Incense in medicine: an early medieval perspective

The production, use, and meaning of incense represent relatively untapped areas of study within early medieval history. In this article, I present evidence for the medical use of incense in the Carolingian world. Using a sample of eighth‐ and ninth‐century manuscripts, I analyse incense recipes, investigating their contexts and ingredients, as well as the use of incense itself as an ingredient in medical remedies. This evidence not only suggests that incense was understood as a multipurpose substance, but also offers a new window into exploring early medieval medical knowledge and practice and relates to the sacralization of medicine in this period.

is entitled Thimiama (see Fig. 1). The recipe lists a handful of ingredients and quantities, but provides neither further instructions for its preparation nor information on its use: Thimiama: cozumber -3, aloeswood, ambergris -3 denarii, confita, camphor -1 denarius, musk -1 denarius. 2 The word thymiama is defined by Isidore of Seville as 'incense', a Greek-derived synonym for incensum, thereby suggesting that this entry is a recipe for incense. 3 Since incense is intended to release fragrant smoke when burned, the recipe's aromatic ingredients fit with this identification. As incense played an important role in early medieval Christian rituals, it might be expected that this recipe would be found among liturgical texts describing the proper application of the substance in religious contexts, or near writings explicating the meaning and significance attached to its use. Instead, the textual environment in which this recipe is found is entirely medical: the manuscript does not contain material relating to the use and significance of incense in the church. As noted above, the My transcription of the Thimiama recipe (csg. 761, p. 66) is as follows: 'cozzunbar~iii aloa arbor denarius iii confitum cafora denarius i musico denarius i'. I suggest that arbor should be interpreted as ambergris, paralleling other remedies' use of ambar, as will be analysed below. Alternatively, this could be emphasizing the arboreal nature of aloeswood. also mixed with antidotes and it is given to asthmatics in drinks.' 8 While he presents these religious and medical uses of incense as distinct, there is no sense that their multiple functions are mutually exclusive or incompatible. The Alphabet of Galen, an anonymous, late antique Latin text that stems from the herbal tradition of Dioscorides, likewise connects cyphi to both medicine and religion, explaining that 'the ancients used to burn [cyphi] for their gods. Doctors place it in other compound medicines.' 9 This comment suggests that the ancient Egyptian religious uses of this type of incense had not continued and that contemporary applications of cyphi were entirely medical. The recipes for cyphi presented by both Dioscorides and The Alphabet of Galen, though also based on aromatic substances, do not share specific ingredients with the recipe from csg. 761. Dioscorides' recipe for cyphi contains galingale, juniper berries, raisins, pine resin, sweet flag, camelthorn, camel hay, myrrh, wine, and honey, while The Alphabet of Galen offers a slightly simpler recipe with galingale, juniper berries, grapes, sweet flag, camelthorn, myrrh, wine, and honey. 10 As noted above, the seventh-century Etymologies of Isidore of Seville, one of the most influential texts circulating in the medieval west, mentions both incensum and thymiama, and Isidore continues the connection between incense and medicine: the topic is discussed in Book IV, De medicina. Incense does not occur in Books VI to VIII, which concern various aspects of the Christian religion, or Book XVII, De agricultura, which includes a sub-chapter on aromatic plants (and even describes some of the ingredients listed above). 11 Is Isidore's categorization of incense as a medical topic simply a vestigial holdover from his classical and late antique sources? While the influence of his classical sources should not be discounted, I suggest that the linking of incense and medicine was not an anachronistic holdover. On the contrary, Isidore's discussion of incense within De medicina fits with the medical context of the incense recipe in csg. 761 and simultaneously adds greater significance to the comments made by Everett, Thurlkill, and Donkin. These connections suggest that the relationship between incense and medicine deserves further investigation.
The following article aims to be an entry point into this topic. Based on a sample of eighteen eighth-and ninth-century manuscripts containing medical recipes (see Appendix 1 for more information on the individual manuscripts), I have identified over a dozen instances of incense, including incense recipes, references to incense recipes in indices of recipe collections (though some of the recipes have not survived), and the appearance of incense as an ingredient within medical recipes. In this article, I shall concentrate on three aspects of incense's medical connections: the contexts in which incense recipes occur, the medical uses of the ingredients listed in these recipes, and the inclusion of incense itself as an ingredient. This body of evidence has major implications for our understanding of incense, its use(s), and its meaning(s). The final sections of the article will build on this topic, exploring how the medical use of incense fits into a wider framework of healing and may serve as a new avenue for studying the relationship between religion and medicine in the Carolingian world. Before exploring the three aspects of medical incense use (incense recipes, their ingredients, and incense as an ingredient), I shall briefly review the current work on incense in the early medieval period to situate the present study.

