¹ Cf. Ideler, , Untersuchungen über Ursprung u. Bedeutung de Sternnamen (1809) 286 ff.Google Scholar; Weber, , Abh. Akad. Berlin (1860) 283 ff.Google Scholar; (1861) 267 ff.; Burgess, E., J. Amer. Orient. Soc. VI (1860), 327 ff.Google Scholar; Steinschneider, , Zeitschr. d. Deutschen Morgenländ. Ges. XVIII (1864), 118 ff.Google Scholar; Ginzel, , Klio I (1902), 14 ff.Google Scholar; id.Handbuch der Chronologie I, 70 ff.; Macdonell-Keith, , Vedic Index I, 427 ff.Google Scholar; Jeremias, , Myth. Lex. IV, 1475 f.Google Scholar; Thorndike, , History of Magic I, 713 ff.Google Scholar; II, 113 ff.; 183; etc.

² The development was, in fact, more complicated: the work of the seasons soon led to the observation of the rising and setting of stars and this to some sort of a solar reckoning; there were also (and are still) different reckonings in use at the same time. But the simplification was necessary for the sake of our argument.

³ Although Reitzenstein 256 ff. raises rather than solves the problems and approaches them only from the religious side, it is astonishing to see how much of the decisive evidence he quotes in the proper context.

⁴ T'oung Pao, N.S., XIII (1912) 699 ff.Google Scholar; see also his ‘Antike Beobachtungen farbiger Sterne’, Abh. Akad. München XXX, 1, 1918, 154Google Scholar; Sternglaube u. Sterndeutung ³ 57.

⁵ I owe much encouragement to the late Professor Cumont, who generously placed his material at my disposal. He does not seem to have formed a final view about the mansions but intended to inquire in greater detail into the problems of iconography. This would have involved research into the medieval art of the East for which I do not feel competent; I have therefore limited my task to the origins and early history of the doctrine. I am further indebted to Professor D. S. Robertson, whose valuable criticism contributed much to the clarification of my argument and saved me from several grave errors; to Professor Last, who read and improved my manuscript; to Dr. J. Schacht, the late Professor B. G. Gunn, and Dr. O. R. Gurney for discussion of the Arabic, Coptic, and Babylonian terms respectively; and to the late Professor F. Saxl for information about illustrated MSS.

⁶ First printed in 1485; then in 1501, 1551, 1571. This text omits the Persian evidence, the horoscopes and the iconographical matter; but it is often so close to our text that it helps to check it and, where necessary, to supplement it.

⁷ See Stegemann, , Beiträge z. Geschichte der Astrologie (1935) 7 ff.Google Scholar; id.Die Fragmente des Dorotheos 48; id. Dorotheos von Sidon u. das sog. Introductorium des Sahl Ibn Bišr (1942) 76, 1; Bidez-Cumont, , Les Mages hellénisés II, 233Google Scholar.

⁸ Another Greek version of this text was published by Camerarius in his Astrologica (1532) 31–36. It is so much shortened that its true character remained unnoticed: it omits the names of the mansions as well as all astronomical data and the references to the Indians, Persians and Dorotheos. It does not depend on our text but is another translation of the same Arabic original.

⁹ Translated by E. Sachau (1879) 343 ff. Albiruni also describes the mansions in his Elements of Astrology, translated by Wright, R. R. (1934) 81 ff.Google Scholar; cf. El-Kazwîni, , Kosmographie, translated into German by Ethé, H. (1868) 87 ff.Google Scholar; Barhebraeus, , Le livre de l'ascension, transl. by Nau, F. (1899) 107 ffGoogle Scholar. The earliest evidence for the Arabic list is found, under the name of Alchandrinus Philosophus, in cod. Paris. lat. 10271, f. 9 (saec. ix) and 17868, f. 5 (saec. x): see Cumont, RA. (1916) I, 17Google Scholar; 20; Gundel, , Dekane u. Dekansternbilder 34Google Scholar; 114, 1. For modern discussions see Ideler, op. cit.; Steinschneider, loc. cit.; Ginzel, loc. cit.; Nallino, , Raccolta di scritti V (1944) 175 ff.Google Scholar; Stegemann, , Handwörterbuch d. deutschen Aberglaubens VII, 49 ff.Google Scholar

¹⁰ The complete list of the Cromwellianus is as follows: 1. . 2. . 3. Ἔξαστρον. 4. . 5. Ἀκρίς. 6. . 7. . 8. . 9. . 10. . 11. Δύο λαγίνια. 12. . 13. Καποῦλα (=Catula? ) τοῦ Λέοντος. 14. Στάχυς. 15. ⟨Velamen?⟩. 16. . 17. Στέφανος. 18. . 19. . 20. . 21. Κάστρον. 22. . 23. . 24. . 25. . 26. . 27. . 28. . There are three further lists in Greek: that published by Cumont, CCAG. VIII, 1, 218 fGoogle Scholar. is of no interest in this context as it only reproduces the Arabic terms. Heeg, , CCAG. V, 3, 90 ffGoogle Scholar. published two lists; the first omitted in seven cases the Arabic terms and replaced them by the following: 3. Πλειάς. 4. . 7. Βραχίων. 14. Ἀρκτοῦρος. 17. Στέφανος. 18. Ἀντάρης. 20. Ἠριδανός. On the second list see below, p. 53 n. 14.

