i. In Old and Middle Iranian

At the time of the Indo-Iranian unity, the Indo-Iranians must have imagined dragons restraining the heavenly waters and causing drought, and not releasing them until slain by a god or hero, as in the Rigvedic myth of Indra and Vṛtra. In the Iranian Zoroastrian literature, however, other than Gandarəβa who lives in the Vourukaṧa Sea, dragons are rarely mentioned in connection with water, though they are sometimes said to dwell by rivers. The demon which causes drought seems not to be a dragon (Av. Apaoša, Mid. Pers. Apōš); instead, the Zoroastrian dragons, materially huge monsters with ravenous appetites for men and horses, have been given their place in the Mazdayasnian view of the world, in which all monsters are the creations of evil and thus antagonists of the true, Mazdayasnian religion. Still but sketched, or briefly alluded to, in the extant Avesta, this aspect of the Iranian dragons is elaborated throughout the later religious writings. In Manichean myths, however, we notice a change in the concept of the monsters, which are now located in the oceans, presumably as the result of Mesopotamian influence.

The most common Indo-Iranian word for dragon, Indian ahi, Avestan aži, originally meant only “snake,” a meaning which Avestan aži still has beside “dragon.” These two words are etymologically related to words in other Indo-European languages such as Latin anguis (hence anguilla, related to Germanic “eel;” see further Mayrhofer, Etymological Dictionary I, p. 68, and III, p. 638). In later Iranian the word aži has mostly been replaced, partly for reasons of linguistic taboo, partly probably for phonetic reasons. Thus only Yidgha and Munji still have (y)īž < aži; Middle and New Persian have mār, which may derive from *marθra “killer” (though there are phonetic difficulties in such a derivation) and kirm/kerm “worm”, Av. kərəma, Sogd. kyrm- (translating Mid. Pers. azdahāg, Parth. aždahāg, see Henning, Sogdica, pp. 21f.); Shughni has sāɣ, which may be from *sušnā “the hisser;” etc. (see Morgenstierne, “An Ancient Indo-Iranian Word for “Dragon”” on this and other words for “snake” and “dragon” in Iranian; on “snake” in Indo-Aryan, see G. Buddruss, “Zur Benennung der Schlange”).

Other dragons or dragon-like monsters in Old and Middle Iranian are the Avestan Gandarəβa (Pahlavi Gandarb/Gandarw), the Pahlavi Kirm (battled and vanquished by Ardašīr I, see below, AŽDAHĀ II), the Zoroastrian Middle Persian Gōčihr and Mūšparīg, and some of the Manichean Middle Persian mazans.

Indo-European and Indo-Iranian connections. Myths of dragons and the slaying of dragons were common among both other Indo-European peoples and the Near-Eastern peoples with whom the Iranians came into contact from the first half of the first millennium B.C. We need only recall the Teutonic myths of the Nibelungen and Beowulf on the one hand, and the Babylonian dragon-slaying myths on the other. The myth which relates how Dahāg was chained to Mount Demāvand by Ferēdūn but is unchained at the end of time (see below) may reflect Indo-European myths of monsters which are vanquished by a god or hero and imprisoned or chained, but sometimes are liberated at the end of time and come forth to wreak havoc among gods and men. In Greek mythology Zeus battles the Titans and imprisons them in Tartarus; according to some authors, he later set them free (see, e.g., Harvey, The Oxford Companion, p. 126a). In the Scandinavian mythology, the monstrous Fenris wolf is chained by the god Týr, but at Ragnarokk (Götterdämmerung) it is unchained and is fought by  Óδinn, whom it swallows, but is itself slain by  Óδinn’s son (see. e.g., Davidson, Gods and Myths, pp. 38, 59). It is of course difficult to establish detailed connections between these various Indo-European myths, and some scholars prefer to see individual developments rather than elements inherited from a distant past (see, e.g., Boyce, Zoroastrianism I, p. 283).

In Indian mythology the only dragon of importance is the snake/dragon (ahi) vanquished by Indra and usually referred to as Vṛtra. The origin of the name has been much discussed and is important for Indian as well as Iranian mythology since in Iranian the standing epithet of Indra, vrṭra-han/Vṛtra-han “smiter of obstacles or defenses/slayer of Vṛtra,” corresponding to Av. vərəθraγan, was thematicized and came to designate one of the most important gods in the Iranian pantheon: Av. Vərəθraγna, Mid. Pers. Wahrām, NPers. Bahrām. The ahi Vṛtra is described in the Rigveda as keeping the (heavenly) waters imprisoned in caves in the mountains. With the vajra (in the Avesta vazra is the chief weapon of Miθra), Indra smites him on the neck, splits his head thus freeing the waters, which immediately rush out like cows and run to the sea. In India the epithet is given also to Agni (the fire-god) and Soma; in the Avesta also to Haoma. In the Avesta and the Zoroastrian Pahlavi texts, however, the epithet is never given to any of the dragon-slaying heroes and no gods (including Vərəθraγna) slay dragons. Nevertheless, this concept may have had a place in the mythology of at least some of the Iranian peoples since in Manichean cosmology the dragon-slaying god Adamas is called Wšγnyy in Sogdian (< OIr. Wrθragna; see further below) and also elsewhere in the later tradition some local counterparts of Vərəθraγna preserve traces of his dragon-killing function: Thus, e.g., Armenian Vahagn kills a dragon (see also Duchesne-Guillemin, La religion, pp. 175-78; and BAHRĀM).

