The sources of Eucharius Rösslin's 'Rosegarden for pregnant women and midwives' (1513).

Few medical authors can unambiguously claim to have written one of the most important works in their field: most important not simply in one language but in half a dozen, and not simply for a few years but for over a century and a half. Yet that distinction has long been given to the work of a largely obscure early sixteenth-century apothecary-turned-physician from Freiburg, Worms, and Frankfurt, one Eucharius Rösslin (c.1470–c.1526).1 His Der Swangern Frauwen und Hebammen Rosegarten (Rosegarden for Pregnant Women and Midwives), first published in Strasbourg and Hagenau in 1513, went through at least sixteen editions in its original form, was revised into three different German versions (each of which went through multiple printings), and was translated into Czech, Danish, Dutch, English, French, Italian, Latin, and Spanish, with almost all of these translations then going through their own multiple editions.2 The Rosegarten is the only work known to have been produced by Rösslin. His son, Eucharius Rösslin Jr, further capitalized on the work by producing in 1526 a German compilation of “marriage texts” which he called Ehestandts Artzney; this included his father's Rosegarten as well as extracts from the Enneas muliebris (Nine-Part Treatise on Women) by Ludovico Bonacciuoli (d. c.1540), a herbal by Johannes Cuba (Johann Wonnecke von Caub, d. 1503/4), and Bartholomeus Metlinger's (born after 1440) tract on paediatrics. Eucharius Jr. also produced a Latin translation of the Rosegarten in 1532. That Rösslin's work was only the third obstetrical text addressed directly to an audience of midwives in a thousand years also places it in an important position in the history of the professionalization of midwifery.3 While it remains to be determined how frequently midwives themselves read the text, it is clear that both physicians and laypersons used the Rosegarten and later adaptations as the basis for medical training and as a reference for information on generation.4

Despite the unquestioned historical importance of this work, its textual sources have never been examined in any systematic way. In large part, this seems to have been due to scholars’ sense that the text was sui generis, an “out of the blue” creation that suddenly revived the long-lost obstetrical practices of the ancients. The one source that scholars have always acknowledged for the Rosegarten is the late antique work of Muscio, the Gynaecia (Gynaecology, itself a Latin translation of Soranus's second-century Greek Gynaikeia), from which Rösslin derived the foetus-in-utero figures that are still the most recognizable feature of the work. Yet, as I will show, while it is clear that Rösslin must have consulted at least one independently circulating fragment of Muscio's text that included the foetus-in-utero images, the Rosegarten owes nothing at all to the full text of the Latin Gynaecia.

In 1994, a philologist, Britta-Juliane Kruse, published an initial analysis of a German manuscript now in Hamburg, dated 1494 and so predating by nearly twenty years the initial publication of the Rosegarten. She argued that it presents an Ur-version of Rösslin's printed text. It lacks a number of features found in the printed work: the imprimatur of Emperor Maximilian; the dedication to Katharina, Duchess of Brunswick-Lüneburg;the rhymed prologue with its viciously critical account of the errors of contemporary midwives; and a closing glossary.5 It also has no illustrations. In nearly every other respect, however, it is the predecessor text of the Rosegarten, which Kruse could now prove had not been created for publication in print. Kruse announced plans to publish this Ur-version (which bears the manuscript title Von Kranckheiten, Siechtagen und zu val der Swangern und geberenden frowen und ihrer neugebornen Kinderen [On the Sicknesses, Illnesses, and Accidents of Pregnant and Labouring Women and Their Newborn Children]), a project that is still much anticipated.

Important as Kruse's analysis was, however, it only pushed back to a manuscript phase the question of the text's origins. Rösslin would have been in his early twenties at the time the Hamburg manuscript was written, and it is by no means clear that he had anything to do with its production. Indeed, it has long been questioned how he could have assembled such a detailed text on a topic on which, as far as we can tell, he had no particular expertise. So where did Von Kranckheiten, Siechtagen und zu val der Swangern und geberenden frowen und ihrer neugebornen Kinderen come from? The present study is meant to contribute to a new understanding of the genesis of the Rosegarten and its antecedent German text by demonstrating that the bulk of the text was not a novel composition by Eucharius Rösslin himself, or even another German physician or apothecary, but a translation of a pre-existing Latin text (composed between 1440 and 1446) from the other side of the Alps by the Paduan and Ferrarese physician Michele Savonarola (c.1385–1466). Savonarola was himself drawing heavily on the obstetrical chapters of the early fourteenth-century Neapolitan physician, Francesco da Piedemonte (d. 1320), but the nature of the correspondences between the German texts and Savonarola's is close enough to prove that the latter was the direct source, not the Neapolitan da Piedemonte.6 This discovery also helps us better understand the relation of Rösslin's 1513 Rosegarten to the one similar printed German text that preceded it, the anonymous Frauenbuüchlein (Women's Manual) that was first published in Augsburg c.1495. Finally, it suggests the importance of interrogating more systematically what was really “new” in the age of print; as this example shows, Rösslin's Rosegarten was as important in disseminating late medieval northern Italian obstetrical practices to the rest of Europe as in capturing local empirical practices in sixteenth-century Germany.

Rösslin's Predecessors

My focus here is on establishing the origins of the obstetrical material, but it should be noted that scholars have for some time acknowledged that the Rosegarten incorporated, as its final three chapters on caring for the child, an originally independent German paediatric text attributed to Bartholomeus Metlinger, the Kinderbuüchlein (Children's Book). The latter was first published in 1473 in Augsburg, where Metlinger was city physician from at least 1470 to 1488.7 Since the Kinderbuüchlein material is also found as part of the Hamburg manuscript, we cannot credit Rösslin with the decision to incorporate it into the Rosegarten. I follow Keil and Kruse in ignoring the question of the sources of this paediatric text or what its incorporation into the Hamburg manuscript might signify; I consider that the task of whoever eventually edits the Hamburg manuscript. However, as I point out below, the incorporation of Metlinger's paediatric text also raises important questions about the genesis of the obstetrical material.

