A Most Sovereign Herb: Pseudo-Antonius Musa on Betony

This essay studies various important aspects of the history of text of the treatise De herba Vettonica, ultimaty attributed to Antonius Musa, Augustus' physician and the brother of King Juba II. The possible existence of an original Greek text, the relationship between the treatise and the writings of Pliny, and the translation of the treatise into Old English are discussed, among other topics. With respect to this translation, the author insists on its importance for the establishment of the Latin text of the treatise, because it dates certainly before the turn of the millennium, and the majority of Latin manuscripts is from a later period.


Introduction
A number of years ago, sheer ignorance might have led me to suggest that in antiquity, all short works dedicated to a single drug (often styled, and not without reason, Wunderdrogentraktate 3 ) were a Roman invention. Then, earlier this year 4 , Vivian Nutton published the first critical edition of a Greek Methodist treatise on the herb centaury 5 in its Latin version by Niccolò da Reggio; its Greek original, however, seems to be lost. This was reason enough to shake my former conviction 6 .
Surely the most important among these tracts is associated with the herbal attributed to Apuleius (Platonicus) 7 . It extols the properties of the herba Vettonica, betony 8 (Stachys officinalis [formerly Betonica officinalis] or a cognate species 9 ). This is the 3 German for 'treatises on miracle drugs'. See Keil 2007b;Brévart 2008. 4 Writing in the summer of 2015 to meet the deadline. 5 Nutton 2014. Altogether lost is another treatise, see Pietrobelli 2014. Plin. nat. 25.80: Celebrauit et Themiso medicus uulgarem herbam plantaginem tamquam inuentor (discoverer, same meaning for inuenire; for reperire, cf. Plin. nat. 25.33) uolumine de ea edito. (Celebrauit: W. H. S. Jones, in the Loeb, translates "has spread the fame", which I prefer to Tecusan's translation "brought into public knowledge" [fr. 263 in Tecusan 2004]): "The physician Themison has extolled the fame of plantain, a plant that is quite familiar, by writing a tract about it, as if he had been its discoverer" (my translation). Might parts of this have been used for Ps.Apul. herb. 1, which runs to 24 sections? One could perhaps also speculate that other Methodist doctors composed such treatises, companion pieces to De uirtute centaureae. -Plin. nat. 22.53 Anthemis magnis laudibus celebratur ab Asclepi ade ("Chamomile is most highly praised by Asclepiades"; translations of Plin. nat. are those of W. H. S. Jones in the Loeb edition, unless indicated otherwise) need not refer to a specific tract, but such a suggestion seems possible, note celebratur in Plin. nat. 25.90 quoted a moment ago, even if uolumine (this seems to be Pliny's usual way of referring to such a tract, cf. Plin. nat. 20.78 (Chrysippus on cabbage) and 22.136 Tisanae, quae ex hordeo fit, laudes uno uolumine condidit Hippocrates ("To ptisan, which is prepared from barley, Hippocrates devoted a whole volume"); Plin. nat. 25.13 de ea edito is absent here, and Iubae uolumen quoque extat de ea herba [sc. Euphorbea] et clarum praeconium, Plin. nat. 25.78 ("But the treatise also of Juba on this plant is still extant"). A treatise on hellebore by Agathinus is attested in Cael. Aur. acut. 3.16.135 (correct Nutton s.v. Agathinos [Agathinus] in Der neue Pauly [Brill's New Pauly]; in the section of Gal. comp. med. sec. loc. 9.5,13.299 Kühn copied from Andromachus, Nutton may be right in reading Ἀγαθί<ν>ου. Orib. coll. med. 10.7 [not 'Syn.'] is a longer Greek excerpt Ἐκ τῶν Ἀγαθίνου ("from the works of Agathinus"). Note that the recipes in Gal. 13.299 and 13.830 Kühn do not contain hellebore and thus are not fragments of the treatise, as the wording may suggest). Archigenes wrote a whole book on castor: Gal. simpl. 12.337 Kühn. 6 The fullest discussion of Greek parallels for the Latin Epistula de uulture is in Möhler 1990: 48-74. Cf. also Keil 2007a, with the bibliography cited there. The Elder Pliny (Plin. nat. 28.112-118) mentions a special, presumably Greek, treatise (peculiari uolumine) on the chameleon, attributed to Democritus (which is lost), and gives a summary of the contents, limb by limb (per singula membra), possibly resembling the Epistula de uulture. 7 Those who are less familiar with these medieval works on herbal medicine will find pertinent information in Collins 2000. There are short notices on medical works in Latin from antiquity and the early middle ages in . Abbreviations of the titles of works in Latin follow the usage of the Thesaurus linguae Latinae. 8 Rufinus (Thorndike 1949: 320) explains: Vetonica. dictum est sufficienter de ea in littera B, capitulo Betonica, sed aliqui mutant B in V dicentes vetonica ("Vetonica. Enough has been said about betony in section B, chapter Betonica, but some people change b for v and call it vetonica"). Bibliographical details for the German translations published in three dissertations for obtaining the degree of Dr. med. at Würzburg University, supervised (if this is the right word) by Gundolf Keil, can be found in the bibliography of Keil 2007b. Keil 1997 is not reliable in many of the details he presents. (The article was also printed in full, but with some errors, in Henning 1998: 32-34.) German versions of the Herbariencorpus were produced only in the 15th century. Arthur Groos (Cornell University) and Bernhard Schnell (Göttingen, Germany) are currently working on a critical edition of the only illustrated translation into German. To date, Schnell has identified some 80 witnesses (for a new list of the mss., see now Schnell & Groos 2018: 24-39). 9 Verhoeven 2011. Heinrich Marzell (an eminent botanist) vehemently doubts (Marzell 1927(Marzell : 1180 whether the bettonica of the ancients is the same as ours: "Ob die vettonica (betonica) der Antike wirklich unsere Art ist, bleibt sehr zweifelhaft. Vielleicht ist darunter ein verwandter Lippenblütler (Stachys alopecurus?) zu verstehen" ("Much doubt remains whether the vettonica [betonica] of the ancients really is the same as our species. It could work I wish to explore a little further here 10 , discussing some aspects of its transmission in Latin and Old English, to make clear how many problems will have to be addressed before a new edition is undertaken. Knowledge of simples 11 surely had a long tradition before such expertise and experience found its way into written tracts; these must have been aimed at users lacking a direct link with the herbarii ('herbalists') or ῥιζοτόμοι (literally 'root-cutters'), persons (of either sex) who might well in the majority have been illiterate 12 , at least in the sense that they would have lacked the skill to compose a treatise. And for them, there would have been no such need.
