The Story of the “Qiulai” Gin Unraveled by Radiocarbon Dating, Material Characterization and Chinese Written Sources

An ancient table zither qin, an emblematic stringed instrument of traditional Chinese music, has been rediscovered in the museum collection of the Conservatoire National des Arts et Métiers (inv.4224, CNAM collection), Paris. This instrument named “Qiulai” qin, whose origin is poorly documented, can claim to be one of the oldest qin preserved in European collections; its state of conservation is exceptional. A thorough examination was carried out based on an innovative approach combining museum expertise, material characterization analyses (optical microscopy, VIS/IR/UV imaging, X-ray uorescence, SEM-EDS, Raman) and advanced radiocarbon dating technology (ECHoMICADAS). Our results highlight the great coherence with the traditional manufacturing practices mentioned in early Qing dynasty qin treatises and poems, in particular the collection of materials with highly symbolic meanings referring to the qin sound, nature and the universe. They also conrm the antiquity of the "Qiulai" qin in Europe by indicating that it was most likely made in a small interval of about 30 years at the turn of the 18th century.


Introduction And Research Aims
The qin or guqin ("old qin") is a seven-stringed Chinese table zither regarded as one of the oldest chordophones in China. Based on surviving early written sources and archeological ndings, the instrument is attested since the Zhou dynasty (11th-5th century B.C.) and its de nitive organological features were established during the Western Jin dynasty (265-316) [1]. Its construction appears simple but allows a wide variety of sounds, from deep bass to crystal-clear high-pitched notes, and nger strokes that can be or gentle or violent. This highly esteemed and mainly solo instrument connotes the re nement and sophistication of the literati, the Chinese scholarly elite, along with the practice of poetry, painting, calligraphy and chess [2]. The qin is considered as one the objects that best epitomizes Chinese thought [1].
Beyond the elaborate sound images it produces, the qin is also revered as a re ned artwork, whose shape is imbued with a strong symbolic meaning. The different parts of the instrument refer to the dragon, the phoenix, the sky and the earth. The making of a qin corresponds to an ideal such as a concentration of the harmony of nature, the energetic ow (qi), and all the forces of the Universe [1]. The materials selected are therefore invested with deep meaning, such as timber from temples that is impregnated with chanting, or from trees growing near water cascades, or in calm and secluded places, etc. Great value is attributed to the numerous cracks (duanwen) on the surface lacquer layer which are considered as a sign of "antiquity", one of the virtues embodied by the qin [3,4].
The ten-year Cultural Revolution (1966)(1967)(1968)(1969)(1970)(1971)(1972)(1973)(1974)(1975)(1976) in China had important consequences on the arts, including on the teaching and practice of qin and listening to this instrument [5]. Like other symbols strongly associated with the elites of the Old Empire, efforts were made to "target for destruction" multi-century guqin [6]. Nowadays, modern qin makers, recognized as the heirs of this ancestral tradition, perform their work with the remaining knowledge that survived the cultural compliance policy under Mao Zedong [7]. Considering its long history, its numerous cultural and symbolic associations and its crucial role in the shaping of Chinese culture, the qin -along with its musical repertoire -was included in 2008 in the UNESCO list of Intangible Cultural Heritage of Humanity.
To our knowledge, the qin rediscovered in 2012 in the collections of the Conservatoire National des Arts et Métiers (CNAM, Paris) is the oldest qin known in European museum collections. In view of the evidence in favour of its authenticity, its value is considerable. Research combining a series of scienti c techniques and early written sources remains relatively inaccessible and can only be found in Chinese scienti c literature or well-informed Chinese expert circles [8]. Likewise, while the literature on Chinese lacquers is abundant [9][10][11][12], few scienti c reports are available on the material characterization of early qin lacquers [8,13]. However, information related to the making of qin appears in the numerous poems praising the instrument and in ancient treatises that shed light on the origin and use of materials [14].
This article combines visual examination of the instrument by qin experts, material characterization analyses and advanced radiocarbon dating technology in an interdisciplinary framework. We will demonstrate that the scienti c investigations on the different parts of the qin from the CNAM collection are consistent with the practices of qin craftsmanship depicted in the following seventeenth and eighteenth-century written sources: the Yuguzhai Qinpu "Abiding with Antiquity" written c.1860 by Zhu Fengjie which is a very popular late Qing dynasty quqin handbook [3] and the Qinfang "Twenty Four Flavours of qin" which is an early treatise written c.1641 by Xu Shangying [4].

