Nigella in the Mirror of Time - A Brief Attempt to Draw a Genus' Ethnohistorical Portrait
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Offa 69/70, 2012/13, 147–169. Nigella in the Mirror of Time A Brief Attempt to Draw a Genus’ Ethnohistorical Portrait By Andreas G. Heiss, Hans-Peter Stika, Nicla De Zorzi and Michael Jursa Introduction1 Nigella (fennel flower) is one of the smaller gen- vated for thousands of years. At least today it is in- era in the Ranunculaceae (buttercup) family: it com- deed a frequently consumed condiment in North prises only about 15 species if considered in the wid- Africa, the Arabian Peninsula, and the Indian sub- er sense, thus including the sister taxa Garidella and continent while also being the object of intensive Komaroffia (Zohary 1983; Dönmez/Mutlu 2004). pharmacological research and more or less reliable In this study, we also treat the various (sub)species phytomedicine vendors (see below). within the Nigella arvensis complex as one single The evolutionary origins of the genus are most species – a rather strong simplification when consid- probably to be found in its centre of species diver- ering the results obtained by Strid (1970) or Bitt- sity, which occurs in the Aegean (Bittkau/Comes kau/Comes (2005; 2008), but sufficient for our pur- 2008) and the adjacent Western-Irano-Turanian re- pose in this paper. gion (Strid 1970; Zohary 1983), as illustrated in All members of the genus Nigella are therophytes Figure 1. N. sativa may thus have come into exist- (annuals that overwinter as seeds) with a short life ence somewhere in this area, although its alleged cycle, requiring open habitats to flourish. This makes long-term cultivation would raise significant obsta- several of them occur frequently in anthropogenic cles to easily proving this hypothesis: When a crop ecosystems. As an example, the well-known orna- has been propagated by humans and traded across mental species Nigella damascena (love-in-a-mist) long distances for long periods, knowledge of (and can nowadays be observed as a rapid colonizer of data on) its origins tends to be obliterated, and only fallow land around the Mediterranean. Taxa from long-term multidisciplinary studies may bring these the Nigella arvensis complex have played a role in roots back into daylight. However, it is precisely this the segetal vegetation of Europe’s agriculture since presumably long history of use, cultivation and/or at least the Late Iron Age (Caneppele et al. 2010; unintentional spread which makes Nigella a promis- Kohler-Schneider et al. forthcoming), but are ing challenge, and an object of considerable interest nowadays endangered or even extinct in wide areas both for archaeobotany/palaeoethnobotany and bo- due to intensification of agricultural technology, and tanical chorology. Also the great popularity which excessive use of fertilizers and herbicides (in central Nigella (and, again, especially N. sativa) has gained Europe, cf. Ludwig/Schnittler 1996; Niklfeld/ in pharmaceutical literature, particularly well-illus- Schratt-Ehrendorfer 1999). Nigella sativa (black trated in M. A. Anwar’s (2005) bibliometric study, cumin), to mention a third example, is probably the makes this kind of research necessary: The amount best-known species of the genus. Lore has it – and we of published literature on black cumin or “black are going to test this against the available evidence – seed” is immense, generating the impression that the that this condiment has been propagated and culti- plant must be a true panacea. Most of these scientif- 1 Our thanks go to Ursula Thanheiser (VIAS, University of most valuable help in improving an earlier version of the Vienna), Otto Brinkkemper (ROB Amersfoort), Marlu Kühn manuscript. – Part of research for this publication was fun- and Patricia Vandorpe (both IPNA/IPAS, Basel University), ded by the Hochschuljubiläumsstiftung der Gemeinde Wien and Marie-Pierre Ruas (Muséum national d’Histoire natu- (HJSt; project number H1888/2008) to which we are greatly relle, Paris) for kindly providing additional information on indebted. Part of the data is from a project (P12049) funded archaeological finds of Nigella species. We thank Cozette by the Austrian Science Fund FWF (Fonds zur Förderung Griffin-Kremer (Université de Bretagne Occidentale) for her der wissenschaftlichen Forschung). 147
Fig. 1. Modern natural distribution of Nigella species in the Old World: 100 % white: 0 species, 100 % dark grey: 9 species. Map source: Spiess 2002; distribution data and species delimitations according to Zohary 1983. ic publications, and even more so the popular “alter- pre-Linnean times (see pp. 158 ff. Table 1–3). Our native” medical literature and “black seed oil” ad- goal is not only to present an “elaborate florilegi- vertisements, frequently refer to historical accounts um”, but to draw a picture of the cultural percep- or “traditional uses” stressing the alleged high med- tions of a group of plants as complete as possible in ical potential of the plant – and usually without dis- these few pages. In addition, the sources we have cit- closing any sources. ed here may prove useful for further research in ar- The current study now aims to compile literature chaeobotany or historical botany, ethnobotany and on Nigella’s possible past uses, by analysing extant ethnomedicine, not only of Nigella but also of oth- data on the genus provided by archaeology and ar- er plant taxa. chaeobotany, with historical written records from Sources of information Archaeobotany For the current work, an archaeobotanical bib- 1999; 2000; 2001; 2004), successively adding more liography of the genus was built, starting with the recent publications on archaeological finds. Infor- comprehensive indices by Schultze-Motel (1992; mation on unpublished data was collected via per- 1993; 1994) and Kroll (1995; 1996; 1997; 1998 a; sonal communication (see Tab. 1). Written Sources The second group of sources used in this work was published posthumously), as well as secondary were numerous written documents from all periods literature. available to us, with a strong focus on the eastern The temporal range covered extends from Meso- Mediterranean – the probable centre of origin of the potamian cuneiform tablets of the mid-3rd millen- genus – and on Europe. Although we tried to lim- nium BC to the 17th century AD (Table 2–3). We de- it ourselves to primary literature from the respec- cided to omit later periods as the development of tive periods, these were not always available. So, in botany as a science on its own was already very ad- some cases, later editions had to be used (as in the vanced in the late 17th century, the literature leav- case of the book by Gaspard Bauhin 1671 which ing only a few questions on the proper identification 148
