First experimental evidence of hop fibres in historical textiles

Hop (Humulus lupulus) has been used in Scandinavia since at least the ninth century AD, as documented through archaeological findings and written, historical records. The written records are mainly focused on the use of cone-shaped flowers for beer brewing and medical purposes, but there are also records, for example, from the famous Swedish botanist Carl von Linne, who mentions the use of hop fibres for textile production. However, until now no experimental investigations have been published on the use of hop fibres in cultural heritage objects. A major reason for this has been the lack of a suitable characterization method. Hop is a bast fibre, just as flax and hemp and bast fibres cannot be distinguished from each other by simple optical inspection. Recently a new identification method for hop fibres was published by the authors of this article. Here we apply the new method in an investigation of two Swedish cultural heritage objects: (i) a woman’s garment from the nineteenth century, which was labelled as having an upper section made from coarse linen and a bottom section made of hemp and hop and (ii) a textile fragment from an eighteenth-century textile sample book, which was labelled as being made from hop. We show that the woman’s garment is made with hop and hemp fibres and the textile fragment from the textile sample book is made with hop. Our work provides the first direct proof that hop fibres were used for textiles in the past.


Introduction
Hop (Humulus lupulus) is an ancient perennial crop plant, native to the Northern hemisphere. The oldest cultivated archaeological findings from Scandinavia, where it is clear that the findings are hop, are macrofossils from Birka, dating back to the ninth century AD (Hansson 1996, 129). Hop is frequently mentioned in historical, written records. The main emphasis is on the use of hop flowers for beer brewing, but other applications are also mentioned: hop flowers were applied for medical purposes (i.e. sleeping draughts) and for embalming and placed in burial coffins, for example, as filling in pillow cases (Strese and Tollin 2015, 263-273).
One of the oldest parts of the Frostathing law (Frostatingsloven), coming from the twelfth century, mentions cultivation of hop Book XIII, no. 11 (Hagland and Sandnes 1994, 93). In the Middle Ages in Norway, it was a duty for all farm owners to cultivate a certain amount of hop plants (Høeg 1976, 385). The same was the case in Sweden, where hop growing was obligatory from 1414 until 1860 (Karlsson Strese et al. 2014). On the other hand, records of cases of restrictions, where hop production was forbidden within certain areas and time periods, can also be found (Lankester 1840, 68).
It is documented through written records that hop fibres extracted from the plant stem were used for textile production in Scandinavia from around 1600 up to the nineteenth century (Bromelio 1687, 66-67;Schissler 1750, 214-216;Hald 1980, 130;Strese and Tollin 2015, 255-256). Carl von Linne mentions in his Flora oeconomica the use of hop for textile production. He writes that if the hop stalks are retted, they can be used for yarn similar to hemp (Linné and Aspelin 1749, 60-61). In 1773, the Norwegian topographer Gerald Shoning describes a travel to Surnadal (Norway). He mentions that in 1758, hemp, hop and also linen goods were imported to Trondheim (Schøning 1778, 10). He also states that flax, wool, hemp and hop were used to make fabrics (Schøning Electronic supplementary material The online version of this article (https://doi.org/10.1007/s12520-020-01171-6) contains supplementary material, which is available to authorized users. 1778,20). In 1781, Fischerström (1781) comments that hop stalks are normally thrown away but that one ought to do as in Jamtland and Medelpad, where they are used to make a weave, which is stronger than flax and hemp. (Fischerström 1781, 486). A fairly recent source mentions that hop fibre quality can vary a lot (Tobler 1938, 84-87). Experiments with substitute materials for textiles were also referred by Freund (1972, 7).
The most widely used textile plants in Scandinavia until the beginning of the twentieth century were flax (Linum usitatissimum) and hemp (Cannabis sativa). Hemp was used for cordage and coarse textiles, but there are also examples of the use of hemp for finer fabrics (Skoglund et al. 2013;Skoglund 2016). A few cases are documented, where stinging nettle (Urtica dioica) has been used for textile production (Hald 1942, 29-49;Bergfjord et al. 2012). Hop fibres were most likely not a very commonly used material compared with other textile materials.
Flax, nettle, hemp and hop are all bast fibres and it is not possible to distinguish them by simple optical inspection (Bergfjord and Holst 2010;Bergfjord et al. 2010;Haugan and Holst 2014). This may well have led to some textiles being incorrectly labelled as made of flax in various museum collections. It should also be noted that during the eighteenth century in Scandinavia, the term linne (Swedish) and lin (Norwegian) became common as a term to describe a plain weave textile irrespective of what it was made of. Earlier, a plain weave textile was often referred to as laerred or lärft in Swedish (Geijer 1979, 17). The terms linne and lin are however also used specifically for textiles made from flax. The modern word for flax is lin in both Swedish and Norwegian. The difficulties in terminology concerning linen also apply to the German Leinwand (Küster-Heise and Mitschke 2011, 159).
In order to find out what plants have been used to produce historical textiles, systematic investigations of objects in cultural heritage collections using the appropriate identification methods are necessary (see, e.g. Lukešová et al. 2017). Precise knowledge of material use in cultural heritage collections is important because it will enable better understanding of resource management in the past.
In this article, we present the first investigation on cultural heritage objects performed with the specific aim of finding out if they are made of hop; we use a very recently developed identification method (Lukešová et al. 2019).

