Return to Born Eunuchs Library

Eunuchs in ancient religion.

by Arthur Darby Nock (Clare College, Cambridge)

Archiv für Religionswissenschaft, vol. 23 (1925-26), pp. 25-33

[p. 25]

Religious castration being Oriental in origin, eunuchs have no place in purely Greek or Roman cults1. They figure prominently in the worship of Cybele, the Dea Syria, Hecate2, the Aphrodite of Aphaca in Syria3, and the Scythian mother goddess; these deities are closely akin, each being a great goddess of fertility. The motive for this self-mutilation is obscure. Professor H.J. Rose has recently supported the view, first put forward by Mr. A. B. Cook (Zeus I p. 394 f.), that it is the desire to give added uires to the goddess4. We cannot deny that the ancients held that man could genuinely magnify god5; moreover, when Prudentius says of Cybele (Peristeph. X 1069f.), 'illam reuolsa masculini germinis | uena effluenti pascit auctam sanguine', he does, I think, indicate that such an idea could find a place within her cult. The severed genitals of her devotees are regarded as an offering by the writer of the Passio S. Symphoriani ch. 6 (p. 82 Ruinart2)6; their dedication to Attis is recorded in the Scholia on Lucian Jupp. Trag. ch. 8 (p. 60 Rabe); they were commonly preserved in a cista.8 At the same time, this evidence hardly bears out the idea that the original aim of the individual's abandonment of generative power was the increase of the deity's.

Nor, again, do the texts bear out the older view, that the eunuch wished to assimilate himself to the goddess.9 It rests chiefly on the fact....

1 The eunuch neocoros of Apollo at Delphi (Hermipp. fr. 13, J.H.G. III p. 39, ap. Anecd. Bekker p. 238.14), is a shadowy figure: cf. Perdrizet Rev. ét gr. XII (1899) p. 40. Poulsen's inferences (Delphische Studien I p. 27 f.) are very hazardous.
2 At Ephesus (cf. Picard Éphèse et Claros p. 179ff., 222ff.), Stratoniceia (Lebas-Waddington Inscr. V 519, 19), and Lagina (Bull. corr. hell. XLIV p. 79n. 11d, p. 84 n. 16).
3 Euseb. Vita Const. III 55.
4 Class. Quart. 1924 p. 14 f.
5 Cf. A.B. Cook Folklore XIV (1903) p. 270f.
6 Cf. Euseb. l. c. teèn daímona hileoûnto.
7 Graillot Le culte de Cybèle p. 297. It should be remarked that in the cult of the Dea Syria religious castration is not connected with anything like the Attis story; her consort is a Zeus, not an Attis (Garstang and Strong The Syrian Goddess p. 11), and the young figure associated with her is a son, not a lover (ib. p. 2, 26; Dussaud Rev. Arch. 1904 II p. 257).
8 L.C. Parnell Cults of the Greek States III p. 300 f., Greece and Babylon p. 257.


[p. 26]

....that he dressed as a woman after his self-mutilation.1 There is, however, no difficulty in accounting for this. The eunuch wished to show that he had abandoned all the active functions of his sex. In the same way, an Albanian girl who refuses to marry is allowed to swear perpetual virginity before twelve witnesses, and, if she cares, to wear man's dress and carry arms; these privileges are not exercised by all who take the oath.2 Further, eunuchs were commonly regarded as of the feminine gender3; Ambrosiaster even says abscisi in mulieres transformantur (XXXV p. 2349C Migne)4. In the Anthologia Latina I 298 (ed. Riese) we read of a spado putting on a mitella, 'bene conscius quid esset'. Moreover, the change of clothes may have in time assumed an apotropaic character and come to be regarded as averting the Evil Eye and unfriendly spirits; such changes of costume are familiar in nuptial and other rites.5 The ....

