The comparative milk-suckling reptile

ABSTRACT Cross-cultural folk beliefs about milk-suckling or milk-drinking amphibians and reptiles have long been noted by scholars. European dialectal folklore, for instance, has countless instances of cow-suckling and milk-stealing animals including butterflies, reptiles, batrachians, hares, hedgehogs and nocturnal birds. These creatures are regularly said to sneak into the domestic space at night to suck life-giving milk or blood from cattle and women. The early documentary evidence for this set of ideas, which relies, in great part, on the motif of breasts or udders suckled by a snake or similar animals such as toads or lizards, has not yet received the study it so richly deserves. Ideally, a comparative study of the milk-suckling reptile (both animal-human and animal-animal) would be carried out across the full gamut of relevant disciplines including ethnology, linguistics, philology, folklore and historicalreligious studies: this would naturally include pre-modern written references and an analysis of their transmission. This paper aims to open up new avenues for research on the traditional fondness of snakes for milk, a truly ‘impossible biology’. It is built around several known and little known premodern literary and iconographic sources — examples can be found from much of Eurasia — and adopts an interdisciplinary and retrospective comparative method.


INTRODUCTION
Some years ago I began a study of pre-modern folklore evidence for the 'bosom (or vaginal) serpent', a phrase adopted in 20 th -century folklore studies about animals entering, living and even growing in the human body. These beliefs are found in psychiatric records -'internal delusional zoopathy' -as well as in a vast range of written and narratives genres including myths, legends, miracle tales, exempla, personal experiences stories, etc. The infesting animals are usually reptiles, worms, batrachians and fishes, but, small mammals like, for example, mice, also feature. These bosom serpents are, almost invariably, disease agents. My research has focused on ancient and medieval written documentation on bosom serpents; but I have also explored and compared oral sources and medical records from different epochs and regions. The invasion of the body by imaginary creatures was, in fact, a common etiological strategy for explaining illness in pre-modern societies and in many places these beliefs persist (Ermacora 2015a, Ermacora 2015bErmacora et al. 2016Ermacora et al. , 2017. In the present article 1 , however, I would like to leave bosom serpents and concentrate on a related antique motif, often connected to folklore about fantastic intruding creatures: the fondness of reptiles for milk. More specifically, I would like to present and briefly discuss some early documentary evidence for this widespread and, indeed, archaic idea, of the cross-cultural motif of mammals being suckled 2 by a snake 1. Unless otherwise stated, all translations are my own and the references to folklore motifs are from Stith Thompson's Motif-Index of Folk-Literature (Thompson 1955-1958. The titles of texts mentioned are given in the original language with translation only with the first occurrence. 2. The authoritative Oxford English Dictionary (OED) Online (s.v. suckle, v.) makes a distinction between 'suckle' for humans taking milk from the breast and 'suck' for other mammals taking milk from their mothers. As the OED, also, notes, though, English usage is loose and 'suck' is sometimes applied to humans taking milk, and 'suckle' to animals (particularly in a domestic set-or another similar animal (lizard, toad, etc.). This folklore is reflected in a wide range of primary sources, which concern both mythical animal-human and animal-animal suckling. As we will see, snakes were traditionally believed to suckle women and there were also snakes which took, it was said, milk from cow udders (Fig. 1). Ideally, a comparative study of the milk-suckling reptile should be carried out across the full gamut of relevant disciplines including ethnology, linguistics, philology, folklore and historical-religious studies, something I have tried to achieve here. A comparative study would naturally, also, include, as below, pre-modern written references and an analysis of their transmission. For reasons of space, my comparative study will only concern a representative sample of documents; I will only hint at the variety and diversity of milk-suckling reptiles. Enough has been done, however, to show the potential for further research.
As I am going to use late-recorded folklore material to shed light on much earlier periods, I hope to demonstrate the applicability, the usefulness and also the limits of this interdisciplinary approach, sometimes called the "retrospective method", recently resuscitated, in particular, in Old Norse studies (Heide & Bek-Pedersen 2014; Sävborg & Bek-Pedersen 2014). One of my aims here, in fact, is to show that a great deal of cross-cultural cognates, hitherto considered unrelated, belong to the same story-complex. They are adaptations of the polymorphic and predominant idea of the (impossible) milk-suckling reptile. The structural adaptation and persistence of both folk beliefs and Märchen or legend plots and their stylistic criteria in many different societies needs, in fact, to be taken into consideration. As Christine Goldberg wrote, when folklore material "evidences comparatively regular regional or ting). As here I will frequently deal with animal-human mixes -the animal that takes from the human breast -and as 'suck' can in some instances be ambiguous, I have employed 'suckle' throughout for milk-taking, save in instances when I refer to the actual mechanics of suction, and, of course, in quotations.
I will not indulge, in this article, in a zoological analysis of the ubiquitous oral traditions on the mythical milk-suckling or milk-drinking reptile; nor will I examine in depth the interesting debate, which took place in 19 th -century natural history, on whether serpents have a special appetite for milk, a view challenged by modern studies 4 . However, it is worth noting here that there is no empirical basis for saying that snakes like mammal milk (or, for that matter, wine). Experiments, indeed, have shown that captive snakes systematically refuse to drink milk 5 . There are also many biomechanical and behavioral implausibilities. Scholars such Gabriella Eichinger Ferro-Luzzi 1987: 125), Ángel Charro (2004 and Margarita Torrez Paz 2017: 160, 170) have remarked how the construction of a snake's mouth (sharp teeth, inflex-4. For Germany, general overview in Olbrich (1904). For France and Italy, pro: Lamare-Picquot 1835; Raspail 1860: 20-23; Hopley 1882: 75-93 (Raspail an author particularly 'gullible', see Ermacora 2015b: 88); contra: Duméril 1835; Gené 1853: 8, 9. 5. See, for example, Olbrich 1904: 70, 71;Fitzsimons 1912: 265, 266;Carr 1963: 148, 149. Hertzog (1967 dissected 3500 snakes in North America in order to find milk or dairy products in the animals. He found nothing. ible lips), together with the absence of diaphragm, makes it impossible for a snake to suck in the same manner as young mammals do. In addition, a snake's digestive system does not permit the digestion of lactose: reptiles, of course, lack the lactase enzyme. Yet, notwithstanding these points, definite scholarly statements are still lacking. The cultural resilience of milk-drinking snakes and the many witnesses of snakes at udders and breasts (on this issue, see Albert-Llorca 1985: 99;Danielson 1996: 193, 194) are, I believe, the explanation. Even herpetologists have been reluctant to write off the milk-drinking snake entirely. Some authors, for example, have noted that snakes do not drink milk habitually but that, if snakes are very thirsty, they can drink small quantities. This presumption, to be intended as a sort of intermediate 'tradition of disbelief ', is always offered as a remote possibility (see, for instance, Viaud-Grand-Marais 1869: 19, 47;Botsford Comstock 1911: 197;Schmidt 1929: 13;Vasu 1946;Eichinger Ferro-Luzzi 1987: 125;Eichinger Ferro-Luzzi 1995: 39;Achille 2015: 176).
The traditional fondness of snakes for milk should be considered an 'impossible biology' or, rather, a "wild biology" (Bettini 2002: 94, inspired by the génétique sauvage of Françoise Héritier-Augé), i.e. a flexible system of shared popular beliefs about animal habits. Indigenous knowledge concerning milk-suckling reptiles is not structured around our distinctions such as that between actual biological or mythical qualities; in cross-cultural ethnozoology, one can hardly find such a thing as a 'paradigm of knowledge' in the modern scientific sense. As Jean Trinquier (2012: 179) put it in regards to traditional representations of ophidians in antiquity, "[i]t would be as artificial as it would be wrong to pretend to establish a clear division between what would fall into 'physical' or medical discourse and what would fall within the realm of myth and religious beliefs". A revealing trope in Greek heroic iconography from the 5 th century BC, shows a snake drinking from wine containers prepared for it by men (Ermacora 2015a: 96, with previous literature).