Incense and the early Middle Ages
Scholarship on incense, in relation to the development of sensory studies, and specifically the increasing interest in the sense of smell, has expanded in recent decades. 12 Much of this work has concentrated on the use of incense in religious contexts, exploring the adoption of incense in world religions, the evolution of practices and rituals involving incense, and the meanings associated with its use. This is particularly true of research that touches on early medieval incense. Two relatively recent, significant works on this topic are Susan Ashbrook Harvey's 2006 Scenting Salvation: Ancient Christianity and the Olfactory Imagination and Thurlkill's Sacred Scents in Early Christianity and Islam (2016), both of which focus on the sense of smell in a religious context. While medicine does feature in these books, especially regarding the meanings attached to good and bad smells (health and purity versus death, decay, and disease), the potential application of incense in early medieval medical treatments receives less attention. Harvey does, however, mention classical and late antique medical uses of incense, and Thurlkill, as noted above, comments on the parallels between the ingredients used in the production of incense and medicines.
Given the religious framework around which research on smell, and by extension incense, has clustered, scholars have tended to overlook potential connections to medicine; sacred, liturgical, and hagiographical texts have instead been studied as sources on the topic. Exodus XXX.34, for example, gives a simple recipe for incense: 'And the Lord said to Moses: Take unto thee spices, stacte, and onycha, galbanum of sweet savour, and the clearest frankincense, all shall be of equal weight.' 13 Eighth-and ninth-century theologians describe the use of incense in early medieval Christian ritual. Incense was essential to the proper performance of the liturgy, and liturgical correctness was a fundamental element of Carolingian efforts to reform and standardize religious practice. Amalarius of Metz, a somewhat controversial Frankish bishop and courtier active in the first half of the ninth century, comments on incense in his work on the liturgy, the Liber officialis. When detailing the entrance of the bishop at Mass, he writes, 'the incense in its censer leads the way; this signifies Christ's body suffused with a pleasing odour'. 14 On the Te igitur, Amalarius explains that 'just as there were two altars in the tabernacle of Moses and in the temple of Solomonone of incense, the other of the burned offering so there are two sacrifices of the holy church'. 15 He thus connects incense use to Old Testament precedent as well as to the performance of sacrifices, relating contemporary church practices to earlier traditions. 16 While evidence from liturgical sources may shed light on the role of incense in the Carolingian church, many questions remain unanswered. The production and preparation of incense is rarely noted in religious writings, though there are exceptions, such as the example from Exodus. Other types of textual evidence, including letters, wills, and charters, can offer insights into the purchase, donation, or import of exotic spices and aromatics, the types of substances used to make incense (as seen in the varied examples above). Yet texts of this nature tend to provide only indirect evidence for potential incense production: the substances these sources document could have been used to produce incense, but also paint, ink, medicine, and perfume, and an explicit link to incense is not generally made. The will of Ursus, a bishop of Benevento in the first half of the ninth century, for example, lists sacks of various exotic products, such as pepper and a certain alivano. 17 Alivano is probably an alternative spelling of olibanum, meaning frankincense. 18 Records of the Abbey of Fontenelle document the annual purchase of pigmenta: the abbot allocated a pound of silver per year for this purpose. 19 The term pigmenta can refer to a range of products including paints, pigments, and their composite parts as well as spices and medicaments. 20 In contrast to the will of Ursus, the abbey records state the intended use of these ingredients, noting that the pigmenta were to be employed in medical remedies for the treatment of sick monks.
With this background in mind, it is possible to turn to the manuscript evidence: in what contexts are incense recipes found?

Incense recipes in Carolingian manuscripts: the medical contexts reviewed
While a single instance of an incense recipe located within a medical text could be an anomaly, the placement of this incense recipe appears to be part of a wider pattern: I have so far identified a further six incense recipes within medical contexts. According to Augusto Beccaria's catalogue of pre-Salernitan medical manuscripts, roughly eighty codices containing medical texts have survived from the eighth and ninth centuries. 21 There is a large variety in the types of medical writings preserved . Unfortunately, two further recipes are noted in an index in a manuscript in the Bibliothèque nationale de France, Paris BnF lat. 6882A, but the recipes themselves have not survived. While more information on the manuscripts can be found in Appendix 1, I shall briefly review the contexts in which the recipes are found before considering their composition. I have grouped them under two categories according to the textual environments in which they are located: a) incense recipes within collections of medical recipes, and b) incense recipes in other medical contexts.

Incense recipes within collections of medical recipes
Three of the seven incense recipes fall within the middle of recipe collections (with one recipe in csg. 44 and two in csg. 759), while one recipe comes at the end of a recipe collection (the opening example from csg. texts can be found on pages 186-368. 24 The medical half of the manuscript has been dated to the second half of the ninth century. 25 A collection of recipes covers pages 228-60, and a recipe for incense entitled Confectio timiame is situated in the middle of this collection on page 247. 26 The recipe sits between two antidotes, one labelled Diaterseron antidotum, the other, Antidotum diaprassiu (see Fig. 2).