An exceptional Chinese zither in European collections
The qin under study is a Chinese table zither owned by the Conservatoire National des Arts et Métiers (CNAM, inv. n° 04224) and currently on loan to the Musée de la musique since 2015 (inv. n° D.04224) (Fig. 1). This qin is composed of a soundboard apparently made from a single block of wood, without a bottom, with a system of bridges, peg pool, a nut, two feet to attach seven silk strings, respectively on and under the forehead, the end (tail) and below the soundboard of the qin [15]. Thirteen round studs or emblems (hui) of mother-of-pearl are aligned to mark harmonic positions. The upper and lower parts of the qin symbolize the sky and the earth; two pillars under the soundboard in a circular and a crescent moon shape are called respectively the pillars of "Heaven" and "Earth" (Fig. 2). This qin was rediscovered in 2012 during the "récolement" of the musical instruments of the CNAM acoustic collection. Philippe Bruguière, the former Curator of non-European collections at the Musée de la Musique, suspected its old age. The Musical Instrument Museums Online database (MIMO) was queried to compare this instrument with other qin listed in partner museums, as well as for qin from private collections [16]. The qin has exceptionally rare features with a leaf banana shape known as jiaoye style and the absence of a bottom. The shape and the carved date situate its manufacturing period during the 17th or 18th century in China. For comparison, the "Gumeihua" qin with a similar shape in China (Forbidden City) was made during the Ming dynasty (1368-1644) and once belonged to the collection of the Emperor Qianlong (1735-1796) [17]. The rediscovered qin is assumed to be one of the oldest qin kept in European museum collections.
Despite previous studies based on the CNAM archives and inventories, its provenance remains mysterious. The CNAM was created in 1794 following decrees issued during the French Revolution to promote national industry. The researchers of the CNAM gathered the following clues [18]: i-the presence of musical instruments is attested since 1814 at the CNAM with the installation of the early "Cabinet de physique de M. Charles" who was elected as a physicist to the Académie des Sciences in 1785 [19].
ii-the qin n°4224 is clearly mentioned in the inventories edited in 1849 at the Conservatoire for the creation of the acoustics gallery in 1851-1852.
iii-objects with inventory numbers around 4000 were presumably kept at the Académie des Sciences prior to their transfer to the Conservatoire. The Académie des Sciences was initially created in 1666 as an informal scholarly circle and kept collections of objects for scienti c studies and education. Some objects were taken from these collections to complete the collection of the Conservatoire after the French Revolution.

Inscriptions and translations
The inscriptions engraved on the back side of the qin are reported in Fig. 2. First, they refer to the name of the instrument "Qiulai" meaning "Autumnal resonances" (Fig. 2.a). The following inscriptions in carved Manchu reproduce an extract from the famous poem "Qinzan" "In praise of the qin" written by Li Bai (701-762) (except the rst verse and some slight modi cations in red ( Fig. 2.b). The bottom part of the inscriptions includes information on the making of the instrument ( Fig. 2.c) and Tang Kai's seal ( Fig. 2.d). Tang Kai was a well-known court zither player and zither maker, as well as a qin collector during the reigns of the Chinese emperors Yonzheng (1722-1735) and Qianlong (1735-1796) of the Qing Dynasty (1636-1912) [17]. The inscription Songtao "gusts of wind in the pines" refers to the nickname of Tang Kai whose complete hao (i.e. pseudonym) is Longwan cunsou Songtao "the so-called old man of the gusting wind in the pines of Longwan". According to the carved inscriptions (Fig. 2.c), the making of the qin is dated 1712, which is consistent with the

Sampling
Micro-sampling was performed on the qin for radiocarbon dating. The size of the samples ranged from 0.2 to 1.4 mg on discrete parts of the back of the body of the qin in order to respect the aesthetic integrity of the object (Fig. 3).

Material characterization analyses
3.2.1. Optical and digital microscopy period of activity of Tang Kai. This maker is known to have owned seven ancient qin with similar inscriptions and the same seal as the "Qiulai" qin [20]. However, no early written sources mentioned this "Qiulai" qin in his collection, even though the name is common for a musical instrument.