In any case, the origins of the remaining three-quarters of the Kranckheiten / Rosegarten, including all the obstetrical and gynaecological sections, are still unaccounted for. It seems likely, on my analysis, that the German author was drawing on his own experience as a medical practitioner (or simply on some other, as yet unidentified text) for some of the groups of recipes now found in the text. For example, many of the aids for delivery in Chapter 5 of both manuscript and print versions of the German text have only a slight correspondence with Savonarola's (far more detailed) Latin work.8 Additionally, there is novel material on surgical repair of the pelvic floor that I have thus far documented (at least in its present form) no where else. Be that as it may, the bulk of the interpretive material—on the structure of the amniotic sac, on the causes of obstetrical complications, and on regimens to avert complications or therapies to rectify them—comes from a single Latin source: the gynaecological and obstetrical chapters of Michele Savonarola's Practica (sometimes called the Practica major, Greater Text of Practical Medicine), a work composed in Padua and Ferrara between 1440 and 1446.

Before addressing the connection to Savonarola, however, let me first explain why the late antique author Muscio should be excluded, or at least minimized, as Rösslin's source. There is no denying the obvious iconographic debt of Rösslin's foetal images to those in Muscio's Gynaecia. Moreover, scholars have always been right to note that Muscio's and Rösslin's texts share a distinction in being among the few premodern representatives of obstetric texts. Although there was an extraordinary amount of mater ial on women's conditions circulating in medieval western Europe, there was virtually nothing focusing solely on assistance in childbirth.9 Not even the famous twelfth-century Trotula ensemble from Salerno offered much information on how normal birth was to be handled, nor was there anything on how to address malpresentations beyond the perfunctory advice to have the midwife reach inside with an oiled hand and “put it in its correct position”.10

Muscio's late antique Gynaecia (thought to have been composed in fifth- or sixth-century North Africa) was always the main exception to this generalization.11 It included extensive sections on the management of both normal and abnormal childbirth, and it viewed midwives as the chief providers of women's gynaecological as well as obstetrical care. By the time medicine experienced a major revival in the eleventh century, Muscio's intended audience of literate midwives had long since disappeared. From the twelfth century on, the practice of gynaecology and obstetrics effectively bifurcated: theoretical gynaecology became more and more the province of learned male physicians, while childbirth management remained in the hands of women. Chapters on gynaecological diseases along with some minimal information on handling difficult birth could be found in most general medical texts (all written by male writers) and there is ample evidence that male physicians (and later, surgeons) treated women for gynaecological and even some obstetrical conditions. Procedures for managing normal childbirth, on the other hand, were part of oral lore rather than written texts.12

In physicians' general medical textbooks, some advice for obstetrical complications could usually be found. Failure of the foetus to exit (especially when it was already dead), retention of the afterbirth, excessive bleeding—all were uniformly treated with medicinal remedies or, at most, recommendations for baths or external massages or other applications to the belly, rather than with anything involving manual intervention. For their part, Latin surgical writers avoided incorporating obstetrical instructions into their texts until 1363, when Guy de Chauliac included extraction of the dead foetus and Caesarean section (to remove a living foetus from its dead mother, primarily with the goal of baptizing it) among the surgeon's tasks.13 The one exception prior to Guy in offering detailed instructions for difficult birth was the obstetrical section in the Latin translation of the Chirurgia of the Spanish Arab physician al-Zahrāwī (d. c.1013, known in Latin as Albucasis). The obvious relevance of Muscio's foetus-in-utero images to Albucasis' text was recognized in the late thirteenth century in northern Italy when the Muscian images were extracted from their original context and appended to the Chirurgia.14 Thus, there was acknowledgement well before Guy's text that obstetrical information was useful for the male surgeon or physician. Even male clerics involved in supervising midwives were interested in Muscio's text. The thirteenth-century Dominican preacher in Liége, Thomas of Cantimpré, for example, included an excerpt from an abbreviated version of Muscio (sans images) in his encyclopaedia for other clerics, Liber de natura rerum (On the Nature of Things), while one of the more famous copies of the foetal images, the Wellcome Apocalypse manuscript, was made c.1420 for a German cleric.15

The full text of Muscio's Gynaecia was no longer copied after the thirteenth century. By that point, it had been superseded either by the new Trotula treatises emanating from the southern Italian centre of Salerno, or by the chapters on women's diseases found in the major Arabic encyclopaedias that had been translated into Latin in the eleventh and twelfth centuries. These newer works were more compatible with the Galenism that dominated medicine overall. Even two abbreviated versions of Muscio made probably in the eleventh century in southern Italy had only brief periods of circulation.16 Rather, as we saw with Albucasis's Chirurgia, only the foetal images from Muscio's Gynaecia continued to attract attention. These often circulated with the accompanying text that explained what the images meant (see Figure 1 below): that is, what the midwife should do if the foetus presented feet-first or hands-first, etc. Besides the five manuscripts where the foetal images were attached to Albucasis's Chirurgia, there are eighteen other extant manuscripts that present the images independently of Muscio's full text. In at least two cases, moreover, the images were embedded into newly composed gynaecological and obstetrical texts.17 In all, there is no part of western Europe save Spain where the Muscian images are not documented in the fourteenth and fifteenth centuries.

Figure 1

Figures 5–8 of the Muscian foetus-in-utero series from a manuscript produced in Bavaria/Swabia around 1485. The copyist has misplaced the image of the small-headed foetus (what should be Figure 6) into the lower right corner. Münich, Bayerische Staatsbibliothek, Cgm 597, f. 260v, reproduced from Karl Sudhoff, ‘Neue Handschriftenbilder von Kindslagen', Archiv für Geschichte der Medizin, Jan. 1908, 1: 310–15, plate III. (Wellcome Library, London.)