MEDIEVAL VARIANTS OF ATU 285A
The close relationship between a snake (as a dead soul?) and a milk offering is very well known in tale-types ATU 285 'The Child and the Snake', ATU 285A 'The Man and the Wounded Snake', and ATU 672 'The Serpent's Crown' (Haavio 1940;Waugh 1960;Lachs 1965: 174, 175, 178-180;Schwarzbaum 1979: 123-136;Albert-Llorca 1985;Uther 2011: 165, 166, 368, 369). There are many variants of ATU 285A with a friendship developing between the (chthonic) milk-drinking snake and man: a snake living in the domestic space or in a cave receives a regular supply of milk in exchange for his generosity; for example, the animal tells farming secrets that lead to prosperity. If people kill the snake, they, thereafter, have bad luck and lose everything: their child (or children) may die. This sometimes happens on the advice of the impudent wife (mirroring, perhaps, the misogynistic vision of the Edenic or power-hungry female), or because the household is blinded by its material abundance. In some versions the snake is ready to reconcile with the man. The moral of the story is, frequently, that a broken friendship cannot be mended. There are instances of ATU 285A in which the food/milk offering is not present, including the earliest known variants of such tale-types: the shorter Greek fables no. 51 and no. 573 of the Perry Index of Aesop's Fables (translations in Gibbs 2002: 39, 40;see Rodríguez Adrados et al. 2000: 57, 107-109, 184, 304, 322, 323, 345, 443, 499, 554, 555, 675;Rodríguez Adrados et al. 2003: 72, 73, 773, 774). In pre-modern Asia, the plot with the milk motif is first documented in a Sanskrit version contained in Pañcatantra 3: 6 (The Five Topics; tale-type 220 in Laurits Bødker 1957: 32). The Indian fable is typically listed under the conventional title "Gold giving serpent" and its plot can be summarised as follows: [a] brāhmaṇa who worshiped a snake [with milk] was rewarded with one gold coin every day. The brāhmaṇa's son attempted to kill the snake, thinking its hole would be filled with gold. The snake bit him and he died. (McComas 2007: 206;full translation in Ryder 1925: 331, 332) Then, there are several 12 th -and 13 th -century European ecotypes, frequently found under the title De homine et serpente (The Man and the Snake; exemplum no. 4251 'Milk and Snakes' in Tubach 1969: 325) (Fig. 2) The milk motif that we find in all these versions has led scholars such as Alexander Huber 2001: 96) to postulate direct Indian influence for ATU 285A on medieval European folklore. The theory of Indian origins for ATU 285A is complex, but the fact is that, as we will see below, the belief that snakes like milk was commonly acknowledged in Europe during the Middle Ages (Mortineau 2003: 96; more wide-ranging documentation in Ermacora 2015a: 273; Ermacora 2015b). In addition, the traditional assumption -first advanced by Theodore Benfey (1859: 360) -that the Pañcatantra is the pro-ANTHROPOZOOLOGICA • 2017 • 52 (1) totype of the whole Hellenistic and medieval tradition of ATU 285A, has been challenged effectively by Francisco Rodríguez Adrado. He, in fact, has proposed that the Indian version was influenced by Aesop's fables (Rodríguez Adrados et al. 2000: 107-109, 567;see Lindahl 1984: col. 984, 985 on the longdebated issue of origins). In Pañcatantra's fable, "the relationship between the Snake and the man is more clear [sic], being based on a definite 'businesslike' arrangement: a piece of gold for a libation of milk" (Schwarzbaum 1979: 126-130, 133; note that Schwarzbaum, like many others, agreed with Benfey). This Indian variant of ATU 285A with the milk and the snake is documented only in Pūrṇabhadra's longer version of the Pañcatantra, the 'textus ornatior'. However, this manuscript dates to 1199 AD and there is no earlier Sanskrit version of the story. It does not appear in other versions of the Pañcatantra known today (like the archaic Tantrākhyāyika, Little Tantra-Stories). Nor does it feature in the reconstruction of the Sanskrit text of the original Pañcatantra, attempted by Franklin Edgerton (1924a, b;see Laurits Bødker 1957: 32;Olivelle 1997: 163 for helpful summaries of the occurrences of ATU 285A in modern editions and translations of the Pañcatantra).
The motif of the milk-drinking snake, which we only find in some Western medieval variants of ATU 285A, together with the roughly contemporary medieval recension of the Pañcatantra by Pūrṇabhadra, seems to confirm Rodríguez Adrado's assumption. It remains to be explained why, if this is not just chance, the pattern of the milk-drinking snake emerged in ATU 285A, in India and in Europe, only between the 12 th and 13 th centuries. In composing his longer text, on the basis of various preexisting Pañcatantra manuscripts, Pūrṇabhadra appears to have used a lost Pañcatantra recension (Olivelle 1997: xli-xlii;McComas 2007: 25, 33, 106). It is impossible now to determine whether the story of the milk-loving serpent was or was not part of this lost version of the Pañcatantra. In any case, Wayland Debs Hand was surely right to observe that scholars have to try to account for tale-types ATU 285, ATU 285A and ATU 672 "in terms of actual happenings or beliefs and legends bearing out the basic affinity between snakes and humans" (Hand 1968: 889). In what follows I would like to take up Hand's challenge.

THE SNAKE IN THE CRADLE
In my research on bosom serpents I have frequently encountered the motif of the thirsty snake that enters the sleeper's mouth and slithers down into the stomach because of its love for milk or wine. The animal is, in fact, according to the storyteller, attracted by the smell of the ingested liquid (see Ermacora 2015b for ancient Greek and medieval sources). The bosom serpent and the milk-suckling serpent are clearly married by a variant of international tale-type ATU 285B* 'The Snake Stays in the Man's Stomach' (Uther 2011: 166, 167), that Giovanni Pizza (2010: 130;Pizza 2012: 176) called the "Snake in the Cradle". Here Pizza relied on Marlène Albert-Llorca (1985). Albert-Lorca had offered a general interpretation for the deep and widespread correspondences (involving the symbolism of corporeal fluids as well as ideas of sexual fertility) that are established, in many European legendary cycles, between a reptile and an infant/child (see also Borghini 1996Borghini , 2005b, with a different approach). Pizza summed up this ATU 285B* variant as follows: a child is sleeping in his crib under a tree or in the grass field, or more rarely in his house, a snake enters the cradle and penetrates into its mouth, the child has no particular ailments but its belly begins to swell. (Pizza 2010: 130;Pizza 2012: 176) Sometimes, these stories end tragically (see Klímová 1973;Pereira Bastos 1988: 21, 22;Boujot 2001 for, respectively, the Moravian, Portuguese and French versions). What distinguishes this hybrid narrative form, however, is the element of milk -in this case, maternal milk -which causes the snake, which is greedy for milk, to creep into the milky-smelling body of a child (or a nursing mother). This connection between bosom serpents, babies and maternal milk has already been noted by several authors for European beliefs about reptiles and amphibians (Riegler 1921: 140;Papi 1988: 37;Giraudon 1991: 73;Nesi 1999: 55;Nesi 2001: 481;Torrez Paz 2017: 156). One will find an example -the oldest of which I am aware 7 -of beliefs and customs surrounding the 'Snake in the cradle' in Olaus Magnus' Historia de gentibus septentrionalibus [etc.] (A Description of the Northern Peoples) 13: 9, an encyclopedic work published in 1555: Magnus includes, there, a chapter entitled "On guarding tiny children from serpents during the harvest" (De infantulis tempore messis a serpentibus custodiendis). Snakes and other reptiles, in Scandinavia,   apparently enter the mouths of infants while their parents are occupied in summer agricultural activities. It is said that the women, while going to harvest their corn, carry with them their infants at the breast. Interestingly, before working in the fields, babies are put in baskets in the branches of trees to protect them from bosom serpents (original Latin text in Magnus 1555: 437, translation in Foote et al. 1998see also Berthold 1850: 5;Bondeson 1997: 45;Hartmann 1998: 68, 69). A dramatic woodcut accompanies the original text, where frustrated snakes look up at a cradle with infant, hanging from a tree (Fig. 3).
The Historia de gentibus septentrionalibus [etc.] contains many other illustrations of 'health risks' in the North (Hair 2000). Olaus, also, tells us of protective Scandinavian snakes, reared on cows' or sheep's milk, that "are regularly seen sleeping in [babies'] cradles, like faithful guardians" (plerumque in cunis, ut fidi custodes, dormire videntur; original text in Magnus 1555: 776, translation in Foote et al. 1998Foote et al. : 1136. This later information seems to be consistent with the well-known pre-Christian beliefs on snakes of Central Europe, the Balkans and the Nordic areas (Scandinavia and the ancient Baltic region of Samogitia) (full discussion and literature in Haavio 1940: 216-218;Ferrari 1997: 13-17;Teijeiro 1999;Luven 2001;Ogden 2013: 162). These notions, involving the worship of domestic snakes, bringing good fortune and wealth to the house and its residents, "probably had the most mentions [of snake myths] and the widest circulation in sixteenth-century works from all over Europe" (Dini 2010: 337, referring to the Lithuanian worship of the serpent). I will return to this below.