Material composition Wood
The monoxylous wooden body of the "Qiulai" qin is made of the light porous brown resinous species from the Taxus genus belonging to the Taxaceae family. The forehead of the qin has been carved in hardwood teak, i.e. Tectona grandis from the Lamiaceae family.
Based on the legacy of the surviving written sources and the practices of modern makers, the commonly mentioned wood species for each part of the instrument are as follows [14] : i-for the soundboard: tong tree, wutong , paotong , qingtong , etc. translatable by sterculia or paulownia; or spruce , or cedar shan or "cloud r wood" yunshan .
ii-for the back side: zi or qiu trees, i.e. catalpa; or qi tree i.e. Rhus vernici ua or lacquer tree.
iii-for the foot: wu tree, i.e. black ebony.
The Taxus wood is not a conventional wood used in the making of qin since the Chinese literature and tradition mainly cite the tong tree as the foremost wood used [3,14,22].

Coating
The wooden body of the "Qiulai" qin is completely covered with black and red coatings, providing a strong impression of the homogeneity and cohesion sought by the qin maker [4]. We will choose the term "lacquer" in the following due to the traditional use of raw tree lacquer as organic binder. Traditionally, the lacquer ash layer huitai is applied on the wood support as the preparation layer. It is usually described as a miscellaneous mixture of Rhus vernici ua lacquer, deer bone powder, deer antler, pig's blood, brick, lime, mother-of-pearl, gypsum, gold, silver, jade, agate, coral, shells, etc. The lacquer was applied in several layers to give the nishing veneer. Great value is attributed to ancient guqin whose lacquer layer is marked by numerous cracks named duanwen . Surprisingly, the lacquered surface of the "Qiulai" qin has no cracks.

Silk
The strings that were found knotted around the feet and the bridge prior to their changes in 2018 are made of silk (Fig. 3). The silk came traditionally from insect cocoons named jiacan or domestic silkworm in Chinese and the species used is Bombyx mori. A coating of dirt, glue or wax on the surface of the strings is visible, ranging from off-white to brown-black. Early recipes suggest the use of sh glue, wax or boiled rice for the impregnation of boiled silk strings that helps hold the bers together and in uences the sound [3].
First, the stratigraphy of the lacquers was observed in-situ under visible (VIS), infrared (IR) and ultraviolet (UV) rays using VIS/IR/UV imaging on a Dino-Lite with DinoCapture 2.0 software. Then, a chip of the coating, which comprises the complete stratigraphy from the surface to the wood support, was investigated using light microscopy (Zeiss Axio Scope.A1 Vario) and digital microscopy (3D Keyence-VHX).

X-Ray uorescence
A preliminary investigation was done using X-ray uorescence spectrometry (XRF) Jetstream M6 from Bruker with a rhodium anode in the X-ray tube at 50 kV and 200 µA with a 100 µm-thick beryllium window. The beam diameter was 450 µm. The target was placed horizontally at a working distance of 2 mm. The spectra recording time was 200 s, with a maximum count rate of 96 kcps. Data were processed with the Esprit software developed by Bruker [23].

MEB-EDS
The coating sample was embedded in a polyester resin ESCIL-SODY 33, pre-polished with SiC papers, to obtain a transversal section. Analyses were then performed using Scanning Electron Microscopy with Energy Dispersive X-ray spectroscopy (SEM-EDS) with the JEOL-JSM IT 300 instrument (mode 'HighVacuum', WD = 10mm, an energy beam of 20keV, mode BED-C) and an X-ray microanalyzer X-MaxN (Oxford). The results were processed with Aztec software (Oxford).

