samples (p=2.51x10 -6 ) was expected. However, the significant decrease in the desiccated samples was unexpected (p=4.15x10 -7 ). Numerous attempts to prevent slippage were tried and abandoned as ineffective.
These attempts included: cleaning the jaws with acetone, inserting a piece of double sided tape into the jaws to hold samples in place, attaching 320 and 220 grit sandpaper into the jaws, grinding the jaws themselves on 320 grit sandpaper with mineral oil lubricant to create a more level surface and even pressure on the samples, and attaching 1 mm thick rubber and Security Friction Tape (¾ inches wide, less than 1 mm thick), manufactured by United States Rubber Company, to the face of either one or both jaws. These methods either did not prevent slippage or resulted in samples that broke at the line of the jaw. Breaking at the line of the jaws also invalidates a sample's results.
Eventually the contact of the surface of the jaws was tested using the carbon and tissue paper method described in ASTM D5034 (2009). Incomplete contact of the jaw faces can allow samples to slip. However, no carbon whatsoever was transferred onto the tissue paper, suggesting a problem with the carbon paper. Contact was then assessed by holding a bright light behind the seam of clamped jaws. If light was visible through the seam, areas of the jaws were not making full contact.
Unsurprisingly, numerous pockets of light were visible through the clamped jaws.
Initial attempts to smooth the jaw surfaces and create full contact using 320 and 220 grit sandpaper and mineral oil were unsuccessful, with light still visible between the clamped jaws.
Permatex Valve Grinding Compound was then used to lap the jaws of machine to one another. A liberal coating of the grinding compound was applied to the inside face of one jaw and its mate was then rubbed against it in a random pattern for some 44 minutes. The grinding compound was removed from the jaws using acetone and the contact of the jaws retested. This process was repeated until the jaws were fully lapped to one another and no light was visible through the seam. This grinding was successful in preventing slippage of untreated habutae, but not for any samples that had been weighted with any application method or recipe. Eventually ¾ inch wide Security Friction Tape was inserted on the inner surface of each of the jaws. The combination of lapped jaws and Security Friction Tape provided the necessary contact and grip to eliminate slipping.
Once a method for ensuring samples did not slip from the jaws during testing was established, unweighted habutae, unaged pink-and dynamite-weighted habutae, aged pink-and dynamite-weighted habutae, and aged pink-and dynamite-weighted habutae with modified moisture content tested.
A two-sided independent t-test at a significance level of 0.05 was used to evaluate the change in mean tensile strength of pink-and dynamite-weighted samples after treatment. Each group consisted of 10 samples. Normality was assumed, despite some dependent variables not falling within the acceptable range of skewness and Kurtosis values (Table 14). Where possible, outliers were deleted and replaced with alternative data. This occurred in pink-weighted habutae at 70% RH, the control dynamite-weighted habutae, and dynamite-weighted habuate at 30% RH. However, outliers remained in the pink-weighted habutae at 30% RH.  (Table 15). Similarly, statistically significant differences were observed on all groups of dynamite-weighted silk (Table 16). Likewise, the significant decrease in tensile strength seen in the completely wetted samples was expected, as observed in tensile testing of unweighted, wet silk. 214 The significant difference observed between the control and humidification at 70% RH in both pink-and dynamite-weighted samples (p=0.04 and p=8.81x10 -11 , respectively) is surprising given the testing conditions' similarity to the control.
Likewise, the increase in strength seen in the dynamite-weighted silk (p=1.59x10 -3 ) at a lowered humidity is unexpected. Conventional preventative conservation holds that drastically decreased relative humidity can lead to embroilment of fibers. 215 A corresponding decrease in tensile strength would be expected, as seen in the pinkweighted samples. The difference between the pink-and dynamite-weighted samples at a lowered RH suggests that the different weighting processes have different sensitivities to humidity. Perhaps most surprising is the positive effect that wetting and allowing samples to dry had on the tensile strength of both pink-and dynamiteweighted silk (p=0.03 and p=0.03, respectively), supporting Hansen and Derelian's work on unweighted silk 216 and suggesting that complete wetting can increase the breaking strength of weighted silks. 213 Ross et. al Tin weighted silk was a common finishing technique in the late nineteenth and early twentieth centuries. What began as a method to compensate for weight loss in degrummed silk developed into weighting levels well above par. Silk that underwent such high levels of weighting is frequently fragile and prone to shattering. Although public outcry and federal regulations eventually led to its phasing out, objects that underwent this treatment are prolific in museum and other cultural institutions. The problems created by the weighting process not only led to dissatisfied consumers, it has led to considerable problems for its long term conservation.
This study developed a method for the even application of the two most common historical tin weighting methods. This enables future research on the conservation of weighted silk that does not utilize prized extant objects. Recreating weighted silk also eliminates any variation that might result from experimentation on objects that might have experienced different life histories that effect their aging process.
This research also suggests a potential future conservation treatment for increasing the tensile strength of tin weighted silks as well as suggesting how pinkand dynamite-weighted silks might have different sensitivities to humidity. Increasing the relative humidity by a mere 10% increased tensile strength, but did not impact the bending on both pink-and dynamite-weighted silks. However, maintaining an increased RH for extant objects is not recommended for well-established reasons. 48 Most intriguing was the effect of complete wetting and then reconditioning at an approximation of conditions more likely to be found in a museum. This treatment, although it altered the drape of the fabric, increased the tensile strength of both pink and dynamite-weighted silks. This suggests that weighted silks might benefit from wet cleaning or a similar process that results in complete submersion. However, this treatment did impact the drape of the fabric, as well as create a more fragile textile during the wetting process itself.
Pink-weighted samples had a reduced tensile strength at a lowered humidity, while dynamite-weighted samples had an increased tensile strength, suggesting that the manner in which an object was weighted, not simply the presence of weighting agents in general, effects silk's sensitivity to humidity. Particular care should be given to extant objects known to have undergone pink weighting to avoid a lowered humidity and the resulting loss of tensile strength.
Further research into how complete wetting effects extant objects with different life histories is necessary to confirm the applicability of this treatment to objects found in museum collections. In addition, care must be exercised when submerging weighted silk in water as, like unweighted silk, it becomes more fragile when wet. Further research into the long term effects of these treatments is also necessary. It is possible that the benefit seen by complete submersion is only temporary. Applying this treatment also requires a judgment evaluation of the benefits of increased tensile strength with the impact on drape. 49

APPENDICES APPENDIX A: PRELIMINARY TESTING RAW DATA
The following tables provide the raw data from the various preliminary tests that informed the main experiment.
Pink-Weighted Weight (g) of 0.5 inch Circles-Dipped Method Dynamite-Weighted Weight (g) of 0.5 inch Circles-Dipped Method