The samples investigated
We investigate two historical objects in this article: the first object is a woman's garment from Jamtland County in Sweden (NM.0131474, left), belonging to the Nordic Museum in Stockholm (Fig. 1). It was donated to the museum in 1917. According to the museum accession record, the donor stated that the upper section is made of coarse linen fabric and the bottom section of hemp and hop (Redogörelse för Nordiska museets utveckling och förvaltning år 1919, p.11). The garment was probably produced in the middle of the nineteenth century. It is written in accession record that it was around 65 years old when it was donated to the Nordic Museum (https://digitaltmuseum.se/011023635901/ overdelssark, downloaded 28.4.2020).
The upper section is made of twill fabric, which is rather greyish and soft in its appearance compared with the bottom section which is made of coarse tabby with a yellow tinge. The object is described in Skoglund (2016). It is stated here that a Herzog test fibre analysis suggests an upper section made with a mixture of flax and hemp and/or hop and a bottom section made with hemp and/or hop. No further details to the analysis are presented (i.e. regarding thickness of the fibres investigated).
The second object is a textile fragment glued onto a sheet of paper in a Swedish textile sample book (NM.0405398+) from 1766 ( Fig. 1, right). The book presents a sample collection of textiles produced in the eighteenth century. The purpose of assembling the collection was to inspire an increase in Swedish textile production.

Methods, including sample preparation
Samples of both objects-the woman's garment from Jamtland and the textile fragment from the textile sample book-were carefully extracted and investigated by white light transmission and polarized light microscopy. In addition, microchemical tests using cuoxam-tetraamminediaquacopper dihydroxide [Cu(NH 3 ) 4 (H 2 O) 2 ](OH) 2 were performed at the end of investigation, in order to investigate the swelling behaviour (Luniak 1953, 80;Wülfert 1999, 281-282, 320;Stratmann 1973, 58-62). The investigations were performed following the recently developed identification method for hop fibres (Lukešová et al. 2019). See also Fig. 2.
For the Jamtland garment, four core samples were extracted since it was made of two different fabrics: two samples of the weft and warp system from the upper section (samples 1 and 2) and two samples of the weft and warp system from the bottom section (samples 3 and 4). For the textile fragment from the textile sample book, we only sampled the thread system of the shorter side (sample 5); it is so small. Both thread systems (warp and weft) show similar thread thickness, spin direction and colour when observed by stereomicroscope. We carefully evaluated ethical issues when sampling and concluded we perform the tests on one thread system only.
The samples were investigated using a polarized light microscope Leica DM750 P. A full wave compensator (λ = 530 nm) was used for the modified Herzog test (Herzog 1922(Herzog , 1943Haugan and Holst 2013). Photographs were taken using the camera Leica MC170 HD and software LAS V4,9. Fibres  were first observed in transmitted white light. Polarized light was used for performing the modified Herzog test. Fibres thinner than 20 μm were not used for the Herzog test since an experience with reference samples has shown that they may give misleading results. A demonstration video on how to perform the Herzog test can be found on https://youtu.be/ sC9GlUKjBDE.

Results
We followed the diagram elaborated for the identification of hop fibres shown in Fig. 2.

