Hemp in Ancient Rope and Fabric from the Christmas Cave in Israel

Our reference: YJASC 2841 P-authorquery-v9

AUTHOR QUERY FORM

Journal: YJASC Please e-mail or fax your responses and any corrections to:

Article Number: 2841 Fax: +31 2048 52789

Dear Author,

Please check your proof carefully and mark all corrections at the appropriate place in the proof (e.g., by using on-screen annotation in the PDF ﬁle) or compile them in a separate list. To ensure fast publication of your paper please return your corrections within 48 hours.

For correction or revision of any artwork, please consult http://www.elsevier.com/artworkinstructions.

Any queries or remarks that have arisen during the processing of your manuscript are listed below and highlighted by ﬂags in the proof.

Location Query / Remark: Click on the Q link to ﬁnd the query’s location in text in article Please insert your reply or correction at the corresponding line in the proof Q1 The following numbers have been identiﬁed as Genbank accession numbers:FJ169596, AF500344, FJ160887, AY958396, FJ160842, AF345317. Please verify if this is correct. Q2 The following references are not cited in the text but presented in the reference list. Please provide them in the text or delete these citations from the list.

Thank you for your assistance. YJASC2841_grabs ■ 28 May 2011 ■ 1/1

Journal of Archaeological Science xxx (2011) 1

Contents lists available at ScienceDirect

Journal of Archaeological Science

journal homepage: http://www.elsevier.com/locate/jas

1 2 11 3 12 4 Highlights 13 5 14 < We analyzed DNA from rope and fabric from the “Christmas Cave” in the Qidron Valley. < Ribulose bisphosphate carboxylase gene DNA was 6 15 primarily from Linum usitatissamum L. < Samples also had variable amounts of hemp (Cannabis sativa L.) DNA. < 14C dating conﬁrmed that samples 7 16 represented both the Roman and Chalcolithic periods in Israel. 8 17 9 18 10 19 20 21

Please cite this article in press as: Murphy, T.M., et al., Hemp in ancient rope and fabric from the Christmas Cave in Israel: talmudic background and DNA sequence identiﬁcation, Journal of Archaeological Science (2011), doi:10.1016/j.jas.2011.05.004 YJASC2841_proof ■ 28 May 2011 ■ 1/10

Journal of Archaeological Science xxx (2011) 1e10

Contents lists available at ScienceDirect

Journal of Archaeological Science

journal homepage: http://www.elsevier.com/locate/jas

1 56 2 Hemp in ancient rope and fabric from the Christmas Cave in Israel: talmudic 57 3 58 4 background and DNA sequence identification 59 5 60 6 Terence M. Murphy a,*, Nahum Ben-Yehuda b, R.E. Taylor c,d,e, John R. Southon f 61 7 62 a 8 Department of Plant Biology, University of California, One Shields Avenue, Davis, CA 95616, USA 63 b Department of Jewish History, CText ATI, Bar Ilan University, Ramat Gan 52900, Israel 9 c Department of Anthropology, University of California, Riverside, CA 92521, USA 64 10 d Cotsen Institute of Archaeology, University of California, Los Angeles, CA 90095, USA 65 11 e Keck Accelerator Mass Spectrometry Laboratory, University of California, Irvine, CA 92697, USA 66 12 f Department of Earth System Science, University of California, Keck Accelerator Mass Spectrometry Laboratory, Irvine, CA 92697, USA 67 13 68 14 69 15 article info abstract 70 16 71 17 Article history: The “Christmas Cave”, a cave in the Qidron Valley near the Dead Sea and Qumran, has yielded a complex 72 Received 18 February 2011 18 collection of plant-derived rope and fabric artifacts. Using polymerase chain reaction (PCR) to amplify 73 Received in revised form DNA of the samples, we estimated the sizes and determined restriction patterns and base sequences of 19 10 May 2011 74 chloroplast genes, primarily rbcL (gene for the large subunit of ribulose bisphosphate carboxylase). DNA 20 Accepted 11 May 2011 75 was successfully extracted from all samples, but was limited to sizes of approximately 200e300 base 21 76 pairs. As expected, the DNA extracted from the samples was identified as coming primarily from flax Keywords: 22 (Linum usitatissamum L.), but two samples had a significant fraction, and all samples had at least a trace, 77 23 Flax Cannabis sativa 78 Linen of hemp ( L.) DNA. Artifacts from the Christmas Cave were thought to date from Roman 24 79 Hemp times, but it was thought possible that some could be much older. Accelerator mass spectrometry (AMS)- 14 25 Rope based C dating confirmed that the samples contained representatives from both the Roman and 80 26 Fabric Chalcolithic periods. This paper provides a synthesis of DNA, isotope, and literary analysis to illuminate 81 27 Israel textile history at the Christmas Cave site. 82 28 Judean desert Ó 2011 Elsevier Ltd. All rights reserved. 83 DNA 29 rbcL gene 84 30 PCR 85 31 Polymerase chain reaction 86 32 AMS-based 87 14C dating 33 88 34 89 35 90 36 91 37 92 38 1. Introduction down DNA. However, even present-day rope is made with natural 93 fi 39 bers that receive a minimum of treatment, and the rope contains 94 40 DNA analysis can identify the biological source of archaeological fragments of tissue with intact organelles (Dunbar and Murphy, 95 fi 41 artifacts. This is true for many plant-based artifacts. Plant cells 2009). We expect that treatments of bers in the past were less 96 42 contain plastids, such as chloroplasts in leaves-often many copies- stringent and the products from which they were made more likely 97 fi 43 and plastids contain DNA sequences that are useful for identi ca- to retain plastids and nuclei. 98 44 tion. There is a great deal of information available concerning the It may be even more surprising that DNA persists in ancient 99 45 base sequences of plastid genes in different plants, much of it objects, since we can expect the rigors of time, with accompanying 100 46 gathered for use in determining evolutionary relationships. This hydration, desiccation, and temperature extremes, to break down 101 47 information can be applied to objects like textiles and baskets. biological molecules. In fact, that does occur (Smith et al., 2003). 102 fi 48 At rst glance, it may be surprising that DNA persists in man- But DNA may show a degree of resistance under certain conditions. 103 49 ufactured objects, and some processes-e.g, mordanting-do break Indeed its structure may have evolved in part to increase its 104 50 stability (Lazcano et al., 1988). There have been many reports of 105 51 ancient DNA isolated from, for example, mammoths preserved in 106 glaciers (Gilbert et al., 2007), human mummies (e.g., Caramelli 52 * Corresponding author. Tel.: þ1 530 752 2413; fax: þ1 530 752 5410. 107 53 E-mail address: [email protected] (T.M. Murphy). et al., 2008), wood (Liepelt et al., 2006), and rope (Mukherjee 108 54 109 e Ó 55 0305-4403/$ see front matter 2011 Elsevier Ltd. All rights reserved. 110 doi:10.1016/j.jas.2011.05.004