As historians we are interested in when such works were first composed and what the written evidence at our disposal can tell us. De herba Vettonica has the advantage of being present in a considerable number of manuscripts as well as being short and therefore easy to manage, with just under fifty single paragraphs or chapters, each usually corresponding to one recipe. The question whether it was from the beginning part of the early Pseudo-Apuleian corpus not preserved in manuscript will be left aside here, but those unfamiliar with this herbal should know that De herba Vettonica is set apart from the chapters in the herbal of Pseudo-Apuleius by having a considerably greater number of recipes for a single item -in Pseudo-Apuleius, few plants run to more than half a dozen, and there are even chapters with just one single recipe. The other difference is that De herba Vettonica, in most cases, specifies the exact amount of the plant to be used and the liquid it is to be taken with. Whether it provided a sort of nucleus for the formation of the Herbariencorpus remains a matter for speculation. As we see in the case of the treatise on centaury mentioned earlier, works of this genre were in circulation. No need to trust the introductory letters which want us to believe that Antonius Musa 13 , Augustus' physician, is here sharing be that a labiate related to it, i.e. perhaps Stachys alopecurus, is meant"). Because in Ackermann's edition of the Herbariencorpus, the first chapter of Ps.Apul. herb. is De herba Vett., Marzell (1927Marzell ( : 1181 Howald & Sigerist 1927: XIV (the date there, "saec. XI", must be an error; first half of the 9th century is the traditional date) but seems to have been overlooked by many scholars interested in Ps.Apul. herb., perhaps because Howald & Sigerist say that in their abbreviated form these excerpts are useless ad textum genuinum restituendum ("for establishing the original wording"). 11 Scribon. Larg. praef. 15: dantes operam, ut simplicia prima ponamus: interdum enim haec efficaciora sunt quam ex pluribus composita medicamenta ("We strove to give pride of place to simples, which at times work better than medicines made up from many ingredients"). 12 As Laurence Totelin reminds me (per litteras), Crateuas may be seen as the exception; Pliny, however, lists him as a medicus. 13 Musa was the brother of king Juba II, who had written Περὶ εὐφορβίου (Plin. nat. 25.77).
his knowledge with M. Agrippa, or that Apuleius (allegedly identical with the author of philosophical works, which explains the epithet "Platonicus", and of the Golden Ass) is writing for the benefit of his dear fellow-citizens, transmitting to them a work put into his own hands by Chiron, the teacher of Achilles, and by Aesculapius, an assertion that raises a few problems, sheer chronology being one of them.
So did the compilers of De herba Vettonica and Pseudo-Apuleius walk hill and dale 14 to collect recipes straight from the horse's mouth, the horse being in this case an old wifie? This claim is not even made 15 , and it may well be that the sources for these compilations were, in the majority at least, already in written form. The most complete among them must have been, in the fourth century AD (a possible date for the composition of De herba Vettonica), Pliny's Natural History, and it is perhaps no coincidence that a work based on herbal and animal drugs listed in Pliny, the Medici na Plinii (Plin. med.) 16 , was composed then, by the way without any acknowledgement to Pliny the Elder. Rather, its author poses as Pliny himself, and his preface with its attacks on doctors just intent on material gain is quite similar in tone to that of Pseudo-Apuleius.
It is all the more surprising that the three books of the Medicina Plinii contain only one recipe featuring uettonica, Plin. med. 3.37.7b, uettonica herba trita morsui imponitur ("ground betony is put on the bite"). The source must be (as indicated by Önnerfors) a passage in Pliny the Elder's Natural History 25.101: morsibus inponi tur Vettonica praecipue, cui uis tanta perhibetur ut inclusae circulo eius serpentes ipsae sese interimant flagellando 17 ("and to the bites is applied in particular betony, the power of which is said to be so great that snakes enclosed in a circle of it lash themselves to death").