of cultivated plants (see e. g. Pavord 2005). For the in the sense of the ICBN (the International Code of most recent periods included (Early Modern Times), Botanical Nomenclature, see McNeill et al. 2006), not all available books or documented plant names only the latter are printed in italics. In contrast, all of the (putative) Nigella species were considered in historical plant names are put in quotation marks, the evaluation due to the great number and the space even if this does not concur with common practice. restrictions of a brief article. Hence, we limited them Quotation marks are not used in Tables 2–3, col- to those which seemed most important to us, and umn 4, where all plant names are written regularly only discussed some characteristic examples in the without quotes, in order to maintain legibility. Orig- text. inal spellings and transcriptions, if missing in the In order to facilitate discrimination between his- literature consulted, were taken from Katzer (2006 torical plant names and modern botanical names onwards). Plant Identification and Interpretation As this work is mainly intended as a compendi- dental thought, interpretation by the researcher is um and bibliographical work, proper identification always necessary. This interpretation must be based (or its refusal) was one of our main goals. on the plant’s properties, the archaeological context, Archaeological plant remains can exactly indi- and all available information on the contemporary cate the presence of a certain plant taxon in a certain society. Analogies found in written sources and in place and period, provided that their identification ethnographic research often assist in better under- is still possible. They thus serve as a valuable source standing a plant’s role and perception (Hansson/ of information on a plant’s distribution, be it natu- Heiss forthcoming). In the case of Nigella species, ral or anthropogenic. But in order to find out about the oil-rich seeds (e. g. Heiss et al. 2011) often bear- past ways of plant utilisation, their reception by past ing a conspicuous aroma such as in N. sativa and societies, or their relation to symbolic or transcen- in the N. arvensis complex suggest a primary use as condiments. However, all other possible aspects of Nigella utilisation need to be inferred from oth- er sources. While the identification of archaeological re- mains mainly depends on their preservation, writ- ten sources often pose several problems to identifi- cation of the plants mentioned, mainly due to vastly differing concepts of what a plant (and a species) is: In a nutshell, in modern botany a species is referred a to by similarities of individuals to a certain speci- men in a herbarium (the holotype). These similari- ties usually involve a broad range of characters, such as morphology, anatomy, karyology (chromosome counts), embryology, cross-breeding, chemical con- b c Fig. 2. The three Nigella species which currently are document- Fig. 3. Agrostemma githago (corn-cockle) flower and seed. Scale ed in the archaeological record by their seeds. a N. arvensis (field bar length (seed): 1 mm. nigella). b N. damascena (love-in-a-mist). c N. sativa (black cumin). Scale bar length (seed): 1 mm. For a detailed seed iden- tification key of the whole genus, please refer to Heiss et al. 2011. 149
stituents, or DNA/RNA sequences – and the range thors to treat Nigella (mainly N. sativa) and the tox- is constantly widening. Although in Linnaeus’ time, ic Agrostemma githago (corn cockle) as more or less far fewer characters were available, the idea behind the same plant – or at least this character has led to the system was basically the same. constant confusion of the two, despite all their con- In pre-Linnean times, however, all kinds of prop- spicuous differences in vegetative, flower, and seed erties were used for classifying the living (plant) morphology (see Fig. 2 versus Fig. 3). Occurrence of world, and some of these characters were not even Nigella arvensis and Agrostemma githago in basically remotely useful for a taxonomical classification: the same ecosystems (in crop fields) makes differen- Seed colour, flower colour, scent, taste, usefulness of tiation in literature even more difficult. its fruits for human nutrition, “tree-like” versus her- Thus, identification of the species mentioned in baceous habit, or even their most common place of written sources had to remain uncertain in many growth, to mention a few. Furthermore, these char- cases. Major exceptions were, on the one hand, il- acters were not generally used in combination, but lustrated herbals allowing identification of the taxa in a quite isolated way. The main consequences for mentioned. On the other hand, several written Nigella can be described as follows: Greek sources sources gave hints on the nature of the plants by de- mention a plant possibly corresponding to Nigella scribing their attributes: A “nigella” plant growing sativa under the names of “git(h)” and “melanthion” in the field and with pungent black seeds was most (e. g. in Fuchs 1895–1900; Berendes 1902; also see probably identified correctly by the ancient author. here Table 2). As will be demonstrated later, the But if any warning of possible toxic properties was common feature of dark-coloured – “black”, so to found, or even a description of red flowers, we could speak – seeds was obviously sufficient for some au- be sure of confusion with Agrostemma. Results Oldest Archaeological Finds In the eastern Mediterranean, the presence and been known and used in the Near East and in An- use of Nigella sativa (black cumin) is more than cient Egypt since at least 2700 BC, the oldest archae- well-documented: In Egypt alone, archaeological ological evidence of Nigella sativa originating from finds from eleven sites help trace the plant from the Saqqara (Rzeuska 2006). 