¹¹ Cf. Gunkel, , Zum religionsgesch, Verständnis d. NT. 48 ff.Google Scholar; Boll, , Aus der Offenbarung Johannis 39 f.Google Scholar; 143; the commentaries of Bousset and Charles; Clemen, , Religionsgesch. Erklärung d. NT.² 405Google Scholar; Bousset, , Religion des Judentums³ 285Google Scholar; Jeremias, , Handbuch d. altorient. Geisteskullur² 110Google Scholar.—On Thebes as a terrestrial copy of the celestial city in Nonn. 5, 63 ff. see Stegemann, , Astrologie u. Universalgeschichte (1930) 230 ff.Google Scholar

¹² I am in possession of two letters written in 1930 by J. K. Fotheringham to Cumont. He writes: ‘I have had some correspondence recently with Mr. W. E. Crum on the Coptic names of the lunar stations. Twenty-four or twenty-five of these appear to be Greek or debased Greek, the remaining three or four being Coptic. These last three are the Coptic equivalents of the Arabic names. Now Mr. Crum assures me that it is impossible that the Egyptians should have turned the Arabic names into Greek. He thinks that by the time Arabic science reached Egypt Greek was forgotten. It would seem then to follow that the Greek names belong to a pre-Arabic period. If so, how did they come to Egypt and why are they not found elsewhere in Greek astronomical or astrological literature? I do not feel inclined to follow up this inquiry, but it would be an interesting inquiry for somebody. I have not examined the relation of the Greek names to the Arabic, Persian, Indian, and Chinese names. The loci classici for the Coptic names are Kircher, Athanasius, Lingua Aegyptiaca restituta (1634), 50–53Google Scholar; 560–563, and the same author, Oedipus Aegyptiacus, tom, ii, pars ii (1653) 241–247Google Scholar. I quote Mr. Crum's last letter on the subject. “Many thanks for your interesting letter of yesterday. I am glad what I said has raised the question of the lunar station-names afresh. In case you have not Kircher at hand, here are those he gives which I take to be indisputably Greek—or, one should say, to have been once Greek. My Greek is too feeble to allow of rightly emending them. So in most cases I don't venture anything. Those marked with * have Coptic article before them, but that is customary whenever a Greek noun is adopted. I have omitted these articles here. I have collated all on the best MS.

Kircher, Ling. Aeg. 560Google Scholar

1. κυτων = κῆος

2. κυτωριον *

3. κυτωριονκολιων = κῶλον

4. ωριας

5. ωριον* (var. ωριων)

6. κλυσος

7. κλαρια

8. (Coptic) (The cubit) * Coptic = Arabic

9. αρμελια*

10. αυτος*

11. Coptic = (the forehead)* Coptic = Arabic

12. χωρων* = χωρίον ?

13. ασφυλια = ἀσφαλεία ?

14. αβυκια

15. χωριτος (p. 52 χοριτος)

16. χαμβαλια

17. (? Coptic) * ?

18. στεφανι= στέφανος

19. καρθιαν = καρδία

20. ογια (a Coptic name added with same Arabic as this one)

21. (Coptic) (quid?)* Ar. ‘ostriches’

22. ογιαπολις = Arabic

23. υπευστως (p. 53 υπευστος)

24. ουπευρυτως

25. υπευιπευτης (p. 53 -τη) leg. υπευιπετης, the ε and υ being written in MS. as alternatives

26. υπευθεριαν

27. αρτυλος

28. αρτυλοσια”