Whereas in the Indian myths the dragon-slayer is the warrior god, in the Zoroastrian myths the dragon-slayers are superhuman heroes: Θraētaona, who slays Aži Dahāka, and Kərəsāspa, who slays Aži Sruuara, the horned dragon. Kərəsāspa seems to have no parallel in Indian myths; however, Θraētaona, son of Āθβiia, appears to be related to another Avestan hero, Θrita, whose name corresponds formally to the Indian Trita Āptya; the patronym Āptya, however, corresponds to that of the Avestan Θraētaona. In the Hōm yašt (Y. 9) the first mortals to press the haoma are enumerated: the first was Vīuuaŋᵛhan, father of Yima, the second Āθβiia, father of Θraētaona, and the third was Θrita (literally “the third”), father of Uruuāxšaiia and Kərəsāspa. In the Rigveda, Trita Āptya is portrayed as the first sacrificer to prepare the soma. Clearly these various mythical persons are related (see Boyce, Zoroastrianism I, pp. 99f.). The Rigvedic Trita Āptya is no dragon-slayer but he does appear in a myth which bears great similarity to that of Indra’s slaying of Vṛtra: In addition to this feat, Indra also liberates some cows which are held imprisoned in a cave by a certain Vala (whose name may or may not be etymologically related to Vṛtra: IE. root *ṷel-; cf. Eng. “wall”) but sometimes this feat is ascribed to Trita Āptya.

It therefore seems clear that although dragon-slaying gods and heroes were part of Indo-Iranian mythology, India and Iran developed distinct myths early, changing, deleting, and adding details. In India dragon-slaying was made a characteristic feature of the god Indra. The notion of a god of victory *Vrtraǵhan “smiter of obstacles/defenses” was probably also common heritage, but whereas the epithet in Iran became the name of the god himself, in India it was given to the warrior god Indra, prompted by his connection with the dragon-slaying (cf., e.g., Rigveda 1.32.2 ahann ahim “he struck/slew the dragon,” with the same verb han-/ghn-, Av. jan-/γn– “to strike, kill,” as in vṛtra-han-/ghn-, vərəθra-jan/γn(a)-). For succinct overviews of the arguments and various theories, see Duchesne-Guillemin, La religion, pp. 175-78; Boyce, Zoroastrianism I, pp. 63f., especially p. 64 nn. 279-80; see also BAHRĀM).

The only other dragon/snake mentioned in the Rigveda is the “dragon of the deep” (ahi budhnya), who is mentioned in lists of lesser divinities and said to dwell at the bottom of heavenly rivers (budhne nadīnāṃ rajaḥsu sīdan; Grassmann, Wörterbuch, cols. 909f.).

In India the dragon-fight was symbolically connected with New-year and the end of drought but in ancient Iran there is no trace of a connection between the killing of the dragon and Now Rūz. Scholars attempting to see such a connection (e.g. Dumézil, Le problème des centaures, pp. 72f., and Widengren, Religionen, pp. 41-49) have failed to prove it (Boyce, Zoroastrianism I, p. 102 n. 110).

In Mithraism a reference to a simulated dragon-slaying is found in a passage from Lampridius (Commodus 9), quoted from Loisy by Widengren, Religionen, pp. 44f. n. 16, but otherwise this myth seems to be quite absent from Mithraism, where the snake apparently was “a symbol of a beneficial, life-giving force” (Hinnells, “Reflections,” p. 295).

Dragons and dragon-like monsters in the Zoroastrian scriptures. 1. Ažis. 2. Gandarəβa. (On Gōčihr and Mūšparīg see Dragons in astrology, below.)

1. Ažis. Physical descriptions are found of several ažis in the Avesta:

Aži Dahāka had three mouths (θrizafanəm), three heads (θrikamarəδəm), and six eyes (xšuuaš.ašīm). (See further on Aži Dahāka below.)

Aži Sruuara, the horned dragon, also called Aži Zairita, was the yellow dragon that Kərəsāspa slew (Y. 9. 1; Yt. 19.40; see on the Legend of Kərəsāspa below) “who swallowed horses (aspō.gar-), who swallowed men (nərə.gar-), the poisonous yellow one, over whom poison flowed the height of a spear;” also described as (Y. 9.30) “the terrifying (sima-), poison-spitting (vīšō.vaēpa-) dragon.” The Dādestān ī dēnīg (71) appears to have preserved two more old epithets of Aži Sruuara no longer found in the Avesta; here we read that Aži Srūwar was one of the seven worst sinners (two of the others being Dahāg and his mother Wadag, see below), being close to Ahriman himself, and that in addition to swallowing men and horses in a terrifying way (sahmgenīhā . . . asp ud mard-ōbārīh kard), it was also a highway robber (rāhdārī ud rāhbīmēnīdārīh . . . kard; cf. the dragon in Aogəmadaēcā below and note Pahlavi Rivayat, p. 69 par. 16, where Kirsāsp tells Ohrmazd that he has killed seven gigantic rāhdārs). In the Dēnkard it is said to be skilled in witchcraft (Dresden, p. 91 [109.6]; Madan, p. 747.20; tr. West, 8.5.23, p. 111; it is spelled slwbl yz and slwblyz). There seem to be no similar creatures in the old Indian mythology; however, in Sumero-Semitic culture, art and literature, horned and multi-headed dragons and other monsters are common-place. The number of heads is often seven (see, e.g., the illustrations in Heidel, The Babylonian Genesis; see also Christensen, Démonologie, pp. 20-23).

A dragon guarding a road is described in Aogəmadaēcā 78, where it is compared with Vayu: “the road which a dragon guards, horse-*crushing (aspaŋhāδō), man-*crushing (vīraŋhāδō), man-slaying (vīraja), without compassion.” (The element –hāδa– may belong to the root had-, OInd. sad– “to treat roughly” according to H. W. Bailey, “Arya”, p. 526, who quotes Yt. 14.56 where we find the parallels jana . . . nōiṱ janən haδa . . . nōiṱ haδən). The descriptions of ažis in the Vidēvdād are of snakes rather than of dragons: “of the snakes that crawl on their bellies (ažinąm udarō.θrąsanąm)” (Vd. 145; on θrąsa– see Hoffmann, Aufsätze I, p. 197 n. 2); “swift snakes (ažaiiō xšuuaēβåŋhō)” (Vd. 18.65).