A few lines earlier (Plin. nat. 25.99), we had read that snake-bite is cured by Britan nica herba (Rumex aquaticus L. according to Jacques André), a plant that Dioscorides treats in the chapter following his account of the Vettonica (κέστρον in Greek, mat. med. 4.1; βρεττανική mat. med. 4.2) 18 -for this reason and because the words are fairly similar, it is hardly surprising that many people confused both plants (the version RV of the Greek Dioscorides says βεττονική: Ῥωμαῖοι βεττόνικαμ 19 ), whose properties were seen as similar 20 . Let us now look at some evidence that could link De herba Vettonica with a possible Greek original, much like the case of De uirtute centaureae. 14 Plin. nat. 26.11 sedere namque in scholis auditioni operatos gratius erat quam ire per solitudines et quaerere herbas alias aliis diebus anni ("For it was more pleasant to sit in a lecture-room engaged in listening, than to go out into the wilds and search for the various plants at their proper season of the year"). 15 Contrast the muliercula quaedam ex Africa ("a certain wifie from the province of Africa") in Scrib. Larg. 122. 16 Cf. Fischer & Kudlien 1993. Add to the bibliography given there Brodersen 2015 andHunt 2020. 17 J. André, commenting on the passage in his Budé edition, also refers to Ser. med. 841, although there betony is taken by mouth, as in Ps.Musa herb. Vett. 42 (not noted in Howald & Sigerist), and other instances. The story about the snakes is also repeated, from Pliny, by Macer Floridus 483-485. 18 Riddle 1983 did not comment specifically on betony, but see his ch. 3 "Drug affinities" (pp. 94 ff.), for Dioscorides' arrangement of chapters. Pelagon. 367 has Vettonicae in the Bobbio palimpsest and brettonicae in Poliziano's 15th-century copy. 19 Cf. also Paul. Aeg. 7.3 p. 200, 19-27 Heiberg, partly quoted in Wellmann's apparatus on Diosc. mat. med. 4.1 p. 167. Heiberg's note "cf. Galen. XII 23" suggests that Galen, On simples has a similar text, which is not the case. Aëtius solves the problem in a different way: 1.72 is the Βετονίκη (i.e. Britannica), 1.196 κέστρον (i.e. Vettonica). Bonet 1991 discusses the various names of betony. 20 Paul of Aegina commented μηδὲν ὅμοιον ἔχουσα τῇ προειρημένῃ πλὴν τῆς ἐνεργείας ("not resembling the plant mentioned earlier except for its effect").
The Greek may well be a version of the Latin above, or of a very similar text, and not the other way round, as we might be inclined to assume at first 25 . The beginning seems to lack a Greek word for 'growing' (corresponding to nas citur); εἰς for in is non-classical, ἡμέρους for opacis 'shady' strange, and even 21 This is not the only link between Dioscorides and the Herbariencorpus; cf. Diosc. mat. med. 3.4.4 (app. crit. on p. 8 Wellmann) and Ps.Apul. herb. 19, both on ἀριστολοχεία 'birthwort', an interpolation in Diosc., starting, as the Latin text does (Howald & Sigerist always print it at the end of a chapter without giving a reason), with the nomina herbae 'names of the plant'. Ps.Apul. herb. 19.2 ad febres acerrimas ("for very hot fevers") is rendered as πρὸς πυρετὸν βαρύν ("for heavy fever"), a wording I do not recall having come across in any other Greek work. 22 Galen. alfab. 290 (Everett 2012: 370) nascitur in pratis et montibus ("grows on meadows and on mountains"). I cannot say if this is a sheer coincidence. Thorndike Cf. Madaus 1938Madaus : 2598Madaus -2602, at 2600: "Die sehr veränderliche Pflanze [Stachys officinalis] wächst auf mäßig trockenen bis nassen Magerwiesen und in lichten Gebüschen namentlich der montanen Stufe" ("Stachys of ficinalis is a rather changeable plant and grows on poor meadows that are not too dry but may be wet and in shrubbery that is not too dense, especially in higher regions"). A German translation for the Latin can be found in Niederer 2005: 67, rendering Sang. 217, p. 309. 25 Howald & Sigerist, on Ps.Apul. herb. 31.1, claim to have spotted another interpolation in Dioscorides (mss. RV) based on the Herbariencorpus. Their reference to II 246, 6sqq. should read I 246, 6-14 Wellmann and concerns Diosc. mat. med. 2.177, but the only parallel I can see in this chapter is Ps.Apul. herb. 31.5 and p. 246, 13-14 Wellm. ἁρμόζει καὶ θηριοδήκτοις σὺν οἴνῳ πινομένη ("it is also good for bites of poisonous animals drunk with wine"), which is not particularly close, replacing the snakes (Ad morsum serpentis "for snake-bite") with poisonous animals in general. stranger is γεννήματα rendering frutices 'shrubs'; frutices meaning fructus (the sense needed for γεννήματα) does not seem to be attested 26 , and my guess would be that the Greek translator did not understand what was meant by frutices and also lacked the necessary botanical knowledge. The reference must be to the shrubbery (more usually frutectum, rarely fruticetum in Latin), found especially on mountain slopes (macchia in Italian), as is suggested by Varro rust. 2.1.16 in montuosis potius locis [quam del. Keil] fruticibus quam in herbidis campis ("more appropriate for grazing goats are hilly places, shrubs, rather than grassy fields"), τόπους ἐπιβλαβεῖς ("dangerous places"?) for loca sancta seems to be a strange turn; loca sancta may refer to "forsaken places", dangerous because in the vicinity of tombs (busta), daemons and spirits dwell and attack passers-by. The Greek has nothing that corresponds to uisus timendos et omnes res sanctas ("frightening apparitions and everything that is numinous") unless one wants to equate bad dreams with uisus timendos.