Raman
Two different Raman instruments were used. The microspectroscope Raman confocal INVIA Renishaw with a laser source of 532 nm was used for the black layer of the transversal section of the coating (see4.1.1.). The microspectroscope Raman with a laser source of 785 nm from BWTek was used for the red layer of the transversal section of the coating (see 4.1.1.). The systems were equipped with an optical microscope (confocal mode, focus x50).
3.3. 14 C analyses 3.3.1. Sub-sampling of the coating Q1, Q2, Q3 and Q4, were sampled using a cleaned chisel (Fig. 3). The multi-layer lacquer sample Q5 (see 4.1.1) required a more elaborate mechanical treatment using microchisels (0.125 mm, 0.25 mm and 0.5 mm, Fig. 4) in order to produce separate 14 C measurements on: 1-the wooden support, 2-the black preparation layer and 3-the reddish nishing lacquer (Fig. 4). Solvents, commonly used by restorers to separate layers, were excluded because they induce carbon contamination. The black lacquer sub-sample Q5-2 was isolated from the wooden sub-sample Q5-1 using a microchisel on sample Q5 under binocular microscope ( Fig. 4.a). The reddish lacquer layer sub-sample Q5-3 was collected in-situ using microchisels under binocular microscope after removal of a light brown lm of dust by scraping with a microchisel (Fig. 4.b).

Chemical pretreatment for 14 C dating
The wooden and silk samples were treated with the classical AAA chemical protocol suitable for wood and charcoal [24] as follows: i-HCl 0.5M (room temperature) and rinsing with ultrapure water until pH = 5, ii-NaOH 0.1M (room temperature) and hot rinsing (80°C) until pH = 5, and iii-HCl 0.5M (room temperature) rinsing until pH = 5. As dust was expected to be the potential lacquer contamination, Q5-2 and Q5-3 underwent acid treatment with HCl 0.5M, at ambient temperature and rinsing until pH = 5.
Chemical glasses and protective aluminium foils were previously baked out at 450°C during 6 hours, making them carbon-free. All chemicals were of ultrapure quality, and water was ultrapure (MilliQ grade).
To control the impact of chemical treatment and of the combustion, we also ran in-house "blank" (AfSud, a 70 kyr charcoal from Border Cave, South Africa) and international standards (wood -SIRI G at 378 ± 39 BP [25]).

CO 2 conversion and physical measurement
Page 6/18 The mass spectrometer ECHoMICADAS was used to measure the 14 C activity of each sample [26,27]. All the samples (from 40 to 160 µgC) were directly injected into an elemental analyser (EA) coupled with the MICADAS gas source through the gas ion source interface (GIS) [28]. The red lacquer sample Q5-3 was an exception and was injected as a gas into the MICADAS gas source through the GIS by tube cracking [29]. CO 2 was obtained off-line by combustion under pure O 2 .
Ages were obtained from 14 C measurements using Bats software by comparing 14 C/ 12 C ratios with OXII standards [30]. Most of the samples provided several measurements, but results were statistically combined. The mean age of the sample then derives from the mean of the individual measurements, which pass the Chi 2 test, and is associated to the maximum between the Chi 2 reduced error and the standard deviation between the median of the individual measurements. Results are expressed in F 14 C as recommended by Reimer et al. [31] and provided as 14 C ages (BP) following Stuiver and Polach's convention [32].
Probability distributions of calibrated 14 C ages were generated using OxCal v4.3.2 [33] based on the IntCal20 calibration curve The optical microscopy revealed that the lacquer of the qin comprises two successive layers: (a) a heterogeneous black layer of around 1 mm with greenish grains applied on the wooden support, (b) a homogeneous reddish surface layer of 35 µm lightly covered with a dust deposit (Fig. 5.A). According to measurements using XRF, the two layers comprise mainly iron and copper inorganic elements.

The black lacquer
This layer is black under visible (VIS) and infrared (IR) rays using VIS/IR/UV imaging. Hence, it contains a black pigment absorbing VIS and IR rays, which cannot be iron oxide since it re ects IR rays. We deduced that the pigment is most likely a carbon black: soot, charcoal, stone black, ivory black, vine black or bone black. Respectively using SEM-EDS and Raman spectroscopy, the matrix was identi ed as being mainly constituted of a ller made of bone black (Ca 3 (PO 4 ) 2 ) with coarse-

The red nishing lacquer
This layer is red under visible (VIS) and slightly re ective under infrared (IR) rays using VIS/IR/UV imaging. Since XRF analyses applied on the sample Q5 mainly detected Cu and Fe, we rst surmised that the red pigment could be an iron oxide. However, using SEM-EDS and Raman spectroscopy (Fig. 5), the red layer was identi ed as being mainly constituted of a mercury sul de (HgS) most likely from the mineral named cinnabar (Fig. 5.C).