(Fotheringham resumes): It will be observed that No. 1 in this list is No. 28 in the Arabic list as given in Ginzel, , Handbuch der mathematischen und technischen Chronologie, Band i, (1906), 72, 73Google Scholar, and with each of the others you must deduct one from the number in the list given above to obtain the number in Ginzel's Arabic list. Kircher gives what he regards as the meaning of the Coptic or Greco-Coptic names and also the Arabic names and their meanings. It is clear even to me, who know no Arabic, that there is a close connexion between the two lists, and this raises the suggestion that the Arabic lunar stations are of Greco-Egyptian origin. Then most of the Coptic names will be debased Greek. The genuine Coptic names will be translations from the Greek. But in view of the existence of lunar stations among the Chinese, Indians, and Persians the question remains how the Egyptian Greeks came to possess this particular piece of astronomical or astrological lore.’—There is a note on this letter in Cumont's handwriting which was, I presume, the essence of his reply: ‘Au moment ou les Coptes empruntèrent les noms des stations lunaires aux Arabes ceux-ci se servaient peut-être encore des noms grecs des constellations. Les noms Arabes ne seraient entrés en usage que plus tard, quand l'astronomie fut cultivée dans l'empire des Abbassides ? ?’ Fotheringham's second letter seems to refer to this: ‘… The difficulty in the way of your suggestion about the Greek names of the lunar stations which have survived in Coptic is that these are peculiar to the lunar stations and are not used generally for the constellations. Hence the Arabs cannot have acquired these names from the Greeks unless there was some Greek list of lunar stations, which there may have been, though, so far as my knowledge goes, no trace of such a list has come down to us.’—I would only add that Cumont at that time believed that the mansions were not known to the Greeks, see his note in CCAG. VIII, 1, 217Google Scholar; further that the second mansion seems to be κεράτιον of Aries, the third κοιλία of Aries, the fourth the Pleiads (the MS. has an alternative name εξαστον), the fifth Orion. The Coptic list is also printed by Seyffarth, G., Systema astronomiae Aegyptiacae (1833) 21 f.Google Scholar; the genuine Coptic names are now incorporated in Crum's Coptic Dictionary; on Abū'l Barakāt see Mallon, , Mélanges de la Faculté Orientate, Univ. Beyrouth II (1907) 261 f.Google Scholar; on the MS. used by Kircher: Hebbelynck–Lantschoot, , Codices Coptici Vaticani, I (1937) 534Google Scholar.

¹³ His authorship was established against that of Pompilius Azalus Placentinus (under whose name the work was later published) by Thorndike, , Isis XV (1931) 31 ff.CrossRefGoogle Scholar; id.History of Magic IV, 150 ff.

¹⁴ CCAG. V, 3, 91 ffGoogle Scholar. (I omit the Arabic part of the list): 1. Ἀνδρομέδα. 2. Ὀιστός. 3. Αἴξ. 4. 4. Ἀετός. 5. Δελφίς. 6. Λαγωός. 7. Ἀρίων ( coni. Heeg). 8. Ἡνίοχος. 9. Λύρα. 10. Ἔριφοι. 11. Στέφανος. 12. Πλειάδες. 13. Ὑάδες. 14. Κένταυρος.19. Ἀντάρης. 20. Ἀρκτοῦρος. 21. Κύων. 22. Ὄρνις. 23. Κρατήρ. 24. Ἀργώ. 25. Ὀφιοῦχος. 26. Ἴκτινος. 27. Ὄνος. 28. Ἴππος.

¹⁵ See Ginzel, loc. cit.; Oldenberg, , Nachr. Gött. Ges. (1909) 572Google Scholar; Vian, R., Ein Mondwahrsagebuch (Diss. Heidelberg 1910) 65Google Scholar.

¹⁶ This is the title of the chapter in CCAG. VIII, 2, 134Google Scholar, but I have changed the accent of the decisive word (μόνων Ruelle); var. lect. to the second passage in CCAG. VIII, 2, 52Google Scholar A third witness may be Horapollo I, 66 ascribing to the Moon Here again I would write μονὰς for the p6vas of the editors which cannot be explained satisfactorily. But the passage is corrupt, and my conjecture does not suffice to repair it.

¹⁷ If the division of the world into twenty-eight klimata (Hermipp. de astrol. II, 12 f., 117 ff.Google Scholar; cf. Häbler, , Wochenschr. f. klass. Philol. 1896, 342Google Scholar; Honigmann, , Die sieben Klimata 10Google Scholar) were genuine Greek doctrine, it would furnish another important proof for the existence of the Greek mansions. But the author of the dialogue Hermippos, Iohannes Katrarios (fourteenth century), often depends on Arabic sources (see Boll, , SB Heidelberg 1912, 18Google Scholar. Abh.), so that his testimony alone is not conclusive. On twenty-eight klimata in China see Eberhard, W., ‘Beiträgez. kosmol. Spekulation Chinas’, Baessler Archiv XVI (1933) 58Google Scholar.

¹⁸ Cf. e.g. Hommel, , Zeitschr. d. Deutschen Morgenl. Ges. XLV (1891) 592 ff.Google Scholar; Ginzel, loc. cit.; some of the contributions make rather painful reading: Röck, , OLZ. XV (1912) 385 ff.Google Scholar; Bork, F., ‘Neue Tierkreise’, Mitteil. d. Vorderasiat. Ges. XVIII, 3 (1913) 1–46Google Scholar; Stucken, E., Der Ursprung des Alphabets u. die Mondstationen (1913)Google Scholar; Lichtenberg, v., ‘Buchstabenreihe u. Mythos’, Memnon VII (1913) 84 ff.Google Scholar

¹⁹ Cf. Bezold, , ‘Zenit- u. Aequatorialgestirne am baby-lonischen Fixsternhimmel’, SB Heidelberg 1913, 11Google Scholar. Abh., 36 f.; 57; Kugler, , Sternkunde u. Sterndienst in Babel, Erg. I, 70 ff.Google Scholar

²⁰ Weidner, , Amer. Journ. Sem. Lang. XL (1923) 192Google Scholar; cf. Jeremias, , Handbuch d. altorientalischen Geisteskultur² 207Google Scholar.