28th/27th century BC up to Roman times, showing Interestingly enough, even older finds of Nigella a significant increase of information since Renate seeds have been discovered in an Eneolithic (Copper Germer’s (1989) comment on the “rare evidence” Age) site in the Spanish province of Murcia (Pre- of Nigella (likewise, see Zohary/Hopf 2000). Scat- cioso Arévalo 2004). Unfortunately they could tered evidence also exists from the Antiquity of pre- not be identified to species level. There is also quite sent-day Syria, Jordan, and Turkey. However, in spite early archaeological evidence for Nigella from Cen- of the vast written evidence on the knowledge of Ni- tral Europe: Pollen grains in miners’ faeces from the gella (sativa) from the Ancient Near East, such as in Late Iron Age Dürrnberg indicate the consumption Babylonian and Assyrian cuneiform tablets (see be- of some unidentified Nigella species (Aspöck et al. low), we still do not know about the plant’s role in 2007). Even older is the Late Bronze Age find of a Ancient Egyptian culture: Textual sources are still partly charred N. damascena seed in an Alpine ore- missing, although frequently – and perhaps contra- processing site (Heiss/Oeggl 2005). Both find con- ry to more reliable knowledge? – stated otherwise texts may suggest import and consumption of Nigel- by companies eager to sell Nigella sativa oil (e. g. la species by metallurgy specialists from the eastern Agentur 2012). What is certain is that the plant has Mediterranean area. Oldest Written Sources: the Ancient Near East This region is basically the richest in very ear- clear, as “kyminon” directly corresponds to Greek, ly historical sources for Nigella, and for N. sativa in Akkadian and Arabic names for cumin (Cuminum particular. We may begin tracing this evidence with cyminum; see Zohary 1995). The Hebrew “qeṣa ḥ” two sources from Biblical texts, which are common- may be somewhat closer (Trever 1959; Zohary ly cited as “clear” evidence of Nigella use: The “qeṣa ḥ” 1995), as the word is used in parallel with “kammōn” from the Book of Isaiah, and the “kyminon” from (cumin) in Isaiah and is understood to refer to black the Book of Matthew. The latter at least is not all that cumin, i. e. Nigella sativa, by ancient translations and 150
Medieval and later commentators alike; also the sur- some scribal traditions of late 3rd millennium BC vival of the word in the form “kizḥa” in Palestinian Sumerian, e. g. in the city of Girsu, the pair “white/ vernacular Arabic points in the same direction (Löw black gamun” is given as (unqualified) “gamun” and 1924, 120–123; see also below). However, it is only in “zibum” or “zibibianum” – the latter corresponding the cuneiform sources from Mesopotamia that one of course to the Akkadian words “zibû”/“zibibiānu” finds good contextual evidence for the cultivation of (e. g. Maekawa 1993, 123 no. 94 ii: 11–12; iii: 4–5; vi: Nigella and for its use for culinary and, to a lesser 5–6). The interpretation of “zibibiānu” as “black ga- degree, medical purposes from the late 3rd millenni- mun/kamūnu” also re-occurs in the native Babyloni- um BC until the late 1st millennium BC. In fact, the an lexical tradition of the 2nd and 1st millennium BC Ancient Near East has produced the richest body of (e. g. Thompson 1949, 70–71; Landsberger/Reiner ancient textual evidence that can be brought to bear 1970, 94: 300–303). The fact that the black seeds were on the present topic, and an unusually strong case the principal product of the plant follows also from for a positive identification of the Ancient (Sumerian medical recipes which indicate that “zibû” had to be and Akkadian, i. e. Babylonian and Assyrian) plant ground with a pestle or a handmill (e. g. von Soden name(s) with the genus Nigella and even specifically 1965–1981, 1525; Salonen 1965, 53). N. sativa can be made – a case that is based not only In the late 3rd millennium BC, the black and white on etymology, as is commonly the case with Ancient “gamun” were cultivated in southern Mesopotamia Near Eastern (and generally ancient Semitic) plant in special spice gardens (Maekawa 1985, 99; 112) to- names, but also based on context. gether with coriander and legumes. From later peri- Starting points are the Akkadian plant names ods there is only some scattered evidence for the cul- “kamūnu” and “zibibiānu”, the latter with its vari- tivation of “kamūnu” in the south (e. g. Stol 1986, 61; ants “zibû”, “sabibiānu”, “šibibānu” etc. (in the follow- 98), and we do not hear of the cultivation of “zibû”, ing, we will always use the variant that is attested in even though the use of the seeds continues in this re- the source in question). These words do not have Se- gion. After the 3rd millennium, explicit attestations mitic etymologies in the sense that they are not of Se- for the cultivation of “zibibiānu” (“sabibiānu”) are mitic origin, rather, they are ancient Wander- or Kul- only available for northern Mesopotamia. Evidence turwörter. The first corresponds to Arabic “kammūn”, comes from the Khabur region (Röllig/Tsukimoto Hebrew “kammōn”, Syriac “kammūnā” (cumin, 1999, 436: note the absence of “kamūnu” in this con- Cuminum cyminum), this word itself being of course text) and from the region around modern Kirkuk a loan of the same origin in all modern European lan- (Zaccagnini 1979, 128). Since the bulk of the evi- guages (via Greek). “Zibibiānu” etc. etymological- dence for the use of “zibibiānu” is also from north- ly corresponds to Syriac “šbūbānā” (and variants), ern Mesopotamia, it is probable that the plant was which is equated in various traditions with Hebrew cultivated and used predominately in this region, “qeṣaḥ” and Arabic “šūnīz”, the latter of which is a while the seeds used for culinary and medical pur- synonym of “kammūn aswad” (black kammūn). Me- poses in the south were possibly imported (cf. Rey dieval Hebrew and Western early modern as well as nolds 2007, 178). modern commentators take all these words to desig- Medical use of “zibû” involves usually braying nate Nigella (locus classicus: Löw 1924, 120–123). The the seeds in a mortar and mixing them with a fatty etymologies of the Akkadian words were recognized substance, or with barley beer. It is used for mouth early on, and they have thus generally been translated or tooth disorders, but also for other internal prob- as “cumin” and “black cumin” (N. sativa) respective- lems (Thompson 1949, 70–71). Culinary use is much ly (e. g. von Soden 1965–1981, 434; 1524; Landsber better documented. In the late 3rd millennium BC, ger/Gurney 1957/58, 336; Thompson 1949, 69–72). the “gamun” varieties are well attested as culinary The context of the attestations of “zibibiānu” etc. ingredients (e. g. Brunke 2011, 198; 219). From lat- bears out the etymological argument. The earliest ev- er periods, rich bodies of evidence come again from idence comes from the second half of the 3rd millen- northern Mesopotamia. In the city of Mari on the nium BC and is in Sumerian. In administrative texts middle Euphrates, “zibû” was an important spice of the period one encounters the condiment “gamun” that was used side by side with “kamūnu”. The two (e. g. Steinkeller/Postgate 1992, 77). This word is were important ingredients of the sumptuous mer- of the same origin as the slightly younger “kamūnu”. su cake, in addition to flour, fat, dates and other “Gamun” comes in two variants: “white” (babbar), fruit, as well as garlic and coriander (Bottéro 1995, and “black” (gi6) (e. g. Schneider 1931, no. 64: 7–8). 23–24), but they were also used for other meat and The cuneiform signs used to write the loan word vegetable dishes (Sasson 2004). Also in the city of “gamun”, viz. (Ú-ŠE+NUN.NUN-SAR or Ú-DIN- Alalakh, further to the west, “zibû” was much in de- ŠE+NUN.NUN-SAR) indicate that garden plants mand (Zeeb 2001, 193–194), and it is also attested (Ú … SAR) characterized by their grainy seeds (ŠE) in the 13th century BC in the city of Ugarit on the are meant, or rather, the seeds of these plants. In northern Mediterranean coast in the local alpha- 151