1 As there is no record of men putting on female attire in connection with these rites but not mutilating themselves, I would suggest that we must distinguish this custom sharply from that tendency to put on the clothes of the opposite sex which is observed in primitive peoples of North America and of other lands (Waitz Anthropologie der Naturvölker III p. 113, 383, J.G. Frazer GoIden Bough3 VI p. 253ff.). Here the individual appears to believe that he has a direct command from a deity to act in this way (cf. G.B.3 VI p. 256f.). This is quite different from the ritual change after enthusiastic self-sacrifice as we know it in the cult of Cybele and in that of the Dea Syria (we have no evidence on this point for Scythian practice, which may have resembled this primitive custom; cf. W.R. Halliday Brit. Sch. Ath. XVII p. 95ff.). Consequently, Waitz' contention that the change of clothes he mentions points to homosexuality (as is obvious at Porto Rico: cf. Rep. Amer. Bur. Ethnol. XXV (1903/4) p. 31), does not, if accepted, oblige us to agree with Ganschinietz (Pauly-Wissowa XI p. 1136 ff.) that the galli acted as they did because of some such psychological abnormality. Lucian De dea Syria 22, Terence Eun. 666, Martial III 81 show that the ancients did not regard eunuchs as necessarily homosexual; the frequent conjunction of galli and cinaedi is unimportant, since in antiquity charges of unnatural vice were lightly made.
2 M.E. Durham Journ. R. Anthr. Inst. XL (1910) p. 400. Mr. B.F.C. Atkinson, of Magdalene College, kindly informs me that the custom was still observed among the Ghegs when he visited Albania in 1923.
3 Cf. Hug P.W. Suppl. III p. 453, [line] 65; they appear as neutral sexually in Valerius Maximus VII 7, 6, Hippol. Refut. V 7, 16 (p. 82. 8 Wendland), Prudent. Perist. X 1071. Catullus LXIII 63 ego mulier is familiar.
4 Hieron. In Macth. XIX 12 says emolliuntur in feminas (XXV p. 135 Migne).
5 Cf M.P. Nilsson Griechische Feste p. 370 ff., W.R. Halliday B.S.A. XVI p. 212 ff., Samter Geburt, Hochzeit und Tod p. 93 ff. (with the modifications of L. Radermacher Sitz.-Ber. Ak. Wien CLXXXII 3 [1916] p. 39ff.; a different explanation of some cases is given by A.B. Cook Class. Rev. 1906 p. 376f.). A curious positive value is assigned to the change in the Bellary district of Madras; men who believe themselves impotent now to dress as women in the hope of recovering their virility (Fawcett Journ. Anthr. Soc. Bombay II. p. 331, 343ff., quoted by J.H. Gray in Hastings' Enc. Rel. Eth. V p. 581b).


[p. 27]

.... belief that one can 'do the devil' by altering one's dress is widespread.1

It must therefore be very doubtful whether any desire to become more like the goddess entered into this religious self-emasculation.2 Professor Rose's objection, that women were available for her service, gains in point when we observe that women played a regular part in that service, as in Caria and at Cyzicus3; in each case virgins participated in the ceremonial, if we accept M. Théodore Reinach's plausible interpretation of parthenoón as 'room of the virgins'4. The Dea Syria had a priestess in the Piraeus5; the great Hittite mother goddess had such from of old6, and her Scythian sister is attended by women as well as by eunuchs on the Karagodeuashkh plaque7; women lived in the temple precincts of Aphrodite at Aphaca.8

The explanation I would offer is as follows.9 The ancients believed that numerous sacred functions could only be performed properly by one who was qualified for them by perfect continence. As a result of this belief, which is well attested in the ancient East, in Greece and Italy, in modern Central and Southern Europe, and in America and Africa10, many cults required as ministers virgins living in chastity or children who had not attained the age of puberty; married people engaged in the service of a temple were often obliged to practise continence while so occupied.11 I would urge that the eunuch has mutilated himself, in the enthusiasm of a great festival, in order that he may be perfectly fitted to serve through his whole life the object of his devotion. Having made the sacrifice he is hagnós12, castus; Isidore writes ....