THE MILK-SNAKE AS A DEFRAUDER OF BABIES
The bosom-serpent variant of the 'Snake in the cradle', with the entry of the snake into the sleeping body, attracted, naturally, by the smell of milk, is connected to another European legend complex: that of the "milk-snake defrauder of babies" (Bermani 1991: 180). The 'defrauder' enters the domestic space to take milk from vulnerable human victims, particularly women and infants. There are two main variants of the legend widely attested among peasants, farmers and shepherds in Europe, India and America. Both feature a snake which steals milk destined for an infant from a woman's breast: without gaining strength from the milk, the infant then stops growing, loses his or her strength and sometimes starves to death 8 (for Europe, see Brandes 1980: 81;1981: 223;Bermani 1991: 179-182;Teijeiro 1999: 310;Borghini 2005a;Torrez Paz 2017; for India, see Basu 1994: 106; the same stories are attested in Mexico, Chile and Argentina, where they do not seem to be indigenous: see Cardozo-Freeman 1978: 10, 11;Jordan 1985: 30, 44;Pereira Bastos 1988: 19-24 9 ). Sometimes, bosom serpent motifs and folktales on milk-suckling snakes are fused together by the modern narrator in the same story (Giraudon 1991: 74, on a Breton folktale about a snake coming out from the mouth of a woman to drink fresh cow's milk), or bound in the same narrative performance (see, for example, the oral tale collected in the Marche region of Italy by Pigorini-Beri 1889: 58, 59). The two variants of the milk-suckling snake, which probably "arose as a culturally shared and codified projection of anxieties concerning infant mortality and the availability of an adequate supply of mother's milk" (Brandes 1980: 81;1981: 223; see also Albert-Llorca 1985: 100), run as follow: in the first one, the snake approaches the infant and makes him or her vomit (the animal, for example, hits the child's stomach with its tail). The milk is, then, devoured by the snake (a contemporary variant from the Venetian Dolomites is in Perco & Zoldan 2001: 104); in the second one, the snake replaces the child during the act of breastfeeding, while putting its tail in the mouth of the child as an improvised dummy. The child notices nothing, and does not cry. The animal, then, suckles the mother's breast (an oral version, collected in Catalonia in 1983, is translated in Pujol 2013: 61). In both cases the infant visibly loses strength until the serpent is finally discovered.
Some anthropologists have adopted a socio-psychological approach to this legend complex, the snake protagonist is a "bad mother": instead of giving milk, it takes the same (Ferraro 1996: 557). Perhaps such tales also evoke fears that the infant would remain an incomplete human being, incapable of socializing and demanding of women's vital energies (Albert-Llorca 1985: 105, 106). Alternatively, the snake offers a symbolic portrayal of masculine attitudes toward women: "just as the serpent deprives the child of milk, woman deprives man of his semen" (Brandes 1980: 81, 82;1981: 222, 223). Others have developed a narrow Freudian interpretation (Bermani 1991: 179-185), or have considered the milk-suckling snake as a symbol of the insatiable and devouring hunger of the child (La Barre 1969: 94;Slater 1968: 89, 90;Albert-Llorca 1985: 105). Of course, it is difficult to adapt scholarly theories of this type, even if they are mutually compatible, to the many different historical and cultural contexts where we find beliefs about snakes and similar milk-suckling animals.

ANCIENT MILK-SUCKLING SERPENTS
The earlier historical evidence for reptiles suckling women's breasts still awaits analysis: most pre-modern texts have gone basically unnoticed (for a first comparative approach of some ancient texts, with the help of folklore, see Devereux 1976: 181-218;Sancassano 1997: 137-139, 166, 167, 173-176, 178-184, 192-194, 341;Martina 2004: 9. Of course, the lack of milkable domestic animals before 1492 (save, possibly, llamas in the Andes) would make any indigenous Latin American legends particularly interesting for the relationship between cattle and human milkdrinking. 138, 148, 149, 155-157;Ronzitti 2011: 34, 39-42, 58). These often revolve around the 'Bite of the Nursing Serpent/Biting of the Nipple' and the distressed reaction of the one who gives her breast to the serpent (see the still fundamental discussion by Devereux 1976: 181-218, even if his psychoanalytic stance is sometimes baffling; one must also take into account the old Romanian ballad 'The Snake' examined in Ispas 1980). Certainly, Trinquier's (2012: 179) claim that the motif of the breastfeeding snake, among Greek-speakers, is only found in the famous episode of the nightmare of Clytemnestra in the Oresteia of Aeschylus, needs to be corrected or, at the very least, hedged around with qualifiers. Clytemnestra dreams of giving birth to a snake that is actually a double of her son Orestes, born from the same womb. She lays this snake in swaddling clothes, and offers him milk and blood from her breasts (Libation Bearers 526-550, 928; original Greek text and translation in Sommerstein 2009: 280-283, 332, 333) 10 .
Scholars such as George Devereux (1976: 193-197, 212) and Rosa Ronzitti (2011: 42) independently compared, for example, this scene with a myth, told by several ancient authorities, involving the infant Herakles and his ambiguous relationships with the goddess Hera (a fine paradigm of the 'wicked stepmother' in Watson 1995: 39, 233, 239 terms). Heracles is so painful to nurse that Hera throws him from her; Heracles, then, wounds her in the breast with a poisoned arrow. Moreover, there is a related tale in which Hera (or Amphitryon) sends two serpents to the cradle of the innocent Heracles, who defends himself by choking them (Citroni Marchetti 2008: 58-60;Ogden 2013: 63-65, 223). The episode from the Oresteia is also of considerable interest from the point of view of comparative mythology, and can be easily paralleled in Vedic and medieval Purāṇic texts. In Indian tradition, the child Kṛṣṇa kills the terrible demoness Pūtanā. Kṛṣṇa sucks out the poisoned milk from Pūtanā's breast, with which she had tried to kill him; then, Kṛṣṇa sucks out her life blood (Preciado-Solis 1984: 56-60;Eichinger Ferro-Luzzi 1987: 64, 137;Ronzitti 2010: 53-59;Ronzitti 2011: 39, 40, 58).