betic cuneiform script in the form “ssbyn” (ibid.). culinary recipes from southern Mesopotamia from Around the same time, Hittite sources from Ana- the first half of the 2nd millennium BC, “zibû” is not tolia make their contribution: There the word “kap- employed, while “kamūnu” is (Bottéro 1995). pani-“ (a loanword, with m > p) renders the Sum- Summing up the Mesopotamian evidence, we erogram “gamun”, and again we find a black and a conclude that the contextual data support the ety- white variety (e. g. Süel/Soysal 2003, 354 §9’: 18’– mology-based identification of “zibû” etc. with Ni- 21’). The black variety is used, inter alia, in a ritual gella sativa. Certainly, nothing in the sources con- where it is sprinkled on dough – a reflex of its being tradicts the identification. We note that the “zibû” used as an ingredient of bread (Hoffner 1974, 103– seeds were in particularly intensive use in, and 104). The richness of the data from northern Meso- therefore perhaps originating from, the hilly regions potamia summarized here makes the dearth of in- of today’s Eastern Turkey, Northern Syria and (espe- formation from the south stand out: In the detailed cially) Northern Iraq. Greek and Roman Antiquity Written sources from Greek Antiquity are – con- Only much later, in the Natural History of Pliny trary to what would be expected – much more am- the Elder from the 1st century AD, does the situa- biguous in their contents than the much older sourc- tion improve: Apart from attributing numerous me- es mentioned above: The works by Hippocrates dicinal effects to the plant (the name “melanosper- contain hints on the utilisation of a plant “melan- mon” = “black seed” corresponds a lot better to the thion” (black flower), a name which seems to be in morphology of Nigella than “melanthion” = “black contrast to Nigella’s blue or white flowers. As no flower”), Pliny describes the extraordinary smell of morphology is described and one of the Hippocrat- its black seeds, as well as their use as a bread con- ic recommendations is that of an abortifacient (cf. diment (Bostock/Riley 1855, lib. 20, 71). Columel- Fuchs 1895–1900), the identity as Nigella seems at la’s recommendation of “git” as an appetizing drug least questionable. von Grot (1887) suggested the for horses (Forster/Heffner 1954) might also be Hippocratic “melanthion” might even be the high- a plausible use for an aromatic plant like N. sativa. ly toxic ergot (Claviceps purpurea), its black sclerotia Pliny’s contemporary author, Pedanios Dioscorides, (the dense mass in which the fungus’ fruiting bodies recommends the plant “melanthion” for roughly the are produced) protruding seed-like from cereal ears. same purposes – with one important difference, as he The advice to “sort it out” (εκλέξας; ibid. 81) is thus warns against the possible lethal effects of the plant’s interpreted as to avoid ergot intoxication. Also the seeds consumed in excess (Berendes 1902). Consid- use as an abortifacient (among other gynecological ering the very low toxicity in all Nigella species (Roth indications, see Table 2) speaks, in von Grot’s opin- et al. 1994), this “mistake” may be due to some ran- ion, in favour of ergot. But rye, which is most suscep- dom misconception, or to the amalgamation of the tible to ergot infection, has never been cultivated as a plant’s properties with those of a toxic plant. Wheth- main cereal in the Mediterranean (cf. Behre 1992), er this species was the corn cockle or not, the “toxic- rendering wide-spread ergotism as well as common ity myth” found in Dioscorides, as we will see later, knowledge about it rather improbable phenomena. was carried forward up to Modern times. And, as mentioned above, Nigella (see Fig. 2) was fre- Archaeologcial evidence of Nigella sativa during quently confused with the toxic – and black-seeded Roman times comes mainly from Egyptian excava- – corn cockle (Agrostemma githago; see Fig. 3) dur- tions (van der Veen 2007; J. Walter/U. Thanheiser ing most periods since Antiquity. It may well be pers. comm.), while only one site in France and one that proper identification of the “melanthion” from in Great Britain document long-distance transport the Hippocratic texts is not possible at all, and it is of the plant across Europe in this period. Also, Ro- very likely that this oldest mention of black cumin man finds of N. arvensis exist from several sites in – the basis for centuries of medical literature deal- central Europe (see Table 1), but currently it is im- ing with this plant – in fact may have described the possible to tell whether at the time this plant was corn cockle. used at all by the local population. 152
Middle Ages The rich archaeobotanical evidence published on as an insecticide. Due to the term “Zizania”, usual- Nigella finds during Antiquity vastly decreases during ly used for a grass-like weed, one translator (Riethe the Middle Ages: Only two sites with Nigella sativa are 1989) suggests it could even refer to poison darnel reported from Late Medieval Germany (Wolf 1991; (Lolium temulentum). In a different chapter, Hilde- Lange 1993), and one site with N. arvensis in Austria gard however does mention something closer to Ni- (unpubl., S. Wiesinger/U. Thanheiser pers. comm.). gella (sativa), when she suggests adding “gith” as a In the south-eastern Mediterranean, archaeologi- condiment to fish (see Table 2). cal and textual evidence has already well document- Much clearer is Albertus Magnus’ 13th-century ed a long tradition in the utilisation of Nigella sati- work “De Vegetabilibus”, although he does not men- va since prehistory (see Table 1). Thus, in spite of the tion anything corresponding to Nigella: Under the scarcity of archaeological finds from the Medieval term “nigella” he deliberately gives a morphology of Arabic world, we judged most Arabic texts as rather Agrostemma githago, with its “long, green and hairy certain documents for Nigella sativa use. Perhaps the stalk […] red flower […] black seeds” (transl. from most influential mention of Nigella sativa is found in Meyer/Jessen 1867, 543). The plant’s properties