²¹ Weidner, , Archiv f. Orientforsch. VII (1931/1932) 174Google Scholar; Albright-Dumont, , Journ. Amer. Or. Soc. LIV (1934) 125Google Scholar; Burrows, , The Oracles of Jacob and Balaam (1938) 5Google Scholar.

²² It was suggested that in the O.T. the word μαӡουρώθ (Job 38, 31; 2 Kings 23, 5) refers to the mansions, see Schiaparelli, , Astronomy in the O.T. 78 f.Google Scholar; but see Cumont, , Syria VIII (1927) 165Google Scholar, 1.

²³ Cf. Macdonell-Keith, , Vedic Index I, 412Google Scholar. It is important to add that the months received their names from selected mansions and kept them even in the period of the lunisolar reckoning.

²⁴ E.g. the headless figure in Capricorn is known under various names, Akephalos Daimon, Eileithyia, Tyche: see Boll, , Sphaera 213Google Scholar; Albumasar ap. Boll 531. The two ostriches in Sagittarius also appear in Albiruni and in the Coptic list but nowhere else (I cannot check Gundel's reference, Neue astrol. Texte d. Herm. Trismeg. (1936) 199Google Scholar, 1 to an Egyptian constellation ‘ostrich at lake’). The woman of the 28th mansion, sitting over two fishes, is Andromeda; that of the 24th mansion, sitting on a throne, much resembles Cassiopeia, cf. Manil. V, 504; Myth. Lex. VI, 911Google Scholar; Boll, , Aus der Offenbarung Johannis 31Google Scholar.

²⁵ Cf. Gundel, , Dekane u. Dekansternbilder (1936) 175 ff.Google Scholar

²⁶ Cf. Stornajolo, , Codd. Urbin. lat. III, 297 fGoogle Scholar. I am using photographs of this MS. kindly lent to me by the late Professor Saxl; on the provenance of the pictures see Gundel, , Mélanges Cumont (1936) 250 fGoogle Scholar. (with a specimen of the text).

²⁷ Described by Hauber, A., Planetenkinderbilder u. Sternbilder (1916) 77 fGoogle Scholar. (reference of Professor Saxl).

²⁸ Cf. Weber, , Abh. Akad. Berlin (1860) 331Google Scholar; Ginzel, , Klio I (1902) 20 ff.Google Scholar; Macdonell-Keith, , Vedic Index I, 415 ffGoogle Scholar. W. Kirfel, Religion der Jaina's (in: Haas, , Bilderatlas d. Rel. gesch. Lieferung 12, 1928)Google Scholar, fig. 19; Different sets in Brennand, W., Hindu Astronomy (1896) 40Google Scholar, plate 7; Mollien, E., ‘Recherches sur le zodiaque indien’, Mém. Acad. Inscr. (1853) 240Google Scholar, pl. 4; cf. also Colebrooke, H. T., Misc. Essays II, 281 ff.Google Scholar

²⁹ The origin of this set is obscure; a vague analogy is found in Greek hepatoscopy: parts of the liver which was considered an image of the universe were called table, nail, sword, basket (Nicand., , Ther. 559 ffGoogle Scholar. et schol.), tongue, mirror, river, grave (Hesych. s. ); further instances in Blecher, , De extispicio 10 f.Google Scholar; cf. JRS XXXVI (1946) 122Google Scholar n. 129.

³⁰ Cf. Schlegel, G., Uranographie chinoise (1875) 583 ff.Google Scholar; Ginzel, loc. cit.; Eberhard, W., ‘Beiträge z. kosmol. Spekulation Chinas in der Han-Zeit’, Baessler Archiv XVI (1933) 55 ff.Google Scholar

³¹ Cf. Boll, , RE VII, 2570 ff.Google Scholar

³² Cf. Langdon, , Babylonian Menologies 85Google Scholar; 89.

³³ Cf. Vett. Val. I, 2 (p. 5 ff. Kr.); Heph. Theb. I, 1; Teucr. in CCAG. VII, 194 ff.Google Scholar

³⁴ See the forthcoming translation by Ritter-Plessner; Thorndike, , History of Magic II, 820Google Scholar.

³⁵ Cf. Weber, , ‘Die vedischen Nachrichten von den naxatra’, Abh. Akad. Berlin (1860) 283 ff.Google Scholar; (1861) 267 ff.; Thibaut, G., ‘Astronomie, Astrologie u. Mathematik’, Grundr. d. indo-arischen Philologie III, 9 (1899) 12 ff.Google Scholar; Oldenberg, H., ‘Naksatra und sieou’, Gött. Nachr. (1909) 544 ff.Google Scholar; Menon, C. P. S., Early Astronomy and Cosmology (1932) 57 ff.Google Scholar; 101 ff. (amateurish).