The two incense recipes in csg. 759, a ninth-century manuscript containing a variety of medical texts, are similarly located within a recipe collection. 27 The recipes, entitled Conpositio thymamatis and Tymiama simplex, represent entry numbers 338 and 340 in the collection, respectively (see Figs 3 and 4). 28 The Conpositio thymamatis is preceded by the Confectio oxymellis, instructions for the preparation of oxymel (a mixture involving vinegar and honey), and followed by a remedy to treat the bite of a rabid dog, Ad canis rapidi morsum. The second incense recipe, Tymiama simplex, comes next, and is followed by another remedy, the Conpositio siromire.
While csg. 761 has been discussed above, its palaeography deserves a comment. Although the location of the incense recipe as the final entry of the collection on pages 51-66 might imply that it was a later addition, a single hand appears to be responsible not only for the entire recipe collection, but for all 290 pages of the manuscript. This consistency implies that the incense recipe, rather than being a later addition, was intentionally placed in this location, and thus represents the last entry in this recipe collection.   Thymiama paltgrimi, is located just below the diagram of the sphere. Book IV, beginning on the following page, appears to be written in the same hand, thereby suggesting that these insertions between Books III and IV were not added at a later date, but deliberately included in this space as information intended to complement the text of the Medicina Plinii. The medical nature of the other inserted texts (the dietetic calendar, remedy group, and Spera Apulei Platonici), albeit of a different genre or style from the Medicina Plinii, reinforces the idea that incense recipes, too, were conceived of within this broadly medical framework. Csg. 878, a manuscript thought to be Walahfrid Strabo's vademecum, contains the final two examples of incense recipes from this manuscript sample. 31 The codex includes a wide variety of writings, ranging from works on computus to Priscian's Institutiones grammaticae, and medical texts are interspersed within the collection. Following several pages of miscellaneous recipes, two lists of ingredients appear on page 334. The first has no title while the second is labelled Item aliter. 32 These lists, covering roughly two-thirds of the page, are the only written material on page 334 (see Fig. 6). The page immediately preceding the recipes, page 333, contains a handful of simple remedies, vernacular glosses, and a later addition. The remedies have several empty lines between them and glosses from the eleventh century provide the names of many of their ingredients in Old High German. 33 Although a number of different hands can be seen within the manuscript, Bernhard Bischoff has suggested that the hand responsible for the widely spaced remedies of page 333 and for the incense recipes on page 334 is Walahfrid's own; he dates the writing on these pages to the same period and the uniformity within this section of text may reflect the planned placement of the incense recipes within a medical context. 34 Without clear titles, it is the ingredients themselves, a topic addressed in more detail below (see Table 1), that indicate that these two lists are, in fact, related to incense. Although, as noted above, many of these ingredients also appear in medical remedies, I have only seen these particular combinations of ingredients within recipes with a thymiama-  based title. This striking similarity among very specific ingredients strongly suggests, therefore, that the two lists on page 334 are also recipes for incense.

The significance of medical contexts
Before considering the ingredients of the incense recipes and their medical uses, the significance of the medical environments in which these recipes are located must be stressed. While the medical contexts of all seven recipes is noteworthy, the appearance of four incense recipes within collections of medical recipes is particularly striking. The recipe collections analysed in this study have been attributed to neither classical nor late antique sources; I specifically targeted recipe collections and individual recipes labelled 'miscellaneous' by Beccaria since these have been less studied (in fact, in almost all cases, they have never been published). 35 The recipes themselves exhibit considerable variation in their presentation and detail. A single collection of recipes may include instructions for preparing prescriptions intended to treat a specific condition, antidotes tackling a host of different ailments, ointments, plasters, and so on. Most significantly for this discussion, recipe collections often contain instructions for the preparation of complex ingredients, such as mixed oils or oxymel (as noted in csg. 759 above). These ingredients could have been prepared separately before being used in medical recipes (or may have been used as treatments themselves and as complex ingredients). The location of incense recipes within collections of medical material, and even alongside complex ingredients such as oxymel, strongly suggests that incense was similarly intended for such a use. The possibility that incense was a multipurpose substance and could have been used as an ingredient in medical recipes must therefore be kept in mind when considering the medical applications of its own constituent elements, the topic of the following section.

The ingredients of incense
In order to address the medical uses of the ingredients of the incense recipes highlighted above, it is first necessary to review the composition of the recipes themselves. An assessment of their ingredients points to many shared features and a heavy reliance on similar, if not identical, substances, suggesting the existence of a number of related recipe traditions.