It seems that ὀρθοκάλαμος 'with an upright stalk' occurs only here, and as a feminine, while the dictionaries list it as a masculine (as we would expect); LSJ translates 'upright stalk.' A very similar word, but an adjective, ὀρθόκαυλος 'with an upright stalk', is attested in Theophrastus 27 and Galen 28 (quoting Theophrastus), and I would rather be inclined to read ὀρθόκαυλον here. If it is indeed an adjective, does it go with a noun? Does εἰς τὴν hide a Latin (h)astam (asta habens tenue longa unius cubiti et quadra, "it has a slender stalk, one cubit long and with four sides"), which is the translation used in the Dioscorides Longobardus 29 (p. 9 Stadler) for καυλὸν ἔχουσα λεπτόν, πήχεως τὸ ὕψος ἢ καὶ μείζων, τετράγωνον ("it has a slender stalk, one cubit long, or even longer, with four sides"), while another Latin version of which we have a fragment right at the end of Ps.  translates tyrso [sic] tenui, ultra cubitum, quadrangulo 30 ("with a slender stalk, more than a cubit long, with four sides")? τριγώνια ('with three angles') cannot be right, also because the adjective would have to be τρίγωνος rather than τριγώνιος. It is surprising that Wellmann printed this text without drawing attention to these problems. The last section talks about the properties and uses of betony: a. ἡ δὲ δύναμις αὐτῆς ἐστιν αὕτη· λειωθεῖσα γὰρ νεαρὰ καὶ ἐπιτεθεῖσα εἰς τεθραυμένην κεφαλὴν κατὰ τῆς πληγῆς ἀνώδυνον ποιεῖ καὶ τὰ τραύματα κολλᾷ καὶ τὰ κεκλασμένα ὀστᾶ ἐκβάλλει καὶ τοῦτο ποιεῖ καθ᾽ ἡμέραν ἀλλασσομένη, ἕως θεραπεύσει ("Its effect is as follows: the fresh plant, ground and put on the wound of the head soothes the pain, closes the wound and extracts the broken pieces of bone; do this every day until the wound has healed").
b. ἀφεψηθεῖσα δὲ μεθ᾽ ὕδατος κεφαλαλγίαν καταντλουμένη καὶ μετὰ ἀσφάλτου περιχριομένη τοῖς κροτάφοις ἰᾶται, ὑποθυμιωμένης αὐτοῖς καὶ τῆς ῥίζης ("Boil it in water and bathe the head with it for headache, and smeared on the temples with bitumen it cures the headache, also when the root is used in fumigation").
Ps.Musa herb. Vett. 1 (l. 35-40): Ad capitis fracturam Herba uettonica contusa et super capitis ictum inposita uulnus mira celeritate glutinatum sanabit; eo quidem efficacius, si tertio quoque die refectam, id est recentiorem, frequentius inposueris, donec sanet. Eius potestas tantam habere fertur utilitatem, ut ossa quoque fracta ui sua extrahat 31 ("For a fracture of the head: Betony crushed and applied to the place where the blow landed on the head will close and heal the wound with amazing speed; it will be all the more efficacious if you apply the poultice on every third day, i.e. renewed, quite often until it has healed. It is believed that its efficacy is so great that by its power it will bring even splinters of bone to the surface").
The first part of the Greek (a.) is sufficiently close to Ps.Musa, although the claim that it dulls the pain (ἀνώδυνον) is not made in other sources I have been able to examine, and τεθραυμένην κεφαλὴν ('head-wound') does not necessarily suggest that the patient has suffered a fracture. In Greek, the dressing with betony leaves is changed daily, the Latin says "every other day" (day 1 is today, in ancient sources). Nothing corresponds to (b.).

De herba Vettonica in Old English
After this look at a Greek parallel for some parts of the Latin De herba Vettonica, we turn to an even more exotic tongue, Old English (OE). Medical texts were translated into Old English long before we encounter translations into other languages, certainly before the turn of the millennium, at a time for which just a few scraps of medical recipes survive in Old High German. It need not be stressed how important this transmission in Old English is because it allows us to control the Latin originals 32 , the Herbariencorpus in this case, since the majority of Latin manuscripts is from a later period. 31 The OE translation is very different and has the patient drink an amount of ground betony (leaves, presumably) in warm beer (þíge hit þonne on hatum beore).