and the beginnings of Islamic culture, where Sahih al- suggested medicinal uses also differ in some impor- Bukhari’s hadith mentions the plant as a panacea tant aspects from what we know from Hippocratic (Stillman 1983; Muhsin Khan 1996). This source traditions for Nigella (or, nota bene, Agrostemma!): may even represent the main driving force for the While the author does recommend it for all kinds of extraordinarily high amount of scientific and pop- skin disorders and toothache, he also indicates the ular publications from Islamic countries (see for ex- use of the seed flour for washing and bleaching laun- ample Anwar 2005; Anonymous 1993). dry. Due to the high saponin content of Agrostemma From Al-Andalus on the Iberian Peninsula, from seeds (Roth et al. 1994) this utilisation as a deter- the 9th century AD, comes the Book of Agriculture gent seems plausible. Similar accounts are also giv- by Ibn al-’Awwam. While mainly explaining meth- en in von Megenberg (1442), who also mentions a ods of how to best cultivate the “well-known garden “Zizania” (see above), but explicitly synonymising it nigella” (Clément-Mullet 1866), the author also as “Lolium” (see Table 2). mentions two species growing in the wild: One sim- Although illustrations, as indicated in the in- ilar to the first (N. sativa), but with “black to ash- troduction, usually are a great help in identifying en seeds” (N. arvensis?), and another one with seeds plants from manuscripts, in some cases the illus- that are “black, round, and rough to the touch”, a trators themselves are also a major source of trou- character combination also found, however not ex- ble. The “Compendium Salernitanum”, as an ex- clusively, in Agrostemma githago (see Fig. 3; Table 2). ample, mentions “nigella”, suggesting its use in the In western and central Europe, the often-cited treatment of ulcers and skin disorders (de Ren- 9th-century “Capitulare de Villis” gives no detailed zi 1852). But while this use conforms to the prop- information on the plant “gith”. It is regarded as be- erties conveyed for N. sativa, its illustration from ing desirable to have in a garden, so at least there is a 14th-century manuscript of the “Compendium” no strong argument against interpreting this as Ni- shows Nigella damascena (Brit. Mus., Eg. MS. 747; gella sativa. The contemporary “Leechbook of Bald” see Pächt 1950). The situation is even more con- from present-day Great Britain is likewise poor in fusing in the case of the slightly younger “Circa In- information on the kind of plant referred to. The au- stans” manuscript: It mentions a plant named “git”, thor’s suggestion of using “gitte” against spleen in- whose properties also basically conform to the tra- flammations is on the one hand autonomous from ditional Nigella uses, as does the description of its the archetypes in Roman/Greek Antiquity, and on seeds as triquetrous (Platearius undated; Dor- the other hand seems to have influenced some Re- veaux 1913). But while one undated 15th-century naissance medical literature: It is found again much French manuscript of the book illustrates this plant later in Parkinson (1629), but also in some of conti- (here named “nielle”) with an image of Agrostemma nental Europe’s herbals (e. g. in Bock 1630). githago (Fig. 4, a), in two others (also undated) the Hildegard von Bingen’s 12th-century “Physica” text is accompanied by illustrations of N. damasce- gives more information than earlier Medieval sourc- na (Fig. 4, b–c). And finally, a fourth manuscript of es, yet still remains unclear. In spite of several transla- the same book, illustrated by the famous artist Robi- tors’ views, the plant “Ratden” (also “Githerum rate” net Testard (Ms Fr. Fv VI #1 fol. 153v), shows a plant and “Zizania”; see Daremberg/von R euss 1855) labelled “Cyminum sylvestre” – a name used for Ni- may ambiguously be referring to Nigella as well as to gella in a later work by Valerius Cordus (Gesner Agrostemma: The author warns against its consump- 1561 b) – but the plate unequivocally shows Consoli- tion for food, but recommends it against ulcers, and da regalis (forking larkspur)! 153
a b c Fig. 4. “Git” or “nielle” in two 15th century French manuscripts of the “Liber de simplici medicina” (or “Circa Instans”) attributed to Matthaeus Platearius (undated). a One manuscript (btv1b6000422n, p. 96r.) is illustrated with Agrostemma githago, while in (b) the same text (btv1b60004232, p. 205v.) is accompanied by an image of Nigella damascena., just as it is in a third manuscript (c) (btv1b9061049h, p. 232v.). Image courtesy (all pictures): Bibliothèque Nationale Française, Paris. And, finally, Johannes von Cuba (1485) tells the al that has come down to us has its own fallacies reader about a plant with “flowers like corn-flower” and strengths, and many of the contradictions in- (= blue) and seeds which are “black, spicy, and with a side the texts, and sometimes even between the au- pleasant odour” (perfectly fitting for Nigella!) but at thor and the illustrator, render tracing a taxon’s his- the same time all editions of his book are illustrated torical perception rather difficult. And in addition to with Agrostemma githago (Fig. 5) with purple flow- the probable confusion/amalgamation of Nigella and ers, its seeds being odourless. Agrostemma in Greek and Roman Antiquity, Lolium As a conclusion for medieval literature, there- temulentum has now been added, as well. fore, no clear conclusion can be made: Every herb- Early Modern Period The Renaissance and Early Modern Period were for three Nigella species (see Table 3), finds that N. the heyday of herbals. This is, on the one hand, due damascena has odourless seeds (which in fact release to the invention of the printing press in the mid- a strong strawberry-like aroma when crushed). The 1450s and the new possibilities of producing a vir- author treats corn cockle in a separate section and tually unlimited number of identical copies of the finds no sensible purpose for it, apart from “some in- same book. On the other hand, the period is char- competent [people] using it instead of nigella, to the acterised by the breaking up of scholastic traditions, great harm of the sick” (transl. from Dodoens 1557). the re-orientation towards Antiquity, and critical Likewise, Valerius Cordus describes Nigella ar- discussion of inherited sources and traditions. The vensis (as “Nigella cornuta”) quite well, even point- philosophical advances in this period often show in ing out an earlier confusion with (maybe) Nigella the schoolmasterly undertone in many of the herb- hispanica (“nigella Citrina”) corrected by Matthiolus als, authors refuting obsolete ideas, and correcting (edition commented by Gesner 1561 b). And he ex- misidentifications. Due to the vast bulk of available plicitly warns against confusing Nigella (sativa) with literature, only a few examples shall be mentioned in a flower easily recognisable as Agrostemma githa- order to illustrate the new attitudes. go, pointing out its characters as “long hairy leaves Dodoens (1557), while giving clear identification […] purple flowers, not unlike a small rose” (transl. 154