Any comprehensive study of the motif of the milksuckling serpent in ancient sources should consider, too, the 2 nd -century Greek rhetorician Lucian of Samosata in his pamphlet attacking religious frauds Alexandros seu Pseudomantis (Alexander, or the False Prophet) 7. Lucian reports that in Macedonian Pella a near contemporary, Alexander of Abonoteichus, the false prophet of the new snake-god Glykōn, saw domestic serpents, kept as house pets, suckling milk from the breast, as babies do (original Greek text and translation in Harmon 1925: 184, 185). Modern com-10. In a version, collected in 1933, of the Spanish narrative poem Mala hija que amamanta al diablo (The Bad Daughter who Suckles the Devil), there are two devils which take the form of serpents to punish the selfishness of a woman, attaching to her breasts and "suckling out milk and blood, blood and milk from her veins" (original text in Catalán & Salazar 1999: 421, 422: this is ballad no. 237, version 'romance C'; see infra note 3 for motifs C15.3. and J2072.7). Carranza Vera 2009; see also Torrez Paz 2017: 152, 153) studied 17 th -century broadsheets (relaciones de sucesos) as possible antecedents of this text, and also compared motifs B765.4.1 and Q452. mentators noticed that this account seems to be consistent with the snake imagery in ancient Macedonian culture, as reflected by Alexander the Great's famous conception from an intercourse of his mother Olympias with a divine snake (probably the disguised god Zeus Meilichios); the same Lucian, after having mentioned the suckling snakes, in his work "then opines that it was this custom that gave rise to the myth that Olympias conceived Alexander the Great by sleeping with a snake" (Ogden 2009b: 176; see also Borghini 1996: 105-107). We do not know whether the historical Alexander of Abonoteichus ever visited Macedonia; but, unlike the modern folklore references discussed above, nevermind Pliny's bova, discussed below (briefly compared to Lucian by Teijeiro 1998: 158; Ogden 2009a: 287), Alexander's snake seems to have positive associations. This fact led Sulochana Asirvatham (2001: 102, 103) to speak of "the absurdity of Alexander of Abonoteichus' claim". But Lucian's description tallies with the ambiguous attitude towards snakes in folklore, in which the animal can be a natural physically repulsive enemy of men or, alternatively, can have a beneficial role and a protective function (Waugh 1960: 156, 157;Albert-Llorca 1985: 100-102, discussing ATU 285). Take the case of local serpent cults in ancient Macedonia and Dalmatia (Šašel Kos 1991), perhaps reflected in the 'house snake' of Greek, Albanian and South Slavic folklore documented from early modern times onwards (Lawson 1910: 327, 328;Nilsson 1940: 70, 71;see Hartnup 2004: 31, 283, 284, 296 for Leo Allatius, c. 1586-1669. The tame and domestic snake (serpent domesticus) living in the house as protector and genius loci (often seen as an ancestor: universal motif F480.2 'Serpent as house spirit') is, of course, perfectly compatible with an ophidian entity fed on milk or wine (an analysis, in this light, of Lucian's passage is in Teijeiro 1998;1999: 310, 311;Boos 2001: 134; the passage was already discussed in comparative terms by MacCulloch 1920: 410 searching out milk-suckling snakes) 11 . 11. It is worth mentioning that Francis Lazenby (1949: 248) put Lucian's Macedonian breast-suckling snakes under the heading "Snakes as Pets" (see also Krenkel 1981: 41, 42 who stressed its erotic implications), and remembered a tame snake, kept by the Locrian Ajax as a companion-dog, in Philostratus of Athens's Heroicus 31: 3 (On Heroes; original Greek text and translation in Rusten & König 2014: 218, 219).  1950: 338-347). A deadly serpent fastens itself around Caradoc's arm. He is only delivered, thanks to the maiden Guigner who, immersing herself in a vat of sweet milk, while Caradoc sits in a vat full of vinegar, offers her breast to the serpent. The snake, of course, detaches itself from Caradoc and attaches itself to Guigner's breast. Caradoc immediately cut off the serpent's head with his sword, along with one of Guignier's nipples; the nipple later being replaced by a prosthetic gold one 12 . This scene reflects motif B765.4.1 quoted infra note 3 -as was noted by Anita Guerreau-Jalabert (1992: 2). Modern oral folklore parallels featuring breast-suckling serpents, in fact, have long been recognized (Harper 1898: 216;Rhys 1901: 689;MacCulloch 1920: 410;Le Menn 1985: 22, 105, 106, 109, 110;Douchet 2002: 121); most interpretative efforts, however, have focused on the motif of the 'golden breast' and/or the mutilation of the breast in Celtic, hagiographic and Indo-European parallels (Le Menn 1985;Laurent 1987;Sterckx 2005: 127-133;Lecco 2010: 35-40 One striking equivalent is to be found in the anonymous Vita Sancti Budici (Life of Saint Budoc) 181-186, contained in the Breton Chronicon Briocense (Chronicle of Saint-Brieuc), written between 1397 and 1416. The "curious legend" (Leduc & Sterckx 1972: 13) runs as follows: the king of Brest has a blood-sucking snake coiled around his arm. His daughter Azenor, then, rubs her chest with oil and milk: the snake jumps onto a breasts and grasps her there, she cuts off the nipple and throws the animal into a brazier. To reward her filial piety, God gives Azenor a golden breast (French translation in Le Menn 1985: 21-24). According to specialists, a simple filiation from the Livre de Carados to the Vita Budoci can be excluded: both sources rely on the 'woman with golden breast' tale-type, which usually features a sick king and a breast obsessed snake (Laurent 1987: 212;see also Le Menn 1985: 28, 29, 35, 42, 43, 108;Lajoye 2005: 82-84, 92, 106, 107). In any case, judging by the many miniatures of the scene which illustrates medieval manuscripts of the Première Continuation de Perceval (Fig. 4), the episode of Guigner and the snake was popular: it "seems to have impressed itself particularly strongly on the memory of contemporary audiences" (Busby 1993: 371;see also Bruckner 2009: 143-145).

LATIN BOU(V)A/BOA(S)
Snakes traditionally, then, like milk. We have looked so far at those that suckled women, but many were believed to suckle animals. Take the giant bovid-suckling snake called bou(v) a/boa(s) in Pliny's Naturalis historia 8: 37 (Natural History; original Latin text and translation in Rackham 1940: 28-31; see also Bona 1991: 72;1994: 430, 431;Teijeiro 1998: 158), and Solinus' Polyhistor or Collectanea rerum memorabilium 2: 33 (Polymath or Collection of Curiosities; Solinus probably also relies on Pliny: original Latin text in Mommsen 1895: 39, 40, translation in Golding 1587; Golding glosses Boa as "a wonderful kind of snake"). Its milk-drinking is a theme that, in Jean Trinquier's words, "we find in other areas" and that "would require a separate study" (Trinquier 2008: 234). In another article Trinquier (2012: 179) wrote: "this belief [in the milk-drinking snake bova] is attested only in the Latin West": something which is not, in fact, true. I will now examine the bova snake in a comparative key as has been occasionally, tentatively attempted by other authors 13 .
The bova, of massive size, is fond of cow milk and, according to the same ancient authors which rely on a folk etymology, takes its name from the domestic ox (Latin bos). According to the Lateinisches etymologisches Wörterbuch, bova has no clear etymology (Walde & Hofmann 1938: 110, s.v 1920: 893-1664, translation in Scanlan 1987. Illuminations with a snake milking a cow appear in manuscripts, as is the case with the 1404 copy of the Liber de natura rerum stored at the Archive of Prague Castle, Library of the Metropolitan Chapter by St. Vitus (see also Zíbrt 1905) (Fig. 5).
North Italian compound bisciabova 'serpent' + 'ox', first attested, perhaps, in an anonymous 13 th -century sonnet (original Italian text in Massèra 1920: 61). Angelico Prati (1954: 205-207;Prati 1968: 18, 19, s.v. boa2) stated that the dialect evidence on bova has been strongly influenced by Latin literature. It is reasonable to envisage, here, instead, an uninterrupted double channel of transmission -both learned and popularfrom the classical beliefs about the bova snake to its modern equivalents. The mono-directional idea that Roman and later writers influenced oral folklore through literature, but that literature was somehow immune to influence from oral accounts, is unhelpful.

GREEK IMPERIAL PARALLELS?
The Palatine Anthology, an eclectic collection of 4,500 short Greek poems and epigrams, contains two interesting epigrams attributed to Polyaenus of Sardis and Tiberius Ilus: these are ancient authors of which little is known save that they lived in the 1 st century AD 14 . Both epigrams (Anthologia Palatina 9: 1, 2) bring us back (consider the case of the snake bova) to udders swollen with milk (in this case, a doe's), perhaps suckled by a snake (here a deadly viper). As that 'perhaps' suggests caution is required here: we have no explicit mention of suckling. We only have the biting snake, the udder and its milk. The two Greek epigrams have already been comparatively evoked, in regards to the milk-suckling serpent, by Manuel García Teijeiro (1998: 158, 159). Apart from that they have been little discussed by modern scholars, though note that Francesco Redi reported them in full, as an epigraph, in his essay dedicated to beliefs on vipers current in his time (original Greek text in Polito 1975: 41-79, translation in Knoefel 1988. Both epigrams tell how the viper injects poison into the doe's udder in order to kill a newborn fawn. The fawn suckles the teat and imbibes poisonous milk: death is, thus, transferred  from mother to child (original Greek texts and translations in Paton 1917: 2, 3). The enmity between the deer/stag and the snake is, of course, a classical and then medieval literary topos, well represented in later European folklore and zoonomastics (Le Quellec 1991;Moulinier-Brogi 2002;Donà 2009: 67, 68;Goudi 2011: 111-113;Lelli 2014: 105, 165, 171, 232).