32
Cf. Adams and Deegan 1992. When only part of the recipes appear in the OE version 33 , as is the case with De herba Vettonica (28 out of 47), we would like to know why this is so: absent are 8 (Ad uomitus et suspiriosos et toracis dolorem, "for vomiting, difficult breathing, and chest pain"), 9 (Ad tisicos et qui purulentum eiciunt, "for pulmonary disease and bringing up pus"), 10 (Ad stomachi dolorem, "for stomach pain"), 11 (Ad iocineris dolorem, "for liver pain"), 12 (Ad lienosos id est spleneticos, "for complaints of the spleen, i.e. splenetics"), 13 (Ad renum dolorem, "for kidney pain"), 18 (Ad colum, "for attacks of colic"), 19 (Ad tussim, "for coughing"), 20 (Ad cotidianas, "for quotidian fever"), 21 (Ad tertianas, "for the kind of malaria called tertian fever"), 22 (Ad quartanas, "for the kind of malaria called quartan fever"), 24 (Ad cauculosos, "for stones in the bladder"), 25 (Ad idropicos, "for dropsy"), 26 (Ad mulieres, quae a par tu laborant, "for women's complaints after giving birth"), 27 (Ad paralisin, "when a patient is paralyzed"), 28 (Ad horrores, "for attacks of shivering in fevers"), 29 (Ad mulieres locosas, "for female patients with private parts that are very spacious"), 32 (Euersis de uehiculo, "for persons thrown off a carriage"), 33 (Ictericis, "for jaundice"), 34a (Ad carbunculum, "for heartburn"), 35 (Qui perfrictionibus laborant, "for patients suffering from cold"). There are unbroken sequences 8-13, 18-22, 24-29, and 32-35. I can find nothing specific in the content of the passages that were omitted that would indicate a reason for an omission and am rather inclined to think of a mechanical problem, i.e. the loss of pages in the course of transmission. Now for some other oddities.While the sequence in the OE and Latin is the same, there are problems with two recipes in the OE, Ps.Musa herb. Vett. 34 and 35 (numbers 15 and 16 in OE). The editor in the EETS Series, de Vriend, does not comment. Ps.Musa herb. Vett. 34 Ad carbunculum comprises two different recipes (the second introduced by Item [although Howald and Sigerist print Idem 'The same', as they do in Ps.Musa herb. Vett. 42] 34 ), for two different kinds of carbunculus: but 34a is for heartburn 35 (missing in OE), while 34b (= OE 15) is for a boil (OE spring 36 )! The OE recipe is much more specific than the Latin text in Howald & Sigerist and calls for 1 dram of betony and old grease to be put on the boil, while the Latin says Vettonicam cum axungia tritam plagae inponat ("he should place betony ground with hog's fat on the wound"). If the weight of 1 dram and old grease rather than grease are not additions of the OE translator (and I see no compelling reason to assume this), the Latin text at his disposition was fuller as well as different from the codices used by Howald and Sigerist. We can, however, check against a manuscript not identified earlier 37  still quite old 38 , 96 (T. 4. 13) in the Hunterian Collection at Glasgow University (CLA 196,fol. 40r): item bittonica cum axundia [sic] trita inponitur et sanabitur ("Likewise, betony ground with hog's fat is placed (upon the wound), and it will heal").
Ps.Musa herb. Vett. 36 Lassis de uia ("for weary travellers") must roughly be the same as OE 17; for one line in Latin, there are three and a half in OE. Discrepancies between both versions are seoð on geswettum wine ("boil in sweetened wine", unspecified amount) for ex oximelli (a mixture of vinegar and honey 39 ) ciatis III, the (Old) English patient tired from a long ride or walk should drink his medicine at night and on an empty stomach, þonne bið he sona unwerig ("then he will quickly be restored = not tired"). This last part is likewise absent from the Latin. I am surprised that these differences between the OE translation and the Latin prototype apparently have not been remarked on so far.
Not that our OE version is always superior. In Ps.Musa herb. Vett. 40 40 Ad ueretri tumorem uel dolorem ("for a swollen or painful penis"), the translator either read uentris 41 'belly' or confused uentris 'belly' and ueretri 'penis' and wrote (in OE 21) wið innoþes sare ("for pain of the entrails"). He goes on to say lege þonne abutan þa wambe 7 þyge hy; þonne eac hraðe cymeþ þaet to bote. ("put it then on the belly and accept it 42 , then the recovery will also be quick"). Did somebody just invent this?
Another case for close comparison is furnished by the preceding recipe, Ps.Musa herb. Vett. 39 = OE 20. The Old English patient gets not just hot water to drink with his pill, but hatum waetere 7 on wine tósomne ("hot water and wine together"), three cups of it instead of two (although one of the older Latin manuscripts, L, also has three), but the amount of honey -1 oz. in the Latin text -is not given.