³⁶ Cf. Kern, H., Verspreide Geschriften II, 3–154Google Scholar: ‘The Brhat-Samhitā or Complete System of natural astrology of Varāhāmihira’, ch. 98–102; 105 (p. 135 ff.); Das grosse Buch der Nativitätslehre des Varāha Mihira (translated by Wulf, W. 1925), c. 16, pp. 120 ff.Google Scholar; English translation ‘The Brihajjatakam’ by H. P. Chatterjce (S. Vijnenanda 1912) 243 ff. On Hellenistic influence on Indian astrology see Sāstrin, V. V. Ramana, CR XXXVI (1922) 20Google Scholar; Boll, , CCAG. V, 1, 156Google Scholar; Gundel, , Dekane u. Dekansternbilder 87Google Scholar.

³⁷ Cf. p. 84 Ludwich p. 87 ; p. 86 . There is a further correspondence when parts of two signs are, as in some mansions, united, p. 80 ; or when the prognostics apply to a part of the sign only, p. 80 .—Cf. Dorotheos π. ἀποδημίας, CCAG. VI, 109Google Scholar; VIII, 2, 119 f.; cf. XII, 195.

³⁸ It was observed that Maximos often depends on Orphic verses (see Kern's Index): this points to the possibility that the source of Dorotheos was the Ἐφημερίδες of ‘Orpheus’ (see below, p. 58).

³⁹ Cf. Henning, , JRAS 1942, 245Google Scholar.

⁴⁰ Cf. ed. by Justi; printed also by Weber, loc. cit. (1860) 327 ff.; Windischmann, , Zoroastr. Studien 59Google Scholar; Henning, loc. cit. 242 ff.

⁴¹ The Greek texts in CCAG. III, 32Google Scholar; IV, 142; VIII, 4, 105; X, 121; XI, 1, 134; XI, 2, 157; the Latin texts (ranging from the ninth to the fifteenth century) were published by Svenberg, E., De latinske Lunaria (1936)Google Scholar with commentary in Swedish which unfortunately I cannot read.

⁴² Varr. r.r. I, 37, 3; cf. Wissowa, , Religion² 247Google Scholar, 2; Otto, W. F., RE Suppl. III, 1188Google Scholar; Svennung, , Untersuchungen zu Palladius u. zur lat. Fach- u. Volkssprache (1935) 260 ff.Google Scholar; Heurgon, J., REL XXV (1947) 236 ffGoogle Scholar.

⁴³ Dessau 7144; cf. 2525; Diehl, , Inscr. lat. Christ. vet. 4377–86Google Scholar; vol. III, p. 311; Mommsen, , Röm. Chronol.² 312Google Scholar; Excavations of Dura-Europos, Fourth Report (1933) 95Google Scholar; 107; Snyder, W. F., JRS XXVI (1936) 12 ff.Google Scholar

⁴⁴ Cf. Bouché-Leclercq, , Astrol. gr. 459 ff.Google Scholar; Schmidt, W., Geburtstag im Altertum (1908) 14 f.Google Scholar; 84 ff.; Nilsson, , Arch. Rel. Wiss. XIV (1911) 438 ff.Google Scholar; id.Entstehung u. religiöse Bedeutung d. griech. Kalenders (1918) 39 f.; the edition of T. A. Sinclair with valuable introduction and commentary. The later history of the Lunaria was reconstructed by Cumont, , ‘Les présages lunaires de Virgile et les “Selenodromia”’, Ant. Class. II (1933) 259 ff.CrossRefGoogle Scholar; see also Lobeck, , Aglaoph. 428 ff.Google Scholar

⁴⁵ Cf. Nilsson, loc. cit. 443; id.Entstehung etc. 44; Wilamowitz, , Glaube d. Hellenen II, 29Google Scholar.

⁴⁶ Fr. 271–9 Kern; cf. Lobeck, , Aglaoph. 411 ff.Google Scholar; Heeg, , Die angeblichen orphischen , Diss. Würzburg (1907) 36 ff.Google Scholar; Ziegler, , RE XVIII, 1, 1402 fGoogle Scholar. (who assumes that they were compiled in the time after Plutarch).

⁴⁷ Fr. 271 K.; cf. fr. 33 with reference to Epigenes' .

⁴⁸ The importance of the passage was first noticed by Heeg, op. cit. 37.

⁴⁹ Pliny XVIII, 321 … Vergilius … Democriti secutus ostentationem; cf. Cumont, loc. cit. 260; Kroll, , Philol. XCIII (1938/1939) 191 ff.Google Scholar

⁵⁰ Cf. Tzetz., , Hes. Erga 800Google Scholar; 820; CCAG. VIII, 4, 102 ff.Google Scholar; Cumont 267.