Southern Germany in particular witnessed something of a revival of Muscio in the fifteenth century. Prior to that point, only one copy of Muscio's complete Gynaecia can be traced to German-speaking areas. This was a manuscript of unknown date that had been held from the early twelfth century by the Benedictine monastery of St Michael's, Bamberg (in northern Bavaria); it was still listed in the library's catalogue near the end of the fifteenth century.18 Whether this copy (long since lost) played any role in the fifteenth-century “revival” of Muscio is unclear. Likewise, it can not yet be proved that an Italian copy of the Gynaecia (again, with the full foetus-in-utero sequence) brought by the reforming cleric Johannes Spenlin to Heidelberg around 1450 (where it was to become part of the Palatine library of Elector Frederick I) was particularly influential.19 Yet we do know of (1) at least three copies of the foetal illustrations series circulating in southern Germany in the fifteenth century;20 (2) the presence of an adaptation of Muscio's obstetrics (text only) entitled Secreta secretorum mulierum (The Secrets of the Secrets of Women) at the Bavarian Carthusian house of Buxheim;21 and (3) a pervasive local interest in “women's secrets”, which manifested itself in the popularity not only of the late thirteenth-century pseudo-Albertus Magnus text of that title, but also in the circulation of gynaecological texts and pictures of “disease women”.22

While the Muscian foetal images, with their accompanying text, could certainly have been helpful in teaching some basic obstetrical interventions, there was something that the full text of Muscio's Gynaecia contributed to obstetrics that neither the images alone nor the abbreviated versions could offer. In all these derivative forms, Muscio's explicit statement in his prologue that his intended audience was midwives had been lost. Even when gynaecological texts were addressed to women (as we occasionally find in France, England, and the Low Countries between the thirteenth and fifteenth centuries), they were addressed to women generically, not to specialist midwives.23 It is therefore notable that in the early 1460s, the Munich physician Johannes Hartlieb (d. 1468) should announce his plans to translate the whole of Muscio's text, wondering why no German translation had yet been made of a book that would obviously be “such a treasure to midwives”.24 Hartlieb ended up abandoning the project because, he said, of the work's excessive length, but he had at least articulated for the first time the virtue of targeting obstetrical writings directly to midwives—a sentiment all the more notable since he addressed his two other writings on women to lay male patrons.25 It is probably no coincidence that Hartlieb conceived of producing such a “treasure for midwives” at precisely the time when towns throughout southern Germany were beginning to draft their own regulations for the licensing of midwives (Regensburg in 1452, with half a dozen other cities following in the next half century). The moment was clearly ripe for the composition of an obstetric text that made available for late fifteenth-century Germany what Muscio had done for late antique North Africa.

In sum, therefore, there is no question that Muscio's Latin Gynaecia—usually just excerpts but also occasionally the whole work—was readily available in southern Germany from the middle of the fifteenth century. There was not only general interest in the topic of reproduction among both male medical practitioners and clergy, but a growing desire to regulate the practice of midwifery. There was, in other words, a “market” for Muscio. But besides offering an obvious iconographic model for the woodcuts that Rösslin commissioned for the Rosegarten, Muscio's Gynaecia played only a minimal role in the Rosegarten's composition. Indeed, it was not even used for all the text that explained what the foetal images were meant to depict.

Originally, Muscio's Gynaecia had had fifteen images of foetal presentations. This number grew to sixteen in manuscripts produced from the twelfth century with the addition of a new image depicting an additional pair of twins, one presenting head-first, the other feet-first. All three of the extant manuscripts of the images that come from southern Germany have this full sixteen-image sequence, and two of them have the accompanying text.26 Yet the first thing we notice when we compare the sequence of images in the Muscio manuscripts with that in the Rosegarten is that, whereas the order of images and topics is quite stable in the two dozen manuscript copies of the Muscian figures,27 Rösslin's printed text deviates not only from the order, but omits presentations that Muscio had included and adds new ones (see Table 1). In Chapter 2, on time of delivery and natural versus unnatural presentations, Rösslin included the first two Muscian images, which illustrated a normal, head-first presentation, and then an abnormal (but not necessarily dangerous) feet-first presentation. In Chapter 3, on deliveries that are hard or easy, he included a new image showing a two-headed baby born in Werdenberg the previous year, in 1512. This latter image, although depicting a child that has already been born, is presented within a uterus that looks like an upside-down bottle, just like all the other “foetuses”. Rösslin has moved beyond the medieval understanding of the images as being aids to obstetrical intervention, into a Renaissance fascination with monsters and marvels.28 The “bottle” of the uterus is now iconographic shorthand for saying “This is a foetus or a newborn”.