Sang Four pills, four days or mornings -this makes sense, and I should say better sense than three pills and three days, the reading in branch β of the Latin text 45 (see the app. crit. in Howald and Sigerist). 38 Bischoff 1998: Nr. 1396, "Wahrscheinlich Narbonensis, VIII./IX. Jh.". 39 So the translator may have read, or understood, ex (o)enomelli "with a mixture of wine and honey" (= ἐν οἰνομέλιτι, "in a mixture of wine and honey"). 40 Niedermann and Liechtenhan mention this as a parallel in Marcell. med. 33.32 CML V p. 564;I cannot understand why. 41 Like Sang. 751 p. 408 Ad uentre tumor. Herba uittonica tere et inpone desuper, "for a swollen belly. Grind betony and put it on top". The relationship between what I take to be excerpts from Ps.Musa herb. Vett. in Sang. 751, especially on p. 408 and 409, and Ps.Musa remains to be examined. More usual is the confusion of uentrem 'belly' (pronounced /uenre/) and uenerem 'Venus (the goddess of love, also used for sexual activities in general)' (/uenɘre/). 42 opponito in the Latin would not mean 'eat', but again 'put on.' 43 All the Saint Gall manuscripts mentioned in this article are online at e-codices.ch, where you can also find a description of the contents, date (all 9th century), and bibliography; see also the entries in Bischoff 2014. 44 Normally, mane is indeclinable; this form of the plural is not listed in the Thesaurus. It also occurs e.g. in Sang. 751 p. 407 per sex manes and in the same recipe (Ps. Musa herb. Vett. 39)  An editor of Ps.Musa herb. Vett. would of course have to answer the question whether the text in Sang. 751 descends directly from the version he wants to edit, or has been subject to more editing in its own turn.
The same recipe (Ps.Musa herb. Vett. 39) also occurs in two collections, Recept. Sang. II (Sang. 44, p. 53 Jörimann 46 ), and Marcellus. The text in Marcellus, a lay person who compiled a collection of recipes "for his sons" (number unspecified; the Latin does not allow us to be sure whether daughters were included) in the early years of the 5th century, is palpably different and has the following (Marcell. med. 20.73 CML V p. 348 Niedermann and Liechtenhan): Mellis Attici p. I, uettonicae tritae -IIII in unum permisce et pone, ut conferueat. Quod cum refrigerauerit, facies pastillos nucis auellanae magnitudine et post cenam singulos hauries, quo facto prime quoslibet concoques cibos ("Honey from Attica, 1 pound; ground betony, 4 oz., mix together and let it boil together. When it has cooled down, make it into pills the size of a hazelnut and take one at a time after your evening meal 47 ; this way, you will wonderfully digest whatever you eat").
In Marcellus, honey and betony are boiled together (Ps.Musa had said "boiled honey"), and the amounts have increased considerably: 1 pound of honey, four ounces instead of four drams of betony. The size of the pills is mentioned (same as a filbert), and you are to take them post cenam 48 , nothing being said about a drink to wash them down with. It is obvious that Marcell. med. 20.73 and Ps.Musa herb. Vett. 39 49 in Latin, OE and in the recipe collections Recept. Sang. II and in Sang. 751 are linked, but through a common ancestor. Only a pharmacopola 50 'drug-seller' would have prepared a batch with one pound (12 or perhaps even 16 oz.) of honey 51 , and I suspect that these weights reflect an error in the transmission (honey as the first item in a recipe must also be rather rare).
It is worth having a closer look at the language of this recipe in Marcellus. It contains words that must be classified as rare: prime 'excellently', conferueat 'boil together', hauries 'you will swallow'. The Thesaurus has, s.v. 1. conferveo, only two attestations, Vitruvius and Palladius; since it is only a variant reading in Vitruvius (7.14.1), the verb does not even figure in the OLD. hauries occurs only once in Marcellus and is not used for swallowing a pill in Celsus, Theodorus Priscianus, or Cassius Felix; in Caelius Aurelianus chron. 1.1.29, it probably refers to the liquid, as it does in chron. 3.2.24. Similarly rare is prime; in Marcell. med., we have it at 9.27, 15.68, 34.21 and in the passage cited above. It also occurs in Ps.Hippocr. ad Antioch. 1 vers. α (rerum mathe 46 The more complete version of this collection in Vat. Palat. Lat. 1088 fol. 31r-50r (online) was not known to Jörimann. It was brought to my attention by Dr. J. Staub, of Mittellateinisches Wörterbuch in Munich. 47 Jutta Kollesch and Diethard Nickel, who provided the German translation, say "nach dem Mittagessen" ("after lunch"). 48 The Testimonialapparat ad loc. leaves us in the dark about discrepancies between Marcell. med. and Ps.Musa herb. Vett. 49 Last recipe of Ps. Musa herb. Vett. in Hunter. 96 fol. 40r. 50 See Totelin 2016. 51 Or was the Greek ὁλκή (meaning 'weight' but also 'dram'), and this was misinterpreted as pondus rather than as drachma / denarius? A similar case seems to occur in Plin. nat. 26.33 Vettonicae tusae pondo libra, mellis Attici semuncia ex aqua calida cotidie bibentibus, "ground betony, one weight/pound/dram, honey from Attica, half an ounce, taken with warm water every day" (my transl.). See above Marcell. med. 20.73 for a possible parallel mistake concerning the weight. sim quoque prime adeptus es, "you are very knowledgeable in every field") and in the Preface (epistula) of Scribonius Largus attributed in Marcellus to Cornelius Celsus, 15: prime dantes operam ("trying very hard"); but here, prime is Heraeus' conjecture! How do we explain the odd phrasing? Was stylistically ambitious Latin transformed into a more mundane form, in Ps.Musa and similar compilations of recipes, or was a simple and more straightforward text embellished? The latter seems much more likely. In the present case, I do not think that Marcellus himself was responsible for the lexical facelift; if that were the case, we would expect to find many more of examples in his work. Nobody has compared Marcellus' sources systematically with his version in De medicamentis, admittedly no easy task because of the scarce manuscript evidence for De medicamentis -just two manuscripts, even if they are 9th century -and for those of his sources with an independent transmission: a manuscript, or rather: the only manuscript of Scribonius Largus, from the early 16th century, was discovered less than forty years ago, Howald and Sigerist's edition of the Herbariencorpus cannot be relied on for subtle textual differences, and it remains to be seen if Pliny, the Medicina Plinii and its later version called Physica Plinii really could provide a firm foundation for such studies.