The Aži Raoiδita, the red dragon (in contradistinction to the Aži Zairita “yellow dragon” = Aži Sruuara), ought to have been one of the most important Avestan dragons (excepting Aži Dahāka, see below) since it was, together with the “daēuua-created winter” (ziiąmca daēuuo.dātəm), Aŋra Mainiiu’s counter-creation (paitiiārəm frākərəṇtaṱ) to Ahura Mazdā’s creation of Airiiana Vaējah (Vd. 1.2; see Christensen, Le premier chapitre du Vendidad, pp. 23, 26-27), and thus should by rights have been the most loathed creature in the original home of the Iranians. However, it is mentioned only here. The Aži Raoiδita was probably not identical with Aži Dahāka. As dragons, the only point they had in common was that they were created by Aŋra Mainiiu as the worst thing in the world, but Aži Dahāka is nowhere said to be red, and is nowhere connected with winter. The Pahlavi translation and commentary has az-iz i rōdīg; was bawēd “and the river snake; there is a large number” without further comments. The corresponding text in the Bundahišn (chap. 31) describes the dragon of the counter-creation as mār ī pad parrag ud ān-iz ī nē pad parrag “the snake with wings and the one with no wings” (TD₁, p. 176.1-2; TD₂, p. 205.7-8; tr. Anklesaria, pp. 264f.; cf. Christensen, op. cit., p. 27).

An Aži Višāpa is mentioned in Nirangistan 48, where it is said that the act of offering libations to the waters between sunset and sunrise is no better than throwing them into the *mouth of the Aži Višāpa. The two manuscripts of the Nirangistan (neither of them old and trustworthy) have vṧāpahe (TD) and viṧāpahe (HJ), the Pahlavi rendering has MYA Y ŠPYL “good water,” which may or may not be a scribal corruption of ʾp Y wš, i.e., az ī wiš “poison snake,” as suggested by Waag (p. 109; see Sanjana’s text p. 197, variants listed on p. 43). The epithet višāpa is commonly interpreted as containing the word vī/ĭša– “poison” (Bartholomae, AirWb., col. 1473 “whose juices are poison,” Boyce, Zoroastrianism I, p. 91 n. 42 “dragon with poisonous slaver,” etc.) and although such an interpretation lies close at hand in view of the other dragons’ association with poison, the Armenian form of the word, višap (Georgian vešapi, also a fabulous serpent; Syriac wšpʾ) must be derived from *vēšā/ăp (Hübschmann, Armen. Etymologie, p. 247; Benveniste, “L’origine”), i.e., Old Iranian *ṷaiš– (Benveniste, p. 7, reconstructs *vā/ăišapa– but still assumes some derivative of “poison”). The etymology of the term is still unclear; besides the older proposals (cf. AirWb.) note that Rigvedic has two verbs viṣ “to pour out” (intr.) and “seize,” both used in the context of “water” and the second in a context involving Indra, the vajra, and the dragon ahi, but neither verb gives a totally satisfactory meaning for the term. Perhaps we should compare Avestan vaēšah “foulness,” so that this dragon was originally the dragon “of foul waters” or the dragon “which fouls the waters.” The importance of the epithet lies in the fact that Armenian višap has become the designation of a whole class of dragons (see Benveniste, “L’origine”) and the hero Vahagn is there called višapakʿal “dragon-slaying.”

In the Middle Iranian period, an aždahā was often depicted on banners to frighten the enemy by its ferocious aspect. Such banners are referred to several times in the Šāh-nāma as aždahā-peykar (e.g., ed. Borūḵīm, II, p. 480 v. 775, IV, p. 924 v. 949). An early reference to them is found in Lucian (De historia conscribenda 29, pp. 42f.), where we are told that the Parthians used banners with different emblems to differentiate divisions of their army, a dragon-banner (drákōn) preceding—Lucian believes—a thousand-man division. (See also AŽDAHĀ II)

2. Gandarəβa/Gandarw (or Gandarb, spelled gndlp). Among the other various noxious creatures depicted in the Zoroastrian Pahlavi literature we find the sea monster Gandarəβa, which may have been a dragon of the sea, though descriptions as to its exact nature are lacking. It was a monster with yellow heals (Zairipāšna-) and living in the sea (upāpa-), which, on emerging to destroy the entire creation of Ašša, was fought and vanquished by Kərəsāspa (Yt. 5, 38, 15.28 19.41; cf. Christensen, Démonologie, pp. 18f., and see below). Like the other dragons it had a ravenous appetite, even more so since it was able to swallow twelve provinces at once (pad ēw-bār 12 deh bē jūd; Pahlavi Rivayat, p. 67 par. 9). Etymologically the name equals OInd. gandharva, a beneficent mythical being, said to be surrounded by the heavenly waters—which flow down at his look—sometimes (in later literature, usually) portrayed as a heavenly musician. It is through the “Iranian polarization” of the inherited Aryan mythological concepts, it seems, that gandarəβa has been turned into a sea monster. Of all the Old Iranian monsters Gandarəβa is the most reminiscent of Near-Eastern, Semitic, sea monsters. (On the OInd. gandharva see, e.g., Oldenberg, Die Religion des Veda, pp. 248ff.; Grassmann, Wörterbuch, col. 378; Dumézil, Le problème, devotes several chapters to this monster.) It also recalls the Manichean water dragons, the mazans (see below). The gandarəβa survives in Sogdian as γntrw (for genuine Iranian *γntṛβ through Indian influence) and the entire Avestan phrase (upāpō gandarəβō) as wpʾpγntrw (P 3.131, see Henning, “Sogdian Tales,” pp. 481f.); it has survived in modern Shughni dialects as the designation of a monster or dragon, but also a werewolf (see Morgenstierne, Etymological Vocabulary, p. 110).