Later material hints that there might be suckling in the background of the two Palatine Anthology poems. In Indian and European vernacular beliefs, for example, there is the common notion that the cow dies after a snake has suckled milk from the udders. Sometimes it is stated that the snake has suckled too much milk so draining out the vital essences of the cow (see, for example, Isidore of Seville and Al-Jāḥiẓ quoted below). Moreover, if the mother is suckled or bitten by a snake, it is said that the snake's poison might pass from the breast into the mouth of the suckling baby or calf through the mother's milk. This idea is perhaps reflected, for example, in the Sanskrit treatise Kāśyapasaṃhitā (see below), and in an interesting commentary -which mixes learned and folklore notions -provided by German physician Michael Maier  In this Chronicle, there are two elements that later recur in Maier and in the pseudo-Aristotle: namely, the cruelty of the execution method (a woman dies after having a breast-suckling toad applied by a jealous Queen) and, possibly, the element of fire (translation in Riley 1863: 232). In the English branch of the Gesta romanorum, there is, similarly, an envious woman who makes her rival put two snakes to her breasts: the innocent woman dies (exemplum emblem, however, also has a much reproduced woodcut: an elegantly dressed man and a woman stand in the deserted alley of a 16 th -or 17 th -century city. The man places a big toad (an animal, note, traditionally believed to be poisonous: motifs B776.2 'Toad considered venomous'; B776.5.1 'Blood of toad venomous') on the woman's exposed breast: her face is a study in terror as the animal begins to suck at her breast (Fig. 6) There are similar stories concerning the milk-suckling serpent from elsewhere in the world. In Newbell Niles Puckett's Folk Beliefs of the Southern Negro, for example, we read the following statement: "[the milk snake] sucks cows dry. You can always tell when this had been done for the cows will then give bloody milk, poisonous to human beings" (Puckett 1926: 43, later taken up by Hand 1964: 437). The poisonous character of milk produced by a cow that has been suckled by a snake is not documented in ancient Latin sources on the snake bova; however, it appears, perhaps, in the 16 th century, according to an iconographic type that is found on medallions on the edge of certain Flemish tapestries. This type is part of a series of images involving learned references to ancient fables on animals mixed with new motifs. It shows a goat or a cow milked by a snake with the following caption -which stress, apparently, the ingratitude of the reptile in respect of the milk-giving animal: "for milk, poison" or "poison instead of milk" (pro lact[e] venenum; see Crick-Kuntziger 1948: 69, 70, 73;Ferrero Viale 1973: 93, 126, 127 on tapestries preserved at Palazzo Borromeo, Isola Bella; Musée Royal des Beaux Arts, Brussels, and Musée des Beaux-Arts, Chartes; Crick-Kuntziger already briefly compared Pliny's passage on bova) (Fig. 7).
One might add further folklore evidence. During the mid 19 th -century, a discussion took place within the medical community in British India. Is it true, the savants debated, that the young of any animal (human beings included) could be poisoned by its mother's milk, should the mother have no. 4281 'Serpents suckled' in Tubach 1969: 327). Popular 16 th -and 17 th -century ballads with the story of Queen Eleanor (confused with Eleanor of Castile spouse of Edward I) also had snakes applied to her breasts (Nørgaard 1964: 166, 167). It is worth noting that Maier traveled extensively in England, and spent a few years at the court of King James VI shortly before the publication of his Atalanta fugiens ], see Allen & Hubbs 1980: 216, 217). For a Junghian approach in which the toad devours the mother and represents the "archetypal fantasy [of ] assimilating [the object of desire] to oneself ", see Fabricius (1994: 57). It might be perhaps useful to recall, here, that "[f ]rom the sixteenth century onwards, the idea that familiars [such toads] sucked blood from witches (often from teats) as part of the diabolical pact was well established in popular print" (Millar 2015: 210). been bitten by a cobra? The poison would, of course, by this logic, pass through the mother's blood into her milk, and, then, fatally, into the young. One of the correspondents, Charles Francis, a surgeon, ended his article reporting two "apparently incredible cases" stated by "two very intelligent native assistants". A Hindu mother and her elder child were bitten by a cobra while sleeping. Both died. "And the infant, whom the mother had taken to her breast to pacify, (for it had begun to cry), died also with symptoms of poisoning" (Francis 1868: 126). The other story is of a calf that died after suckling milk from its mother who had recently been bitten by a snake. The mother remained unaffected, but the calf imbibed the poison which had been left upon the udder. The vernacular explanation given by the informant was the following: "the snake had (as is alleged to be the custom of such snakes) entwined itself around one of the hind legs of the cow, and suckled its milk" (Francis 1868: 126) 17 . According to this opinion, the Indian milk-suckling snake, after having suckled the cow, left its venom on the udder, which was later transmitted to the calf thus killing it. So, perhaps, this happened with the doe and the viper described by the two Palatine Anthology epigrams: the allusion, if that is what we have, was quite possibly straightforward to, say, a 1st-century AD reader or listener.

A BELIEF OF THE INDO-EUROPEAN AGE?
Are these tales, recorded in Europe and elsewhere, "completely independent of history, economy, level of technology and empirical observation" (Eichinger Ferro-Luzzi 1987: 125)? For example, Indo-Europeanists tend to consider the belief in milksuckling and drinking reptiles as "a belief of [the] Indo-European age" (Olsen 2011: 25). Similar claims by anthropologists and philologists have evoked a "common cultural background of the countries of the Indo-European area" (Albert-Llorca 1985: 102), or "ancient notions of Indo-Europeans" (Teijeiro 1999: 311). One should keep in mind, however, that the theme of the snake attracted out of its hole by a bowl of wine/milk/ beer, which will intoxicate the drinking-creature, is found in a New Kingdom Egyptian folktale (the Tale of the Doomed Prince, corresponding to miscellaneous tale-type 934 'Tales of the Predestined Death' in Uther 2011: 574, 575) included in the Papyrus Harris 500 (= British Museum 10060). This tale was probably composed between the late XVIIIth Dynasty and the end or the middle of the XXth Dynasty, that is, 1185 to 1078 BC, and is clearly outside the Indo-European homelands. In regards to the nature of the liquid which tempts the greedy snake from its hole, the text is mutilated. More recent translations venture "wine" and "beer" (translations in Lichtheim 1976: 202;Simpson 2003: 78) 18 ; old but reliable translations such as William Flinders Petrie (1895: 26) 17. Note that Francis suspected, rationally, that all the animals and the humans, mothers and children, 'simply' died after having been bitten by snakes. 18. Additional translations have the offering of "something […] probably wine" (Goodwin 1874: 159, 160) or "jars" of liquids (Grayson & Redford 1973: 63;Fisher 2010: 40). and Gaston Maspero & Hasan El-Shamy (2002: 157, 158), instead, supply "milk". El-Shamy created motif B763.1.1 § (see infra note 3) for this episode 19 .
Phylarchus' 3 rd -century BC FGrH 81 F27 (= Aelian's De natura animalium, On the Nature of Animals, 17: 5) (already referenced in a footnote by Gaston Maspero), describes the very similar method which Alexandrians use to attract their tame Agathos Daimon snakes from their holes. They snap their fingers and offer the snakes barley with wine and honey (original Greek text and translation in Scholfield 1959: 326-329;see also Ogden 2013: 243, 290, 291, 348). The conviction that serpents can be enticed and captured with wine is, also, attested in Greek-Roman world: the locus classicus is Aristotle's Historia animalium (History of the Animals) 8: 4, 594a, where it is recorded that vipers can be caught, after drinking themselves to intoxication, on wine laid out on pottery shards in stone walls (original Greek text and translation in Balme 1991: 109-111). The notion that serpents are drawn to milk and wine appears, meanwhile, too, in the tale of Bulūqiyā included in the Macnaghten or Calcutta II edition of the medieval Egyptian recension of the Alf laylah wa-laylah (Thousand and One Nights; see Marzolph et al. 2004: 130-132). Here the Queen of the Vipers is enticed into an iron cage by two containers, one full of milk, the other of wine: she is attracted by the scent of milk and then falls asleep as she drinks the wine (translation in Lyons & Lyons 2008; see 19. Forbes (1905: 416) briefly compared the Tale of the Doomed Prince to European folklore on milk-and breast-suckling snakes. Older commentators of the tale remarked that the use of pure milk or milk mixed with honey to entice serpents (seen as protectors) was still well known in 19 th -century Egypt (Maspero 1886: 21; Petrie 1895: 34). also El-Shamy 2006b: 40;Kuehn 2011: 60). The similarity with the Egyptian Tale of the Doomed Prince is sound proof, I believe, of the prolonged circulation of these traditional folk narrative motifs in the Near East.