I would like to illustrate such minor editing activity with the example of Plin. med. 3.37.6, two recipes for snakebite 52 : (a) hyssopi semen bibitur 53 ("hyssop 54 seed is drunk"). (b) eius qui percussus est uertex inciditur eoque additur euphorbium: medetur quacumque parte corporis periculum est ("The crown of the patient's head is incised and milkwort [Euphorbia resinifera Berg] rubbed in. This will work for whichever part of the body where there is danger").
The source given for (a) is Plin. nat. 25.136: Putant et (sc. hysopum) serpentium ictibus aduersari, tritum cum melle et sale et cumino ("Pounded with honey, salt, and cumin it is also supposed to counteract the poison of snake bites"). Jacques André in his Budé edition does not comment on the differences, which are far from minor: in Pliny, nothing is said about whether the plant (probably the leaves) or the seeds are used (as specified in (a)), honey, salt, and cumin seed are totally lacking in (a). Can it be called the same recipe?
Now for (b), whose source is Plin. nat. 25.78: Contra serpentes medetur quacumque parte percussa uertice inciso et medicamento (sc. euphorbio) addito ibi ("In whatever part of the body the bite may be, an incision is made in the top of the skull and the medicament inserted there").
Here, we have some moderate rewriting, but no difference in the content 55 .

Pliny and De herba Vettonica
Pliny has many recipes containing betony, although not quite as many as De herba Vettonica, and we should expect a significant number of more or less identical recipes in both works. One look at the Testimonialapparat in Howald and Sigerist's edition shows that this is not the case. Still more surprising is that the Medicina Plinii, which derives almost wholly from Pliny's Natural History, has one single instance where uettonica appears, for snakebite (Plin. med. 3.37.7): uettonica herba trita morsui imponitur ("ground betony is put on the bite"). It comes from Pliny, Natural History 25.101: morsibus inponitur Vettonica praecipue 56 , cui uis tanta perhibetur, ut inclusae circulo eius serpentes ipsae sese interimant flagellando. datur ad ictus semen eius denarii pondere cum III cyathis uini 57 uel farina drachmis III sextario aquae -farina et <in>ponitur -("to the bites is applied in particular betony, the power of which is said to be so great that snakes enclosed in a circle of it lash themselves to death. For the bites is given its seed, the dose being a denarius with three cyathi of wine, or else it is ground and three drachmae of the powder are given in a sextarius of water; the powder is also applied locally").
Paragraph 42 may correspond more or less to what Pliny says, but paragraph 43 is less straightforward, and indeed, for neither of the recipes do Howald and Sigerist offer parallels or sources. tritum 'ground' in paragraph 43 suggests that semen 'seed' has dropped out. With weights and measures, we need not quibble too much, since this was a problem well known in antiquity, and doctors who insisted on precision preferred writing these in words (ὁλογραμμάτως 62 'written in full') rather than using special characters. If a mixture of crushed plant and dark wine were to be spread on the bite, why specify the amounts with such precision 63 ? The part touching the bite itself would hardly take 3 cyathi of wine! This, however, is how the Old English translator took it: 24. Ef(t w)ið naedran slite genim þaere ylcan wyrte a(ne) tr(ym)esan g(e)waege, gecnid on r(e)ad w(ín), gedo þonne ðaet þaes wines syn þre(o) ful fulle, smyre ðonne mid (þ)am wyrtum ða wunde 7 mid þy wíne, þo(nne) byð hio sona hal ("Another recipe for snake bite: Take the same plant, 1 dram, grind it with red wine, then take three cupfuls of this wine and anoint the wound with this plant and the wine, then will it soon be cured").
Our next example is Pliny's Natural History 25.127, a remedy for poisoning: Vettonicae semen in mulso aut passo uel farina drachma 64 in uini ueteris cyathis quattuor; uomere cogendi atque iterum bibere ("... the seed of betony taken in honey wine or in raisin wine, or drachma doses of the powder may be taken in four cyathi of old wine; but the patients must be made to vomit and take a second draught").