A comparison of the ingredients in the seven incense recipes Table 1 presents the ingredients of each of the seven recipes. These ingredients have been rearranged from their original order to underline both similarities and differences between the recipes (see Appendix 2 for a comparison of their original order). Since incense depends on ingredients that produce a fragrant smell when burned, all of these recipes are unified by their almost exclusive reliance on aromatic, plant-based substances, and especially gums and resins, such as myrrh, frankincense, ladanum, storax, and camphor. Spices, such as saffron, cinnamon, and clove, also appear relatively frequently. Musk and ambergris, both animal products, are notable exceptions to the otherwise herbal ingredients, but as strong-smelling substances, they are not unlike the other components. The recipes in csg. 759 might also incorporate an animal product in the form of a mollusc (listed as ungiculas marinas in csg. 759a and onetus idest ungellas in csg. 759b), although additional research is needed to interpret these terms more precisely in this context. 36 I suggest that ungellas may be linked to the term onycha, seen above in the quotation from Exodus. Onycha has traditionally been interpreted as a mollusc, although it has been suggested that it may represent ladanum, a gum resin. 37 Finally, csg. 759a also includes two ingredients quite unlike the gums, resins, spices, and other dry substances involved in these recipes: wine and honey. The use of these ingredients would have altered the consistency of the end product, creating a liquid and suggesting that the standard way in which incense was used, by incineration, may not have been applicable in this case. Although none of the seven recipes is identical, Table 1 reveals not only a large degree of overlap, but also suggests the existence of two families of recipes, Groups A and B. Both groups can be further subdivided into complex and simple recipe variants. Group A consists of the two recipes found in csg. 759. While these are most unlike the other recipes, they do appear to be related to each other: as its name suggests, Tymiama simplex is a simpler recipe, using only five ingredients in contrast to the eleven ingredients listed in the Conpositio thymamatis. Of these, the recipes share three key ingredients, myrrh, storax, and ungiculas marinas or onetus/ ungellas. It must be noted that they also share certain ingredients with some recipes of Group B (especially the more complex recipes), such as myrrh, storax, saffron, and mastic, but their mutual context within csg. 759 and inclusion of ungiculas/ungellas, an ingredient entirely unattested in other incense recipes, argues for a stronger link between these two recipes.
The five recipes of Group B represent a tightly knit cluster that can, like Group A, be broken into complex and simple variations. All five recipes share five core ingredients: cozumber, confita, aloeswood, camphor, and musk. These ingredients are the primary components of the three simple recipes, csg. 752, 761, and 878a. Csg. 761 and 878a share one (ambergris) and two (ambergris and storax) additional ingredients with the longer recipes, respectively. The complex recipes of csg. 44 and csg. 878b share twelve of the same ingredients (the basic five listed above, plus ambergris, storax, frankincense, myrrh, mastic, spikenard, and saffron), though csg. 878b contains an additional three ingredients (cloves, cinnamon, and galingale).
Across all seven recipes, it is important to note that many ingredients represent exotic products that would have travelled great distances to reach western Europe. 38 Many of the plants that produce aromatic gums and resins, such as the trees bled for myrrh (genus Commiphora) and frankincense (genus Boswellia), can be found in the Arabian Peninsula and along the east coast of Africa. Even further afield, the laurels that produce camphor grow in south-east Asia, and particularly on the islands of Sumatra and Borneo, while the musk deer, which secretes musk, is native to mountainous regions of southern Asia, such as the Himalayas. 39 The recipes in Group B are especially noteworthy in regard to their exotic nature. The precise identifications of cozumber and confita remain unclear. In Origins of the European Economy, Michael McCormick classes cozumber as an 'exotic substance' without elaborating further. 40 Both appear in few other sources from this period; to date, I have only found references to them in a letter from Boniface in the mid-eighth century and several other recipes from manuscripts involved in this study. on their contexts in these recipes as well as later descriptions of the substances, it appears that cozumber and confita are also gums or resins imported from the Middle East or East Africa. 42 Whatever their exact origin, cozumber, confita, ambergris, and camphor all seem to be substances newly introduced to the Latin west; these ingredients do not appear in earlier Biblical, Egyptian, Greek, or Roman incense recipes or medicines. There is, however, some debate as to when exactly a number of these ingredients first arrived in western Europe: while certain ingredients, such as cozumber and confita, are unattested before the eighth century, other products may have been present in the late antique Mediterranean. Musk, for example, is first noted in Jerome's (d. 419) Adversus Iovinianum, but does not reappear for at least a century and only then is mentioned by Greek sources, such as Paul of Aegina (fl. seventh century). 43 The next Latin references to musk come from these very incense recipes. The presence of these ingredients in Carolingian manuscripts appears, therefore, to testify to the arrival (or, in some cases, reintroduction) of new products from the east, highlighting not only the existence of long-distance trade, but its continued development in the early medieval period. 44 Moreover, the fact that all of these new products are found together in Group B adds further weight to the relatedness of the recipes within this cluster, suggesting the spread of a shared tradition alongside the arrival of these exotics. With this more detailed understanding of the recipes and their ingredients, it is now possible to return to Thurlkill's observation that medicines use many of the same ingredients as incense.