Aži Dahāka. Aži Dahāka (Pahl. Az[i]dahāg, spelled ʾcydhʾk, or Dahāg) belongs to the realm of mythologized history or historicized mythology. He is depicted in the Avesta as a dragon-like (aži) monster with three mouths (θrizafanəm), three heads (θrikamarəδəm), six eyes (xšuuaš.ašīm), with a thousand viles (hazaŋrā.yaoxštīm), very strong (aš.aojaŋhəm), a demoniac devil (daēuuīm drujim). For the rest he behaves like the other heroes and non-heroes of the Avestan mythological prehistory, and it is not clear whether he was originally considered as a human in dragon-shape or a dragon in man-shape. The former alternative is suggested by his epithet dahāka– if it means “man(-like)” (Schwartz, Orientalia 49, pp. 123f., who for the word formation compares maṧiia– and maṧiiāka-, and for the meaning compares Khotanese daha– “male,” Wakhi δāi “man,” and translates Aži Dahāka as “he hominoid serpent, the Snake-man”). However, the traditional connection of Av. dahāka with the OInd. dāsas and dasyus, who are also among the opponents of Indra and who are usually assumed to be the indigenous pre-Aryan inhabitants of northwestern India can not be wholly discarded. (Note that Pashto lōy “big” is likely to be from dahāka, which may point to an original meaning “big, huge.”) (See also Christensen, Démonologie, pp. 20ff.)

A number of elements of the myth of Aži Dahāka have been preserved in the Zoroastrian texts: In their struggle to regain the xᵛarənah after it left Yima, Ahura Mazdā and Aŋrō Mainiiu each employed their best helpers. Ātar and Aži Dahāka with his brother Spitiiura here faced one another and threatened one another, but the threats of Ātar were the most efficient ones, and Evil’s attempt was foiled. According to the Avestan myth Spitiiura sawed Yima in half (Yt. 19.46 yimo.kərəṇtəm “the Yima-cutter”), but the Bundahišn states that he did it together with Aži Dahāka (TD₁, p. 196. 13-14; TD₂, p. 228.12; tr. Anklesaria, pp. 292f.; tr. West, p. 131; see also Darmesteter, Zend-Avesta III, p. 629 n. 76). According to the Pahlavi Rivayat (46.35, p. 136), it was as a recompense (pāddāšn) for his success against Dahāg that Ādur Farnbag was established victoriously in Ḵᵛārazm, and the Bundahišn says that when Jam was cut in half, the xwarrah of Jam saved Ādur Farnbag from the hand of Dahāg (TD₁, p. 102. 15-17; TD₂, pp. 124.14-125.1; tr. Anklesaria, pp. 158f.; tr. West, p. 63).

In the Avesta (Yt. 5.29-35 and 15.19-21) we are told that Aži Dahāka worshipped Arduuī Sūrā in the land of Baβri and Vaiiu in the inaccessible (dužita) Kuuiriṇta. The tradition has interpreted Baβri as Babylon, Old Persian Bab(a)iruš (Mid. Pers. Bābēl) and Kuuiriṇta as Aži Dahāka’s castle in Babylon. The Bundahišn, in a list of dwellings (mānīhā) made by the kays, reports that Dahāg made a dwelling in Babel called Kuling dušdīd (TD₁, p. 179.11-12; TD₂, p. 209.8; tr. Anklesaria, pp. 268f.; see further Darmesteter, Zend Avesta II, pp. 584f. n. 16). In the Dēnkard (7.4.72) it is told that Dahāg by sorcery had made many wonderful things (widimās) in Bābēl, which induced people to idolatry, in order to destroy the world but that Zardošt recited the words of the Religion and thus rendered nought Dahāg’s efforts (Dresden p. 119 [54]; Madan, p. 639.5-10; Molé, La légende, pp. 56f.). These interpretations are of doubtful historical and geographical value, not least because of the Avestan form of the name, Baβri, which differs from the others; nevertheless it is understandable that the Iranians, after they came into contact with Near-Eastern, especially Semitic culture, located Aži Dahāka, the big dragon, in Babylon, which must have been notorious for its dragons, in literature and artistic representations. Another mansion was constructed by Dahāg in *Šambarān (wr. yʾmblʾn; cf. Šāh-nāma Šambarān) and one in India (TD₁ , p. 179.15; TD₂, p. 209. 11-12; tr. Anklesaria, pp. 270f.). Only once, it seems, is Dahāg associated with a river, namely in the chapter on rivers in the Bundahišn where he is said to have asked a favor from Ahriman and the demons by the river Sped in Azarbaijan (TD₁, p.82.15-17; TD₂, p. 88.3; tr. Anklesaria, pp. 108f.; tr. West, p. 80).

Aži Dahāka’s prayer to Arduuī Sūrā and Vaiiu was for them to give him the power to render unpopulated (amaṧiia-) the seven climes, i.e., the entire world; Arduuī Sūrā and Vaiiu of course did not grant this prayer; on the contrary, when Θraētaona subsequently worshipped them, asking them to grant him the power to overcome Aži Dahāka, it was him they granted his wish. The same two passages contain another fragment of the myth: At the same time, Θraētaona asked and was granted the power to lead away Sauuaṛŋhauuāci and Arənauuāci, the two most beautiful women in the world, whom the later tradition represented as Yima’s sisters or daughters and as having been captured and detained by Aži Dahāka (further on the capture of beautiful women by dragons, see AŽDAHĀ II). More information originally contained in the Avesta is to be found in the later Pahlavi texts, especially in the resumés of the nasks given in the Dēnkard (books 8 and 9, tr. West, SBE 37, and elsewhere); in particular, the twentieth fragard of the Sūdgar nask, called Vohuxšaθrəm, was devoted in its entirety to the rule of Dahāg.