2 T.M. Murphy et al. / Journal of Archaeological Science xxx (2011) 1e10

111 in 1960 (Allegro, 1965, pp. 6e15). In 2007, the cave was surveyed 176 112 again by Roi Porat and Hanan Eshel (Porat et al., 2007). They 177 113 confirmed that the finds have no connection to the Qumran Caves 178 114 (see also Shamir and Sukenik, 2010). This cave served as a refuge, 179 115 beginning in the Chalcolithic Period and afterwards at the end of 180 116 the Great Revolt in 73 CE and again in the Bar Kokhba Revolt in 135 181 117 CE. (Porat et al., 2009) Among the archaeological finds from this 182 118 cave are wool and linen textiles from various periods. This assort- 183 119 ment of textiles, in contrast to those found in Qumran e which are 184 120 exclusively linen - is similar to those found in Masada (Belis, 2003, 185 121 p. 211, 219, Sheffer and Granger-Taylor, 1994). In the opinion of 186 122 Dr. Orit Shamir of the Israeli Antiquities Authority, comparison to 187 123 the textiles found in The Cave of Letters (Granger-Taylor, 2006; 188 124 Yadin, 1963) is more exact, those being generally coarser than the 189 125 Masada textiles (personal correspondence). 190 126 Because this site is not considered part of the Qumran complex 191 127 of caves, Humpert and Gunneweg’s (2003) inclusion of these 192 128 textiles in the Qumran group of findings is confusing. The intro- 193 129 duction (p. XIX) under the subheading “textiles"” does not mention 194 130 the CC at all, notwithstanding its textile finds being included in the 195 131 subsequent chapters. The CC finds all appear numbered as category 196 132 “QCC”(¼“Qumran Christmas Cave”), and some have been given 197 133 a parallel “QUM” number (Belis, 2003,p.221;Müller et al., 2003, 198 134 p. 277). In Humpert and Gunneweg (2003) Walton Rogers reports 199 135 her analysis of some of these fibers, referring to them simply as 200 136 coming from “a site in the Dead Sea region”, which is correct under 201 137 any circumstances. 202 138 These cordage and textile samples were stored since their 203 139 discovery at the Rockefeller Museum in Jerusalem, examined at 204 140 École Biblique et Archéologique Française de Jérusalem (EBAF), and 205 141 only recently relocated to the Israel Antiquities Authority. The 206 142 articles had been bundled in batches from the various loci within 207 Fig. 1. Evidence of ancient uses of hemp in Europe and Asia Minor (Fleming and fi 143 Clarke, 1998). the CC. Due to the dif culty of dating these samples, either by 208 144 physical observation or by genetic testing, we chose to confirm the 209 145 ages of selected samples through 14C dating (see below). 210 146 et al., 2008). The degradation of DNA results in a gradual reduction It is generally accepted, based on both literary sources and 211 147 in its length as the polymeric strands become fragmented, but even archaeological finds, that the primary fibers in use in the 1ste4th 212 148 relatively short fragments retain useful information in their base centuries CE in the Land of Israel were sheep wool, goat and camel 213 149 sequences. hair, and flax-linen1. Additional fibers, such as hemp, cotton and silk 214 150 With those considerations in mind, we resolved to identify DNA are mentioned, albeit infrequently, in contemporary rabbinical 215 151 in fibers of samples of rope and cloth that have been found in an literary sources, but have not been previously corroborated by 216 152 archeological site near Qumran and the Dead Sea. Microscopic archaeological finds of the period (Baginski, 2001; Shamir, 2001; 217 153 observations have identified various samples of cloth from caves Crowfoot and Crowfoot, 1961). This research project focuses on 218 154 above Qumran as flax, cotton, and wool (Walton Rogers, 2003 and the genetic identification of vegetable fibers constituting cordage 219 155 personal communication). Muller et al. (2004, 2006, 2007), using and textiles found in the CC. It has always been assumed that the 220 156 X-ray micro-diffraction, identified flax and cotton in Qumran fiber used in articles of this type is flax, which is well known as 221 157 samples. But there have been indications of the early use of hemp in being widespread in the Land of Israel in this period, and is 222 158 East Asia (Mukherjee et al., 2008) and in Europe and Asia Minor mentioned numerous times in the contemporary rabbinic literature 223 159 (Fig. 1). Our objective was to learn whether the DNA of materials (Mishnah; Toseftah and the Jerusalem Talmud)2. As stated in the 224 160 from an archeological site near the Dead Sea could confirm the introduction, this assumption has been validated by optical 225 161 presence of hemp or other fibers. We expected the use of DNA microscopy (Walton Rogers, 2003) and X-ray diffraction (Müller 226 162 sequence information to confirm the identity of the major et al., 2004, 2006, 2007). Nevertheless, it is difficult to distinguish 227 163 component (as flax), but also to indicate whether fibers from hemp cellulosic bast fibers such as flax and hemp by these means 228 164 or another plant species form a detectable fraction of one or more 229 165 samples. As will be shown, our data do indicate that flax-linen 230 166 dominates in every sample tested and that there is a small 231 167 amount of hemp DNA in most samples. 1 See: ASTM D 6798-02 Standard Terminology Relating to Flax and Linen. 2003. 232 “ ” fi “ ” 168 West Conshohoken, PA. Flax refers to the plant and its ber. Linen refers to the 233 products produced from spinning onwards in the production process. Both the 169 2. Background Hebrew and Aramaic languages do not discern between these two designations, 234 170 often causing ambiguity. 235 171 Samples described in this paper came from the “Christmas Cave” 2 The Mishnah, Toseftah and Jerusalem Talmud are all works redacted in the Land 236 172 (herein abbreviated “CC”). The CC is located in the Judean Desert on of Israel. Documentation of material culture appearing in all of them should be 237 relevant to our research. The Mishnah (and probably Toseftah) represents 225 CE 173 the west bank of the Dead Sea, in the Qidron Valley, 1 km south of 238 terminus ante quem and probably long before 70 CE terminus post quem. The Jer- e 174 Qumran (ICS coordinates 189887/121095)(Fig. 2). Its name usalem Talmud, which is a work expounding on the Mishnah, is 350 CE terminus 239 175 commemorates the day on which it was discovered by John Allegro ante quem. 240