The use of incense ingredients in medical recipes
Carmélia Opsomer's Index de la pharmacopée du Ier au Xe siècle, an index of the ingredients found in the recipes of over fifty medical texts written between the first and tenth centuries, presents a useful overview of the frequencies with which particular products are named. 45 Of the recipe 42 The terms confita and cozimbrum are both included in the thirteenth-century medico-botanical glossary Alphita. According to this text, the two substances are related to each other (with cozimbrum described as fex confite) and represent derivatives of storax. A. García González (ed.), Alphita (Florence, 2007); confita is discussed on pp. collections involved in the present study, only some remedies from csg. 44 are included in Opsomer's detailed analysis; the recipe collections from the other manuscripts assessed here, lacking published transcriptions, were not part of the index. I shall highlight a handful of examples, concentrating on the exotic gums and resins, to give a sense of how often these ingredients occur in the texts. Storax, seen in five of the incense recipes, is listed in Index de la pharmacopée 336 times, while mastic, which occurs in three incense recipes, is recorded 484 times; both ingredients are widely distributed within the sample of medical texts and are found in classical, late antique, and early medieval material. 46 Frankincense is a more complicated case because it went by several names, including thus and (o)libanum; while it only occurs in two of the seven incense recipes, a summation of the entries from Index de la pharmacopée for the multiple terms results in a total of 763 instances. 47 Myrrh, found in four incense recipes, is listed as an ingredient 1017 times, and, similarly, the term resina, meaning an unspecified resin, is found 1016 times. 48 In contrast, Index de la pharmacopée records only six occurrences of cozumbrium, cozumber, one of which is from the incense recipe in csg. 44. 49 Ambergris is only noted once, and, again, the reference comes from the recipe in csg. 44. 50 This, however, should be no surprise since the writings sampled for Index de la pharmacopée include many texts that pre-date the posited arrival of these exotic substances in the early medieval west. Furthermore, the majority of the manuscripts in which the incense recipes are found were not included in Index de la pharmacopée. Only those substances that were known since antiquity, therefore, such as myrrh and frankincense, reflect the full extent to which these types of products were included in medical recipes.
The sample of eighth-and ninth-century manuscripts consulted in this study conforms to the trends seen above: many of the ingredients listed in the incense recipes also occur frequently in medical remedies. Given constraints of space, only a small selection of remedies will be presented to illustrate some of the many ways in which these ingredients are recorded. The recipe collection of csg. 759 that produced two of the incense recipes analysed above also contains an entry to expel demons, Ad demonio expellendo. This 46 Opsomer, Index de la pharmacopée du Ier au Xe siècle, pp. 432-6, 746-9. 47 Opsomer, Index de la pharmacopée du Ier au Xe siècle, pp. 397-9, 530, 769-73. 48 Opsomer, Index de la pharmacopée du Ier au Xe siècle: for myrrh, see pp. 471-80; for resina, see pp. 641-9. 49 Opsomer, Index de la pharmacopée du Ier au Xe siècle, p. 222. 50 Opsomer, Index de la pharmacopée du Ier au Xe siècle, p. 48. treatment on page 68 names myrrh and frankincense while also listing some ingredients not seen in incense recipes, such as sulphur and castoreum, the secretion from a beaver's castor sac. 51 The instructions, which direct that the ingredients be turned into a powder and then burned, indicate that this was intended to be a fumigation. This method of treatment, burning a mixture of dry ingredients, parallels the way that incense is most commonly consumed in a religious setting. While a connection to the Christian uses of incense and the symbolism attached to its use is never made explicit in this remedy, the question of religious elements in medical practice will be considered in more detail below.
Csg. 217 presents a remedy for ear pain, Ad aurem dolorem, on page 267. 52 The list of ingredients begins with myrrh and also contains aloe, poppies, water, and honey. Once combined, this mixture was applied directly to the ear, demonstrating that the ingredients found in incense recipes could be used in multiple ways and not exclusively as burned substances. Similarly, a treatment for liver and spleen problems, De epar et de spleno, found on page 365 of csg. 751, makes no mention of fumigation, despite including a number of the ingredients frequently used in incense recipes, such as myrrh, frankincense, and mastic. 53 Overall, the textual evidence records the extensive use of aromatic plant-based substances in medical contexts. At times these ingredients were employed in a manner paralleling the religious use of incense, i.e. fumigation, while in many cases the treatments were administered in other ways. Given the medical uses of many incense ingredients, and their frequent use together (such as the combination of myrrh and frankincense listed in several of the selected recipes), it seems logical that incense preparations may have been similarly employed in healing practices on medical grounds alone; I shall return to the question of Christian influence below. Such a use supports the idea suggested above to explain the appearance of incense recipes in medical contexts: incense may represent a complex ingredient that could have been used in medical treatments. This raises the question, is incense listed as an ingredient in remedies?