1 Cf. J.G. Frazer J.R.A.I. XV p. 98ff., G.B.3 VI p. 260ff.
2 I may note that the Nigerian youth who plays the part of the Great Mother of Ekong is not castrated (P. A. Talbot Life in Southern Nigeria p. 181).
3 B.C.H. XXXII p. 499ff.; Lobeck Aglaophamus p. 10117.
4 B.C.H. l. c.
5 I.G. III 1280a.
6 Garstang and Strong op. cit. p. 9.
7 Rostovtzeff Iranians and Greeks in South Russia p. 105 pl. XXIII (= Minns Scythians and Greeks p. 218 fig. 120).
8 Euseb. l. c.
9 Anticipated, but not developed, by H. Hepding Attis p. 162. Picard (op. cit. p. 228) suggests that the conception found its way into the cult of Hecate, but was not original there.
10 Cf. E. Fehrle Die kultische Keuschheit im Altertum (R.G.V.V. VI) passim, H. H. Bancroft Native Races of the Pacific States II p. 204, III p. 433, 435, 495, 498; J. Roscoe The Bakitara or Banyoro p. 92, 96, 98, 105, The Banyankole p. 41, 43, 111, 147.
11 Cf. the Phrygian inscription (Journ. Hell. Stud. VIII p. 382 n. 13 = Steinleitner Die Beicht p. 47 n. 23, cf. p. 86), in which a man who did not respect his wife's duty confesses his sin and records its punishment. A. similar obligation rests on the sacred dairyman of the Todas while in office (W.H.R. Rivers The Todas p. 99).
12 Attis as thalameepólos of Cybele is hagnós (Anth. Pal. VI 220. 3), sanctus (Dessau Inscr. lat. sel. 4117): the religious service of the galli is sanctum (S. Aug. Ciu. dei VI 7).


[p. 28]

... (Orig. X 33), 'castus primum a castratione dicitur; postea placuit ueteribus etiam eos sic nominare qui perpetuam libidinis abstinentiam pollicebantur', and Claudian uses the phrase (In Eutrop. 1 468), 'exsecti ueneris stimulos et uolnere casti'. So Paulinus of Nola says of such folk among his heathen flock (XIX 186 Hartel), 'ebrietas demens, amor impius illos | sanctificent'.1 All this confirms the conclusion we could draw from Catullus LXIII 17: 'Et corpus euirastis ueneris nimio odio'.

The eunuch, by his deliberate act, put himself on a footing with the virgin and the pure child. If it is asked why his devotion did not take the form of a life of voluntary continence, the answer must probably be that the Greeks did not in general regard such a life as possible for men. The Olympian Pantheon has its Artemis, but no similar god; does not this reflect popular opinion? Orphism gave a definite idealisation to abstinence, but the muses probably felt for this new-fangled claim to special purity the contempt which Theseus expresses to Hippolytus. Such abstinence remained something involving a peculiar sanctity2, as in the Thracians described by Posidonius ap. Strab. VII 3 p. 296, in Apollonius of Tyana, and in adherents of Mithraism (Tertull. De praescr. haer. 40).3 In Virgil's Elysium sacerdotes casti dum uita manebat received high honour (Aen. VI 661).

We can now, I think, understand what Prof. Rose justly calls a curious and parodoxical thing, 'that a goddess of fertility should be worshipped by the unfertile'. Dr. Fehrle has shown how often the pure child or maiden appears in ceremonies intended to promote fertility, as in ploughing, in, rain-making, in harvesting, and in marriage ritual.4 A virgin priestess is associated with the phallic Heracles of Thespiae5; chastity after appointment was required of the priestess of Anaitis at Ecbatana6, and of the neocoros of Aphrodite at Sicyon.7 We have seen earlier that in all probability virgin priestesses served Cybele at Cyzicus and in Caria, and we know from Strabo XIV 23 p. 641 that virgins and ...

1 Cf. further Apul. Met. VIII 29 insuper ridicule sacerdotum laudantes. castimoniam (of the eunuch priests of the Dea Syria), and Hieron. In Matth. XIX 12 quibus castimoniae necessitas, non uoluntas est (XXV p. 185 Migne).
2 Cf. of Christian abstinence Hieron. Adu. Jouin. I 34 si sacerdotibus non licet uxores tangere, in eo sancti sunt quia imitantur pudicitiam uirginalem (XXIII p. 257 Migne).
3 If the passage is rigthly so interpreted, but cf. Cumont-Latte Myst. d. Mithra 3. Aufl. p. 151, 4.
4 Op. cit. p. 63 ff.: on the last point I may be allowed to refer to my paper Eros the Child, in the Classical Review 1924 p. 152 ff.
5 Pausanias IX 27.6; here a Hieros Gamos may well be implied.
6 Plut. Artax. 27.
7 Pausan. II 10.4.