In the Pahlavi texts Dahāg is portrayed as the embodiment and originator of the bad religion, i.e., the opposite of the Good Mazdayasnian Religion. In Dēnkard 3.229 we are told that the bad religion and the non-law was codified by Dahāg in the writings of Judaism (the ʾwlytʾ, Syriac Urāyθā, i.e., the Pentateuch), and that from Dahāg it went to Abraham, the dastūr of the Jews (tr. Menasce, p. 243; cf. Zaehner, Zurvan, p. 30). In the Sūdgar nask Dahāg was said to have possessed five defects (greediness, want of energy, indolence, defilement, and illicit intercourse), the opposites of the best qualities wisdom, instructed eloquence, diligence, and energetic effort (according to the resumé in Dēnkard 9.5.1-2, DH, p.172.8-9; Madan, p. 789. 1518; tr. West, p. 177). In Dēnkard 3.308 Dahāg, destroyer of the world is said to have been of Arabic (tāj) race. Dēnkard 3.287-88 lists ten good counsels to mankind given by Jam followed by ten bad, counter-counsels by Dahāg, which are referred to also elsewhere in the Dēnkard (tr. Menasce, pp. 243, 283-85; the ʿOlamāʾ-e eslām contains a note that the name Dahāk actually means “ten sins,” see Persian Rivayat, p. 454, and Zaehner, Zurvan, p. 413). In the Dādestān ī dēnīg (71) Dahāg is said to have been one of the seven worst sinners ever, i.e., those who are close to Ahriman himself (two more being the Az ī Srūwar and Dahāg’s mother Wadag); here Dahāg is said to be the first who lauded (stāyīd) sorcery (jādūgīh) (cf. also Dēnkard 8.35.13 and 9.10.2-3; tr. West, pp. 111, 185). He is often referred to as Bēwarasp in the Pahlavi texts (e.g., Dēnkard 9.21.7; tr. West, p. 214; Mēnōg ī xrad 7.29, 26.34, 35, 38; tr. West, pp. 35, 60f.; Bundahišn TD₁, p. 66.7-8; TD₂, p. 80.6-7; tr. Anklesaria, pp. 98f.; tr. West, p. 40).

A curious note is found in Mēnōg ī xrad, chap. 27, in which the sage asks about the benefit of all the ancient rulers. About Azdahāg the Mēnōg ī xrad says that the advantage (sūd) of Azdahāg Bēwarasp and Frāsyāg the Turanian was that if they had not received the rule it would have gone to Xešm (wrath) and then it could not have been taken from him till the end of the world because Xešm has no bodily existence.

The Pahlavi texts moreover provide Dahāg with a mother, who is the embodiment of evil and sinfulness and is one of the seven worst sinners ever (see Dādestān ī dēnīg 71, 77). Her name is variously given as Ōdag, Wadag, etc. She was the first to have practiced whoredom (rōspīgīh) and incestuous adultery, having intercourse with her son while her husband Arwadasp/Urwadasp (or Xrūdasp as West, tr., Bundahišn, p. 131?) was still alive, and without his sanction (adastūrīhā) and unlawfully (adādestānīha) (TD₁, p. 196.17-197.3; TD₂, pp. 228.15-229.4; tr. Anklesaria, pp. 292f.; tr. West, pp. 131f.). (In the same vein one might expect Dahāg to have instituted the heinous sin of sodomy, but that had of course already been done by Ahriman, when he performed it on himself to create “demons and lies and other abortions,” see Mēnōg ī xrad 7.10, tr. West, pp. 32f.; Zaehner, Zurvan, pp. 368f.) She is described in some detail in Dēnkard 9.21.4-5, which seems to allude to some relationship between her and Jamšēd (DH, p. 188.18-21; Dresden, p. 58 [175]; Madan, p. 810.19ff.; tr. West, pp. 212-13 n. 5). Being the most sinful of females, by one commentator of Vd. 18.30 Ōda is identified with the druj that tells Srōš who are the four males who make her pregnant. However, this is probably a late interpretation since the summary of Vd. 18 in the Dēnkard has only druz (see Darmesteter, Zend-Avesta II, pp. 248f. and n. 43) in Vd. 19.6, Aŋra Mainiiu promises Zaraθuštra a boon such as the one the ruler (daiiŋˊhupaiti) Vaδaγana got, if he forswears the Mazdayasnian religion. The passage is found also in Mēnōg ī xrad where Wadagān dahibed is glossed by Dahāg (56.24-25, pp. 154f.; tr. West, p. 103 with n. 3). This means that the tradition took Av. vaδaγana to be a matronymic referring to Aži Dahāka, and it is possible, of course, that the whole character of Wadag/Ōda is built upon this interpretation.

The rule of Aži Dahāka. From the Pahlavi texts onwards, Dahāg is inserted into the list of mythical Pīšdādīān (Pēšdādīān) rulers of Iran, i.e., the rulers descended from Hōšang ī Pēš-dād (Haošiiaŋha Paraδāta), succeeding Jam ī Xšēd (Yima Xšaēta) and preceding Frēdōn (Θraētaona). The Avesta does not say explicitly whether Dahāg was a king or not, but from the way he was mentioned among the early rulers of the Iranians, it was quite natural that he should be considered as such. An attempt was made by S. Wikander to trace the triple succession of Yima-Aži Dahāka-Θraētaona back to Indo-European patterns, comparing the Greek myth of the Ouranides, according to which Zeus conquered Kronos and the Titans and established a reign of order; however, Duchesne-Guillemin (La religion, pp. 336f. with references) has convincingly argued against such a connection, pointing out the differences between the Greek and Iranian myths and, most importantly, pointing out the fact that there is no trace of this kind of triple succession in India, which precludes an Indo-Iranian date for it. It seems better to assume that it is a post-Avestan creation, due to a certain interpretation of the Avestan texts.