THE FAMILY OF DEMONIC MILK-SUCKLING ANIMALS
Whatever the historical or, rather, polygenetic origins of the fondness of snakes for milk, the contribution of philological and linguistic inquiries is of greatest importance. Certainly, modern European dialectal folklore has countless instances of cow-suckling and milk-stealing animals including butterflies, reptiles, batrachians, hares, hedgehogs and nocturnal birds (for the well-known folklore implications of the ancient compounds Greek Αἰγοθήλας and Latin Caprimulgus 'goatsucker/goatmilker ', see Riegler 1921;Pollard 1977: 50, 51;Capponi 1979: 126-129;Beccaria 1995: 44, 45, 184;Teijeiro 1999: 309;Martínez 2003;Bracchi 2009: 135, 136;Ronzitti 2011: 32, 33, 51-54;Poplin 2016) 20 . This is also shown in many popular romance zoonyms. Etymologically, they employ the fabulous action, attributed to animals, of sucking or stealing milk. These creatures are regularly said to sneak into the domestic space at night to suck life-giving milk or blood from cattle and women. The pioneer in the studies on milk-suckling animals in European dialects was Richard Riegler (1910;1921: 139-141, 143), whose survey is still without equal. In the 1980s, Mario Alinei revisited some of these questions in his journal Quaderni di Semantica, dealing with zoonomastics or zoonymia in its magicoreligious context (Mooijman 1993: 192-197;Caprini 1998). As an expression of traditional etiologies, the folklore on milk-suckling may serve to explain phenomena such as milk scarcity or sudden death syndrome. As we have seen, in the cycle of legends on the milk-snake defrauder of babies, the human victims of these noxious animals are, in fact, the more vulnerable members of the community: those at the "beginnings of life […] and especially in the nursing period" (Djéribi 1988: 35-39). Children, for instance, deprived of milk, stop growing and then die.
Milk-suckling animals were, in post-medieval Europe, usually sent by witches to cause milk to dry up and to ruin the production of butter and other dairy products. Raluca Betea (2015: 445) recently observed that the capacity of witches to rob milk represented a pan-European theme associated with village sorcery. On occasion, milk-suckling animals were associated with the archaic image of the supernatural witch, the 'night witch' with her oppressive, devouring, cannibalistic, and vampiric qualities. This is type 'C' witchcraft (= the mythological witch) in Eva 20. Lévi-Strauss (1988: 34, 41) explained one of the French names of the nightjar, crapaud-volant 'flying-toad', thus: both toads and nightjars have wide mouths. Now note Southern Dutch dialects have: kikvorsnachtegaal 'frog-nightingale', paduil 'toad-owl' and vliegende pad 'flying toad' (Swanenberg 2002: 37); while, in Southern Italy, the terms for the toad and the nightjar are sometimes interchangeable (Ronzitti 2011: 33, apud Plomteux 1987. Given the alleged fondness of toads and nightjars, over a wide area, for milk, a multi-causal explanation might work better than Lévi-Strauss' claim for an oversized mouth. Pócs 1999: 37-57) model of witchcraft-human conflict and interaction. The model in question was created for Hungary, but it could be efficiently employed in a historical perspective and works for other cultural contexts as well. One might usefully recall the traditions in Scandinavia (borrowed too into Sami culture), Iceland and the Baltic areas, where, from early modern times onwards, the 'supernatural milk-stealer' is a magical being associated with the village witch. She often takes on the form of a toad, snake, cat or hare to steal butter and milk from neighboring farms (Tatár 1987;Nildin-Wall & Wall 1993;Van Gent 2009: 111-117;Būgienė 2011).

INDO-EUROPEAN LINGUISTIC EVIDENCE
Curiously enough, the early Indian and Indo-European linguistic evidence for this widespread set of ideas on milksuckling animals has not yet received the study it so richly deserves. Many Tamil and (pan-) Indian sthala purāṇas 'temple-histories', vernacular tales connected to a specific sacred-place, have a snake or a deity appearing as a snake which suckles a cow's udders dry. These stories, studied in particular by Gabriella Eichinger Ferro-Luzzi, are, however, extremely difficult to date: some of them are perhaps medieval (Crooke 1920: 418, 419;Eichinger Ferro-Luzzi 1987: 6, 21, 48, 51, 57, 91, 125, 134, 141, 176, 186;Nagar 1990: 182). One can find traces of several similar beliefs, of demon snakes or animal-like sorcerers that assault and suckle the breasts of women or the udders of cattle, in Vedic and medieval Purāṇic texts. Cattle are safeguarded through hymns to Agni, the Vedic fire and sky god. Agni is "called to protect cows from yātudhāna-'sorcerers' […] that suck the cow milk" (Ronzitti 2010: 43;2011: 37, 38, with an Italian translation of the relevant sources). The presence of milk-suckling reptiles, however, is never explicitly mentioned. More relevant is the Kāśyapasaṃhitā (Kāśyapa's Compendium), a classical Sanskrit text on gynaecology and obstetrics. There is a general agreement that this obscure Āyurvedic text, from the 6 th or 7 th century AD, presents material which is much older than the late Indian classical era. The treatise, arranged in the form of dialogue and attributed to the sage Kāśyapa, dwells at length on the demon Jātahāriṇī, an aggressive terrestrial avatar of the Revatī goddess. In a section of the compendium devoted to the demons' role in miscarriages and the early death of children, Jātahāriṇī is described as taking the form of a 'Mother-Cow'. She "destroys the offspring [i.e. cattle] of a cowherd by tying up and milking the cows, and by killing them, and breaking their legs". The same female demon is, then, said to take the shape of a vengeful serpent, attaching herself to women (to their breasts?) and poisoning their children. We are perhaps dealing with early evidence of the mythical milk-suckling serpent, affecting both cows and women (translation in Wujastyk 2001: 217, 221, 227;see also Wujastyk 1999;Cerulli 2011;Gordon White 2013: 19).
The archaic Sanskrit tradition, in addition, includes names of reptiles such as go-sarpa-'cow snake' and go-lattikā 'cow liane [i.e. serpent]' (for contemporary Indian materials, see Das 1998: 56 for the Assamese name of snake gwala hap 'cowherd'). It, also, includes a reptile, with which it is associated in various lexicons, the godhā-. The godhā-(go 'cow' + dhā 'sucking', 'cow milker/sucker'), certainly more than a simple "punning on cow, go" (pace Doniger 1991: 284), is etymologically comparable with Armenian kov(a)diacʿ, and probably refers to a monitor lizard or to an iguana-like creature, as was demonstrated by Heinrich Lüders (1942;see also Turner 1966: 230, 232, s.v. gōdhā-and gōlaka-;Monchi-Zadeh 1990: 45, 46;Mayrhofer 1992: 498, s.v. godhā;Martirosyan 2010: 372, 373, s.v. kovadiac';Ronzitti 2011: 36, 37, 40, 41). This reptile godhā-is mentioned, for example, in Ṛgveda (Knowledge of the Stanzas of Praise) 10, 28: 10cd-11ab, and Manusmṛti In an enigmatic passage of Ṛgveda Indra is, instead, compared to a thirsty buffalo: these are obscure and difficult to translate words, but it could be that the reptile godhā-traps the hooves of a buffalo just as mythical milk-suckling reptiles do in European folklore (see infra note 3 for motifs B765.4 and B765.4.2). Here I give two possible translations of the Ṛgveda. For the first one, which favours a folklore interpretation, see also Lüders 1942: 48-50;Joachim 1978: 70, 71;Ronzitti 2011: 36 (all with further bibliography): 21. Ronzitti combined this etymological proposal with another: *guēb(h)-'slimy, tremulous, burbot, toad', which would lead to a root *guēb(h)ah2 / *guōbho(n)-(and variants) 'toad' or 'aquatic animal', according to the principle that there may be multiple etymologies behind a single historical form (Ronzitti 2011: 15, 16, 24-26, 62; some critical linguistic remarks were raised in Garnier 2012). In other words, there are two etymological possibilities behind Latin būfō, both of equal formal and semantic validity. the powerful thirsty buffalo was surely imprisoned. A lizard trickily drew (caught) it; (translation in Velankar 1952: 8, slightly modified) [t]he buffalo also got trapped, when it was thirsty. The monitor-lizard plowed this way for him. (translation in Jamison & Brereton 2014: 1420 However, a different translation of the passage might run: withheld indeed is the thirsty buffalo, from [the buffalo] the alligator [godha] takes away this foot. According to this third interpretation, Ṛgveda would suggest, instead, that the buffalo is thirsty and approaches the water. Then, an alligator come up and bites his leg, preventing him from leaving: a credible enough event in the swamps of India. Note, also that Ṛgveda 10, 28: 11ab, soon after, wishes this fate on those who stuff themselves with the food of the Brahmin priests: "[t]he monitor-lizard will plow this way for those who taunt the brahmins with their food" (Jamison & Brereton 2014: 1420). Does this means, then, that if you are thirsty but steal from the priests, an alligator will seize you by the leg?