⁵¹ FGr Hist. 328 F 85 ff. Jacoby, (FHG. I, 413 ff.)Google Scholar; Reitzenstein, , Nachr. Gött. Ges. (1906) 40 ff.Google Scholar; Tresp, , Die Fragmente der griech. Kultschriftsteller 78 ff.Google Scholar

⁵² The last date is mentioned only in Schol. Aristoph. Plut. 1126 (Tresp 79 f.) : here ἐνάτη is my conjecture for Tphri for the following reasons: (1) the enumeration is with one understandable exception progressive; (2) Philochorus attributed the 3rd to Athena (fr. 189 Jac. = Procl., , Hes. Erga 778Google Scholar; Tresp 83); (3) the 9th is in fact given to the Charites in CCAG. III, 33Google Scholar (the 19th to the Muses); cf. Cumont 263.

⁵³ Cf. Tresp 105 f.

⁵⁴ Cf. Reitzenstein, loc. cit. 40.

⁵⁵ CCAG. III, 32Google Scholar; cf. XI, 1, 134 ff.; Cumont 263 f.

⁵⁶ E.g. in CCAG. XI, 2, 157Google Scholar and in some Latin texts published by Svenberg, op. cit.—A different tradition is preserved in CCAG. X, 75Google Scholar where an angel and a demon with unintelligible names are attributed to every day: 1. ; 2. , etc.; cf. the magical demons of the hours, below, n. 102.

⁵⁷ CCAG. XI, 1, 134 ff.Google Scholar; IV, 142; VIII, 4, 105.

⁵⁸ Cf. W. Schmidt, op. cit. 12 ff.; 24 ff.; on the monthly festivals of the Greeks see also Nilsson loc. cit. 439, 3; id.Entstehung u. religiöse Bedeutung d. griech. Kalenders (1918) 33 ff. (among these, celebration of birthdays, e.g. Plato's on the 7th, Epicurus’ on the 20th). In Rome, the first of the month was a festival of the Lares (Tib. I, 3, 34; Hor. c. III, 23, 2; Prop. IV, 3, 53; also the Nonae and Idus: Cato agr. 143, 2); on the Kalendae there was also a sacrifice to Iuno, on the Idus to Iuppiter (Macrob. I, 15, 17 f.). The anniversaries of the temples of Iuno and Iuppiter, natales templorum, were celebrated on the Kalendae and Idus respectively (see Wissowa, , Religion² 568 ffGoogle Scholar.).

⁵⁹ Cf. SIG. 1024 ff. and Philochorus' work περὶ ἐορτῶν; Laqueur, , RE XIX, 2435 f.Google Scholar

⁶⁰ Cf. Reitzenstein, , Poimandres 287 f.Google Scholar; Boll, , Aus der Offenbarung Johannis 23 f.Google Scholar

⁶¹ What we call folkloristic tradition is nothing else but scattered fragments of those lunar calendars. There are many advices regulating the farmer's work: what should be done or avoided crescente or decrescente luna in agricultu (Varr. r.r. I, 37; Pliny XVIII, 321 ff.), for cutting timber, even hair- and nail-cutting (Cato agr. 31, 2; Varro I, 37, 2; Pliny XVI, 194; XXVIII, 28); there are even exact dates: Cato agr. 37, 4 advises cutting timber diebus vii proximis quibus luna plena fuerit, and Pliny suggests it a xx in xxx, and hair-cutting (XXVIII, 28) xvii luna vel xxviiii; see also Tavenner, , TAPA XLIX, (1918) 67 ff.Google Scholar Another fragment of calendary lore rendered the day of the main phases of the Moon, the Jewish Sabbath, adverse for any undertaking: Mt. 24, 20; Hor. sat. I, 9, 69; Tib. I, 3, 18; Boll 134, 1. It seems possible that this observance of the Moon-phases was transferred to the Nundinae: this is perhaps meant with Plutarch's words (Q.R. 42) that the Nundinae were sacred to Kronos; see also Kroll, , RE XVII, 1469 f.Google Scholar

⁶² Cf. Eudoxus ap. Cic. div. II, 87; Theophr. ap. Procl. Tim. III, 151 D. However, such manipulations with the lunar reckoning were not unknown in Egypt: there is a hemerology of the 12th dynasty (c. 1800 B.C.) recording the good and bad days of the month and a more elaborate menology (c. 1200 B.C.), similar to those of Babylonia, see Meyer, E., Gesch. d. Alt. II, 1², 418 f.Google Scholar

⁶³ Cf. Langdon, , Babylonian Menologies and the Semitic Calendars 73 ff.Google Scholar; Labat, R., Hémérologies et ménologies d'Assur (1939) 50 ff.Google Scholar

⁶⁴ Langdon, op. cit. 77.