T.M. Murphy et al. / Journal of Archaeological Science xxx (2011) 1e10 3

241 306 242 307 243 308 244 309 245 310 246 311 247 312 248 313 249 314 250 315 251 316 252 317 253 318 254 319 255 320 256 321 257 322 258 323 259 324 260 325 261 326 262 327 263 328 264 329 265 330 266 Fig. 2. A, Map of the Dead Sea with principal sites from the 1st century CE (Humpert and Chambon, 2003). B, Map of the refuge caves in the Qumran area. (Porat et al., 2009). 331 267 332 268 333 (Bergfjord and Holst, 2010) all the more so when they are extremely mentioned in the Diocletian Edict (26:23e28:30) by merit of its 269 334 old, having suffered the damaging and deteriorating effects of time. superior linen products, referred to there by its Roman name 270 335 Furthermore, optical and X-ray techniques may miss minor “Scythopolis”. Beth Shean, approximately 80 km north of the Dead 271 336 components of mixed fibers. Sea, is frequently mentioned in Talmudic sources (E.g. Jerusalem 272 337 Hemp (cannabis) is mentioned in the Talmudic literature as an Talmud tractate Qiddushin 2:62:3, Midrash Beresheit Rabbah 19 273 338 existing, marginal, textile product in the Land of Israel, similar- and 20, Midrash Tanʿumah Tazriaʾ 7) as the famous center in this 274 339 albeit inferior-to flax. Both flax-hemp blend and wool-camel’s hair period for flax growing and manufacture of fine linen garments. 275 340 blend are mentioned in Mishnah Tractate Kilʾayim 9:1 and Flax is mentioned in the Biblical period (c.1200 BCE), in nearby 276 341 a parallel in Tractate Negaʿim 11:2. The indication is that the two Jericho (Joshua 2, 6) and Crowfoot (1960, 1965, 1982) cites evidence 277 342 fibers blended together are quite similar. Hemp as a similar, albeit of earlier finds there, too. In adjacent Hesban (Platt, 2009, 278 343 invalid, substitute for flax for the sacerdotal vestments is pp.163e196. Mitchel, 1992 pp. 3e38. Personal correspondence with 279 344 mentioned in Toseftah Tractate Menaḥot 9:17. In modern literature, Paul Ray, Associate Curator, Horn Archaeological Museum at 280 345 mention has been made of the adulteration of flax with hemp is Andrew University), ʿEin Gedi (Shamir, 2007 “The Loomweights” 281 346 a possible means of deceit, as flax is more expensive (See Oakley, and 2007 “The Textiles”) and Masada (Sheffer and Granger-Taylor, 282 347 1928, pp.167e9). But blending flax with hemp may be practiced 1994), textile manufacturing implements from the relevant time 283 348 to obtain rope that is both soft and strong (Weindling, 1947, p. 286). periods have been unearthed, some of them possibly related to 284 349 These two practices again indicate the similarity between hemp rope manufacture. Another possible source is Jerusalem, approxi- 285 350 and flax. mately 40 km away. 286 351 Although linen cordage (rope or twine)3 is mentioned in Talmudic literature mentions rope-making in several contexts, 287 352 rabbinical literature (Jerusalem Talmud Tractate Sukkah 1:52:2 and one of which is the halakhic injunction against using (even) 288 353 Tractate ʿEruvin 5:22:4), it is possible that hemp was also used in a derelict synagogue (which is particularly suitable, being a long 289 354 these products and as well in other coarse textiles (Barber, 1991, building) as a rope walk (Mishnah Tractate Megillah 3:3) due to the 290 355 p. 15; McKenna, 2004, pp. 4e5; Weindling, 1947, p. 286). site’s holiness. Others include rope making done by two halakhi- 291 356 As opposed to the fabric and nets found in the Land of Israel, cally impure individuals (Mishnah Tractate Zavim 3:2) and the 292 357 which have typically been the focus of extensive research, there is injunction against making rope in a city of refuge (Toseftah Tractate 293 358 less information on the cordage finds (although see Shamir, http:// Makot 3:10) (See also: Herzberg, 1924, pp. 140e146.). 294 359 www.atnfriends.com). Where were the ropes and textiles manu- Hemp in an agricultural context is mentioned only once in the 295 360 factured, and where did the raw materials come from? This being Mishnah (Tractate Kilʾayim 2:5). According to Felix (1967, pp. 296 361 a cave of refuge, it is highly unlikely that the materials were 220e222) the text is corrupt4 and should be read “caraway”. Amar 297 362 manufactured in situ. A probable candidate is Beth Shean, which is (2000, p. 336, and personal correspondence) believes that the 298 363 printed text can be accepted as is, but in any case hemp was defi- 299 364 nitely a very marginal crop. The Mishnah and Jerusalem Talmud 300 365 301 3 See Denton and Daniels: “Rope” is an article of cordage more than approxi- 366 302 mately 4 mm in diameter", and “twine” is twisted cordage less than 4 mm in 367 “hevel” “meshiha” 303 diameter. The Talmudic term for rope is , and for twine is . Our 368 examples are border line in their diameters. Toseftah Tractate ’Eruvin 2, 2 mentions 4 The printed text reads “knbs”, while variants including the Jerusalem Talmud 304 rope of 8 cm in diameter, for measuring the t’hum. In Tractate Zavim 1, 11 a rope read “krbs”. A perennial plant is necessary for the Mishnaic textual context. This 369 305 50 m long is mentioned. designation fits caraway (“krbs”), not cannabis. 370

4 T.M. Murphy et al. / Journal of Archaeological Science xxx (2011) 1e10

371 436 372 437 373 438 374 439 375 440 376 441 377 442 378 443 379 444 380 445 381 446 382 447 383 448 384 449 385 450 386 451 387 452 388 453 389 Fig. 3. Photographs of selected rope (928, 931, 795, 796) and fabric (786, 955, 019) samples used in this analysis. 454 390 455 391 (Tractate Kilʾayim 9:7) mentions that garments made of hemp Indeed, rope (both laid and braided) was certainly used in 456 392 (possibly blended or combined with sheep’s wool) were imported antiquity for a multitude of purposes. Many rope finds have been 457 393 from “overseas” perhaps from the Aegean and Asia Minor regions retrieved from CC and other sites in the Land of Israel. The vast 458 394 (Fig.1) (Fleming and Clark,1998; Herzberg, 1924), through the ports majority of these ropes are made from date-palm fiber; additional 459 395 of Caesaria and Tyre and their respective vicinities. Perhaps fibers ropes are of goat and camel hair; and the distinct minority is of bast 460 396 and/or yarn for weaving and rope manufacture were also imported, fiber, which has always been assumed to be flax. Up until now, no 461 397 not locally grown. Pliny the Elder mentions hemp in a textile archaeological textile finds in the Land of Israel have been identi- 462 398 context (Book 19, ch. 8, and ch. 56.9). It is particularly suitable for fied as hemp, but up until now DNA technology has not been 463 399 rope and hunting nets. Two localities of production are mentioned: employed. 464 400 Alabanda and Mylasa, both in Asia Minor (Fig. 1). 465 401 In order to position these ropes’ usage in their contemporary 3. Methods 466 402 and local context, the following are the various uses for rope 467 403 mentioned in Talmudic literature, any or all of which may very well 3.1. Samples 468 404 have existed at CC. 469 405 Mishnah: Samples of cordage and textile, part of the National Treasures 470 406 Collection of the Israel Antiquities Authority, were identified by 471 407 Used as a guideline for harvesting straight rows of grain curator Dr. Orit Shamir as coming from the “Christmas Cave” 472 408 (Peʾah 4:5) collection. Cordage samples were numbered I.A.A. 582928 (herein 473 409 To support the branches of a grapevine (Kilʾayim 6:9) abbreviated 928), 582931 (931), 585795 (795), 585796 (796), and 474 410 A leash for an animal (Shabbat 5:3 passim) 637538 (538). Textile samples were numbered 582812 (812), 475 411 A handle for a basket (Shabbat 8:2 passim) 582955 (955), 583019 (019), 585440 (440), 585786 (786), 577061 476 412 A sailor’s knot (Shabbat 15:1) (061), and 577289 (289). Two samples of modern rope were 477 413 A cameleer’s knot (ibid.) Tied to a bucket for drawing water from included: one stated to be of linen from Japan, and one of flax from 478 414 a cistern (Shabbat 15:2 passim) the United Kingdom. A positive control sample of flax, Linum usi- 479 415 To secure a burden to an animal’s back (Shabbat 24:1) tatissimum L., was obtained from the garden of the Plant Sciences 480 416 To measure the “teḥum”, the distance permitted to walk on the Department, University of California, Davis, California, USA. Positive 481 417 Sabbath (ʿEruvin 1:9) control samples of Cannabis sativa L. were obtained from the police 482 418 Rope bed or stool (Pesaḥim 4:9 passim) department of Bakersfield, California, USA. 483 419 To measure a parcel of land (Bava Batrah 7:2e3) 484 420 To bind an animal’s legs (Parah 3:9) 3.2. Extraction of DNA 485 421 486 422 Toseftah: Samples of cordage and textile, ranging in weight from 0.04 g to 487 423 <0.01 g, were each ground to dust in ceramic mortars and pestles 488 424 To secure a person who is going to immerse in water - a “lifeline” or (less effectively) in disposable microfuge tubes with plastic 489 425 (Shabbat 1:18) pestles; 0.7 ml of buffer containing 1% cetyltrimethyl ammonium 490 426 To bind bundles of branches (Betzah 3:10) bromide, 0.7 M NaCl, 10.5 mM ethylenedinitrilotetraacetic acid, and 491 427 Used for shade, placed on the roof of a booth (Sukkah 1:4) 55 mM Tris buffer, pH 8 was added, and the slurry was ground to 492 428 To climb up to the roof (Makot 2:6) homogeneity. The mixture was heated at 65 C for 10 min or more 493 429 To check the water level in a cistern - a “dipstick” (Makhshirin 9:6) and then cooled; 4 ml of RNAase containing 107 units were added; 494 430 and the solution was incubated at 37 C for 20 min. The mixture 495 431 Additional maritime uses of rope are detailed in Sperber (1986). was extracted with 350 ml of chloroform, and DNA in the aqueous 496 432 Shipping in the adjacent Dead Sea was well developed in the layer was precipitated with an equal volume of isopropanol. After 497 433 Roman period (Porat e personal correspondence). Porat et al. centrifugation, the pellet was washed with 70% ethanol, dried, and 498 434 (2009) notes that entry and mobility in the CC require the use of dissolved in water. The DNA was further purified by adsorption on 499 435 rope, as in rappelling, or abseiling. and elution from glass filters (Qiagen or Promega). Samples of fresh 500