Incense as an ingredient
A ninth-century manuscript located in the Biblioteca Apostolica Vaticana, BAV reg. lat. 1143, provides an answer to the above question. This manuscript, which contains several recipe collections unattributed to earlier traditions, includes a remedy for epilepsy titled Ad cadiuo homine potionem probatam. 54 The treatment presents a list of twenty-four ingredients (see Fig. 7), recording both locally available herbs, such as pennyroyal, bettony, parsley, and lovage, as well as exotics. These include a number of the gums and resins seen in the recipes above, such as frankincense and myrrh, and fragrant, imported  spices, such as cinnamon and clove. 55 Most significantly, however, the recipe lists timiama itself, thus indicating that a pre-prepared incense mixture, such as the product of one of the seven recipes discussed above, could have been used within a medical context. Looking at the rest of the remedy, its instructions direct that the ingredients should be made into a powder, mixed with wine, and then given to the patient to drink. That this remedy is to be liquefied by the addition of wine and then consumed by drinking fits with one of the recipes of csg. 759, the Conpositio thymamatis, the only recipe to include a liquid. Returning to Opsomer's Index de la pharmacopée, thymiama appears thirteen times as an ingredient within the selected medical texts, such as the works of a variety of late antique authors, including Marcellus of Bordeaux (fl. c.490-510), Oribasius (c.320-400), Caelius Aurelianus (fl. c.400), and Paul of Aegina (c.625-90). 56 I am, however, unconvinced that all of these references actually reflect incense. The entry listed in Marcellus' De medicamentis liber, for example, records not thymiama but ammoniaci thymiamatis, or gum ammoniac. 57 Indeed, of the 494 references to ammoniacum documented by Opsomer's Index de la pharmacopée, sixty-three include the added descriptor thymiama or thymiamatis. 58 Moreover, Index de la pharmacopée does not record any references to incensum within its selected texts, whereas I have identified this term as an ingredient in the manuscript sample involved in this study. Specifically, a treatment for lethargy entitled Ad litargigum qui nimis dormiunt located on page 326 of csg. 44, a section of the manuscript transcribed by neither Sigerist nor Jörimann and therefore not included in Opsomer's Index de la pharmacopée, recommends a fumigation with incense, incense (see Fig. 8 independently of ammoniacum) and incensum, and c) that their use may represent a particularly Carolingian development and one linked to Christian influences, topics to be addressed below.

Summary of manuscript evidence
The evidence presented above indicates that incense must be understood as a multipurpose substance. The textual environments in which incense recipes are found and the appearance of individual components of these recipes as ingredients within medical remedies initially suggested that incense could have been used medicinally. The discovery of timiama and incense listed as ingredients within remedies, however, provides undeniable evidence for its application in medical contexts. This medical side to incense does not detract from its use in Christian ritual but adds another layer to our understanding of the ways in which it was used and viewed by early medieval individuals. The potential for multiple applications of incense connects to a wider study of health, healing, and spirituality during this period. The remaining sections of this article will aim to explore this area of intersection, first situating the use and production of incense alongside the users and producers of the manuscripts analysed above.
Locating incense use and production: manuscripts, medicine, and incense The significance the Carolingian court placed on the correct performance of the liturgy would suggest that efforts were made to ensure that sufficient amounts of incense reached churches for use in Christian ritual. The wealth of the church, moreover, would have made possible the purchase of the exotic ingredients involved in incense production.
The aforementioned records of the Abbey of Fontenelle documented that a pound of silver was allocated annually for the purchase of spices, reflecting the importance placed on importing foreign ingredients and the church's ability to finance such purchases. 60 A letter in the Collectio sangallensis provides more details of the types of exotic goods in circulation. Attempting to appease Louis the German, a ninth-century bishop, probably Salomon II of Constance, sent the king exotic goods acquired overseas. 61 The letter records some of these gifts, such as fine textiles, an ivory comb, and exotic fruits and spices. 62 While the fruits and spices, including dates, figs, pomegranate, cinnamon, galingale, cloves, mastic, and pepper, all appear in medical texts, some of the products, such as cinnamon, galingale, and mastic, are also found in certain incense recipes. Although the letter makes no explicit connections to either incense or medicine, it is noteworthy that these exotica were in the possession of a Frankish bishop and speaks to the evident wealth of the ecclesiastical elite (or at least of certain individuals). Ecclesiastical communities therefore represent sites that would have required incense for religious purposes and would have been able to finance its production. It must be remembered that royal residences, and perhaps even those of the leading aristocratic families, often had chapels, meaning that lay individuals, too, would have been involved in funding the supply of incense. In addition to the funds, these overlapping elite environments would have offered not only the spaces in which incense would have been used as part of Christian ritual, but also the networks that sustained the intellectual traditions underpinning its multipurpose nature. Given that all extant early medieval manuscripts containing medical texts are 'likely monastic products', I shall concentrate on incense use in this context. 63 Indeed, surviving manuscripts, including remarkable ninth-century library catalogues from both Reichenau and St Gall, indicate both the range of medical texts to which certain monastic communities may have had access, and their work in copying and adding to these writings. 64 Additional textual evidence, such as The Plan of St Gall, csg. 1092, suggests that herbal medicine was not simply copied but put into practice. Although the plan depicted in this early ninth-century manuscript looks much like an architectural blueprint, it is now thought to represent an idealized vision of a monastic centre. 65 Detailed labels provide a wealth of information, specifying even the plants in the gardens and the livestock in the fields. Significantly for this study, it includes a medical area which contains a bloodletting room, physician's quarters, infirmary, and infirmary garden. 66 That the ideal monastic design features spaces dedicated to medical practice indicates the importance placed on medicine and the maintenance of good health and suggests engagement with medicine on both intellectual and practical levels. Monastic communities thus represent sites in which the duality of incense could have existedenvironments in which incense could have been afforded, described, produced, and used to serve both spiritual and medical needs.