[p. 29]

.... eunuchs alike assisted in the cult of Hecate at Ephesus. Both classes had the requisite ritual purity.

To this it may be objected that the eunuch is not really on a footing with the virgin and the pure child, as their efficacy may rest in their untouched uires; in short, their chastity may be something effective in virtue of its positive character. hagnós, argue Dr. Pfister1 and others, means 'awakening religious respect and filled with strength'. Yet, even if this is etymologically correct, the word connotes in practice rather the negative conception of 'abstaining from'; from Aeschylus onwards it is constantly used with a simple genitive of separation.2 It is in fact almost the equivalent of katharós. Further, there is much evidence to show that chastity in cultus was commonly regarded as something negative and as an abstention, parallel to abstention from specific kinds of food and drink, from the pollution of death, from wearing woollen clothes, and so on. He who would enter the shrine of Aesculapius recently discovered at Thuburbo Majus in Africa was thus warned, 'a muliere, a suilla, a faba, a tonsore, a balineo commune custodiat triduo', and similar prohibitions are common in Greek.3 hagneúein, which in papyri commonly means 'to fast'4, is defined by Hesychius s. v. as kathareúein apó te afrodisíoon kaì apò nekroû. In the same way Strabo says of the Arabians (XVI 1, 20 p. 745) hoósper apó nekroû tò loutrón estin, hoútoo kaì apò sunousías. The requirement of chastity in ....

1 Phil. Woch. 1923 p. 860. His treatment of this word is somewhat imaginative; hagnós is quite distinct from hágios on which see his remarks P.W. XI p. 2116. The negative view is upheld by O. Schrader E.R.E. II p. 51b.
2 Cf. E. Williger R.G.V.V. XIX p. 45 ff.
3 A. Merlin C. R. Ac. Inscr. 1916 p. 262 ff.: cf. Dittenberger Syll.3 983, 1042; Wächter R.G.V.V. IX i; Deubner De incubatione p. 14 ff.; Zingerle Zu griech. Reinheitsvorschriften (Bulicev Zbornick, Strena Buliciana p. 171 ff.); Hopfner Griechisch-Ägyptischer Offenbarungszauber I p. 233 § 838ff.; Oldenberg Die Religion des Veda p. 412; and further Pausan. VIII 13.1 oú mónon es tàs míxeis allà kaì es tà álla agisteúein; Acta Philippi I 3 (vol. II ii p. 2. 25 ed. Bonnet-Lipsius) tácha mallon ou sumférei moi gameîn kaì meedèn esthíein tà es hústeron donoûnta tò soôma, <hoîon> oînon kaì kréa ... (where I restore <hoîon> as implying a more natural corruption than is postulated by Batiffol's oînon for -os) hagneûsai eè hudropoteêsai, and in medicinal magic Marcellus Empiricus De medicamentis XXIX 26 abstineat uenere, et ne mulierem aut praegnantem contingat, aut sepulchrum ingrediatur omnino obseruare debebit, and in alchemy Archelaus perì teês auteês hierâs téchnees l. 41f. (ed. Goldschmidt R.G.V.V. XIX ii p. 51). The same conception of abstinence is found in the Neopythagorean ideal sketched by Reitzenstein Historia monachorum und historia Lausiaca I p. 94ff.
4 Cf. W. Otto Priester und Tempel im hellenistischen Ägypten I p. 25 3. So the norm castus means 'fast' and is derived from careo (Walde s.v. careo), and castitas (Thes. l.l. III p. 542. 48 ff.) and castimonia (ib. p. 537. 56) are used of self denial.