Polarization microscopy and the modified Herzog test
All samples except sample 1.5, which had no suitable region for testing, show Orange I in 0°and Indigo II in 90°position according to Michel-Levy birefringence chart when performing the modified Herzog test (Fig. 3).
Numerous crystals, probably calcium oxalates or other phytoliths, were clearly visible in all sub-samples except sample 1.4 (this is not used as an identification criterion).

Microchemical tests using cuoxam
Cuoxam, also called Schweizer's reagent, is an established tool for fibre identification since it causes swelling typical for species. All tested sub-samples show irregular undulation when swelled in cuoxam (Fig. 4, upper left and right) which together with the Herzog test result indicates hop. Samples 3.5; 4.3 and 4.4 show in addition harmonica-like folding of the middle lamella on some fibres, which indicates hemp (Fig.  4, lower left and right).
All sub-samples except 3.5 show clearly visible remains of protoplasm in the lumen. Sub-samples 1.3; 1.5; 2.4; 3.4; 4.3 and 5.4 showed a typical rounded edge of a fibre with a plasma thread sticking out (Fig. 4, upper right).

White light microscopy
All samples show strong, irregular thickness variations along the fibre lengths. This is one of the most characteristic features for hop (Fig. 5, lower left). There are wide flattened regions without cross marks that are even and smooth (Fig. 5, upper left and right). These often alternate with regions containing frequent cross marks and dislocations. All original samples show frequent flexions (Fig. 5, lower right). Undulated fibres (many twist flexion after each other) that might remind one of cotton fibres are also common.
We conclude that the upper section of the woman's garment NM.0131474 is made with hop (Humulus lupulus) and the bottom part is made with a fibre blend of hop and hemp (Humulus lupulus and Cannabis sativa). The textile fragment from the textile sample book (NM.0405398+) is made with hop (Humulus lupulus)-only one of the thread systems was investigated, because of the limited amount of original material.

Discussion
We have conducted a fibre identification analysis of two historical objects: a woman's garment (NM.0131474) and a textile fragment from a textile sample book (NM.0405398+).
Based on the behaviour of fibre samples from the two objects in polarized light, characteristic swelling in cuoxam and distinctive fibre morphology using the identification method (Lukešová et al. 2019), we conclude that for the garment NM.0131474, the upper part is made with hop fibres and the bottom section is made with a fibre blend of hop and hemp fibres. The quality of the upper section is rather soft and fine compared with the bottom section. This shows that the textile quality is a result of fibre processing and selection and is not an inherent quality of the plant fibre species used.
Our fibre analysis result for the upper section differs from that of Skoglund (2016, p. 88), who claims that flax is also present. Of course, we cannot exclude that sampling on two different sections of the garment may contain different fibres. Alternatively, if very thin fibres (less than 20 μm diameter) were investigated, a false result is possible as investigation on reference samples have shown. It is important to take into consideration that methods in microscopy, such as fibre analysis and microchemical tests, are comparative studies that build upon each other. Note also that the identification method used here (Lukešová et al. 2019) is for cultivated hop (Humulus lupulus). Wild hop has not been investigated. It is very probable that the objects investigated here are made of cultivated hop, but we cannot exclude wild hop completely.
We note that the original museum accession record states that the upper section is made of coarse linen and the bottom section of hemp and hop. Strictly spoken this is not wrong, since linen just refers to the weave, but as mentioned in the introduction, linen is often taken to mean flax, and when other types of fibres are explicitly mentioned, flax is the natural association. Another recent report states that the upper section is made of hemp (Frankow 1992, p.75), which is incorrect. The fabric sample from the textile sample book (NM.0405398+) is made with hop fibres. This agrees with the information in the textile sample book.

Conclusion
In this paper, we present the first experimental evidence of hop fibres in historical textiles. The fibre identification is based on the behaviour of fibres in polarized light, characteristic swelling in cuoxam and fibre morphology following (Lukešová et al. 2019). Both objects are investigated: the woman's garment (NM.0131474) and the textile fragment from the textile sample book (NM.0405398+) confirm the use of hop fibres. Our results highlight the importance of careful material analysis of cultural objects. Precise knowledge of material use in cultural heritage collections is crucial because it is necessary for understanding resource management in the past.