Fig. 5. “Nigella” or “raten” in von Cuba (1485), Cap. 277: In Fig. 6. Nigella species in Besler (1613), Plate 174: “Melanthi- contrast to the name used, the illustration shows Agrostemma um Damascenum flore pleno” as well as “Melanthium Sativum githago (corn cockle). Image courtesy: Bayerische Staatsbiblio- flore simplici”, both with the characteristic flower involucre, thek, München. correspond to Nigella damascena, while “Melanthium Hispani- cum maius” is what is known today as Nigella hispanica. Image courtesy: Taschen Verlag, Köln. from Gesner 1561 a, Cap. 48). Konrad Gesner, in his Hieronymus Bock (1630), as another example, own book “Horti Germaniae”, identifies the dissect- does not differentiate either between the effects of ed involucre beneath the flowers of N. damascena the three Nigella species he mentions, although he as a diagnostic trait, but adds a bit of confusion in strictly separates Agrostemma githago from them, mentioning “Gith vulgaris”, identifying its seeds as attributing no useful effects to it. But also in his book fragrant, but as yellowish (the latter a typical feature we find the warning about excessive consumption of for N. hispanica). It may be that in fact he was refer- Nigella seeds, carrying forward the “toxicity myth” ring to N. arvensis, whose seeds are brighter in col- falsely attributed to Nigella since Dioscorides. our than those of N. sativa (Gesner 1561 c). Some further differentiation worth mentioning is Gerard (1597) is an interesting source inas- found in Parkinson (1629): As already anticipated much as the author explicitly only attributes med- in Gerard (1597), this author clearly differentiates ical properties to the “Nigella Romana” (N. sativa), three kinds of Nigella “not fit to be used” in medi- although also giving descriptions for other spe- cine from N. sativa which he deems a useful remedy. cies (see Table 3). Still quite in a Dioscoridic tradi- Among other purposes, he recommends the plant tion, he warns against the toxicity of excess Nigella against the “swollen spleen” as originally found in use while in his chapter on corn cockle he does not Bald’s Leechbook of the 9th century. mention anything of the like. He even claims that In the case of Nigella, at least the medical recom- confusion of Agrostemma githago with darnel might mendations during early Modern times remain more have been the reason for possible intoxications. In or less the same as in the Middle Ages, and most of Mattioli (1590), no differentiation at all is found the authors do not differentiate between the quali- in terms of medical properties: Three Nigella species ties of the different species as remedies. At any rate, and Agrostemma are subsumed including a pletho- most of these works serve as examples of the emanci- ra of indications – but also the Dioscuridic warning pation of botany from medicine and pharmacology: against toxic effects. The mention of medicinal effects of the plants gradu- 155
ally move into the background, while in many books and thus without any information on medicinal use, large portions of the texts are devoted to plant char- may serve as an example for the shift in the percep- acters. The illustrations from the “Hortvs Eystetten- tion of plants in the period (Fig. 6). sis” by Basilius Besler (1613), a horticultural work Conclusions and prospects It may be obvious that textual sources on plants logical evidence of a plant (Nigella sativa, that is) in are to be used with caution, and proper identifica- order to find out about its cultural reception, and re- tion of the plants treated is still the main problem. sulting in a rather sound verification of its use as a In the limited space of this article, we hope to have condiment and medical plant for approximately exemplified these difficulties for the genus Nigel- 5,000 years in the Levant. la as thoroughly as possible. Apart from this “sim- It has to be emphasized that assessing the actu- ple” diagnosis, we have also tried to demonstrate al benefits (or none) of the medical use of Nigella (N. that in written sources, properties attributed to cer- sativa) is far from the scope of this paper and the ex- tain plants frequently shifted around among the pertise of its authors. Still, we would like to empha- taxa described, and – as in the case of Nigella and size that the multitude of medical effects attributed Agrostemma – sometimes even seem to have existed to N. sativa by modern “black seed oil” vendors may separately from the physical plant itself, but rather not even be based on historical accounts of Nigella at as independent ideas (see also the chapter on man- all, but maybe refer to those of the toxic corn cockle drake in Heiss/Kohler-Schneider 2011). This may (Agrostemma githago). And also the “toxicity myth” be due to authors who were more focused on the tra- itself is still found in modern literature (e. g. Duke et dition (and repetition) of ancient diagnoses and of a al. 2008, 302; Tahraoui 2007), uncritically passed certain symbolism instead of a real striving for bet- down for more than two millennia. ter knowledge about the plants. Modern medical research has, however, started to Especially for the Eastern Mediterranean region test these old myths against reality, and has indeed during Antiquity, it has proved highly effective to revealed some promising medical effects of Nigella combine historical, linguistic, and actual archaeo species (e. g. Anwar 2005; Landa et al. 2009). 156