Table 1

The foetus-in-utero images in Muscio, Gynaecia, vis-à-vis Rösslin, Rosegarten1

Sequence of images in Muscio, Gynaecia	Sequence of images in Rösslin, Rosegarten
Image 1: hic est secundum naturam primus et melior ab omnibus partus (normal presentation, head down, arms at side, legs straight)	Image 1: Chap. 2, f. Ci verso (die natürlich geburt)
Image 2: et hic secundum naturam est, sed secundus partus (head up, both feet straight down, arms down by sides)	Image 2: Chap. 2, f. Ci verso (wan das kindt uß müter leib kompt zü ersten mit den füessen)
[no correspondence]	Image 3: Chap. 3, f. Cii verso, new image of two-headed baby born in Werdenberg (ein kindt mit zweyen heuptern geboren)
[no correspondence]	Image 4: Chap. 4, f. Diii verso, repetition of image 1 (Dises alles ist gesagt von der natürlichen geburt)
[no correspondence]	Image 5: Chap. 4, f. Diiii recto, repetition of image 2 (Wo aber das kind erscheynt und kompt mit unnatürlicher geburt mit beden füessen)
Image 3: Si in divexum2iacet (transverse lie, usually with arms extended straight out to side, i.e., straight up/down axis of uterus)	Image 8: Chap. 4, f. Ei recto (Wo aber das kind ka¨me mit einer seiten an die geburt [here with both hands swinging to the foetus's right side])
Image 4: Quotiens manum mittit (one hand leading, extended through mouth of uterus; head down, legs together, other arm at side)	Image 11: Chap. 4, f. Ei verso (Item ob das kind ein hand erzeugte)
Image 5: Si ambas manus eius foras invenerit (two hands leading)	Image 12: Chap. 4, f. Eii verso (Ob aber das kind mit beyden henden erschyne)
Image 6: [S]i brevissimum caput habeat et ambas manus foras eiecerit (small head, with two hands alongside)	[no correspondence]
Image 7: Si in pedibens descendens in aliquam partem vulvae relicum corpus inclinaverit (feet descending first, but “stuck” in sides of uterus, arms extended straight to sides)	[no correspondence]
Image 8: Et si unum pedem foris eiecerit (one foot presenting, other bent; arms extended to sides)	Image 7: Chap. 4, f. Diiii verso (Wo aber das kind züm ersten ka¨m mit einem füß allein)
Image 9: Si ambos pedes foris eiecerit et manus super caput contortas habuerit (feet first, arms extended to side or above head)	Image 6: Chap. 4, f. Diiii recto (Wo aber das kind erscheint mit beyden füessen)
Image 10: Si divisis pedibus duabus partibus vulvae plantas infigat (feet first, legs separated, arms extended down and out)	Image 9: Chap. 4, f. Ei recto (So aber das kind kem mit geteilten füessen)
Image 10A: Si caput transversum aut tortum habeat (head first, neck twisted, arms extended out)	Image 14: Chap. 4, f. Eii verso (Item ob das kind mit gebognem geneygtem oder krümem haupt erschyne)
Image 11: Si genua ostenderit (knees first)	Image 10: Chap. 4, f. Ei verso (Und so das kind sich mit den kneüen erzeugt)
Image 12: Si vero in naticas sedet (buttocks first)	Image 13: Chap. 4, f. Eii recto (Item ob das kind sich mitt dem hindern erzeugte)
Image 13: Si duplicatus fuerit . . . volo duobus modis pecus in vulva duplicari (folded in half, either forwards or backwards, buttocks or head leading)	Images 15 and 16: Chap. 4, f. Eii verso (Und so das kind ka¨me mitt beiden oder einem füß und mit dem haupt [bent forwards]) and Eiii recto (Item ob das kind geteilt lege oder uff seinem angesicht [bent backwards])
Image 14: Si in divexum iacet (transverse, with arms bent)	[no correspondence]
Image 15: Et si plures ab uno fuerint, tres vel quattuor (twins (or more), both head up/feet down or head down/feet up)	Images 17 and 18: Chap. 4, f. Eiii recto (Item ob der kinde mer dann eins wer als zwyling und sich gleich erzeugten mit den ho¨uptern [heads down]) and Eiii verso (Wo aber die zwyling ko¨men mitt den füessen [heads up])
Image 16: (2nd set of twins, one up one down, or both head down)	Image 19: Chap. 4, f. Eiii verso (So aber der zwyling einer kompt mit den haupt, der ander mitt den füessen [one up, one down])

1 Muscio, Gynaecia, as edited in Valentin Rose (ed.), Sorani Gynaeciorum vetus translatio latina, Leipzig, Teubner, 1882, pp. 84—9; and Eucharius Rösslin, Der Swangern Frauwen und hebammen Rosegarten, facsimile reproduction of the 1513 Strasbourg edition, ed. Huldrych M Koelbing, Zürich, Verlag Bibliophile Drucke von J Stocker, 1976.

2 The term here, divexus, means ‘‘spread out’’. Some manuscripts read dextrum.

In Chapter 4, Rösslin repeats the first two images of “natural” positions and then fills in the rest of the chapter with fourteen more images, all of them corresponding to models from Muscio, but differently arranged and sometimes presenting two different versions of parallel conditions. For example, images 15 and 16 show foetuses doubled over, one forward, the other backward. In fact, both versions can be found in Muscio manuscripts, but never together. Rösslin seems to have had at least two manuscript exemplars that he was working from. But more to the point, he had a different text than Muscio's that he was trying to illustrate with this expanded sequence of images.

Compare Muscio's description of what should be done in the case of a foot presentation (Figure 1, lower left figure) with that of Rösslin (Figure 2):

Figure 2

Depiction of a foot presentation in Rösslin's Rosegarten. Reproduced from Eucharius Rösslin, Der schwanngeren Frawen und Hebammen Rosengarte [sic], [Augsburg, Heinrich Steiner, 1529], f. Ciiii verso. (Wellcome Library, London.)

As is obvious, Muscio's recommendations and Rösslin's have no overlap other than the injunction to push the presenting foot back inside. Whereas Muscio recommends a podalic extraction pure and simple, Rösslin recommends that the parturient should perform a sort of gymnastic self-massage in order to get the child back into a head presentation. Only after that has been attempted does he advise (in the subsequent passage) podalic extraction. There is no direct textual correspondence here.

One might think that Rösslin was simply offering an original elaboration on Muscio—based, presumably, on his own experience—except that there is another text that presents the same instructions as the Rosegarten's. Here is Savonarola's recommendation:

If [the foetus] should come out upon one foot, it is necessary that the woman not sit, but lie with her thighs elevated, and with her head erect. And the midwife with her hand should gently reposition its feet. Then, the patient should turn herself around with many turnings, until the embryo turns itself around, its head inclined downward toward the port of exit, and then let her sit again.31

Like Rösslin, Savonarola then goes on to recommend podalic extraction if the foetus will not move sufficiently. In other words, the only significant difference between Savonarola's version and Rösslin's is that in the former the woman's head is up and in the latter it is down—the kind of minor difference readily attributable to scribal error (whether in the Latin or German tradition).

After the footling presentation, Rösslin's text diverges both from Savonarola's Latin text (which did not go into any further detail on different malpresentations) and from the unillustrated manuscript text of Kranckheiten as found in the Hamburg manuscript. The text accompanying Figures 4, 6, and 9 to 19 of the Rosegarten (as well as smaller passages in between) reflects to varying degrees the Latin text that had traditionally accompanied the Muscian images. Whether Rösslin translated the Latin text himself or drew on a previously existing German translation is unclear.32 Beyond the foetal images and their brief explanatory text, however, there is no evidence that Muscio served as the source either for the Kranckheiten or for Rösslin's printed Rosegarten. Rather, it is Savonarola's Latin Practica that underlies the bulk of these two German texts.