In the Pahlavi texts the reign of terror of Dahāg is described in some detail (see, e.g., Dēnkard 9.21.12-16). His genealogy is given in the Bundahišn chap. 31.6, which traces it back to the Evil Spirit himself (TD₁, pp. 196.17-197.3; TD₂, pp. 228.15-229.4; tr. Anklesaria, pp. 292f.; tr. West, p. 131f.). About his rule we read in the Bundahišn that a hundred years after the xwarrah left Jam the millennium reign (hazārag xwadāyīh) came to Scorpio (Gazdumb) and then Dahāg ruled for a thousand years, until the millennium rule came to Sagittarius (Nēmasp) and the five hundred-year rule of Frēdōn (TD₁, p. 206.1-2; TD₂, p. 239.4-5; tr. Anklesaria, pp. 306f.; tr. West, p. 150; briefly mentioned in Dēnkard 3.329, Menasce, p. 308; Mēnōg ī xrad 57.25). In the ʿOlamāʾ-e eslām we are told that after Jamšēd became deranged, he was seized and slewn by the Arab Dahāk who made himself king and reigned a thousand years, mixing men and demons, and working much sorcery, until Ferēdūn, son of Ātfī, came and bound him (Persian Rivayat, p. 454; Zaehner, Zurvan, p. 413).

Aži Dahāka and Θraētaona. The Avesta contains several references to Θraētaona’s victory over Aži Dahāka (Y. 9.8, Yt. 5.29-35, 14.40, 15.23-25, 19.37, 92, Vd. 1.17), but there is little detail, except in the case of his liberating Arənauuāci and Sauuaŋhauuāci. The Pahlavi texts contain some further details: In the Dēnkard (7. 1.25-26) we are told that by the power of the xwarrah, which came to him while he was still in the womb of his mother, Frēdōn was able at the age of nine to go forth and vanquish Dahāg and to deliver Xwanirah from the ravages of the lands (dehān) of Māzandarān (Dresden, p. 359; Madan, p. 596.2-12; tr. Molé, p. 9; see also Dēnkard 9.21.17-24, DH, pp. 190.8-191.21; Dresden, pp. 55-57 [178-82]; Madan, pp. 812.19-815.1; tr. West, pp. 217ff.). The Dēnkard (9.21.8-10) relates how Frēdōn first struck Dahāg with his club upon the shoulder (frēg), the heart, and the skull, without killing him, and that he then hewed him with a sword three times, which caused the body of Dahāg to turn into (gaštan) various noxious creatures. Seeing this Ohrmazd told Frēdōn not to cut Dahāg so that the world should not become flooded with reptiles and other noxious creatures (DH, p. 189.8-14; Dresden, p. 58 [176-77]; Madan, p. 811.13-21; tr. West, p. 214). This curious episode is hard to explain, but one is reminded of cosmogonical myths in which a giant is partitioned to give rise to the various elements of the world, as the Indian primeval Man, Puruṣa. Maybe the episode reflects some early element of the story of the creation by the Evil Spirit.

But most importantly the Pahlavi texts attribute to Aži Dahāka and Frēdōn/Kərəsāspa an eschatological role: Aži Dahāka is not killed by Frēdōn, but captured and chained “with awful fetters, in the most grievous punishment of confinement” at Mount Demāvand (Dēnkard 9.21.10, tr. West, p. 214; see also, e.g., Bundahišn TD₁, p. 66.7-9; TD₂, p. 80.6-7; tr. Anklesaria, pp. 98f.; tr. West, p. 40; Mēnōg ī xrad 26.38; tr. West, p. 61). At the beginning of the millennium of Ušēdarmāh, the druz of the seed of dragons (az-tōhmag) will be destroyed. Aži Dahāka breaks loose from the fetters and rushes out to terrorize the world, devouring one third of men, oxen, sheep, and other creatures of Ohrmazd, and smiting the water, the fire, and the plants. The last three then request from Ohrmazd that Frēdōn should be resuscitated to combat him. In the event, however, it is not Frēdōn but Kirsāsp (in the texts variously called Sām or son of Sām) who is reawakened and kills the dragon. The reason for this is not clear though it may be connected with the statement (see above) found in the Pahlavi texts that Ohrmazd refused Frēdōn permission to slay Dahāg because as a result the earth would have been flooded with noxious creatures (xrafstar). (See Bundahišn TD₁, pp. 188.12-189.1; TD₂, pp. 219.14-220.5; tr. Anklesaria, pp. 180-83; tr. West, p. 119; and see also, e.g., Bahman yašt, p. 128; Dēnkard 6.B4 [ed. Shaked, pp. 134f.], 9.15.2; Šāyest nē šāyest 20.18, ed. Kotwal, pp. 87f.; A. Christensen, Démonologie, pp. 20-25, 52). That this role was early assigned to Kirsāsp/Sām is clear from Yt. 13.61 and Mēnōg ī xrad 61.20-24 (tr. West, p. 110), where it is told that the body of Sām is protected by 99,999 fravašis of the righteous so that demons and fiends may not harm it.

The legend of Kərəsāspa/Kirsāsp. Kərəsāspa’s slayings of the Aži Sruuara and the Gandarəβa are only alluded to in the Avesta itself, but elaborated in the later Pahlavi writings. His slaying of Dahāg is recounted only in the Pahlavi and later texts. The Pahlavi texts contain two versions of the legend of Kirsāsp (spelled krssp; also Krišāsp, Grišāsp spelled klyšsp, glyš(ʾ)sp), one in the 14th, fragard of the Sūdgar nask as retold in the Dēnkard (DH, pp. 182.7-83.2 [klsʾsp. glsʾsp]; Dresden, pp. 65-66 [klyšʾsp]; Madan, pp. 802.14-803.12; tr. West, 9.15, pp. 196-99), another in the Pahlavi Rivayat (ed. Dhabhar, pp. 65-74). Both accounts were edited and translated by Nyberg, “La légende.”