14 C dating results
The sample Q5-2 on the ash lacquer was thick enough to perform four measurements using radiocarbon dating (Table 1, Fig. 6.a.). The heterogeneity of these measurements with a weak A comb of 5.4% is consistent with the composition of the lacquer ash layer as a mixture of several solids. The 14 C dating mean value obtained [69AD − 245AD] is incompatible with the historical information gathered on the "Qiulai" qin.
The wooden samples Q1, Q2, Q3 and Q5-1 yielded similar 14 C results. The combination of the results using the "combine" option of Bayesian modelling [33] gave a high agreement index A comb =114.1%. It can be concluded that the wood pieces chosen for the manufacturing of the qin are contemporary. The tree or trees used grew and were cut during the [1458AD − 1528AD] or [1552AD − 1634AD] intervals (Table 1, Fig. 6.b.).
The silk string sample Q4 and the red lacquer sample Q5-3 present similar 14 C results ( Table 1). The combination of these measurements, acknowledged by a high A comb value of 110.9%, highlighted that the silk strings and the red lacquer are contemporaneous, which might provide the last period of lacquering and changing of the strings on the qin. Considering that the presence of the qin in the CNAM collections is attested since 1849 (terminus ante quem in the Bayesian analysis), this probable last maintenance in playable condition might be included in the following intervals: [1659AD − 1699AD] (25.0%), [1721AD − 1814AD] (66.6%) or [1834AD − 1848AD] (3.8%) (Fig. 6.c). Table 1 Conventional, calibrated and modelled 14 C ages obtained for the "Qiulai" qin using OxCal4.3.2 [33] based on IntCal20 [34].
Statistical results of Bayesian modelling are shown in the last four columns.

Characterization of the wooden body
The radiocarbon results are consistent with a traditional practice of reusing a symbolic wood when making a qin, as explained in "Abiding with Antiquity", c.1860 [3]. The radiocarbon dating results provide a signi cant hiatus, that could reach several centuries, between the dating of the wood in the    (Fig. 6). The hiatus is relevant to the Chinese tradition which symbolically associates the material quality of "a very old paulownia wood" with the quality of the sound. This virtue of "antiquity" is cited in the "Twenty-Four Flavours of qin", ca. 1641, and detailed in the introduction by Goormaghtigh, as part of the aesthetic canons [4]. Such a music appreciation is still controversially present in the modern practice of carving qin in a wood that has been stored for more than a hundred years [35]. However, the storage of wood through several generations of qin makers is not attested in the past.
The dating hiatus between the wood on the one hand and the red lacquer and string on the other cannot be due to the sundrying of a multi-century old tree trunk. The wood used to make the qin should be porous, light, even and smooth without the knots or distended and twisted bers caused by the in uence of the sun or the wind on the growth of the wood, as mentioned in early written sources. Therefore, secondary branches are preferentially selected according to an appropriate age and height [22]. The Tang period poems (618-690, 705-907) studied by You [14] con rm the preference for secondary branches rather than the trunk, except in the case of a tree less than a decade old.
"As for the material that should be used, it should be from younger branches; that is, branches that are close to the top of the trunk. If the tong is not old, then its young branches will certainly not be big. Choose wood far away from the ground, with the top blown by wind and dried by sun. It should be moistened by frost and dew with plenty of pure air. Furthermore, it is good if the material is facing southeast, (because) "the Great Brightness (the sun) is born in the east".
The Book of Odes states: "The wutong grows there on those slopes and faces the sun" [3] The interpretation of the 14 C results as due to the reuse of wood is reinforced by the fact that the "Qiulai" qin is not made from a conventional paulownia species, but from a porous resinous wood from the Taxus family. Considering that the changes in Li Bai's poem carved on the back side are a subtle and skilful adaptation to this particular qin, the replacement of the sinogram tong for paulownia by bai for resinous woods of cedar or cypress may be linked to this manufacturing choice. Other wood species than paulownia can be used, especially old building timbers: "As for good quality in old materials, (…) one cannot be certain that it will be tong wood. Some other kinds of wood are also all right. The wood is useable as long as it is light and porous, crisp and smooth. (…) As for beams, pillars, and router rafters, the ancients used tong for some of these. But it has to have survived for a couple of hundred years. If at an old monastery or temple on a high precipice or cliff, or by a waterfall, or stream shore in high, spacious and quiet regions, absolutely cut off from the cacophony of the everyday world, there one can nd good material for the qin. Naturally it will have a special sound. " [3] As a testimony of this practice during the 19th century, the wood of the "Yun He" qin kept at the Musical Instruments Museum, Brussels (Coll. MIM, n° inv. 0760) has been identi ed as originating from Jing Hui Temple in the Guangzhou province (Canton) thanks to ink inscriptions on the back side [36]. This unful lled epigraph also indicates that the qin was built by the maker Li Menggeng whose life is poorly known, but who partly coincides meets the reign of Daoguang (1820-1850). The entry of this qin in the MIM collections is dated between 1881 and 1886 according to the inventories.
The 14 C results for the wood of the "Qiulai" qin are consistent with a long tradition in qin making that consists in reusing old wood with a highly symbolic and spiritual value, whether during the period of activity of Tang Kai or in the early 19th century before the recorded presence of the qin in the CNAM collection in 1849.