[p. 30]

....sacred ministers was based much more on the impurity involved in sexual intercourse than on any peculiar powers resting in the pure.1 This is emphasised by the Phocian preference for old men in the priesthood of Heracles2, by the similar preference for old women as priestesses at Dodona3, and by the choice of the aged as guardians of the sacred fire at Delphi and at Athens.4 After Echecrates did violence to a Delphic prophetess a woman fifty years of age, dressed as a maiden, filled the post.5 More remarkable than this is what Pausanias tells us (VIII 5. 12) of the cult of Artemis Hymnia at Orchomenus. After Aristocrates outraged the virgin priestess, the Arcadians arranged that Artemis should have in future hiéreian gunaîka homilías androôn apochroóntoos échousan. The choice of those in whom age had destroyed sexual desire and energy6 and the use of such an expedient as the narcotic administered to the hierophant at Eleusis7 show clearly enough that the chastity demanded was purely negative. Otherwise the young and pure would alone have been satisfactory. On this view of the matter we understand the temporary continence of married people at sacred seasons, before ploughing, sowing and reaping8 (all emphatically sacred acts in a primitive society), and in war. Abstinence helps people to approach the deity; Athenagoras uses (Legatio 93) as an argument which he hopes is self-evident, ei dè tò en partheníai kaì eunouchíai ménein parísteesi toôi theoôi, tò dè méchris ennoías kaì epithumías eltheîn apágei ...9

This negative chastity the eunuch has in common with the pure child and the virgin. The importance of the individual's self-castration lies not so much in the act as in the life which follows it.10 We do not ...

1 Here I agree with Ziehen P.W. VIII 1418. Impurity may be dangerous: of a spell in cod. Vindob. 98 fol. 113 v. prius consideret salutem suam et uadat purus ab omnibus (Heim Fleck. Jahrb. Suppl. XIX p. 608).
2 Plut. De Pyth. orac. 20 p. 403 F.
3 Strabo VI 11 p. 329.
4 Plut. Numa 9.
5 Diod. Sic. XVI 26.
6 Cf. for further examples Fehrle op. cit. p. 95 1.
7 Fehrle p. 104.
8 Cf. J.G. Frazer G.B.3 V p. 104ff., VII p. 1092, VIII p. 14f.; J. Roscoe The Banyankole p. 31 (rainmaking).
9 Well quoted by E. Hahn Demeter und Baubo p. 50: cf. Porphyr. ap. Euseb. Praep. euang. V 6. 3, Macrob. Sat. I 23.13 (nobles at Heliopolis longi temporis castimonia puri are inspired). The assertion of positive strength resulting from purity occurs in the ending of the Acta Pauli et Theclae, as added in three Paris Mss.: élegon gàr hóti teêi Artémidi douleúei parthénos oûsa, kaì ek toútou ischúei pròs tàs iáseis (Ed. Bonnet-Lipsius I p. 270), and in folklore (cf. Siebs Schlesiens volkstümliche Überlieferungen VI [= Sagen IV) p. 176 s.v. Reine). But the. negative nature of virginity is very closely stated in such a text as Plut. Num. 9: tò ágonon kaì ákaron teêi partheníai sunoikeioûntos.
10 So the rite, as practised by the Skoptzy in Russia in modern times, was originally a means of ensuring sexual purity, although it has acquired sacramental value with them (Grass, Die russischen Sekten II p. 687 ff., 694, 714 ff., 731 ff.).


[p. 31]

.... find inscriptions in which eunuchs assert that they have sacrificed their manhood1 or that they have given their uires to the goddess; there is no clear parallel to the taurobolium records, as we should expect if the castration was regarded as an act of value in itself and affecting the goddess. On the other hand, we have an inscription in which a gallus boasts of the amount he begged for the Dea Syria2; it is the life of devotion which counts. The Megabyzoi at Ephesus may have been bought eunuchs; our texts do not mention their self-castration.

The value of the eunuch's chastity in ritual concerned with fertility appears clearly in the custom recorded by Eusebius, Vita Constantini IV 25. Eunuchs worshipped the Nile, and, when Constantine stopped their rites, it was believed that the Nile would not rise and fertilise the fields3; the river's intimate connection with fertility is obvious.4

The eunuch's sacred functions are quite in conformity with this view of his nature. They are, above all, religious begging, self-mortification, and assistance in worship. The first is a practice not infrequently associated with children, as in the Rhodian chelidonismós, in the Athenian and Samian bearing of the Eiresione, in the English rite of the Maylady, and in the Greek rite of December 31 at Pergamon in modern times.5 The principle underlying it is, no doubt, do ut des, as appears in the promise made in the lines attributed to Homer (l. 16f.): ploûtos gàr épeisi (v.l. éseisi) | pollós, sùn ploútooi dè kaì eufrosúnee tethaluîa.5 The ....