The Rosegarten and Michele Savonarola's Practica

Michele Savonarola was born in Padua c.1385, and studied and taught medicine there until being called to the court of the d'Este family in Ferrara in 1440. He remained in Ferrara for the rest of his career and died there around 1466. Although not as famous now as his firebrand grandson, the Florentine preacher Girolamo, he certainly enjoyed local fame, producing some seventeen different medical writings on topics from fevers to medicinal baths to physiognomy as well as various other political and moral works.33

One of his medical writings was a text little studied in the history of women's medicine, but pertinent to this story as a model, perhaps, for Rösslin (or rather, the German translator from whom Rösslin borrowed). This was a work called Ad mulieres ferrarienses de regimine pregnantium et noviter natorum usque ad septennium (Regimen for the Ladies of Ferrara for Pregnancy and Care of the Newborn Child, hereafter simply Regimen), written c.1460.34 This vernacular work, composed in the local Ferrarese dialect, is currently known in only three manuscript copies, which does not suggest that it enjoyed any broad popularity. Nevertheless, recognizing that he had foreclosed international fame by composing it in the local vernacular, Savonarola mused out loud that the work might some day be translated into Latin and thus “foreigners, who are unfamiliar with the vernacular used in this volume, will be able to have it translated [into their own tongue] with great ease”.35 While I find no evidence that Savonarola's Italian Regimen directly influenced the Rosegarten, knowledge that a work written specifically for an audience of laywomen and their midwives could find a market may have given Rösslin the idea to publish his own work on obstetrics. The German text of Kranckheiten as it exists in the Hamburg manuscript does not address women; as Kusche has already noted in her 1994 study, it instead speaks to “the common man” (gemeine man).36 Rösslin had served in 1508 at the court of Duchess Katharina of Brunswick-Lüneburg, and in dedicating his work to her and then explicitly addressing both pregnant women and midwives, he was doing in German exactly what Savonarola had done for the women of Ferrara. After all, Savonarola himself recognized the value of patronage by women, whom he called “the trumpets of physicians” (medicorum tube).37

It was not the Italian Regimen, however, but rather Savonarola's Latin magnum opus, his Practica (sometimes called the Practica major) that underlies the German Kranckheiten. A huge work, the Latin Practica consisted of short sections on proper decorum for the physician, general sections on pharmaceutics and regimen, and then, as its longest part, an extended treatise on diseases, addressed in head-to-toe order. As was normal in such works, the Practica included a section on gynaecological and obstetrical conditions; this was far more detailed (about causes, symptoms, and therapies) than the vernacular text Savonarola would compose twenty years later for women. The Practica found its way to Germany along with a large body of other works by northern Italian physicians. Medical student traffic from Germany to northern Italy had been heavy since at least the second third of the fifteenth century, and these students brought back copies of the lectures and writings of their teachers. For example, we know that the Munich physician Johannes Hartlieb studied in Padua in the 1430s, when Savonarola himself would have still been teaching there.38 We also know that the Nuremberg physician Hermann Schedel (1410–1485) made his own copy of Savonarola's Practica when he was studying in Padua in 1442.39 Furthermore, printed copies of the entire Practica were available in editions published in 1479 and thereafter.40 Given this traffic, it seems likely that the Practica would have been readily available in Germany by the 1490s. In what follows, I demonstrate how the author of Kranckheiten manipulated Savonarola's Latin text. Four examples should suffice.

A Prenatal Regimen to Avoid Difficult Birth

The author of Kranckheiten devotes the fourth chapter of his text to “how a woman should behave during, before, and after childbirth, and how one should come to her aid in a difficult delivery”.41 This chapter includes the bulk of the work's instructions for regimen during the latter part of the pregnancy and for ministrations during active labour, whether it was difficult or not. It also includes a description of the obstetrical chair (about which more later) and the sequence of foetus-in-utero figures with accompanying instructions on how to deliver the malpositioned foetus.

In the first third of the chapter, the author explains that one should take steps to avoid a difficult labour even before it happens. The woman should have all potentially complicating conditions of the genitalia (boils, ulcers, warts) attended to by a surgeon. Likewise she should try to remedy any urinary tract conditions or problems of the anus. Constipation should be avoided by changing her diet, by use of enemas, or by medicines. The woman should also use various lubricating applications and stretching techniques to prepare the birth canal. Further recommendations about baths, foods, wine, etc., follow. Not a single authority is cited.

Comparing these instructions to Savonarola's Latin text, we find in the latter half of his long Tract VI, chapter XXI, rubric 32, De difficultate partus (On Difficult Birth) virtually identical recommendations. Savonarola begins by clarifying that the regimens he is about to describe are really only for women of the governing classes, “for the physician does not busy himself very much with the poor”.42 We find here all the same elements as in the German text—the recommendation to take care of complicating secondary conditions prior to the birth, foods and drinks to avoid or employ, baths that are recommended or not, oils to apply to the genitals to lubricate them. The differences are slight: Savonarola recommends frequent intercourse in the days prior to birth, a topic on which the German author is silent. The latter says that in older women, the genitals and uterus are drier and harder, and so less likely to be stretched by the softening regimens;43 Savonarola makes no age distinction. Overall, however, the correspondences between the two texts leave no doubt that the German author must have been relying on his Italian predecessor.

The Obstetrical Chair44

The second birth regimen in both Savonarola and Kranckheiten, that to be followed during labour itself, shows the same relationship. The description of the obstetrical chair merits close examination. The use of a specially designed chair for the labouring woman—distinguished by a hole in its centre and its low back—was described by Muscio in his Gynaecia, which in turn was a spare abbreviation of Soranus's earlier, more detailed description.45 Some form of Muscio's Gynaecia, as noted above, was indeed available throughout most of the Middle Ages, but in the more commonly accessible abbreviated versions, reference to the obstetrical chair had been omitted. Nor did any version of the popular Trotula texts or any of the major encyclopaedic works that included chapters on difficult birth make reference to a special chair. In his Chirurgia, Albucasis had said only that, if the foetus was stuck at the back of the uterus, the midwife should make the labouring woman sit on a chair, its form unspecified (“fac eam sedere super sedem”), and lean to the right side to help dislodge the child.46 In his Canon medicine (Canon of medicine), Avicenna merely says “it is better that what she sits on during labour is a stool with a cushion placed behind it”.47

In the manuscript copy of Kranckheiten, there is only a passing reference to ein banck on which the labouring woman should sit, with a cushion behind her.48 It may well be that the brevity of this passage reflects a first-draft attempt by the author to address this issue, for in the printed text of the Rosegarten the obstetrical chair is very clearly described:

[I]n the lands of upper [that is, southern] Germany [Vnd in hohen teutschen landen], and also in Italian regions [auch in welschen landen], the midwives have special chairs when the women are to give birth, and they are not high but cut out and hollow inside, as shown here [see Figure 3]. And they are to be made in such a manner that the woman is able to lean backwards with her back. The chair should be stuffed with cloths at the back, and when the time has come, the midwife is to reposition the cloths on the right or the left side as needed. And she is to sit before the woman, sedulously watching the movements of the child in the womb.49

Figure 3

Image of the birth chair in Eucharius Rösslin, Der schwanngeren Frawen und Hebammen Rosengarte [sic], [Augsburg, Heinrich Steiner, 1529], f. Cii verso. (Wellcome Library, London.)