Composition of the lacquer ash layer
The radiocarbon dating and material characterization of the "Qiulai" qin lacquer ash layer revealed a mixture of two major solid ingredients, bone black and malachite, consistent with the use of powdered deer antler and crushed stones mentioned in the Qinfang c.1641 and the Yuguzhai Qinpu c.1860 [3,4]. The lacquer ash layer is located between the wood body dated from The detection of carbon and phosphorous in the lacquer ash layer of the "Qiulai" qin does not differentiate deer antler ash from deer or another animal bone ash. Li et al. [13] also reported that the detection of Ca 3 (PO 4 ) 2 merely indicates the presence of bone ash regarding the "Yu Quan" qin from the Yuan Dynasty. However, traditional recipes comprising the raw sap of the lacquer tree and the ash of deer antlers lujiao shuang are mostly cited [14,37]. According to the Yuguzhai Qinpu c.1860, "Whenever others put on the powder and lacquer they use a cattle horn comb (...). The powder; that is, deer horn shuang , (…) should be ground ne." The detection of the coarse grains of malachite is suggestive of the qin makers' practice of occasionally adding fragments of metal or stone to the lacquer ash layer in accordance with the symbolism of the carillon and lithophone so as to obtain the ideal acoustic properties, as discussed by Goormaghtigh [4] based on the following citation in the Qinfang c.1641: "The ngers struck the strings as if they drummed metal or stone, without any interference from the neighboring strings. It is in this search for limpidity that one goes beyond all sounds. ". Chang [38] mentioned that the highly estimated lacquer of qin made in the Tang Dynasty (618-690, 705-907) corresponded to the Asian lacquer technology consisting in adding scraps of metallic or stone llers. A mixture, commonly named "eight precious ash" lacquer baobaohui by qin makers, is made of fragments of gold, copper, bronze, porcelain, silver, mother-of-pearl, malachite, pearl, jade, agate, coral or seashells [8].
Powdered deer antler and crushed stones are not ingredients inherent to the qin but can be found in Asian lacquers [39]. For example, a lacquered tray from the Ming period (1368-1644) presents surface "snake-belly cracks" as for qin from the same period, which could be due to the use of similar compounds such as powdered deer antler, bone, horn, ash, sandstone, or ceramics [40]. Our results are consistent with the traditional composition of the lacquer ash layer of qin; however, due to the long tradition spread over several dynasties, a more accurate attribution is not possible.