T.M. Murphy et al. / Journal of Archaeological Science xxx (2011) 1e10 5

501 L. usitatissimum L. and dried C. sativa L. were treated similarly after 566 502 having been ground in disposable microfuge tubes with plastic 567 503 pestles. 568 504 569 505 3.3. DNA amplification 570 506 571 507 Identification of the plant species contributing fibers to the 572 508 cordage and textile samples began with the polymerase chain 573 509 reaction (PCR)-amplification of the ribulose bisphosphate 574 510 carboxylase-oxygenase large subunit (rbcL) gene (GenBank refer- 575 511 Q1 ences: L. usitatissimum, FJ169596; C. sativa, AF500344). This gene is 576 512 located in chloroplast DNA. Thus, it is specific for photosynthetic 577 513 organisms and its use avoids confusion from the presence of animal 578 514 or fungal DNAs. Initial attempts at amplifying the gene used 579 515 primers recommended by Dunbar and Murphy (2009), which 580 516 produce a DNA fragment of approximately 771 base pairs. These 581 517 attempts were unsuccessful, generating DNA that was identified as 582 518 coming from a Salvia species, a laboratory contaminant. We 583 519 recognized that amplifying a large sequence would select against Fig. 5. DNA analysis of rope and fabric samples. Following PCR amplification using rbcL 584 BamH 520 ancient DNA templates, which tend to be fragmented. We then primers, DNA was subjected to cleavage by 1 restriction enzyme. As shown in the 585 lower frame, control DNA from C. sativa was completely cleaved, whereas DNA from 521 tested other combinations of primers giving smaller fragments L. usitatissimum was not. DNAs from modern linen (Japan) and flax (UK) ropes were not 586 522 (Fig. 4 below) and adopted two primers that generated a 184 bp cleaved. Rope 931 and fabrics 786 and 019 showed significant cleavage. Numbers at the 587 523 DNA fragment: rbcLF2, TGTTTACTTCCATTGTGGGTAATG, and left show the size of DNA fragments in base pairs. 588 524 rbcLR3a, TTCGGTTTAATAGTACAGCCCAAT. The DNA segment 589 525 defined by these primers included an EcoR1 restriction site in the 590 526 Salvia gene, but not the Linum or Cannabis genes. Treating the contaminating Salvia. A second set of primers was specific for matK, 591 527 templates with EcoR1 blocked amplification of the contaminant. which codes for the maturase enzyme for lysine tRNA (GenBank 592 528 A separate set of samples extracted under conditions that avoided references: L. usitatissimum, FJ160842; C. sativa, AF345317). Primers 593 529 the contaminant confirmed the results obtained with the first set. (matKF1, AAAAGGTTGGGGTCGGAATT; matKRL1, CCTGCGTCCTTTA- 594 530 Two other sets of primers, representing two other sections of the TATAATCTTTG; matKRC1, TCATTCATAATTGACCAGATCG) were 595 531 chloroplast genome, were tested. One was specific for trnL, the chosen to give amplified fragments of 109 base pairs from Linum and 596 532 leucine transfer RNA gene (GenBank references: L. usitatissimum, 162 base pair from Cannabis. No product was expected from Salvia, 597 533 FJ160887; C. sativa, AY958396). Primers (trnLF1: AGCTGTTCTAA- since Salvia lacked a complementary base sequence, at least in that 598 534 CAAATGGAGTTG; trnLR1: GGACTCTATCTTTGTTCTCGTCC) were region of the genome. 599 535 chosen to give amplified fragments of 311 base pairs from PCRs were conducted in 20 ml of solution containing 12.1 mLof 600 536 Linum, 180 base pairs from Cannabis, and 283 base pairs from any water, 4 mLof5 Green Go Taq Buffer (Promega Corporation, 601 537 602 538 603 539 604 540 605 541 606 542 607 543 608 544 609 545 610 546 611 547 612 548 613 549 614 550 615 551 616 552 617 553 618 554 619 555 620 556 621 557 622 558 623 559 624 560 625 561 626 562 627 rbcL fl L. usitatissimum C. sativa fl 563 Fig. 4. Selection of primers for the analysis of ancient samples. Extracts of modern ax ( ) and hemp ( ) plants were compared to those of ax rope and 628 three ancient samples, using primers designed to give amplified DNA of 184 (a), 390 (b), 378 (c), or 584 (d) base pairs. Under the PCR conditions used, modern flax gave products 564 only with primers a and c; modern hemp gave products with all four primer pairs. Archeological samples gave products with primers a and b. Numbers at the left show the size of 629 565 DNA fragments in base pairs. 630