Especially because this passage seems to be a novel addition of the printed text vis-à-vis the manuscript version, we might be inclined to assume that it was Rösslin, in his role as editor, who added this apparently attentive observation of current practices in Germany. However, an even fuller description of the birthing chair is found in the printed Frauenbüchlein of c.1495 and its manuscript version, both of which mention the practices of “Italian regions” and “our lands”.50 I suggest that these several German descriptions all derive ultimately from that in Tract VI, Chapter XXI, rubric 32 of Savonarola's Latin Practica. He describes the birthing chair and its use in much the same way, but with considerably more detail:

First, the midwife ought to prepare a chair above which the parturient ought to stand, or rather, in relationship to which the parturient ought to position herself so as to make the birth quicker. And in diverse regions and cities women have diverse inventions, which are not possible for me to enumerate. But I should touch on their common [features] which are applicable in all cases.

When she is finally in the act of giving birth, let the midwife order the pregnant woman to sit for the space of an hour or thereabouts. I say “thereabouts” because there are some women so accustomed to giving birth that they give birth in one hour. For if it is not her first birth, the midwife ought to inform herself right from the start so that she knows how she ought to regulate herself vis-à-vis the pregnant woman. Then she should make her walk around, jumping sometimes on one foot, sometimes on the other, which is exceedingly helpful, or she should shout out forcefully, or she should hold her breath so that it presses on the lower parts. Likewise, she should have her hips rubbed and pressed in order to expel the foetus. And when the woman senses that the foetus is descending and the mouth of the womb is opening up with intensifying pains, and that the fluids begin to flow out in greater quantity, then let the midwife order that the pregnant woman sit upon the high seat with a cushion on its outer edge. And behind let [another] cushion be placed, and another woman to whom she can cling; or, if it is possible, let her stand on her own feet and let her suspend herself from the neck of a strong woman who holds her up. Or let her squat on her knees on a bed where she is supported by other women. And some women, such as the Greeks, have a seat made in this manner, like this. [Presumably an illustration was meant to appear here.]51

While the parturient sits upon the first semicircular outer edge [of the chair], behind her stands the [woman] who supports her, and she holds on to the cushion, and behind her is another woman slightly above her, holding and controlling her, on whom the woman leans for support. And this is a good method, although it is not used everywhere. But you can be certain, as I have heard tell from [women], no single procedure works [in all situations] because it is necessary to adapt according to the pains and the causes impeding the exit of the foetus.52

In the section of Kranckheiten that follows the passage describing the birthing chair (at which point its readings reconverge with those of the Rosegarten), the German author includes much of Savonarola's additional description of the positions the parturient and her attendants should take as the labour pangs increase; he simply does not structure so much of the action around the chair itself. Comparison of the Kranckheiten and Rosegarten passages with Savonarola thus shows that both versions of the German text are little more than an abbreviated and somewhat rearranged translation of Savonarola's Latin. Rösslin may indeed be offering a local observation of German practices when he exchanges Savonarola's cushions for the sheets of cloth. But nearly every other detail of this birth scene is the same: having the woman sit for an hour; having her hold her breath or shout; having the midwife anoint her hands with lily or almond oil; having her encourage the woman by telling her she will bear a son; breaking the amniotic sac with her nails or a knife or scissors. Instead of jumping on one foot and then the other, the two versions of the German text have the woman go up and down stairs. And all the elaborate supports offered by the other attendants have disappeared. Finally, there is no mention in either version of the German text of the use of a speculum to open the uterine mouth if it does not do so spontaneously, perhaps because no such instrument was then in use in Germany.53 These are the signs of intelligent adaptation, not original composition. The addition of the apparently local reference in Rösslin's Rosegarten, Vnd in hohen teutschen landen, causes the reader to overlook his source of information on practices observed auch in welschen landen.

Causes of Difficult Labour

As noted, the original German author of Kranckheiten was not an unthinking compiler. There are points where he clearly abbreviates and significantly rearranges Savonarola's much longer and more structurally challenging text. For example, in his third chapter he provides an itemized listing of the causes and signs of impending difficult labour. With efficient concision, he lays out eighteen situations that will complicate birth. Thus, for example, if the uterus is small and the woman less than twelve years old; if the opening of the uterus is too small; and so forth. In Savonarola, in contrast, we find a much longer analysis with more dialectical organization.54 There are eight general kinds of factors: those having to do with the pregnant woman herself; with the foetus; or the uterus; or the afterbirth; or an adjacent part of the body; or the time of the birth; or the midwife; or underlying causes. Each of these headings is then broken down further.55 The German author does retain the heuristic model of the enumerated list from Savonarola, but he fashions a much more concise summary. Savonarola had listed those factors of difficult birth attributable to the foetus as (1) it is female (which according to him is harder to give birth to than a male); (2) its head is too big (especially difficult if the woman herself is small); (3) its body is too big; (4) its body is too small (and thus feeble); (5) it has two heads or is otherwise deficient; (6) it is dead; or (7) it is malpositioned in the womb. Savonarola provides a personal anecdote for his second category, describing a one-year-old boy he saw in Padua with an adult-sized head who was mentally deficient, unable to talk, and who eventually died at the age of nine or ten. The German author likewise has seven factors attributable to the foetus, but in the manuscript version of the text Savonarola's anecdote is omitted and there is no local marvel to take its place. In the printed text, Rösslin, in contrast, adds back in his own local “wonder”: the baby born “in this twelfth year [i.e., 1512] in the county of Werdenberg [Württemberg?], with two heads”. Rösslin, as noted above, employs the “baby in a bottle” frame of the Muscian foetal images to present a woodcut of the two-headed baby. Such localized touches have long persuaded scholars that Rösslin was an original writer. What comparison with Savonarola and the Hamburg text shows, rather, is that he had an astute eye for nuanced adaptation.