From the Avesta we only learn that Kərəsāspa for some reason or other had settled upon the back of the dragon—presumably thinking it to be a hill—to cook his midday meal (Y. 9.1, Yt. 19.40) and that the heat from his fire made the dragon hot and sweaty and finally woke him up, whereupon he jumped up from underneath the cooking pot, scattering the boiling water, and frightening Kərəsāspa, who fled, but eventually slew the dragon. The account in the Sūdgar only states that Kirsāsp killed the horned dragon, but the one in the Pahlavi Rivayat adds a few details (this version is also found in Dârâb Hormazyâr’s Rivâyat, p. 62, not translated in Unvala, Persian Rivayat): Kirsāsp tells Ohrmazd that there was a horned dragon (Pers. aždahā-ī), swallowing men and horses, which has teeth as large as his arm, ears as large as fourteen nmt’s, eyes as large as a chariot, and a horn as large as a šʾk (Pers. haštād araš “eighty ells”). He ran after it (pad pušt hamē tazīd; Pers. bar pošt-e vey “on its back”) for half a day until he caught up with its head, struck his mace at its neck and killed it. The Persian Rivayat adds that when he looked into its mouth he saw men hanging from its teeth, a feature which the Pahlavi Rivayat reserves for Gandarəβa/Gandarw. Kirsāsp tells Ohrmazd that Gandarw, large enough to devour twelve provinces at once (and so tall that the sea reached him to the knee and his head reached the sun according to the Persian Rivayat), seized him by the beard and pulled him into the sea where they fought for nine days and nights, when Kirsāsp managed to seize him by the foot and promptly pulled off his skin from his feet to his head and used it to bind the monster, which he left to his friend Axrūrag to guard. Then, after eating fifteen horses, Kirsāsp fell asleep under a tree and Gandarw pulled Axrūrag and Kirsāsp’s wife and parents into the sea. All the people came and roused him and he ran down to the sea taking a thousand strides in one. Arriving at the sea he delivered those abducted, seized Gandarw, and killed him. The eschatological role played by Kirsāsp is not mentioned in the Pahlavi Rivayat but the Sūdgar tells how, when Dahāg runs free of the fetters to destroy the world, he (i.e., Kirsāsp) is awakened (hangēzīhēd) to vanquish the powerful demon (Dēnkard, DH p. 182.16-18; Dresden, p. 65 [161]; Madan, p. 803.3-6; tr. West. pp. 198f.; see also Christensen, Démonologie, pp. 17f., 51).

Dragons in astrology. Zoroastrian and Manichean astrology know of several dragons or snake-like monsters. Gōčihr and Mūšparīg. In the Bundahišn the snake-like (mār homānāg) Gōčihr and Mūšparīg with the tail (dumbōmand) and wings (parrwar) are said to be the evil opponents of the sun, moon and stars. These two harmful beings were bound to the sun so as not to run free and cause harm (Bundahišn TD₁, p. 43. 11-17; TD₂, pp. 52.12-53, 13; cf. Bundahišn TD₁, p. 159.12ff.; TD₂, p. 188.4ff.; tr. Anklesaria, pp. 242f.; tr. West, pp. 113f; MacKenzie, “Zoroastrian Astrology,” pp. 513, 516; Zaehner, Zurvan, pp. 159, 164). Both are probably derived from Avestan concepts: In the Avesta gaociθra is an epithet of the moon and Mūš Pairikā “the witch Mūš” is found in Y. 16.8, where she is mentioned in connection with Āzi, the demon of greed (see ĀZ) and the heretic (aṧəmaoγa).

In the Pahlavi cosmogony Gōčihr is described as “similar to a snake with the head in Gemini (dō-pahikar) and the tail in Centaurus (Nēmasp), so that at all times there are six constellations between its head and tail.” It runs retrograde, so that every ten years the head and tail have changed place. It is said to be standing in the middle of the sky, an expression which may refer to the polar region of the sky and so perhaps contain a reminiscence of the constellation Draco, which circles the pole (see MacKenzie, “Zoroastrian Astrology,” pp. 515f. with notes). In the ʿOlamāʾ-e eslām it is stated that the heaven of Gōčihr is below the heaven of the moon (Zaehner, Zurvan, p. 417; cf. Persian Rivayat, p. 429 bottom; see also Škand-gumānīg wizār, ed. Menasce, pp. 47, 55, 60).

At the end of time Gōčihr will fall down on the earth, which it will terrify like a wolf does a sheep; its fire and halo will then melt the metal of Šahrewar in the hills and mountains, thus providing the river of molten metal necessary for the purification of men. (Gōčihr appears to be the only fiery dragon in ancient Iran.) At the end, after Ohrmazd himself has come down to earth to send Āz and Ahriman back to the Darkness whence they had come, Gōčihr the serpent burns in the molten metal and the pollution of Hell burns and Hell becomes pure (Bundahišn TD₁, pp. 193.11-16, 195.17-196.2; TD₂, pp. 225.3-8, 227.12-15; tr. Anklesaria. pp. 288-91; tr. West, pp. 125f., 129).

Mūšparīg may originally have been considered the demon who causes the eclipses of the moon, as is indicated by its name Mūš meaning “mouse” but originally also probably “thief,” cf. OInd. muṣ “to steal” (see Darmesteter I, Zend-Avesta, pp. 144 n. 15).

In the Manichean cosmogony these two dragons are simply called “two dragons” (dō azdahāg); they were hung up (āgust) and fettered (gišt) in the lowest heaven and two angels (frēstag), male and female, were put in charge to make them revolve ceaselessly, i.e., presumably to make the firmament turn so as to keep the Manichean salvation machine going (Jackson, Researches, 30f., 31f.; Boyce, Reader, p. 60 text y 1 with note).