1 Picard op. cit. P. 227 regards this as substituted for self-immolation.
2 B.C.H. XXI p. 60.
3 Cf. Heinichen ad. loc. for other evidence.
4 We may note that children (Seru. ad Georg. IV 363) and virgins (Plut.] De fluuiis 16) were regarded as suitable offerings to the Nile. To quote a distant parallel, H.H. Johnston J.R.A.I. XIII p. 473 says that in the Congo region youths castrate themselves 'in order to more fittingly offer themselves to the phallic worship, which increasingly prevails as we advance from the coast to the interior'. Here again the castration aims at rendering the devotes suitable for service, and is not performed for any special value of its own. The eunuch does not often appear in ancient magic, as does the pure boy; the only example I can quote is Sextus Placitus Papyriensis De medicamentis ex animalibus XVII 19 (p. 56 Ackermann): AD FEBRES ACERRIMAS, A uestigio spadonis discedentis a ianua si sustuleris quodlibet, dicens: tollo te ut ille Gaius (v. l. Graius) febribus liberetur. Nominabis eum, ad cuius brachium suspensurus es, ad id uero loqueris [ad] quod sustuleris' (this ad seems superfluous). It is to be observed that this charm occurs in the section headed De puello et puella uirgine, though at the same time that section does contain irrelevancies.
5 Cf. H. Hepding Hess. Bl. f. Volksk. VII p. 40ff. Mr. H.T. Deas has reminded me of the begging of Scotch children in the Clyde valley on All Hallowse'en.
6 ap. Suid. II1 p. 1106.12 Bernhardy: cf. Lucr. Il 627 largifica stipe ditantes, and on the principle R. Wünsch Rh. Mus. LXIX p. 127 ff., G. Van der Leeuw Archiv XX p. 241 ff. The priests of Isis begged for her (Suid. I p. 52); from this systematic mendicity we must distinguish the private activity of Orpheotelestae. For Oriental parallels cf. Graillot op. cit. p. 3132; Bancroft op. cit. III p. 436f. records that certain Mexicans who practised severe asceticism for a year lived by begging.


[p. 32]

.... second is not peculiar to eunuchs1, though the blood of consecrated persons may have been particularly acceptable to the deity. Their assistance in worship was of various kinds. The galli of Cybele sang litanies in honour of Attis, shouted, danced, ran, parried Cybele in procession, and prophesied.2 Under the Empire, if not earlier, they were sharply distinguished from the sacerdotes proper. The galli of Bambyce were the ritual ministers of the temple. To turn to Scythia, the eunuch on the Karagodeuashkh plaque is handing to the goddess what Professor Rostovtzeff (1. c.) regards as 'a round vase containing the sacred beverage'4; further, Herodotus tells us (IV 67) that the Enáreës, who are probably identical, prophesied by the use of linden bark.

Having made himself an acceptable minister to the goddess5, the eunuch emphasised his sacred character by his outward appearance; his long flowing hair, his necklace, and his earrings all marked his self-dedication6. The ceremonies at Bambyce which followed his death show how carefully he was protected against evil spirits.7 His sanctity was effective; he could enter the Plutonium at Hierapolis in Phrygia, deadly as its fumes were to all other living things.8

But we must not be surprised that the eunuch priest, in spite of his special holiness, often failed to receive public respect9. The assistant ....