Menstrual Retention/Lochial Flow

A final example will suffice to show one other element of the German author's compiling technique. Savonarola had structured the gynaecological and obstetrical chapters of his Practica major in the same way most medical encyclopaedists did: he included the diseases and conditions unique to the female as part of his overall a capite ad calcem (head-to-toe) ordering. Thus, women's conditions follow sections on urinary conditions and andrological problems, and precede those of the joints and lower extremities. The author of the German Kranckheiten (and Rösslin following him), in contrast, was not writing a textbook on women's conditions generally but only on obstetrical concerns. Thus, he has no separate chapters on menstrual problems, uterine cancer, etc. But certain of those topics were still pertinent and so he extracted some sections from Savonarola's other chapters and embedded them unobtrusively in the obstetrical chapters. Thus, for example, in his chapter on excessive lochial flow after birth, Savonarola had referred his reader back to his earlier chapter on treating excessive menstrual flow.56 The German author culls from Savonarola's rubric 7 (De fluxu non naturali menstruorum) a ten-item list of causes of excessive menstrual/lochial flow. He then skips over the many symptoms that Savonarola had described, and moves to his injunction that, because the condition becomes increasingly grave the longer it lasts, the physician should be very concerned about it. “Lastly, you should know that honour rarely devolves on physicians from treatment of this, because women are ashamed to reveal this suffering and hence very often it only comes into the hands of the physicians when it has already endured a long time.” 57 The German author, interestingly, turns this warning to the physician into a gentle admonition that “women in distress [should] not have too much shame, but rather should disclose and explain their sufferings to the doctor”, for it is on the basis of his questions and her frank answers that he can determine his advice.58 Then, rather than wrestling with Savonarola's extraordinarily lengthy treatments for menstrual flux (they occupy four and half columns in the printed edition of the Latin), the German author just returns to the few remedies laid out in the original rubric 34 on the lochial flow.

Rethinking the Rosegarten

While several more sections of the German Kranckheiten and Rösslin's Rosegarten can readily be shown to have parallels with Savonarola's Practica (for example, the description of the three “skins” that surround the foetus in utero in Chapter 1, which is a very brief excerpt from Savonarola's rubric 22), the Italian physician's text was clearly not the German author's or Rösslin's only source. The author of Kranckheiten seems to have taken most if not all his references to the Persian authority Avicenna from Savonarola, but all references to the German authority Albertus Magnus are new.59 While so much of the advice in the first half of Chapter 7 on how to conduct a labour comes, as we have seen, directly from Savonarola, there are other passages that may eventually show the German author's originality. For example, the German author includes towards the end of the chapter instructions for repairing an ano-vaginal fistula.60 Part of this section derives ultimately from the Treatments of women of the twelfth-century Salernitan healer Trota, which became the central piece of the so-called Trotula ensemble. Although incorporated into the Compendium medicine (Compendium of medicine) by Gilbertus Anglicus in the mid-thirteenth century and, in somewhat more abbreviated form, into Francesco da Piedemonte's medical textbook, Trota's instruction ultimately had little impact on learned discussions of this condition. Even Savonarola's treatment of the condition was perfunctory at best.61 Yet here in the German we find Trota's treatment, not simply in something close to its full form, but elaborated in such a way as to suggest yet another layer of authorship.

The instruction to sew the tear between the vagina and the anus with four or five stitches with a silk thread comes directly from Trota, but then the German author describes another procedure I have seen nowhere else in a medieval text. He instructs that pieces of linen cloth should be plastered on the two sides of the tear. Then, it is these linen strips, not the woman's own flesh, that is sewn together—for, our German author notes, “as the surgeons should know, thus the lips of the rupture go together”.62 It is very interesting, therefore, that whereas Trota had attributed the affliction of ano-vaginal fistula to the inadequacies of those assisting the parturient, the author of Kranckheiten less harshly credits it to the unfortunate combination of a vagina that is too small and a foetus that is too big. This may suggest that these novel elements come not out of some other textual source but out of the writer's own surgical practice.

Also, as Kruse herself noted, there is still need to work out the relations between the Rosegarten and its earlier printed predecessor, the anonymous Frauenbüchlein of c.1495 (for which Kruse also found a manuscript version that predated the printed text).63 Kruse showed that where the Frauenbüchlein and the Rosegarten overlap in the chapter on prenatal regimen, the former has the fuller readings, implying that Rösslin then incorporated select passages of the Frauenbüchlein into the Rosegarten. Comparison of the passage describing the birthing chair, for example, in Savonarola's Latin text with the three German texts does not easily resolve the lines of affiliation: the text in the Frauenbüchlein (which is essentially identical in both manuscript and print versions here) presents a fuller reading than either the Hamburg text or the printed Rosegarten, though it is not closer to the Latin. Rather, the author/translator is here trying to explain more fully the arrangement of the chair and the sheets draped around it, as if trying to visualize a scene that was less familiar to him than it had been to Savonarola.64 The same is true of a passage recommending the use of apples baked with sugar and the drinking of sweet apple juice as an aid for delivery.65 As with the surgical innovations, it may well be that the four German texts show successive layers of creative translation and other interventions on the core Latin text, demonstrating an effort not only to understand the words but also the social practices being imported into Germany from northern Italy.