The text in Ps.Musa (herb. Vett. 41) seems to be quite close: Ad uenenum qui sumpserit. Vettonicae dragmas III ex uino ciatis IV, statim dato, bibat, reiciet uenenum ("For people who have ingested a poison. Give at once three drams of betony with four cyathi of wine, let him drink this, he will bring up the poison").
There are two points to consider here: statim dato 65 "give it at once", and the promise that this will make the patient spew out the poison (þonne aspiweð he þaet attor 66 ), whereas Pliny says that the patients should be forced to vomit before drinking the mixture again 67 . Pliny also wants old wine; did ueteris then drop out before cyathis in Pseudo-Antonius Musa The St Gall botanicus prefers the oral route (p. 70): 24 Ad serpentum morsum. Herba uitonica dragmas VI (III ante corr.), uino nigro ciatus III bibas, miraueris ("For snake bite. Betony, 6 drams, dark wine, 3 cyathi, drink this and you will be amazed"). 64 This is the text printed by André; I wonder if farina drachma can stand, and would opt for farinae drachma. 65 There is the remote possibility that dato should be read as an ablative instead of an imperative: "as soon as it (the poison) has been given", the meaning of the passage would not be affected. 66 The boiling (wylle tosomne, "boil together") is only in the OE. 67 Recept. Sang. II p. 53 has tra<h>itur uenenum, "the poison is drawn out".

Betony recipes similar to those in De herba Vettonica
We started with the question how recipes for tracts like De herba Vettonica or the individual chapters in Pseudo-Apuleius, herbarius were collected. When one looks at the 47 sections of De herba Vettonica plus the three additional recipes in the Appendix (p. 287 Howald & Sigerist), one might be under the impression that they represent most of the material on betony current and known in antiquity. A look especially at some collections of recipes in manuscripts (not yet published), however, shows that this is not the case. It is certainly much easier to check De herba Vettonica against Pliny's Natural history. As we noted above, the overlap of the two works is only partial: Pliny has a number of recipes not in De herba Vettonica, and vice versa. This can easily be verified (and the same applies to Marcellus, De medicamentis), leading me to present now some of this unpublished material 68 . We start with Sang. 751 (p. 380), a recipe for earache: CXII Item: uettonicę sucus cum lanae subela 69 in aure stilla et si uermis fuerit cadit ("112 Likewise: Let juice of betony with wool fat drip into the ear and if there is a worm, it will fall out").
Pliny  gives recipes for betony to be used on epileptics 70 , different from what we read in Vat. Reg. lat. 1004 fol. 75 r . It comes from Book 2 of Pseudo-Petroncellus (also referred to as Tereoperica) 71 : Item. betonice puluis cum melle datur. sanat.
On p. 166a, we read a recipe Ad dolorem et tumorem testium: Item: bettonicę folia siccata in umbra in puluerem redacta ϶ II. in aqua tepide cyatum II. potui data [ad] dolorem sanabunt ("Likewise: Betony leaves dried in the shade and pulverised, two scruples, in two cyathi of luke-warm water, given to drink will heal the pain").
All these recipes would have been perfectly at home in either Pliny or De herba Vettonica, and they must come from recipe collections featuring simples unknown 68 Plin. phys. Bamb. 42.3;62.4;65.5;65.18;82.42;86.6. 69 I have no idea how this should be corrected or interpreted. I translate cum lana sucida. 70 Diosc. mat. med. 4.1.3 θεραπεύει δὲ καὶ ἐπιληπτικοὺς καὶ μαινομένους μεθ' ὕδατος πινομένη, "drunk with water, it also cures patients suffering from epilepsy and mania". 71 De Renzi 1856 published only a few chapters from this second book, pp. 287-90. López Figueroa 2011 (online) publishes only book one and does not discuss whether there was an (original) book 2. What de Renzi 1856 counts as book 3 is Ps.Democritus, Liber medicinalis, see Fischer 1994. In general, see also Fischer 2013 for the way these compilations were made and structured. For book 2, unpublished, we can only advance the dates of the manuscripts as a terminus ante quem, e.g. London, British Library, Sloane 2839, probably written in England, late 11th or early 12th century. to us 73 . The genre itself is attested, for instance, in Pseudo-Dioscorides, De her bis femininis (edited by Kästner 74 ), in the St Gall botanicus (edited by E. Landgraf and recently by M. Niederer), and in the herbarius Parisinus (edited by A. Ferraces Rodríguez 2012). While De herbis femininis has Dioscorides as its base, Pseudo-Apuleius (including De herba Vettonica) provides the foundation and framework for the St Gall botanicus and the Paris herbal, and additions from other sources, known and unknown, are grouped around them. Also highly significant as potential sources of more plant recipes are the pseudo-Hippocratic Dynamidia and the Curae herba rum 75 , both last examined in detail by Ferraces Rodríguez 1999 76 .
All of the examples discussed above show that a new edition of De herba Vet tonica will not just have to take into account more than the only seven manuscripts used by Howald and Sigerist (and three of these, L, Vr, and C, lack various parts of the text!) but cast its net much wider and address other questions of a more general character as well. Accordingly, future editors of any of these texts might well sigh, with "Nanki-Poo, Here's a pretty mess!". This notwithstanding, entertainment will certainly be guaranteed.