With this monastic and medical background in mind, the significance of the medical use of incense deserves consideration.

The significance of incense as a multipurpose substance
The identification of incense as a substance with multiple, overlapping uses is significant for a number of reasons. I shall highlight several aspects that are particularly important to consider, including the potential Christian influence on, and symbolism attached to, medical incense use as well as its specifically Carolingian dimension.
Sacralizing medicine? A consideration of the potential Christian influence on, and symbolism attached to, medical incense use How should the religious significance of incense be understood when it is used medicinally? There are numerous ways to answer this question, ranging from an areligious interpretation on one end of the spectrum, to an application of incense that fully embraces and engages with Christian symbolism, on the other. First off, there appears to be a long tradition of healing practices that use incense. Consider, for example, the references to cyphi in Dioscorides' De materia medica or The Alphabet of Galen noted earlier. The medical uses of many of the individual incense ingredients have also been showcased above. Might the incorporation of incense as an ingredient itself simply represent an extension of the use of its constituent parts, without Christian overtones? While it is possible that a medical use of incense was devoid of any religious significance and was based entirely on contemporary medical knowledge or earlier traditions, this possibility seems highly unlikely, especially given the monastic environments in which many of these recipe collections were composed and housed. Indeed, the aforementioned cyphi examples appear to be entirely distinct from the Carolingian recipes for incense. In contrast, the reliance on incense in Christian rituals and its associated symbolic meanings may have encouraged its application in medical contexts during this period. Yet the degree to which contemporary liturgical uses of incense influenced its use in healing practices remains highly complex, as the following examples indicate. a) Purification: an ancient blend of spiritual and medical practices Incense has long been linked to purification. 67 Indeed, the burning of particular substances may produce the desired cleansing, purifying effect. Citronella candles, for example, are well known to clear the air of insects. The burning of aromatics would have covered up or eliminated noxious odours, thereby appearing to purify the air. 68 Ideas of purity are intrinsically linked to sanctity, and this relationship bridges the divide between spiritual and corporeal purification, crossing the boundaries of body, mind, and soul. The presence of sweet odours, particularly when inexplicable or unexpected, was regarded as a sign of sanctity. 69 This is perhaps most frequently associated with the opening of tombs: a burial, a site of death and decay, would never, under normal circumstances, produce a sweet-smelling odour. Early Christian and medieval saints' Lives, however, record many such tomb openings. 70 The aromatic odour was interpreted as a confirmation of the individual's sanctity and holiness (though it is likely that the heavy use of incense in preparing the dead for burial resulted in these sweet smells, creating a self-fulfilling perfuming-cycle). 71 The association between good fragrances, sanctity, and purification was also recorded in other contexts, often making explicit connections to health. Gregory of Tours, for example, wrote that a paralytic brought to the tomb of St Albin was cured thanks to the intervention of St Martin and St Albin, whose presence was 'manifested by a sweet odour in the church'. 72 Conversely, devils, disease, and sinfulness were all associated with the smells of death, decay, and other putrid stenches. 73 Given these relationships, it is understandable that the burning of aromatics was seen as a potential cure for many diseases, and especially those thought to be related to the supernatural or linked to spiritual causes, such as demon possession, the topic of the following example. 74 Yet it must be remembered that the burning of aromatic substances, including incense, as part of purification and healing rituals has a long history and antedates the liturgical use of incense. However, while these deep-rooted traditions are essential to recognize, I argue that they are not directly responsible for the appearance of incense in Carolingian medical texts. b) Christian influence: an explicitly Christian connection The treatment in BAV reg. lat. 1143 that listed timiama as an ingredient, Ad cadiuo homine potionem probatam, targets epilepsy, a condition that has 'often been considered in premodern civilizations to be the work of indwelling, malicious spirits'. 75 Similarly, one of the remedies noted above for its use of exotic aromatics, Ad demonio expellendo of csg. 759, represents a demonifuge: the demon possessing the afflicted individual is to be smoked out. While incense is not listed as an ingredient in this treatment, some of its constituent parts, such as frankincense and myrrh, are used. Paris BnF lat. 9332, a manuscript from the sample that did not include any specific references to incense, does contain a treatment essentially identical to the demonifuge in csg. 759, listing the same title, ingredients, and instructions for its preparation and use. 76 While the two remedies appear to stem from the same source, their textual environments are considerably different, showing that knowledge of this specific treatment circulated during this period. (The remedy in csg. 759 represents the ninety-first entry in a recipe collection, while the fumigation of Paris BnF lat. 9332 follows an excerpt of the Practica, a Latin translation of Alexander of Tralles' (c.525-605) Twelve Books on Medicine.) Such examples highlight that many of the treatments that incorporate a fumigation of incense or its aromatic ingredients relate to conditions associated with spiritual, supernatural causes. I suggest that this is no coincidence: the incineration of aromatic substances (whether specifically incense or its ingredients) directly parallels the liturgical use of incense.