1 There is no direct evidence that the priests of Bellona mere mutilated, though cf. E. Strong Brit. Sch. Rome IX p. 211. In Central America consecrated persons and the general public alike scarified themselves with sharp thorns (Bancroft p. 436, 471, 486, 494).
2 Cf. Graillot p. 301ff.
3 Wissowa R.K.2 p. 3208.
4 Garstang had earlier detected traces of communion in the cult of the Hittite Mother (Land of the Hittites p. 111f., 164f.). '
5 I should perhaps emphasise the point that the motive I suggest for his self-mutilation is not exactly asceticism (as Gruppe Gr. Myth. p. 1542ff.), though ascetic ideas could easily find their way into the cult; we see them in Jamblichus' explanation of the story of Attis, as reproduced by Julian V p. 167C, Sallustius Perì theoôn ch. IV.
6 Cf. Hepding op. cit. p. 1625, E. Strong l. c. p. 2084. Graillot op. cit. p. 299.
7 Lucian De dea Syria ch. 52 ff.
8 Strabo XIV 14 p. 630; cf. what Pomponius Mela has to say (III 6. 48) on the miraculous powers of the nine virgins of Sena (discussed by S. Reinach Cultes, Mythes et Religions I p. 195 ff.). Drechsler Schles. volkst. Überl. III p. 222 records a Silesian belief that a pure boy or girl can rekindle by his or her breath an extinguished torch that is still glowing.
9 He was sometimes held in honour. Hippocrates Perì aéroon ch. 22 speaks of popular respect for the Anarieîs (as Gomperz restores); an inscription from Lagina (B.C.H. XLIV p. 84 n. 16) mentions hoi semnótatoi teês theâs eunoûchoi, and a myth in [Plut.] De fluuiis ch. 12 makes Cybele take vengeance on Sagaris for insulting the galli. Cf. also Juvenal VI 511 ff.

[p. 33]

... in rites was not necessarily held in honour; etéleis, egoò dè eteloúmeen is a taunt Demosthenes flings at Aeschines (XVIII 265). The virgin priestess, who by continual continence remained able to render fit service, was honoured in antiquity and is honoured in countless places1; the child and the eunuch do not inevitably receive similar respect. The insults which the latter had to bear were not ritual mockery or maltreatment for religious purposes2; they were the Greek comment on an ungreek custom.

To venture for a moment outside the classical field, I would suggest in closing that on this view of castration as a means of ensuring ritual purity we can account for the sporadic self-mutilation recorded as happening at the yamplanting in Nigeria.3 That may well be a drastic extension of the principle that continence must be observed at such seasons, a principle which can find expression, as in the Motu tribe in New Guinea, in the setting apart of one of the chief men as holy or taboo, and causing him to live apart from his wife and to obstain from certain kinds of food.4 In intense enthusiasm an individual may go further.

1 The virgin's merit is expressed by the word egkrateusaménee (= partheneúsasa), which became a fixed epithet among the Christians (as Ath. Mitt. XIII p. 272 n. 138: for its earlier use cf. Steph. Thes. III p. 102D, and for the 'athletic' ideal in Christian asceticism cf. E.C. Butler The Lausiac History of Palladius I p. 237ff.). It is noteworthy that Prudentius Adu. Symmachum II 1071f. denies merit to the Vestals because their abstinence was something chosen for them in childhood, not a self-restraint deliberately accepted by adults: 'Nec contempta perit miseris sed adempta uoluptas | corporis intacti, non mens intacta tenetur'. On respect for virgin priestesses cf. Westermarck History of Moral Ideas II p. 417 ff.
2 Ritual maltreatment of a sacred minister is not unknown, as for instance his drenching with water to secure rain (G.B.3 I p. 277, II p. 77), or his rolling on the fields to secure their fertility (G.B.3 II p. 103), or pulling out some of his hair and burning it to make the flax grow (H. Marzell Hess. Bl. XI p. 23). Pope Gregory VII protested to the King of Denmark against the popular habit of treating the clergy as responsible for such evils as intemperiem temporum, corruptiones aeris, quascumque molestias corporum (Mansi Concilia ed. 1759 XX p. 305). On religious abuse in general cf. Usener Rh. Mus. XXX p. 225 (= Kl. Schr. IV p. 139f.), J.G. Frazer Commentary on Pausanias II p. 492, Nilsson op. cit. p. 176.
3 P.A. Talbot ap. G.B.3 V p. 2710. The Ekoi ceremony he describes in In the Shadow of the Bush p. 74ff. is different, as it involves the mutilation of an individual against his or her wishes and the offering of the severed parts to the Eja.
4 G.B.3 II p. 106.