Summary Nigella (Ranunculaceae) is a small genus of an- has hardly ever been documented basing on the crit- nuals mainly of Irano-Turanian distribution, but ical use of sources, and employing a multidiscipli- some species also extending across most of Europe, nary approach. northern Africa, and western Asia. The genus com- The current paper now attempts to combine a prises several taxa of ethnopharmacological interest, wide range of written sources and the full archaeo as their seeds are used in a wide range of medici- logical record in order to find evidence of this ge- nal contexts – those of N. sativa have even developed nus’ possible use in history and prehistory, and to into a plant drug of prime commercial interest. serve as an example of the combination and inter- However, the historical dimension (the “traditional linking of research tools from historical botany and uses”) of Nigella species, nowadays used as a mar- archaeobotany. keting tool and often a justification of modern use, Zusammenfassung Nigella (Ranunculaceae) ist eine kleine Gattung gezogen wird, wurde kaum jemals anhand textkri- annueller Pflanzen mit vor allem Irano-Turanischer tischer Untersuchungen oder interdisziplinärer An- Verbreitung. Einige Arten erstrecken sich auch über sätze dokumentiert. große Teile Europas, Nordafrikas und Westasiens. Die vorliegende Arbeit versucht nun, ein mög- Die Gattung umfasst einige Taxa mit ethnophar- lichst breites Spektrum von Textquellen zusammen makologischer Nutzung, ihre Samen finden in ei- mit einer vollständigen Dokumentation archäolo- ner Vielzahl volksmedizinischer Indikationen An- gischer Funde heranzuziehen, um Hinweise auf die wendung – jene von N. sativa haben sich sogar zu tatsächliche Nutzung dieser Gattung in historischer einer pflanzlichen Droge von großem kommerziel- und vorgeschichtlicher Zeit zu finden und um als lem Interesse entwickelt. Die historische Dimension Beispiel für die Möglichkeiten interdisziplinärer Ar- jedoch (die „traditionelle Anwendung“) von Nigella- beit von historischer Botanik und Archäobotanik zu Arten, die heute oft als Vermarktungsargument und dienen. Rechtfertigung für medizinischen Gebrauch heran- Abbreviations in Tables 2–3 Languages: AR = Arabic; AS = Assyrian; DE = German dental disorders; DET = detergent and bleaching agent; DIA = (Deutsch); EN = English; FR = French; GR = Greek; HE = He- diabetes; DIU = diuretic; EGA = emmenagogue, galactagogue; brew; IT = Italian; JA = Judaeo-Arabic; LA = Latin (Latin or La- EVE = „evil eye“ protection; FRA = heals fractures; FUM = fu- tin-like names are usually not explicitly labelled); NL = Dutch migant; HEM = haemostatic; HEP = hepatical disorders, jaun- (Nederlands); SR = Serbian. dice; HYC = hypercholesterinemia; HYH = hypertension, heart Species: Ag = Agrostemma githago (corn cockle); An = Ane disorders; INS = insectifuge; MAP = male potency, prostate thum graveolens (dill); Cc = Cuminum cyminum (cumin); Cp = disorders; MEN = mental and nervous disorders, headaches; Claviceps purpurea (ergot); Cr = Consolida regalis (forking lark OPH = ophthalmic disorders; ORN = ornamental; OTO = oto- spur); Lt = Lolium temulentum (darnel); cf. = “confer”, identi- logic disorders; PAN = panacea (covers all mentioned uses ex- fication of a plant specimen incomplete; * = ornamental culti- cept veterinary); RES = respiratory disorders; SKI = skin disor- vars of Nigella. ders; SNA = drives away snakes; SPL = “swollen spleen”; TOS Uses: ABO = abortifacient; ADO = antidote; ARI = antirheu- = tonic, stimulans; TOX = toxicity warning; VET = veterina- matic, antiinflammatory; ATH = anthelminthic; BIT = heals ry use (diverse). bites; CAL = carminative, laxative; CON = condiment; DEN = 157
Table 1. Archaeological finds of Nigella species. a Identified as Nigella damascena in Ruas (1995), but revised as Nigella sativa by Heiss/Ruas (2007). b Initially identified as Nigella cf. 158 sativa (Vandorpe 2010), but according to the published images safe to be identified as Nigella sativa. c Initially identified as Nigella arvensis (Kroll 1998 b), but revised as Nigella sp. by H. Kroll in 2012 (pers. comm.). Species Date/Period Archaeological site Number and preservation of finds References N. arvensis N. damascena N. sativa Nigella sp. Early Modern Times 1975–1610 AD Netherlands: Heveskesklooster terp 1 waterlogged seed x – – – Cappers 1995 ca. 1900–1600 AD Libya: Wadi el-Agial unreported number of desiccated seeds – – x – Pelling 2003 16th cent. AD France: Rigny-Ussé, rectory 1 mineralised seed from – – xa – Ruas 1995 a cesspit 1530–1580 AD Germany: Göttingen, Johannisstraße 20 mineralised seeds from – – x – Hellwig 1997 a cesspit Middle Ages 14th/13th cent. AD Germany: Cottbus 7 waterlogged seeds from a well – – x – Lange 1993 14th/13th cent. AD Germany: Höxter mineralised seeds from – – x – Wolf 1991 a cesspit 12th–8th cent. AD Spain: Lleida, Costa de Magdalena (Islamic) 3 mineralised seeds – – – x Alonso i Martínez 2005 10th–5th cent. AD Austria: Leobendorf 2 waterlogged seeds x – – – S. Wiesinger/U. Thanheiser pers. comm. 8th–7th cent. AD Syria: Qasr al-Hayr al-Sharqi unreported number of mineralised seeds – – – x M. Kühn pers. comm. Antiquity (Common Era) and Late Antiquity ca. 600 AD Tunisia: Carthage (Late Antiquity) pollen grains in the harbour sediments – – ? – van Zeist et al. 2001 150–125 AD Netherlands: Maastricht 1 waterlogged seed x – – – Bakels/Dijkman 2000 3rd cent. AD Germany: Groß-Gerau 1 seed x – – – Kreuz/Stika 2009 3rd cent. AD Germany: Friesheim unreported number of waterlogged seeds x – – – Knörzer 1971 3rd–1st cent. AD UK: Carlisle, Annetwell Street 1 waterlogged seed – – x – Huntley 1989 3rd–1st cent. AD Egypt: Dakleh oasis unreported number of seeds – – x – J. Walter/U. Thanheiser pers. comm. 2nd cent. AD Germany: Butzbach 1 waterlogged seed x – – – Knörzer/Baatz 1973 2nd cent. AD France: Biesheim/Kunheim 2 mineralised seeds in a pit – – xb – Vandorpe 2010 2nd /1st cent. AD Egypt: Mons Claudianus (Roman) 62 desiccated + 2 charred seeds – – x – van der Veen 2007 2nd /1st cent. AD Egypt: Mons Porphyrites (Roman) 31 desiccated + 12 charred seeds – – x – van der Veen 2007 180 AD–460 BC Germany: Hessisches Ried pollen grains in a peat bog ? ? – – Singer 2006 1st cent. AD France: Biesheim/Kunheim 2 waterlogged seeds x – – – Vandorpe 2010
Table 1, continued. Archaeological finds of Nigella species. a Identified as Nigella damascena in Ruas (1995), but revised as Nigella sativa by Heiss/Ruas (2007). b Initially identified as Nigella cf. sativa (Vandorpe 2010), but according to the published images safe to be identified as Nigella sativa. c Initially identified as Nigella arvensis (Kroll 1998 b), but revised as Nigella sp. by H. Kroll in 2012 (pers. comm.). Species Date/Period Archaeological site Number and preservation of finds References N. arvensis N. damascena N. sativa Nigella sp. Antiquity and Prehistory (1st millennium BC) 330–578 BC Austria: Dürrnberg mine (La Tène) pollen grains in human faeces – – – x Aspöck et al. 2007 110–250 BC Austria: Roseldorf (La Tène) 1 charred seed in a sanctuary x – – – Caneppele et al. 2010 7th–11th cent. BC Egypt: Gebel Roma (Intermediate Per. 3) 1 desiccated grain from a midden layer – – x – Cappers et al. 2007 7th–16th cent. BC Egypt: Gebel Qarn el-Gir (New Kingdom/Intermediate 4 desiccated seeds – – x – Cappers et al. 2007 Per. 3) 7th–16th cent. BC Egypt: Wadi el-Huôl (New Kingdom/Intermediate 1 desiccated seed – – x – Cappers et al. 2007 Per. 3) 7th–16th cent. BC Egypt: Wadi el-Huôl (New Kingdom/Intermediate 1 desiccated seed – – x – Cappers et al. 2007 Per. 3) 810 BC Bosnia-Herzegovina: Klisura Kadića Brdo 1 waterlogged seed – – – ? Kučan 1995 ca. 800 BC Jordan: Tell Deir ‘Allā 9 desiccated seeds – – x – Neef 1990 920–1410 BC Austria: Brixlegg, Mauken (Late Bronze Age) 1 partly charred seed – x – – Heiss/Oeggl 2005 Antiquity and Prehistory (2 nd millennium BC and earlier) 11th–16th cent. BC? Egypt: Thebes (New Kingdom? Passalacqua Collect., unreported number of desiccated seeds – – x – Germer 1988; 1989 Berlin, Inv.