6 T.M. Murphy et al. / Journal of Archaeological Science xxx (2011) 1e10

631 696 632 697 633 698 634 699 635 700 636 701 637 702 638 703 639 704 640 705 641 706 642 707 643 708 644 709 645 710 646 711 647 712 648 713 649 714 650 715 651 716 652 717 653 718 654 719 655 720 656 721 fi rbcL fl 657 Fig. 6. Base sequences of PCR-ampli ed DNA from representative samples of rope and fabric, with corresponding ax and hemp sequences for comparison. Sequences 795 and 722 931 are from rope samples; sequences 786 and 955 are from fabric samples. All samples were amplified using forward (rbcF2) and reverse (rbcR3a) primers. The sequencing 658 procedure was performed with each primer; the results from both primers are shown. Base differences in the hemp and flax DNA are shown in boldface black and the corresponding 723 659 superpositions in sample 931 are indicated by code letters (M,Y,R,K). 724 660 725 661 726 Madison, WI, USA),1.6 mL dNTPs (2.5 mM of each dNTP), 0.125 mL Taq BamH1 (Promega) and 1 ml of appropriate buffer at 37 C for at least 662 727 DNA Polymerase (Go Taq, 5m/mL, Promega), 0.6 mL of each primer 1 h. The mixtures were then subjected to acrylamide gel electro- 663 728 (forward and reverse, 20 mM), and 1 mL of template DNA. Normal PCR phoresis for 2 h at 110 V on 10-cm long gels. The separating gel 664 729 conditions were 96 C for 1 min; 35 cycles of 94 C for 45 s,60C for contained 10% acrylamide (37.5:1 acrylamide:bis-acrylamide, Bio- 665 730 45 s, and 72 C for 1 min; 72 C for 5 min; 4 C hold. For study of rad) and TriseHCl, pH 8.8; the stacking gel contained 4% acrylamide 666 731 amplification kinetics (Fig. 9), samples were removed after the 72 C and TriseHCl, pH 6.8. Electrode buffer contained 0.004 M Tris and 667 732 step of the 24th, 28th, 32nd, 36th, 40th, and 44th cycles. 0.192 M glycine. The gels were stained with ethidium bromide and 668 733 images of their fluorescence recorded. Quantification of fluores- 669 734 cence (Fig. 9) was performed using an Alpha Innotech FluorChem 670 3.4. Analysis of PCR products 735 8900 imager (Cell Biosciences Corp., Santa Clara, CA, USA). 671 736 Base sequence analysis was used to confirm the identity of the 672 The products of PCR were identified in two ways. Restriction 737 reaction products. Products were purified by separation on 1.5% 673 analysis and gel electrophoresis were used to distinguish Linum 738 agarose gels; the bands were excised, and the DNA isolated using 674 from Cannabis products. The sequence of the Cannabis DNA frag- 739 glass filters (Promega). The concentrations of nucleic acid in the 675 ment contained a BamH1 restriction site, which was missing in the 740 samples were measured by their fluorescence when combined with 676 Linum fragment. PCR solutions, 8 ml, were treated with 1 mlof 741 677 742 678 743 679 744 680 745 681 746 682 747 683 748 684 749 685 750 686 751 687 752 688 753 689 754 690 755 691 756 692 757 693 758 Fig. 7. Chromatograms showing superpositions in the DNA base sequence of samples 931 (top) and 786 (bottom). Bases marked M indicate superpositions of A (green) and C (blue); 694 bases marked Y indicate superpositions of T (red) and C. These chromatograms show partial sequences using primer rbcLF2; the chromatograms produced by primer rbcLR3a 759 695 showed additional superpositions. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) 760

T.M. Murphy et al. / Journal of Archaeological Science xxx (2011) 1e10 7

761 826 762 827 763 828 764 829 765 830 766 831 767 832 768 833 769 834 770 835 771 Fig. 8. DNA analysis of fabric samples using trnL primers. Numbers at the left show the 836 772 size of DNA fragments in base pairs. “Linum” and “Cannabis” were modern control 837 samples. Under the conditions of this experiment, only hemp and no flax DNA was 773 838 amplified from archeological samples. 774 839 775 ethidium bromide, and the base sequences of samples were 840 776 determined by the UC Davis Gene Sequencing Facility, using both 841 777 forward and reverse primers. 842 778 843 779 3.5. AMS-based 14C analyses 844 780 845 781 846 Radiocarbon measurements were performed at the Keck Carbon 782 847 Cycle Accelerator Mass Spectrometry Laboratory, University of 783 848 California, Irvine (Southon et al., 2004). Samples were chemically 784 849 pretreated by successive heating in 1 N HCl (to remove carbonates 785 850 and acid-soluble humates), 1 N NaOH (to remove base-soluble 786 851 humates), 1 N HCl (to neutralize NaOH), and ddH O (to remove 787 2 852 HCl) before combustion. 788 853 789 854 790 4. Results 855 791 856 792 The study focused on five samples of cordage and five of textile, 857 793 examples of which are shown in Fig. 3. 858 794 859 795 4.1. PCR 860 796 861 797 As noted in the Methods section, the choice of primers was 862 798 a critical part of the study. Following the discovery that no rbcL DNA 863 799 product 771 base pairs long could be amplified from the archeo- 864 800 logical DNA templates, we tested various rbcL primer combinations 865 801 to find a successful set (Fig. 4). Using template DNA from modern 866 802 hemp, amplified DNA was obtained with all four primer pairs 867 803 tested. However, templates from flax plants and from modern flax 868 804 rope only gave product using primer pairs rbcLF2/rbcLR3a and 869 805 rbcLF2/rbcLR4a. Three template DNAs extracted from archeological 870 806 samples gave product corresponding to the smallest (184 bp) band 871 807 obtained with flax (and hemp). Two archeological DNAs gave 872 808 a product with primer pairs rbcLF1/rbcLR3a that likely represented 873 809 hemp. For subsequent tests, we concentrated on rbcLF2/rbcLR3a, 874 810 which gave the smallest product and thus was least likely to 875 811 discriminate against the gene from a minor species. Fig. 9. Analysis of the relative amounts of flax and hemp DNA templates in three 876 812 The rbcL PCR products obtained using the template DNA archaeological samples: 931 (rope), 019 (fabric), 928 (rope). See Table 1 for sample 877 n 813 extracted from all the archeological cordage and textile samples ages. Estimates shown were calculated by the amplification factor (2 , where 878 n ¼ fl 814 contained strong bands of approximately 184 base pairs (Fig. 5). To number of cycles) needed to give equal intensity of ax and hemp bands. 879 815 distinguish between flax and hemp templates, we noted that the 880 816 band produced using authentic C. sativa DNA template was cut over showing (in the experiment presented) 44% cutting and sample 786 881 817 90% by BamH1, yielding fragments of 115 and 69 base pairs. The PCR showing 39% cutting. However, repetitions of the PCR reaction and 882 818 product of the modern UK rope did not show a detectable amount restriction digestion, particularly of sample 786, did not consis- 883 819 of cutting and thus was entirely flax. Interestingly, the PCR product tently show the smaller bands produced by BamH1. 884 820 of the modern Japanese rope showed a small amount of cutting, The base sequences of DNA from the archeological samples 885 821 indicating the presence of some C. sativa DNA. Most of the confirmed their identity as primarily L. usitatissimum Fig. 6. Within 886 822 archeological samples showed only faint bands at 115 and 69 base the 184-base pair amplified DNA, there was a stretch of 86 base 887 823 pairs after BamH1 treatment, indicating that, like the Japanese rope, pairs in which accurate sequence determinations could be 888 824 they contained little C. sativa DNA. Rope sample 931 and textile obtained from both primers. Within that region were nine sites at 889 825 sample 786 were the most notable exceptions, with sample 931 which the sequences of the Linum and Cannabis genes differed. The 890