Clearly, then, there is need for more systematic comparative study of the Frauenbuüchlein (in both its manuscript and print versions), the Hamburg manuscript of Kranckheiten, and Rösslin's Rosegarten, in comparison to one another and with other gynaecological, pharmaceutical and surgical texts, German and Latin, that may have been available to these German writers. Even without a full understanding of all of the sources used by the still anonymous author of Kranckheiten and by Eucharius Rösslin after him, however, recognition of their reliance on Savonarola's Practica suggests that we should look beyond the young apothecary Eucharius Rösslin to find some other figure who can claim responsibility for all this work of translation and adaptation. We must assume that by the early 1490s at the latest—when Rösslin was beginning his career and before the Hamburg manuscript was written in 1494 and the Frauenbuüchlein was published in c.1495—there was already circulating in southern Germany a fairly comprehensive translation/adaptation of Savonarola's Latin obstetrics. The Hamburg manuscript, the sole extant copy of the Kranckheiten, is unlikely to have been the original copy of the text; as Kruse noted, it is the work of a professional scribe and bears several errors of inattention unlikely to have been allowed to stand had the composing author been supervising.66 We know that the Hamburg manuscript was in the hands of one Constantin Rösslin, apparently an older relative of Eucharius Rösslin, and it stayed in the family for at least two generations.67 We do not know, however, who this Constantin was nor what his competence might have been to have produced a text such as this.

Although we thus cannot rule out Constantin Rösslin from consideration, there are some other likely candidates for the role of translator/compiler of Kranckheiten. Bartholomeus Metlinger, the physician whose paediatric text was absorbed into the Hamburg manuscript and afterwards into the Rosegarten, took his medical doctorate at the University of Bologna in 1470. He resided throughout most of his career in Augsburg, where the Frauenbüchlein was first published c.1495.68 Most interestingly, we know that Metlinger had a pronounced interest in gynaecological matters. He composed a brief text (perhaps for his own wife) for strengthening the womb, with instructions in German keyed to Latin recipes that followed. He also copied with his own hand a series of Latin chapters (apparently drawn from an as yet unidentified Latin compendium) on gynaecological conditions.69 Metlinger therefore merits further examination as a possible collaborator in the translation of Savonarola and in the composition of the Ur-version of the Rosegarten.

Another possible candidate, perhaps even more likely than Metlinger, is Bartholomeus Scherrenmüller (b. c.1450 and active until at least 1493), a physician trained at Erfurt, Tübingen, and probably some northern Italian university. He was associated with the court of Duke Eberhard im Bart in Tübingen. Known to have translated both a regimen from Guglielmo da Saliceto's Latin Summa conversationis et curationis and Pietro d'Argellata's Chirurgia, he also claims to have translated a work entitled Wie sich die kindendenn frawenn in dem geberen der kind halten sollent (How Pregnant Women Should Comport Themselves in Bearing Their Children).70 This work has been assumed to be lost, but as both Kruse's findings and mine suggest, because the origin of the Frauenbüchlein and the Rosegarten's Ur-text must now be situated in at least 1494 if not earlier, we are in fact in need of finding something like Scherrenmüller's text. As its title suggests, it matches the topic of the Savonarola translation perfectly. The text in the Hamburg manuscript may, therefore, be Scherrenmüller's “lost” text.

Whomever we eventually identify as the “author” of the Hamburg text, Kranckheiten, it is clear now that Savonarola, the professor from Padua and Ferrara, and not Muscio, must be recognized as the key source for what would prove to be the foundational text of early modern European obstetrical literature. Remaining puzzles notwithstanding, the present study should establish that Rösslin's Der Swangern Frauwen und Hebammen Rosegarten is neither sui generis nor a work created out of thin air. Nor, aside from some quite novel traditions in pelvic floor surgery and perhaps some individual pharmaceutical therapies, does it reflect a particularly German tradition of women's healthcare. Eucharius Rösslin, “author” of the Rosegarten, can be credited with very little originality; as we have seen, not even such local details as the recommendation of baked apples as an aid for delivery can be attributed to him. His imprint on the text was really that of a salesman, someone who recognized the possibilities of print, both as a mechanism to cheaply reproduce the Muscian images and as a way to curry favour among women who might offer some kind of patronage. Less clear is whether he also had larger “public health” motives in mind, as did Johannes Hartlieb several decades before him, in intending to address an obstetrical text to midwives. Surely the most important finding of this analysis, however, is to show how intimately linked the manuscript culture of the late Middle Ages was with the print culture of the Renaissance. Savonarola's concern that he would not get due credit for his obstetrical writing was justified, since his name is found nowhere in the German Kranckheiten or the Rosegarten, and it has been lost to the historical record until now. Yet in fact, the obstetrical practices of Italy that he described proved the basis not only for German obstetrics, but for developing traditions all over western and central Europe as the Rosegarten was repeatedly translated and republished over the next century and a half.71

Muscio, Gynaecia, Bk II, chap. 18, ¶ 18:	Eucharius Rösslin, Rosegarten, chap. 4:
Et si unum pedem foris eiecerit? iubemus ut numquam eum obstetrix teneat et conetur adducere, ne reliquo corpore infantis plus cludatur. [S]ed prius infixis digitis ad inguina infantis susum eum revocet et post inmissa manu sua pedem alterum corrigat et si fieri potest manus eius lateribus iungat, et adprehensis pedibus foris adducere conetur.29	Wo aber das kind zum ersten käm mit einem füss allein, so soll man die müter do an rucken legen, die bein vber sich, das haupt vnder sich vnd den hindern wol erhäben. Vnd sol die hebame mit ir hand des kindes füss wider hinder sich senfftiklichen schyben. Vnd soll die müter sich züm dicker male vmbschyben vnd voltzen so lang biss das kind sein haupt vndersich gekeret zu dem aussgang. Darnach soll die müter widerumb sitzen auff iren stül vnd sol ir die hebame wider helffen als obstat.30
And if one foot should come out? We command that the midwife should never take hold of it and try to pull it out, lest the rest of the infant's body is stuck inside. Rather, having first fixed her fingers on the groin of the infant, let her reduce it back upwards. And after, putting in her hand, let her correct the other foot. And if possible, let her join its hand to its sides. And grabbing hold of the feet, let her attempt to draw it out.	But when the child comes at first with one foot alone, one should have the woman lie on her back with her legs over her, her head below her, and her hind parts quite elevated. And then the midwife should gently put back the child's foot. And the mother should push herself around and roll around many times until the child has turned his head around toward the exit. Then, the mother should once again sit on her stool and the midwife should assist her again, as [described] above.