-No. 6997) 11th–16th cent. BC Egypt: Gebel Roma (New Kingdom) 2 desiccated grains from a midden layer – – x – Cappers et al. 2007 9250–10450 BC Syria: Tell Qaramel 1 charred grain – – – x Willcox et al. 2008 ca. 1323 BC Egypt: Tutankhamun tomb (New Kingdom, unreported number of desiccated seeds in a pot – – x – Germer 1985; Hepper 1990; 18th dynasty) de Vartavan 1990 late 14th cent. BC Turkey: Ulu Burun shipwreck unreported number of waterlogged seeds – – x – Haldane 1993 15th–19th cent. BC Serbia: Feudvar 1 charred seed – – – xc Kroll 1998 b ca. 1650 BC Turkey: Boyalı Höyük (Old Hittite) ca. 180,000 charred seeds in a flask – – x – Salih et al. 2009 1771–1991 BC Egypt: Kahun (Middle Kingdom, 12th dynasty) unreported number of seeds – – x – Bienkowski/Southworth 1986; Germer 1989 2216–2707 BC Egypt: Saqqara (Old Kingdom, unknown number of charred seeds – – x – Rzeuska 2006 4th–6th dynasty) 2300–2900 BC Spain: El Prado (Jumilla, Murcia) unreported number of seeds – – – x Precioso Arévalo 2004 159
Table 2. Written sources on Nigella species from Antiquity to the Middle Ages. Abbreviations see p. 157. 160 Species Date/Period Work (original title and author where available) References Name used Uses (original name where available) N. arvensis N. damascena N. hispanica N. sativa others Middle Ages 15th cent. AD liber de simplici medicina / circa instans (by Matthaeus Platearius?): Platearius undated; Dor- Git; nielle (FR) – x – ? Ag, ARI, CAL, DIU p. 96r; p. 232v; p. 205v; veaux 1913 Cr 15th–12th cent. AD Chilandar Medical Codex Jarić et al. 2011 mačkov brk, grahor (SR) – x – – – ATH, DIU, SKI 1485 AD Gart der gesuntheit (Johannes von Cuba): Cap. 277 von Cuba 1485 Nigella; melanthion, git (GR); – – – – Ag ADO, ATH, CAL, raden (DE) DEN, DIU, FUM, INS, OPH, OTO, RES, SKI, SNA, TOX 1442 AD buch von den naturlichen dingen (by Konrad von Megenberg) von Megenberg 1442; No. 55: Nigella; rôteu korn- – – – – Ag CAL, DEN, Pfeiffer 1861 pluom (DE) DET, TOS No. 89: Zizania; ratenkraut (DE) – – – – Lt TOX ca. 1440 AD Kräuterbuch (by Johannes Hartlieb) Rumpler 1980 nigella; ratten (DE) – – – – Ag CAL, DEN, DET, SKI 14th cent. AD Nihayal al-Arab fi funun al-adab (by Al-Nuwayri) Lev 2002 ? (AR) – – – x – ADO, ARI, BIT, DEN, SKI 14th cent. AD Fasl fī ma’rifat al-matānim wa-al asiqā (royal crop register of Varisco 1991 ? (al-ḥabba al-sawdā‘) (AR) – – – x – no uses mentioned Rasulid Yemen, by Al-Afdal) ca. 1350 AD compendium Salernitanum (by Bartholomaeus Copho?) Pächt 1950; Pavord 2005; Nigella – x – x – SKI de Renzi 1852 1309–1304 AD De ruralia commoda (by Petrus de Crescentiis) de Crescentiis nigella – – – – Ag? ATH, CAL, DIU, 1477–1483 OTO, SKI 13th cent. AD de vegetabilibus (by Albrecht von Bollstadt / Albertus Magnus): Meyer/Jessen 1867 nigella – – – – Ag? ARI, DEN, DET, book 6, tract. 2, ch. 13 Lt? FUM, MEN, OPH, SKI, SNA 13th–10th cent. AD medical prescriptions from the Cairo Genizah Taylor-Schechter Lev 2007; Lev/Amar 2006; ? (JA) – – – x – ADO, ARI, OPH, Collection 2008 SKI 12th cent. AD liber simplicis medicinae / “physica” (by Hildegard von Bingen) Daremberg/von Reuss book 1, cap. 12: Zizania; Rat- – – – – Ag? INS, SKI, TOX 1855 den (DE) Lt? book 5, cap. 24: gith – – – x – CON 12th cent. AD al-Aqrābādīn (The Dispensatory by Ibn at-Tilmīd) Kahl 2007 ? – – – x – ARI, CAL, MEN, SKI 12th cent. AD Kitāb al-Filāha (The Book of Agriculture, by Ibn al-‘Awwam): Clément-Mullet 1866 ? x – – x Ag CON ch. 26, art. 3
Table 2, continued. Written sources on Nigella species from Antiquity to the Middle Ages. Abbreviations see p. 157. 12th cent. AD Kitābu’l Kulliyāt fī al-Tibb (General Rules of Medicine, by Zimara 1542 ? – – – x – FUM, OTO Ibn Rushd / Averroës): p. 5 11th cent. AD Diwan al-Filāha (Treatise on Agriculture, by Ibn Bassal) Harvey 1975 ? – – – x – no uses mentioned 11th/10th cent. AD Mishnaic glossary Hebrew-Greek: No. 73 Starr 1934/35 ΜΕΛΑΝΘΗ (melanthe) (GR); – – – x – no uses mentioned ( זהבלךHE) 10th/9th cent. AD Kitāb al-Qanûn fī al-Tibb (The Canon of Medicine, by Ibn Sina / al-Shaykh al-Ra‘is 1593 ( شونيزHabbet as-suda) (AR) – – – x – ? Avicenna): book 2, ch. 21 9th cent. AD Bald’s Leechbook: f. 93r Deegan 1988 gitte(r) (GR/EN); sutherne – – – ? – ARI, CON, SPL wyrt (EN) 9th cent. AD capitulare de villis vel curtis imperii (by Alcuin of York? von Fischer-Benzon 1894 git (GR) – – – ? – no uses mentioned Ansegisus of Saint-Wandrille? Louis the Pious?) 9th cent. AD Kitāb al‘Sumūm (Book on Poisons, by Ibn Wahshīya) Levey 1966 ( شونيزḤabba sawdā‘) (AR) – – – x – ADO 9th cent. AD Sahih al-Bukhari (Al-Bukhari‘s hadith collections): vol. 7, book 71, Muhsin Khan 1996 ? (AR) – – – x – PAN no. 591 f. Antiquity 2nd /1st cent. AD “Book of Matthew” 23:23 Dt. Bibelges. 1992; κύμινον (kyminon) (GR) – – – – Cc no uses mentioned Zohary 1995 1st cent. AD de medicina (On Medicine, by Aulus Cornelius Celsus): lib. 4 Lee 1863 gith – – – ? – DIU, EGA, SKI 1st cent. AD de re rustica (On Agriculture, by Lucius Iunius Moderatus Colu- Forster/Heffer 1954 git – – – ? – VET, TOS mella), lib. 6, 23 1st cent. AD de materia medica (On Medicaments, by Pedanios Dioscorides) Berendes 1902 μελάνθιον (melanthion) (GR) – – – ? – ATH, BIT, CON, DEN, DIU, EGA, FUM, MEN, OPH, RES, SKI, SNA, TOX 1st cent. AD naturalis historia (Natural History, by Plinius Maior / Pliny the Bostock/Riley 1855 git; μελάνθιον (melanthion), – – – ? – ADO, BIT, CON, Elder): lib. 20, 71 ? (melanospermon) (GR) DEN, EGA, FUM, INS, MEN, OPH, RES, SKI, SNA 4th cent. BC “corpus Hippocraticum” (by Hippokrates of Kos?) Fuchs 1900; μελάνθιον (melanthion) (GR) – – – ? Ag? ABO, CAL, EGA, von Grot 1887 Cp? MEN 7th cent. BC various medical prescriptions on Assyrian cuneiform tablets Thompson 1923; 1925; 1930; ? (zibûm) (AS) – – – ? – ARI, DEN, MEN, (or older) 1937; Richardson 2007 OPH, SKI 7th/8th cent. BC “Book of Isaiah” 28:27 Moldenke 1955; Trever ( ק ֶצ ַחquéṣaḥ) (HE) – – – ? An? no uses mentioned 1959 8th cent. BC description of the contents of the garden of Mesopotamian Leach 1982 ? (AS) – – – ? Cc? no uses mentioned king Merodach Balada 20th cent. BC Hurrian text from Nuzi (Nuzu): tab. HSS 13,353 Farber 1977 ? (AS) – – – ? – no uses mentioned 161
You can also read