8 T.M. Murphy et al. / Journal of Archaeological Science xxx (2011) 1e10

891 Table 1 956 892 Radiocarbon ages of representative samples of rope and fabric. 957 893 Laboratory number Accession number Material d13C(&) 14C Age (yrs BP) Calibrated age (cal BCE/CE)a 958 894 UCIAMS-79816 585440 Fabric (flax) 24.3 1870 15 CE 80e215b 959 895 UCIAMS-79814 582928 Rope (flax) e 1950 15 CE 5e85 960 896 UCIAMS-79813 582931 Rope (flax, hemp) 25.3 1900 15 CE 65e130 961 fl b 897 UCIAMS-79817 583019 Fabric ( ax, hemp) 24.3 4760 15 BCE 3635-3520 962 UCIAMS-79815 585786 Fabric (flax, hemp) e 4770 15 BCE 3635e3625 (0.12) BCE 3605e3525 (0.88) 898 963 a 14 s 899 Calibration of the C age for each measurement utilized CALIB 6.0 program protocols employing the Intcal09 data set (Reimer et al., 2009). Single interval 2 range 964 calibration values are expressed for intercepts representing 0.95 of the relative area under the probability distribution. In cases of multiple intercepts, the 2s ranges with 900 965 relative areas under probability distribution of 0.05 are noted in parenthesis for intercept separations of >20 yrs. Age ranges are rounded to nearest 5 yr increments. 901 b Multiple intercepts with probability distribution separations of <15 yrs. 966 902 967 903 sequence of rope sample 931 showed superpositions of two bases A few other base-sequence superpositions, e.g. K (¼G/T) or 968 904 at all nine sites (from at least one reading direction)(Figs. 6 and 7), a present/deleted base (¼T/), occurred near the ends of the 86- 969 905 confirming that this sample contained a significant amount of base pair stretch, but probably represented sequencing errors 970 906 Cannabis DNA. Chromatograms of the other samples, including rather than the inclusion of a variant Linum or another species, 971 907 textile sample 786, were not interpreted by the computer as since in each case the superpositions were found with only one 972 908 having a significant amount of Cannabis DNA, but small peaks primer. However, one of the superpositions in sample 931 indicated 973 909 corresponding to Cannabis bases could be seen in the chromato- C/T (confirmed in three independent determinations), whereas the 974 910 gram for sample 786 (Fig. 7). sequences of Linum and Cannabis at that position were C and A, 975 911 976 912 977 913 978 914 979 915 980 916 981 917 982 918 983 919 984 920 985 921 986 922 987 923 988 924 989 925 990 926 991 927 992 928 993 929 994 930 995 931 996 932 997 933 998 934 999 935 1000 936 1001 937 1002 938 1003 939 1004 940 1005 941 1006 942 1007 943 1008 944 1009 945 1010 946 1011 947 1012 948 1013 949 1014 950 1015 951 1016 952 Fig. 10. Comparison of base sequences of PCR-amplified rbcL DNA from six old-world fiber plants: flax (L. usitatissimum L.), hemp (C. sativa L), date palm (Phoenix dactylifera L.), 1017 Gossypium hirsutum Musa textilis Boehmeria nivea 953 cotton ( L.), abaca ( Nêe), and ramie ( (L.) Gaud). Aqua- and yellow-colored segments represent the sequences binding to the forward 1018 and reverse primers, respectively. The unshaded central area represents the region in which our sequences of the DNA from archaeological samples were confirmed (see Fig. 6). 954 Bold-face letters within the central area highlight differences from the flax sequence. The BamH1 site in the hemp sequence is shown in blue and underlined. (For interpretation of 1019 955 the references to colour in this figure legend, the reader is referred to the web version of this article.) 1020

T.M. Murphy et al. / Journal of Archaeological Science xxx (2011) 1e10 9

1021 respectively. In this position, the codons containing C, T, and origin. These include deliberate incorporation of hemp into flax 1086 1022 A (GGC, GGT, GGA) all code for glycine, so this is a “silent” substi- rope and linen textiles in situ and the importation of hemp- 1087 1023 tution. It is possible that the rbcL gene of the archeological Cannabis containing rope and textiles from other places. Samples 931 and 1088 1024 differed from the modern species used for identification. 786 most likely acquired their substantial amounts of hemp DNA in 1089 1025 In an attempt to confirm the data obtained from the rbcL anal- their fabrication. Samples that had much smaller amounts of hemp 1090 1026 ysis, we used the extracts of DNA as templates to amplify fragments DNA might have acquired it as dust, through their storage over 1091 1027 of two other chloroplast genes, trnL and matK. Using the trnL several centuries in contact with some (perhaps a small number) of 1092 1028 primers, we obtained Cannabis fragments from rope sample 931 hemp products. In the same way, samples might have acquired 1093 1029 and most of the textile samples, especially 786 (Fig. 8). There was small amounts of hemp dust during their excavation and transport 1094 1030 no indication of Linum template DNA in the ancient samples, to museums, particularly if their bundles were bound with hemp 1095 1031 although the modern control gave a good band. Using the matK rope or twine. Finally, we cannot ignore the possibility of 1096 1032 primers, we could not obtain Linum or Cannabis fragments from any contamination in the analytical laboratory, although controls did 1097 1033 archeological DNA templates, although again the modern control not indicate this. The lack of contemporary hemp bands in the 1098 1034 templates worked well (data not shown). Assuming that the archeological samples in Fig. 4 strongly suggests that the hemp 1099 1035 primer-complementary sequences of the trnL and matK target DNA, whether incorporated deliberately or through contamination, 1100 1036 genes have not changed drastically over the last 2000 years, these was ancient. 1101 1037 results suggest that different regions of chloroplast DNA fragment What is the possibility that the results we interpret as repre- 1102 1038 at different rates. senting hemp actually reflect another fiber? A comparison of the 1103 1039 Finally, we estimated the relative amounts of flax and hemp amplified segments of the rbcL genes from six old-world fiber 1104 1040 DNA using a semi-quantitative “kinetics” technique in which we plants, flax, hemp, date palm, cotton, banana palm, and ramie, 1105 1041 compared the amount of rbcL PCR product as a function of the shows that the base sequences of the central regions (Figs. 6 and 10, 1106 1042 number of replication cycles. Although under appropriate condi- unshaded) differ among all six. The BamH1 site in the hemp 1107 1043 tions, the amplification of DNAs can give amounts of products that sequence occurs only in that species, so that the minor bands seen 1108 1044 represent the relative amounts of different templates, many prob- in Fig. 5 must represent hemp. The specific base superpositions 1109 1045 lems with the technique can confound the analysis. It is more shown in Fig. 7 occur only in the flax-hemp pairing. The lack of 1110 1046 accurate to compare the number of cycles that give equal band other minor bases indicates that, if present, the amounts of other 1111 1047 intensities upon staining, even given the uncertainty inherent in fibers (or strictly speaking, their DNA) must be very small. 1112 1048 the relationship between staining intensity and DNA size. We The 14C dating of the samples confirms a suggestion by Dr. Orit 1113 1049 applied this technique to three samples that appeared to show Shamir (personal communication) that they represent two distinct 1114 1050 three different levels of hemp DNA, 931, 019, and 928. In fact, as periods of use, one prehistoric and one from the time of the Roman 1115 1051 shown in Fig. 9, the levels of hemp varied from ca one-fourth that of conquest. Previous 14C dating of two wood samples from CC found 1116 1052 flax (sample 931) to 1/4000 (sample 928). even earlier dates, 3670 and 4830 BCE (Rasmussen et al., 2005), but 1117 1053 the authors pointed out that the dates might have represented 1118 14 1054 4.2. AMS-based C dating “cultural activity 6000 years ago or the use of old wood.” Samples 1119 1055 from the “Cave of the Warrior,” dated to approximately the same 1120 1056 The ages of five samples, including two cordage and three textile times, ca. 3800 BCE and ca. 4400 BCE (Jull et al., 1998), included 1121 1057 samples, were determined by AMS-based 14C measurements. textiles and baskets, a clear indication of prehistoric culture. Our 1122 1058 Radiocarbon ages are expressed as conventional 14C dates three dated samples from the Early Bronze period, though some- 1123 1059 normalized to 25& d13C(Stuiver and Polach, 1977). The calibra- what younger, confirm the suggestion of very old cultural activity 1124 1060 tion of the 14C age for each measurement utilized CALIB 6.0 and add the information that this activity included the use of hemp. 1125 1061 program protocols employing the Intcal09 data set (Reimer et al., In conclusion, this work points out the value of PCR in deter- 1126 1062 2009). Calibrated values are expressed as a 2 sigma range. The mining the plant-fiber composition of textile and cordage, partic- 1127 1063 data (Table 1) revealed that the samples represented two distinct ularly when there are mixtures of fibers. Our multiple-disciplinary 1128 1064 periods. The two rope samples, 928 and 931, and textile sample 440 approach, combining the PCR data with literary-historical analysis 1129 1065 were produced in the first two centuries CE; two textile samples, and AMS-based 14C dating of the textile samples, has enabled us to 1130 1066 019 and 786, were much older, from the fourth millennium BCE. unearth this little known aspect of textile history e i.e. ancient 1131 1067 hemp in the Land of Israel. 1132 1068 5. Discussion 1133 1069 1134 Uncited references 1070 The PCR data confirmed the identities of the contents of all the 1135 1071 cordage and textile samples as primarily flax-linen. The sequence 1136 Jerusalem Talmud, 2001; Midrash, 1875; Midrash Beresheit 1072 data revealed the presence of hemp DNA, and by inference hemp 1137 Mishnah, 1907; Rabbah, 1903; Toseftah, 1975; Lauffer, 1971; 1073 fibers, but only in one rope sample, 931, and one textile sample, 1138 Rackham, 1938; Denton and Daniels, 2002; Eshel, 2009; Felix, 1074 786. The gel electrophoresis patterns, more sensitive to minor 1139 1967; Shamir, 2007. 1075 components, indicated the presence of hemp DNA in those samples 1140 1076 and all the other samples, with the exception of the control Linum 1141 1077 DNA and the DNA from the contemporary flax rope. Acknowledgment 1142 1078 The lack of linearity inherent in PCR made it impossible to 1143 1079 estimate the relative amounts of flax and hemp accurately from the The authors are indebted to Dr. Orit Shamir, curator of organic 1144 1080 initial analyses, which were performed with a fixed number of materials of the Israel Antiquities Authority, for allowing us access 1145 1081 amplification cycles, but a modification of the PCR protocol (Fig. 9) to the textile collection and for very courteously answering our 1146 1082 revealed that the fraction of DNA from hemp varied widely, from numerous questions. We wish also to thank Will Beaumont for his 1147 1083 25% to 0.025% that of the amount of flax DNA. The ubiquity of the dedicated laboratory work. Support to undertake the AMS-based 1148 1084 hemp DNA, particularly in the small amounts found in most of the 14C dates were provided, in part, by the Gabrielle O. Vierra 1149 1085 archeological samples, forces us to consider the possibilities for its Memorial Fund to RET. 1150