THE SLAVIC LEG TRAPPER SNAKE
The Kitāb al-Ḥayawān (Book of Animals) of the Arab Mutazilite writer  shows that, during the Early Middle-Ages, the folklore theme of the milk-suckling snake trapping cows' legs, which perhaps we find centuries before in the Ṛgveda, is attested in Slavic culture. (Note that this motif is not attested in Latin sources on the bova snake). A vague sense of disbelief permeates Al-Jāḥiẓ's ethnographic account of the Slavs, an account which likely depends on Slav slaves he had met in his home country, Iraq. This may be a clue that these folk beliefs were not known in the Arab world, otherwise we would not expect such incredulity. Al-Jāḥiẓ's passage is traditionally considered as the earliest historical evidence for the later, well-attested, Slavic body of folklore surrounding the milk-suckling snake (Lewicki 1961: 282;Klímová 1973: 205, 206;Pompeo 2000: 92). Al-Jāḥiẓ writes: [a]ccording to what castrated and intact Slavs claim [a verb that indicates uncertainty is used] of the snake (hayya), in their countries it goes to the cow, it wraps itself around the hind legs [lit. 'thighs'] to the knees and hocks, then lifts its chest towards her udders with its mouth to reach the nipples. The cow cannot, despite her strength, move. [The snake] continues to suck the milk; whenever he sucks [the cow] becomes weaker. If she nears the point of death, it lets her go. They argue that the cow dies or suffers from a serious illness in the udder that makes it difficult to cure. The snake is greatly attracted to the milk. If the vipers find a container [of milk] without a cover, they dive to drink it and sometimes vomit what they have in their bowels: serious damage then affects those who drink this milk. The devastating powers of the milk-suckling snake is also attested in Isidore of Seville's Etymologiae (Eymologies) 12, 4: 28. Isidore wrote, in the 7 th century, that the milk-suckling snake "attaches itself to the udders of the [cows] with plenty of milk, and kills them by suckling on them" (plurimo lacte riguis se uberibus innectit et sugens interimit; original text in Valastro Canale 2004: 50, translation in Barney et al. 2006: 257). In Al-Jāḥiẓ's Arab account the cow, instead, usually survives. The detail, in Al-Jāḥiẓ, of the illness affecting udders is interesting, as it is also widely attested in rural European folklore. In oral narratives, the cow milked by a snake typically carries marks or abscesses on its udders or legs (see, for example, Boujot 2001 for France). The name bova as we have seen was, in antiquity, connected by popular etymology to Latin bos. In Isidore (12, 4: 22), in fact, it also denoted a kind of enigmatic disease provoked by the bite of an aquatic serpent called, in Greek, hydros (ΰδρον, also assimilated to the snake bova in a 8 th -century gloss by Paulus diaconus ex Festo: original Latin text in Lindsay 1913: 27). According to Isidore, who records a homeopathic cure, the bova disease, which causes victims to swell up, can be cured with cow (bos) dung (original Latin text in Valastro Canale 2004: 49, translation in Barney et al. 2006: 256). This additional sense of Latin bova as 'infection, disease', usually provoking a swelling of legs or a skin condition colouring the body with red, like measles, is recurrent in ancient sources (Ernout 1909: 123;Mata Oroval 2015: 135). Perhaps the general idea of bova as 'disease' derived, by extension, from diseased udders (i.e. the inflammatory hardening of the udders?), traditionally ascribed, by local rural populations, to the harmful milk-suckling snake.
As to Al-Jāḥiẓ's final sentence -"[snakes] dive to drink [milk in containers] and sometimes vomit what they have in their bowels", poisoning those that drink the milk -does this perhaps come from some ancient author? In Greco-Roman, medieval Latin, Jewish, Arab, Byzantine and even classical Chinese literature there are many comparable stories and reports, told as simple facts, but with a clear folklore background. These all have the (uncovered) bowl or jar of milk or wine and the gluttonous snake, salamander or toad. The basic pattern relates how a thirsty reptile or amphibian approach a bowl of liquid (wine or milk), a dangerous action that leads to a deadly contamination of the liquid, because of the poisonous nature of the animal (for Galen, for instance, see Sordi 2003: 243;Mattern 2008: 35, 38;Boudon-Millot 2009: 49, 50; compare with motifs B765+, N332.3, N332.3.3 §, N346.0.1.1 and Q597.1.1 § quoted infra note 3, and with the probable Jewish ecotype ATU 285A* 'The Adder Poisons the Children's Food' in Schwarzbaum 1979: 131, 132;El-Shamy 2004: 92;Uther 2011: 166). In a related narrative cycle influenced by the Greco-Roman pharmaceutical theory on the theriac, a homeopathic medical concoction produced with the flesh of snakes, the contaminated liquid, instead of being poisonous, has strong healing properties (Gaillard-Seux 2012: 280, 281;Trinquier 2012: 195, 196). Indeed, so well attested are these stories in the classical world that they could usefully be catalogued and classified in a motif and tale-type index of ancient oral narrative. In Zhang Zhuo's late 7 th -century Chaoye qianzai (The Complete Stories of Court and Countryside), quoted in Li Shih-chen's 16 th -century massive Pen-ts'ao kang mu (Compendium of Materia Medica), there is the famous tale, well-attested in antiquity, of the leper healed by drinking wine in which, unknown to him, a snake has drowned (translation in Xiwen 2003: 35-40) 23 . Lorenzo Pompeo (2000: 91, 92) briefly mentioned Al-Jāḥiẓ's "curious note" while examining some verses from the Ukrainian ethnographic poem Roxolania (Ruthenia) 509-564, written in Latin, in 1584, by the humanist Sebastian Klonowic . This episode was inspired by both learned and folk materials and shows us, again, the Western Slavic pattern of the reptile greedy for cow's milk. There is even the motif, already seen in Al-Jāḥiẓ, of the snake which traps the ruminant's legs: it is suggested that the legs are blocked because of the cold of the entwined snake. The cow is then unharmed and it cannot run away (original Latin text and Italian translation in Pompeo 2000: 101-104; see also Moore Coleman 1963: 9) 24 . Identical popular notions were widespread in Europe in 16 thand 17 th -century natural history (see the sources discussed in Olbrich 1904: 68; one can usefully mention Gessner 1558: 526: "sometimes [snakes] suck the cows, wrapping their tail around their legs" [v]accas aliquando sugunt cauda cruribus earum circumplicata).

ADDITIONAL WESTERN PARALLELS
On this matter, I will now consider, as an additional Western counterpart, the wonder (prodigio) told by Giovanni Francesco Poggio Bracciolini (1380-1459) in his very popular Confabulationes (Confabulations) or Liber facetiarum (Book of Jests), a collection of brief narratives, witticisms, jests and jokes, completed between 1438 and 1451 and printed around 1470. Poggio worked in Rome as apostolic secretary and, for many years, he kept a private notebook in which he annotated stories that circulated among the papal Curia. In Liber facetiarum 31, part of a series of tales focusing on reports of monstrous beings, there is related an event communicated to the Pope in a letter from Ferrara in 1438 (for dating, see Pignatti 1999: 264, 265 2006: 92), in his index of Polish folktales, listed Roxolania as the earliest historical evidence for the widespread Polish tale-type T140 'The Cow and the Snake'. The plot of T140, often, runs as follows: a snake sucks a cow, the farmer kills the reptile and the cow stops giving milk or dies. T140 should be obviously examined together with ATU 285, which can be summarised thus: "[a] child shares its milk with a snake. When the mother sees this, she fears for her child and kills the snake. Soon the child becomes ill and dies" (Uther 2011: 165; possible Greek-Roman parallels of ATU 285 are discussed in Citroni Marchetti 2008).