⁶⁵ Langdon 73 ff. On the thirty divine patrons in Persia see Zoroastr. fr. S 5; 7 (Bidez-Cumont, , Les Mages hellénisés II, 102; 109)Google Scholar.

⁶⁶ Pap. Mus. Brit. 121, 757–790 = PGM. II, p. 34Google Scholar Pr. It will be noticed that one of the twenty-eight symbols is missing. Preisendanz makes up the number by punctuation ; Reitzenstein, , Poimandres 263Google Scholar and Boll, , T'oung Pao XIII, (1912) 711Google Scholar n. 13 are wrong in adding at the beginning σῦς.

⁶⁷ Cf. Reitzenstein, , Das iranische Erlösungsmysterium 171 ff.Google Scholar; Nock, Harv. Theol. Rev. XXVII (1934) 84Google Scholar.

⁶⁸ Plut. Is. 42 f.; Procl., Hes. Erga 817Google Scholar; Ox. Pap. 886 (= Preisendanz, PGM. II, p. 151 fGoogle Scholar.); Horap. I, 10. I am dealing here with the festival of Osiris in so far as it reflects lunar symbolism; for other evidence see Gressmann, , ‘Tod u. Auferstehung des Osiris’, Der alte Orient XXIII, 3 (1923)Google Scholar.

⁶⁹ Damasc. vita Isid., Phot. Bibl. 343a4 B.; Reitzenstein, , Poimandres 266Google Scholar.

⁷⁰ Horap. I, 10; Plut. Is. 74; Aelian. h.a. X, 15; Porph. abst. IV, 9; cf. Lyd. mens. III, 11 (p. 50, 14 W.).

⁷¹ Pap. Leid. J 395, l. 207 f. (= PGM. II, p. 97Google Scholar); 557 f.; 561; 701 f.; cf. Dornseiff, , Das Alphabet in Mystik u. Magie 35 ff.Google Scholar

⁷² Cf. Dornseiff 67 f. etc.

⁷³ Cf. Dieterich, , Mithrasliturgie³ 40Google Scholar; 228; Eitrem, , Symb. Osl. XXII (1942) 70 ff.Google Scholar

⁷⁴ Pap. Leid. J 395, l. 776 f. ( = PGM. II, p. 122Google Scholar); Berlin Pap. gr. 9566, 13 f.(= PGM. II, p. 146Google Scholar); Reitzenstein, , Poimandres 263Google Scholar; Dornseiff 37; Nilsson, , Bull. Soc. Royale des Lettres de Lund (1947/1948) II, 88Google Scholar.

⁷⁵ Diod. II, 57, 4; Dornseiff 38. Cf. also Lucian, . Vera hist. I, 13Google Scholar where the army of the Moon-dwellers consists of twenty-eight myriads: Boll, , Aus der Offenbarung Johannis 146Google Scholar.

⁷⁶ Exc. ex Nicomacho 4–6, 9; p. 275–80 Jan (see esp.; p. 276); Dornseiff 52.

⁷⁷ The history of this magical practice begins with the evidence of papyri, concerning numbers and letters, sounds and names, etc., mentioned above. In a further text (Ox. Pap. 886 = PGM. II, p. 151 fGoogle Scholar.) twenty-nine palm-leaves were used, each inscribed with the names of the gods (apparently of the days of the month): these were lifted two by two, and the leaf which remained last contained the answer to the query; reference is made to the symbolism of Isis and Osiris. Another papyrus, PGM. I, p. 194Google Scholar (l. 370 ff.; cf. Festugière I, 294), mentions twenty-eight laurel-leaves, no doubt for a similar manipulation. These are just casual scraps of a widespread practice. Albertus Magnus (thirteenth century) or whoever wrote under his name the Speculum astrologicum quotes a liber Praestigiorum of Hermes Trismegistos and another book of (Pseudo- ) Apollonios of Tyana, describing the amulets of the mansions (cf. CCAG. V, 1, 98Google Scholar; Thorndike, , History of Magic II, 223Google Scholar; 699; Festugière, , La Révélation d'Hermès Trismégiste I, 106Google Scholar): each of the amulets contained the pictorial representation of the mansion, seven divine names in correct or contrary order, to achieve or to avert something. They were fumigated with two kinds of spice; two angels were in the service of each of them. The anonymous Arabic work, Picatrix, following Indian sources (which in turn will have been based on Greek texts) gives a long description of the amulets, e.g. (p. 13 of the forthcoming translation of Ritter-Plessner, kindly made available to me by the late Professor Saxl): ‘… und in dieser Station (i.e. the first) macht man Talismane, um Ehegatten und Freunde durch Entfremdung oder Feindschaft auseinanderzubringen, und ferner Talismane, dass ein Sklave entfliehen und entlaufen kann, zu wem er will und wünscht, und um die Genossenschaft von Genossen zu zerstören …’ (it will be noticed that the aim of the amulets corresponds to the day-to-day prognostics such as are given in the Lunaria, also in the codex Cromwellianus); on the general character of Picatrix see Ritter, , Vorträge der Bibliothek Warburg I, 94 ff.Google Scholar; Thorndike II, 820. The cod. Coll. Corp. Chr. (Oxon.) 125, saec. xiii, f. 62 contains a short extract from such a work, ascribed to Hermes and Apollonius; a longer text is in the cod. Urbin. lat. 1384, saec. xv, f. 7–26, with illustrations of the mansions and amulets. Later works, registered in the various volumes of Thorndike's History of Magic, are based on these and similar texts; it will suffice to mention De occulta philosophia by Agrippa von Nettesheim (1486–1535), which contains a chapter, II, 46, De imaginibus mansionum Lunae.