Would such overtly Christian elements have been expected to heighten the healing potential of a treatment? Or were they included primarily in the types of treatments seen above because of the contemporary understanding of the spiritual/supernatural causes of certain conditions? While a comprehensive assessment of the occurrence of Christian elements in early medieval remediessuch as the use of holy water, inclusion of prayers, as well as the incorporation of incensewould be needed to address these questions, it is important to note another feature of the demonifuge in Paris BnF lat. 9332. Unlike the version listed in csg. 759, Ad demonio expellendo of Paris BnF lat. 9332 includes an additional approach to expelling demons that can help our understanding of the issues at stake. Although this recipe involves neither fumigation nor incense ingredients, it does include a number of other Christian features, such as the Lord's prayer. 77 This suggests that those conditions associated with spiritual and supernatural causes may have been more likely to have been treated with preparations involving explicitly Christian elements.  78 While early medieval individuals working with incense in medical contexts may have conceived of its use and potential symbolic meanings differently, I suggest that it is likely that it was deployed with at least some level of sacred significance. Leja's research on the sacralization of Carolingian medicine, although concentrating on different types of textual evidence, adds weight to this argument. In particular, Leja highlights 'a vision of medicine in which the health of the body and the health of the soul ran together and reinforced each other', an expression of early medieval healing that aligns perfectly with the duality of incense. 79 Furthermore, Leja suggests that the absorption of 'classical medical knowledge within a Christian tradition' was an element of Carolingian correctio. 80 It is essential to place the medical use of incense within this bigger picture. When Carolingian reform efforts are considered more broadly, including the importance placed on liturgical correctness as well as the growth of churches during this period, it seems likely that the trade in incense (or the raw ingredients used to produce it) also expanded. I suggest that the appearance of incense in medical contexts was specifically linked to this development and crucially supported by several other related changes. In particular, the arrival of new exotic aromatic substances from the east, such as cozumber, camphor, and ambergris, corresponds to the appearance of the recipes for thymiama. As these recipes represent among the earliest references to many of these ingredients (as noted above), it suggests the linked movement of the substances themselves and knowledge surrounding their use. 81 Perhaps most notably, the recipes from Group B appear to share a common tradition and thus speak to the exchange of information and the movement of manuscripts fostered by the intellectual networks spanning Carolingian Europe.
Given that the appearance of incense within this study's manuscript sample appears to be distinct from the earlier cyphi-based traditions, I suggest that the inclusion of incense recipes in medical writings and the use of incense as an ingredient in medical treatments is strongly linked to the aforementioned Carolingian developments. While the sacralization of medicine offered an intellectual environment in which incense could be incorporated, Carolingian incense requirements helped to expand the numbers of incense recipes available (and knowledge associated with them). Simultaneously, the exchange of information and texts supported the spread of the newly introduced thymiama recipes and their wider application.

Conclusion
Ultimately, incense, as a multipurpose substance, offers a new perspective on early medieval attitudes towards healing, connecting physical, mental, and spiritual health in the broadest of terms. Incense should not be seen as a substance reserved exclusively for liturgical uses, but as one with multiple, complementary applications. Moreover, the production of incense and the significance of its use appear to be topics directly related to major developments in the Carolingian world. Incense thus represents an inroad into a much wider discussion of health, healing, and spirituality, as well as networks of exchange in this period, and one much deserving of future research.

Appendix 1
The research presented in this article is based on a sample of eighteen eighth-and ninth-century manuscripts. This appendix presents a list of the eighteen manuscripts and information on their dating, place of origin, and contents. Manuscripts containing references to incense have been marked with an asterisk. The material presented here has been taken from B. Bischoff, Katalog der festländischen Handschriften des neunten Jahrhunderts (mit Ausnahme der wisigotischen), 3 vols (Wiesbaden, 2014) and A. Beccaria, I codici di medicina del periodo presalernitano (Rome, 1956).
Note: manuscripts in the Stiftsbibliothek St Gallen are paginated rather than foliated.