In later literature, the seven planets are sometimes called aždahā (see Eilers, Sinn und Herkunft, p. 11).

Dragons and dragon-like monsters in Manichean writings. The Manichean texts mention dragons in general terms (Mid. Pers. azdahāg, Parth. aždahāg) but they play no prominent role. Thus in Mir. Man. I (p. 22[194] = Boyce, Reader, p. 72 y 39) we read that the female and male mazans (q.v.) and āsrēštārs copulate to produce dragon brats (ʾwzdhʾg zhg) and in Mir. Man. III (p. 30[875]) the Living Soul complains that it was seized and mangled by innumerable demons, including “dark, ugly, stinking, black dragons” (tʾryg ʾjdhʾg dwrcyhr gndʾgʾwd syʾw).

More important in Manichean cosmology is the class of demoniacal beings called mazans. In the Zoroastrian writings these are clearly only “giants,” whom the ocean reaches to the knees. In the Manichean myths, however, they are definitely connected with the ocean and though most of them are indeterminate “giants” (e.g., the mazans which usually accompany the āsrēštārs, some are sea dragons or dragon-like sea monsters (see below).

For the history of Iranian mythology, however, the most important fact is that the dragon-killing episode has been fitted into the Manichean cosmological scheme: Here we find the third son of the Living Spirit (Mihryazd), Adamas of Light (Syr. Adamos Nuhrā), being sent by his father (Mir. Man. I, p. 10[1 82], Boyce, Reader, p. 65 text y 14), to throw down the mazan, stretching it out from east to west, and putting his foot upon it so that it could do no harm, and, in another text (M 472v, Boyce, Reader, p. 80 text z 16), we are told that he suppressed (nyrʾft) the giant dragon (azdahāg ī mazan), which corresponds to “giant of the sea” (mzṇʿy zrhyg) in a Manichean Middle Persian text; this is the gígas tēs thalássēs in the Kephalaia (I, pp. 113ff.), which is responsible for ebb and flow (see Henning, “Book of the Giants,” p. 54 and n. 3, and cf. Boyce, Reader, pp. 6, 62 text y 5 with note; and Sundermann, Parabelbuch, pp. 21f. with n. 30). Adamas’s Iranian heritage is evident in his Sogdian name, Wšγnyy Bγyy (M 583 I r8 in Waldschmidt and Lentz, Manichäische Dogmatik, p. 68 [545], with comm. 88 [565], from *wṛθragna-); whether this name implies that the Sogdians had preserved traditions of a demon-slaying Vərəθraγna or whether it is due to Indian influence, has not yet been investigated. The Greco-Syriac name Adamas/Adamos probably means “indomitable, adamant” (Cumont, “Adamas”). Augustine calls him “the belligerous indomitable hero (or “the hero Adamas”) who holds a spear in his right and a shield in his left” (adamantem heroam belligerum; Contra Faustum quoted by Cumont, p. 79). In the Chinese text edited by Waldschmidt (Waldschmidt and Lentz, pp. 9, 33 [486, 510]) he is called “the brave, strong, equipped with the ten powers, demon-subduing emissary.” This god is also called, for reasons as yet not understood, Wisbed Yazd (three other sons of the Living Spirit being the Dahibed, Zandbed, and Mānbed Yazds), and the god with four forms (yazd ī taskirb; see Sundermann, “The Five sons,” passim, and “Namen von Göttern,” pp. 101f., 127 n. 166, 131 n. 226). The mazan which he suppresses is that formed by that part of the ejected seed of the Archons which fell on the moist, and became a horrible monster (Mir. Man. I, p. 10 [182]; Syriac hywtʾ snytʾ “a terrible beast” in the likeness of the King of Darkness (see Theodore Bar Konay in Jackson, Researches, p. 247); note that Ebn al-Nadīm reports that the King of Darkness had “the head of a lion” and a “body like the body of a dragon (tannīn)” (Fehrešt, tr. Dodge, II, p. 778; Taqīzāda, Mānī wa dīn-e ū, p. 151 infra). In the Manichean Psalm-book (p. 138.41-42) it is stated explicitly that Adamas subdues the Hyle (i.e., Matter, in Iranian texts called Āz), and in the Kephalaia Adamas conquers a “sea giant” which had been formed out of those elements of the Darkness which were thrown into the sea by the Living Spirit and caused its saltiness and bitterness (Kephalaia, pp. 114-15). In another Middle Persian text concerning Adamas and his fight with this sea-monster, the monster is not named (Sundermann, Kosmogoniche Texte, pp. 47ff.).

The battle between a god and a sea monster is of course well known from the Babylonian creation myths (see e.g., the translation by Heidel, The Babylonian Genesis, pp. 40ff.). In these the young god Marduk conquers Tiʾâmat, the primeval salt-water ocean personified (cf. the Manichean myth just quoted), by shooting an arrow into her mouth which she was unable to close because Marduk had let the evil wind loose into her face (cf. the Persian versions of the dragon-slaying described below in AŽDAHĀ II). It is also found in the Bible, where the Lord is described as having slain Leviathan, the serpent and the tannīnīm (crocodiles) in the sea (see, e.g., Heidel, op. cit., pp. 102ff.). With Hebrew tannīnīm compare Ebn al-Nadīm’s use of Ar. tannīn in his description of the Evil Spirit quoted above and note Middle Persian *TNYNA: possibly attested in Dēnkard, Madan, p. 816.13, and Pahlavi Rivayat, p. 22.10 (see Henning, “Two Manichean Magical Texts,” p. 42 and Bailey, Zoroastrian Problems, pp. 29f., 2nd ed., p. xxxiii.)

Clearly a large number of elements from different sources, literary and oral, combined to form the various concepts of dragons. These elements and their various connections and interactions, especially those stemming from the Iranian and the Semitic traditions still have to be investigated in detail.