The silk string and the red lacquer dating
The analyses revealed that the remaining silk string and the red lacquer are contemporaneous, meaning that the last lacquering and maintenance in playable condition occurred in China during the Qing dynasty (1644-1912) since the radiocarbon dating yielded the [1659-1699],  or [1834][1835][1836][1837][1838][1839][1840][1841][1842][1843][1844][1845][1846][1847][1848] intervals. According to our investigations, three hypotheses are currently available for the attribution of the period when the "Quilai" qin that entered the CNAM collection before 1849 was built: i-it was collected, dated and signed by Tang Kai in 1712, ii-it was made by Tang Kai in 1712, or iii-it is a wellinformed late 18th-19th century copy using an old wood carved in China.
The absence of typical cracks duanwen on the red lacquer is surprising. Cracks may begin to appear a hundred years after the making, and the shape of the cracks can be used for the relative dating of the qin. Re-lacquering or restoration of the "Qiulai" qin by successive owners can be considered during the [1721-1814] or [1834-1848] intervals, as such practices have already been observed on other ancient qin . For example, the "Mingfeng" qin is known to have been recut and re-lacquered by Tang Kai [20]. The cracks on the "Xianren you" qin from the Song period (1127-1279) have been wiped out by restoration [41]. However, the cracks are not necessarily proof of antiquity even if this is the reason why Chinese connoisseurs estimate them. Like the "Qiulai" qin, the "Yun He" qin that entered the MIM collection c.1880 presents no surface cracks either.
The "Qiulai" qin shows 18th-century features, and material characterization analyses as well as 14 C results are consistent with the traditional practices of the qin makers from this period as described in written sources [3,4]. The authenticity of this qin is credible. Indeed, the Qing dynasty period was rich in cultural exchanges between China and France under the Ancient Regime. French sinophilia during the 18th Century Enlightenment is also well-known under the reign of Qianlong (1735-1796). China's tolerance toward Christianity allowed the establishment of Jesuits whose exchanges with the Académie Royale des Sciences provided information on Chinese techniques of the time. For example, the Jesuit Pierre Nicolas Le Chéron d'Incarville (1706-1757) sent a memoir to the Académie in 1760 mentioning the use of resin tree and calcined deer bones in black Chinese lacquers [42]. The sending and the studying of qin in the eighteenth century in France is attested. For example, the Jesuit priest Joseph-Marie Amiot (1718-1793) sent a guqin to M. Bertin for his cabinet of curiosities [43], as reported in his correspondence with the Académie des Sciences [44]: "The rst is a Kin with seven strings, unlike those of today (…). It is made of a single piece of wood".
If the "Qiulai" qin was made by Tang Kai and kept its original lacquer, then it was manufactured in the [1659-1699] interval, since the 14 C dating results on the red lacquer provide intervals that do not include the date of 1712 carved on the back side of the qin. The carving of inscriptions on a guqin traditionally took place after the instrument had been made and even played for several years.

Conclusions
The most signi cant ndings that emerge from this study are that the qin rediscovered in the CNAM collection has exceptional characteristics that are consistent with current knowledge about eighteenth-century qin manufacture: the shape, Tang Kai's seal and pseudonym, the multi-layered lacquer composition, and the wood and lacquer dating. While interdisciplinary investigations on early Chinese zithers are seldom used, our results were achieved thanks to the combination of material characterization techniques, advanced radiocarbon dating technology and museum expertise confronted with early Qing dynasty written sources. This research project constituted a unique opportunity to study one of the most remarkable qin kept in Europe.

Declarations
Availability of data and materials All data generated or analyzed during this study are included in this published article. Raw data (including spectra) are available upon request from the authors. The Chinese zither or qin from the CNAM collection (inv. n°04224). The silk strings were strung in 2018 at the Musée de la musique, Paris. Pictures by Claude Germain © Musée de la musique, Cité de la musique -Philharmonie de Paris.

Figure 2
Pictures of the front and back sides of the qin. The inscriptions carved on the back side of the qin and their transcriptions: a) the name of the qin: 1lai is a polycalamus ute of the yue type with three holes. The qin tends to reproduce the soft and nostalgic sounds of this "antique" ute suggesting an autumnal atmosphere b) "In praise of the qin", Li Bai's poem (translation by Lucie Rault): the original text is in black, the modi ed passages carved in the qin are in red and under the corresponding sonograms. 2 tong (in original version) translated by paulownia or wutong [21]. 3 hui studs that mark harmonic positions, c) information on the making of the qin: 4 dated from 1712 according to the era mentioned under the Qing Dynasty, 5 dated from the spring of the year renchen , which is the name of the sexagesimal year corresponding to 1712 (according to the cyclic calendar), d) zoom in of Tang Kai's seal.   wooden samples of the back side of the foot Q1, the pillar Q2, the foot Q3 and the wood support under the lacquer Q5-1 with their respective Bayesian modeling combination in black, and the upper diagram in blue of the combination of all wooden parts; c) sub-sample of the upper lacquer in red Q5-3 and sub-samples of the silk string in blue Q4 with their respective Bayesian modeling combination in black, and the upper diagram in black of the combination of the red lacquer and the silk string parts taking into account the event of the entry of the qin in the CNAM collection prior to 1849.