10 T.M. Murphy et al. / Journal of Archaeological Science xxx (2011) 1e10

1151Q2 References Jull, A.J.T., Donahue, D.J., Carmi, I., and Segal, D., 1998. The cave of the warrior: 1208 1152 Ancient Rabbinic Sources a fourth millennium burial in the Judean Desert, In: Schick, T. (Ed.), IAA Reports 1209 No. 5. Israeli Antiquities Authority, Jerusalem, Israel. 1153 Lazcano, A., Guerrero, R., Margulis, L., Oro, J., 1988. The evolutionary transition from 1210 1154 Jerusalem Talmud, 2001. Academy of the Hebrew Language edition, Jerusalem. RNA to DNA in early cells. Journal of Molecular Evolution 27, 283e290. 1211 Midrash Tanʿumah, 1875. Warsaw edition. 1155 Liepelt, S., Sperisen, C., Deguilloux, M.-F., Petit, R.J., Kissling, R., Spencer, M., De 1212 Midrash Beresheit Rabbah, 1903. Theodor - Albeck edition, Berlin. Beaulieu, J.-L., Taberlet, P., Gielly, L., Ziegenhagen, B., 2006. Authenticated DNA 1156 Mishnah, (Vilna edition, 1907). from ancient wood remains. Annals of Botany 98, 1107e1111. 1213 1157 Toseftah, 1975. Zukermandel edition, Jerusalem. Lieberman edition, New York. McKenna, H.A., 2004. Handbook of Fibre Rope Technology. CRC Press, Boca Raton FL. 1214 e 1158 1955 1988. Mitchel, L., 1992. Tell Hesban Stratum 15; ca. 198e63 B.C. In: Hesban 7. Andrews 1215 University Press, Berrien Springs, pp. 3e38. Ancient general sources 1159 Mukherjee, A., Roy, S.C., De Bera, S., Jiang, H.-E., Li, X., Li, C.-S., Bera, S., 2008. Results 1216 1160 of molecular analysis of an archaeological hemp (Cannabis sativa L.) DNA 1217 ’ e 1161 Lauffer, S. (Ed.), 1971. Diokletian s Preisedikt. De Gruyter, Berlin, pp. 170 179. sample from North West China. Genetic Resources and Crop Evolution 55, 1218 Rackham, E. (Ed.), 1938. Pliny the Elder. The Natural History, Loeb edition. Mass, e 1162 481 485. 1219 Cambridge. Müller, M., Papiz, M.Z., Clarke, D.T., Roberts, M.A., Murphy, B.M., Burghammer, M., 1163 Riekel, C., Pantos, E., Gunneweg, J., 2003. Identification of the textiles using 1220 Modern Sources 1164 microscopy and synchrotron radiation x-ray fibre diffraction. In: Humbert, J.-B., 1221 ’ 1165 Gunneweg, J. (Eds.), Khirbet Qumran et Ain Feshkha II. Etudes d anthropologie, 1222 Allegro, J.M., 1965. Search in the Desert. W.H. Allen, London. de physique, et de chimie. Academic Press, Gottingen, pp. 277e286. 1166 Amar, Z., 2000. Agricultural Produce in the Land of Israel in the Middle Ages (in Müller, M., Murphy, B., Burghammer, M., Riekel, C., Roberts, M., Papiz, M., Clarke, D., 1223 1167 Hebrew). Yad Izhak Ben Zvi, Jerusalem. Gunneweg, J., Pantos, E., 2004. Identification of ancient textile fibres from 1224 Baginski, A., 2001. Later islamic and medieval textiles from excavations of the Israel Khirbet Qumran caves using synchrotron radiation microbeam diffraction. 1168 e 1225 antiquities authority. Textile History 32, 81 92. Spectrochimica Acta Part B 59, 1669e1674. 1169 Barber, E.J.W., 1991. Prehistoric Textiles. Princeton University Press, Princeton. Müller, M., Murphy, B., Burghammer, M., Snigireva, I., Riekel, C., Gunneweg, J., 1226 1170 Belis, M., 2003. Des Textiles Catalogues et Commentaires. in French. In: Humbert, J.-B., Pantos, E., 2006. Identification of single archaeological textile fibres from the 1227 ’ Gunneweg, J. (Eds.), Khirbet Qumran et Ain Feshkha II: Etudes d anthropologie, de Cave of Letters using synchrotron radiation microbeam diffraction and micro- 1171 e 1228 physique, et de chimie. Academic Press, Gottingen, pp. 207 276. fluorescence. Applied Physics A83, 183e188. fi 1172 Bergfjord, C., Holst, D., 2010. A procedure for identifying textile bast bers using Müller, M., Murphy, B., Burghammer, M., Riekel, C., Pantos, E., Gunneweg, J., 2007. 1229 fl e 1173 microscopy: ax, nettle/ramie, hemp, jute. Ultramicroscopy 110, 1192 1197. Ageing of native cellulose fibres under archaeological conditions: Textiles from 1230 1174 Caramelli, D., Milani, L., Vai,S.,Modi, A., Pecchioli, E., Girardi, M., Pilli, E., Lari,M., Lippi, B., the Dead Sea region studied using X-ray microdiffraction. Applied Physics A89, 1231 Ronchitelli, A., Mallegni, F., Casoli, A., Bertorelle, G., Barbujani, G., 2008. A 28,000 877e881. 1175 years old Cro-Magnon mtDNA sequence differs from all potentially contaminating Oakley, F.I., 1928. Long Vegetable Fibres. Ernest Benn, London. 1232 1176 modern sequences. PLoS ONE. http://e2700.doi:10.1371/journal.pone.0002700. Platt, E., 2009. The Textile Tools from Tell Hesban and Vicinity. In: Hesban 12. 1233 1177 Crowfoot, E., 1960. Appendix A: aextiles, matting and basketry. In: Kenyon, K.M. Andrews University Press, Berrien Springs. 1234 (Ed.), Excavations at Jericho I: Tombs Excavated in 1952-54.London. British Porat, R., Eshel, H., Frumkin, A., 2009. The Christmas Cave in the lower Qidron 1178 e 1235 School of Archaeology in Jerusalem, pp. 519 526. stream (in Hebrew). In: Eshel, H., Porat, R. (Eds.), Refuge Caves of the Bar 1179 Crowfoot, E., 1965. Textiles, matting and basketry. In: Kenyon, K.M. (Ed.), Excava- Kokhba Revolt, vol. 2. Israel Exploration Society, Jerusalem, pp. 31e52. 