given birth to a deformed snake that drank all its milk after having trapped its legs (Fig. 8). This disturbing event seems, apparently, to reflects a world of supernatural wonder, but the cowherds, Poggio says, saw this with their own eyes (see infra note 3, again, for motifs B765.4 and B765.4.2). The text runs as follows: Monstra hoc anno plura diversis in locis natura edidit. In agro Senagaliensi, in Piceno, bos  [This year, nature has brought a number of monsters into the world in various places. In the territory of Sinigallia, which is in the Picentino country, a cow gave birth to a dragon of amazing size. Its head was larger than that of a calf, and its neck was as long as a man's forearm, while its body was doglike, only longer. When the cow saw what she had spawned, she bellowed as loud as she could and tried to flee in terror. But the dragon arose and wrapped its tail around her hind legs, put its mouth to her udders, sucked out all the milk, then left the cow and fled to the nearby woods. After this, the cow's teats and other parts of her body touched by the dragon's tail became blackened as though burned, and remained that way for a long time. The cowherds (in whose fields the cow had been) affirmed what they had seen, and that the cow has since had calves. All of this was in a letter from Ferrara.] (Original text in Pignatti 1999: 265, 266, translation in Hurwood 1968 Here we have, once more (think of Al-Jāḥiẓ's record of Slav beliefs), the illness in the cow's udder which follows on from the snake's harmful suckling. It may be useful to reconstruct here the textual diffusion of Bracciolini's Facetiae. This account of the cow and the snake was later inserted -together with another eight of Bracciolini's fables and short stories -into the influential and expanded bilingual  & Goldberg 1990: xiii;Pignatti 1999: 259) 25 . Bracciolini's snake story, therefore, certainly had many early modern European readers.
Bill Ellis (2001) studied Poggio's work in folkloristic terms and considered several of Poggio's late medieval tales to be literary versions of (what were then) contemporary legends. We find in Poggio's stories, including the cow and the monstrous milk-suckling snake, in fact, the same narrative context and the same stylistic, functional and rhetorical features that we usually associate with legends collected in recent times. Though, in stylistic terms, the old legends about the milksuckling serpent might have undergone literary and genre adaptation, they still maintain the character of a folk narrative with certain recurrent stylistic elements: localisation, the naming of witnesses, etc. (see Henken 2001 for additional case studies on the shared characteristics of contemporary legends and medieval legends). Ellis, in addition, compared Poggio's story to a modern legend about an udder-suckling snake. An English newspaper (The British Chronicle) reported, in 1770, that a farmer had seen "a most enormous overgrown adder, or hag worm, crawl out of the bush, and winding up one of the cow's legs, apply its mouth to one of the paps" (Ellis 2001: 81, apud Simpson & Roud 2000: 2, s.v. adders). But, as we have seen, it is easy to provide folklore parallels for such widespread belief-tales, including the way that snakes take hold of ruminants' legs. For contemporary materials one may consider, for instance, the Calabrian and Sicilian zoonym pasturavacche/'mpastura-vacchi ('cowsherders'), a dialectal name of the common four-lined snake, which is said to suckle cow's milk while blocking a cow's legs (Riegler 1921: 140;Vinja 1977: 17;Beccaria 1995: 183;Ronzitti 2011: 37).  Ermacora D.

CONCLUSION
Folkloristics should ideally employ composite data in order to make comparative observations. Based on several well known, little known and unknown pre-modern sources, and adopting an interdisciplinary and retrospective comparative method, this study has offered some new research directions on the fabulous propensity of reptiles to consume milk, a truly 'impossible biology'. Employing, in particular, the methodological tools developed by folklore studies, I have followed evidence for milk-suckling reptiles intertextually back through time, and examined the cultural ramifications and various adaptations of relevant traditional anthrozoological themes. A folkloric analysis combined with a philological-contextual study has allowed me to isolate and analyse a series of parallel variants on milk-suckling reptiles that are linked to each other in terms of typological similarities. In this paper, I have begun to illuminate the complex web of interactions between the deep cultural groundings of these traditional themes and their relationship with orally-derived written texts. There are more cross-cultural parallels than previous scholars have acknowledged, and most of the written sources from earlier times are supported by relatively recent and, in many cases, contemporary oral and oral-based folklore material.
Even if similar strands of snake-lore might figure in a variety of narrative genres, including vernacular beliefs, the number and range of narrative forms about snakes and bovine and human milk-stealers in pre-modern and present-day folklore should be read by scholars, I believe, as an enduring storycomplex. Any discussion of the development of any particular tale connected with milk-suckling reptiles would profit from an objective and extensive parallel consideration of pre-modern sources. But, as Patricia Kirkpatrick once noted, "[t]o designate a specific written text as a legend, a fable, or any other tale type, because it shows similarities with oral genres, does not necessarily mean that the written text was originally oral, only that it may resemble oral forms" (Kirkpatrick 1988: 97). Since our knowledge of sources is now richer than before, we must, therefore, try to understand what lies behind these similarities. The clear affiliation of pre-modern written traditions to international folklore types and motifs of oral narratives give them an undeniable family resemblance mostly centered on the notion of 'narrative content': they can very reasonably be termed 'parallels'. It would, otherwise, be impossible to fit cross-cultural evidence together in a meaningful pattern. However, it is not only a matter of continuity of content: as seen in Poggio's story, stylistic and rhetorical narrative features have continued through the centuries as well. In Ellis' words, certainly "these legends are part of an unbroken performance tradition" (Ellis 2001: 87), which has retained, with great stability, the folkloric pattern of the milk-suckling serpent from ancient times up to the present day. The setting of these tales, involving livestock or nursing women, was deeply rooted in the everyday world and in the natural and zoological notions of community members.
It is not easy to find an all-embracing interpretation (always assuming that one exists) for the extraordinary persistence and diffusion, in many periods and places, of the oral traditions of the milk-suckling reptile, nor of its deep socio-cultural meaning(s). I have indicated, above, common themes and fundamental problems that considerably change our historical perception. Nevertheless, I have privileged the description and careful mapping of sources at the expense of their ultimate interpretation(s). "True grand theories", Alan Dundes provocatively said in his celebrated 2004 plenary address to the American Folklore Society, "allow us to understand data that would otherwise remain enigmatic, if not indecipherable" (Dundes 2005: 389; see also the heated debate, starting from this article, in Haring 2016). One must ask, for example, why the snake is the typical protagonist in these stories, which are told with such consistency through time and space. There are over-simplistic psychoanalytic interpretations where the snake is a penetrating phallus universally connected to genital symbolism, sexual allusions, desire and oral fantasies (a snake is the obvious phallic animal). But I doubt that the application of Freudian psychoanalytic concepts focused on sexual symbolism is a useful interpretative key for understanding the entire milk-suckling reptile complex. Indeed, I suspect that this interpretation may prove somewhat arbitrary (see, for example, Charuty 1992 on the "unfinished dialogue" between anthropology and psychoanalysis).
I hope to have opened the way here, in any case, to some new research possibilities for the historical and evolutionary problem of the man-serpent relations in popular tales and in life. Here we clearly see two poles active in the symbolic treatment of the snake conceived as a member of a synanthrope species: wild/repulsion and domestic/attraction (e.g., the snake as domestic spirit) 26 . These possibilities are there in the many different genres of folklore that circulated widely in agricultural and pastoral societies in Europe and elsewhere, and in the complicated relationship between this system of categories and the written record. Supplementary empirical inquiries should concentrate, for instance, on the processes of polygenesis and the cultural and historical dynamics that made these ideas so popular across such large swathes of the world; not to mention the morphological and functional connections between ancient, medieval and modern material. And here we must ask about possible mutual cultural derivation and influence. Only when this has been done, will we have innovative theoretical generalisations about the milk-feeding serpent; generalisations that build on those already made along rationalist, anthropological, philological, psychological and historic-religious lines. ANTHROPOZOOLOGICA • 2017 • 52 (1) tions on the revised paper. I warmly thank Giovanni Canova and Hanna Urbańska, respectively, for providing Al-Jāḥiẓ's Italian translation and the English translation of her article, Krishna Del Toso, for having discussed with me a tentative retranslation of the passage from the Ṛgveda, and Hana Šedinová, for help in identifying the Czech manuscript with the illumination of the bova snake. Finally, I would like to express my gratitude to Rosa Ronzitti, who kindly read and commented on a draft of the article: without the knowledge she has shared in a number of publications, this study would have been impossible.