1236 1180 tions at Jericho II: Tombs Excavated in 1955-58.London. British School of Porat, R., Eshel, H., Frumkin, A., 2007. Remnants from the time of the revolts against 1237 Archaeology in Jerusalem. ” 1181 the Romans from Caves in the Lower Nahal Qidron region (in Hebrew). 1238 Crowfoot, E.,1982. Textiles, matting and basketry. In: Kenyon, K.M. (Ed.), Excavations at Mehqarei Yehuda v’Shomron 18, 231e264. e 1182 Jericho IV. London. British School of Archaeology in Jerusalem, pp. 546 550. Reimer, P.J., Baillie, M.G.L., Bard, E., Bayliss, A., Beck, J.W., et al., 2009. INTCAL09 and 1239 1183 Crowfoot, G., Crowfoot, E., 1961. The textiles and basketry. In: Les Grottes De Marine09 radiocarbon age calibration curves, 0e50,000 years cal BP. Radio- 1240 e Murabba at. Clarendon, Oxford, pp. 53 62. carbon 51, 1111e1150. 1184 fi 1241 Denton, M.J., Daniels, P.N. (Eds.), 2002. Textile Terms and De nitions. The Textile Shamir, O., 2001. Byzantine and Early Islamic Textiles Excavated in Israel. Textile 1185 Institute, Manchester. History 32 (I), 93e105. 1242 fi 1186 Dunbar, M., Murphy, T.M., 2009. DNA analysis of natural ber rope. Journal of Shamir, O., 2007. The loomweights. In: Stern, E. (Ed.), Ein Gedi Excavations I: Final 1243 Forensic Sciences 54, 108e113. 1187 Report. Israel Exploration Society, Jerusalem, pp. 381e390. 1244 Eshel, H., 2009. On the new research in the caves of refuge in the Judean Desert and Shamir, O., 2007. The textiles. In: Hirschfeld, Y. (Ed.), Ein Gedi Excavations II: Final 1188 Judean Hills. In: Eshel, H., Porat, R. (Eds.), Refuge Caves of the Bar Kokhba Revolt Report. Israel Exploration Society, Jerusalem, pp. 587e594. 1245 e 1189 (in Hebrew), vol. 2. Israel Exploration Society, Jerusalem, pp. 1 9. Smith, C.I., Chamberlain, A.T., Riley, M.S., Stringer, C., Collins, M.J., 2003. The thermal 1246 Felix, Y., 1967. Mixed Sowing Breeding and Grafting (in Hebrew). Dvir, Tel Aviv. history of human fossils and the likelihood of successful DNA amplification. 1190 Cannabis sativa 1247 Fleming, M., Clarke, R., 1998. Physical evidence for the antiquity of L. Journal of Human Evolution 45, 203e217. e 1191 Journal of the International Hemp Association. 5, 80 92. Shamir, O., Sukenik, N., 2010. The Christmas Cave textiles compared to Qumran 1248 1192 Gilbert, M.T., Tomsho, L.P., Rendulic, S., Packard, M., Drautz, D.I., Sher, A., textiles. Archaeological Textiles Newsletter 51, 26. 1249 1193 Tikhonov, A., Dalén, L., Kuznetsova, T., Kosintsev, P., Campos, P.F., Higham, T., Sheffer, A., Granger-Taylor, H., 1994. Textiles. In: Masada, I.V. (Ed.), The Yigael Yadin 1250 Collins, M.J., Wilson, A.S., Shidlovskiy, F., Buigues, B., Ericson, P.G., e 1194 Excavations 1963 1965 Final Reports. Israel Exploration Society, Jerusalem, 1251 Germonpré, M., Götherström, A., Iacumin, P., Nikolaev, V., Nowak-Kemp, M., pp. 153e243. 1195 Willerslev, E., Knight, J.R., Irzyk, G.P., Perbost, C.S., Fredrikson, K.M., Harkins, T.T., Southon, J., Santos, G., Druffel-Rodriguez, K., Druffel, E., Trumbore, S., Xu, X., 1252 1196 Sheridan, S., Miller, W., Schuster, S.C., 2007. Whole-genome shotgun sequencing Griffin, S., Ali, S., Mazon, M., 2004. The Keck carbon cycle AMS laboratory, 1253 of mitochondria from ancient hair shafts. Science 317, 1927e1930. 1197 university of California, Irvine: initial operation and a background surprise. 1254 Granger-Taylor, H., 2006. Textiles from Khirbat Qazone and the cave of letters, two Radiocarbon 46, 41e49. fi 1198 burial sites near the Dead Sea: similarities and differences in nd spots and Sperber, D., 1986. Nautica Talmudica. Brill, Leiden. 1255 In Situ 1199 textile types. In: Textiles Riggisberger Berichts 13. Abegg-Stiftung, Stuiver, M., Polach, H.A., 1977. Discussion: reporting of 14C data. Radiocarbon 19, 1256 Riggisberg, pp. 113e131. e 1200 355 363. 1257 Herzberg, A.S., 1924. Cultural Life in Israel in the Mishnaic and Talmudic Periods (in Walton Rogers, P., 2003. Fibres in miscellaneous samples from a site in the Dead 1201 Hebrew). Sztybel, Warsaw. Sea region. In: Khirbet Qumran et Ain Feshkha II. Academic Press, Gottingen, 1258 1202 Humpert, J.B., Gunneweg, J. (Eds.), 2003. Khirbet Qumran et Ain Feshkha II. pp. 287e288. 1259 Academic Press, Gottingen. 1203 Weindling, L., 1947. Long Vegetable Fibers. Columbia University Press, New York. 1260 Humpert, J.B., Chambon, A. (Eds.), 2003. The Excavations of Khirbet Qumran and Ain Yadin, Y., 1963. The finds from the Bar Kokhba period. In: Cave of Letters. Israel 1204 Feshkha IB. Vandenhoeck & Ruprecht, Gottingen. Exploration Society, Jerusalem (in Hebrew). 1261 1205 1262 1206 1263 1207 1264