⁷⁸ Cf. Heracl. fr. 6 D.; Manil. I, 184; Serv. Aen. IV, 584; Boll, , SB Heidelb. (1910, 16Google Scholar. Abh.), 42 n. 35; Nilsson, , Arch. Rel. Wiss. XXX (1933) 144Google Scholar, 3; Nock, , Harv. Theol. Rev. XXVII (1934) 76Google Scholar.

⁷⁹ Cf. Macrob. I, 18, 10; Mart. Cap. I, 76; Boll, loc. cit.; JRS XXXVIII (1948) 40 f.Google Scholar

⁸⁰ Iambl. myst. VII, 3; Porph. ap. Euseb. pr. ev. V, 10, 4; Procl., Tim. I, 107 DGoogle Scholar.

⁸¹ Procl., Tim. III, 56 D.Google Scholar; cf. in remp. II, 58 Kr.; Reitzenstein, , Das iranische Erlösungsmyst. 168Google Scholar, 4; Gundel, , Dekane 234Google Scholar, 1.

⁸² See also above, p. 58, the evidence of the Orphic Ἐφημερίδες.

⁸³ Cf. Boll, , Sphaera 295 ff.Google Scholar; Reitzenstein, , Poimandres 256 ff.Google Scholar; Sethe, , Nachr. Gött. Ges. (1920) 103 ff.Google Scholar; 118 f.; Gundel, , Neue astrol. Texte d. Herm. Trismeg. 229 ff.Google Scholar

⁸⁴ PGM. I, p. 54Google Scholar (l. 501 ff.); cf. Reitzenstein 257; Boll (see n. 85) 710; 718.

⁸⁵ T'oung Pao XIII (1912) 711Google Scholar n. 13.

⁸⁶ Cf. PGM. I, p. 144Google Scholar (l. 2302) ; (l. 2335) , ; Mich. Pap. 154, i, 90.

⁸⁷ T'oung Pao XIII (1912) 704Google Scholar; 711 n. 13.

⁸⁸ PGM. I, p. 152Google Scholar (l. 2549 ff.); p. 162 (l. 2800); cf. Eitrem, loc. cit.

⁸⁹ Porph. abst. IV, 16; cf. Cumont, , Die Mysterien des Mithra³ 138 ff.Google Scholar; Reitzenstein, , Das iranische Erlösungsmyst. 168Google Scholar; i d. Hellenist. Mysterienrel.³ 43.

⁹⁰ Apul., Met. XI, 24Google Scholar; cf. Reitzenstein, , Das iran. Erlösungsmyst. 167 f.Google Scholar

⁹¹ Chavannes, , T'oung Pao N.S. VII (1906) 106 f.Google Scholar; Boll, loc. cit.; a different development is suggested by Jeremias, , Handb. d. altorient. Geisteskultur² 243Google Scholar.

⁹² Cf. Boll, loc. cit. There was also a third set, the thirty-six animals of the decans, in Egypt as well as in China quoted by Boll in support of his argument (see also below n. 117).

⁹³ Boll, , Sphaera 296Google Scholar.

⁹⁴ Cf. CCAG. V, 4, 172 ff.Google Scholar; Zoroastr. fr. O 42 B.-C.; Bidez-Cumont, op. cit. I, 122.

⁹⁵ Cf. Peiser, , Zeitschr. f. Assyrol. IV, 361 ff.Google Scholar; Cuneiform Texts … in the Brit. Mus. xxii, 48Google Scholar; Boll, , T'oung Pao XIII (1912) 713Google Scholar n. 25.

⁹⁶ CCAG. VII, 157, 6Google Scholar.

⁹⁷ Cf. Chavannes, loc. cit.; Boll, loc. cit. 700; id.Sphaera 328.

⁹⁸ CCAG. VII, 194 ff.Google Scholar; cf. VIII, 4, 198, 16; 217, 7.

⁹⁹ Boll, 295 f.; Gundel, , Neue astrol. Texte 229 ff.Google Scholar

¹⁰⁰ Ps.-Callisth. 12, 7.

¹⁰¹ It may well be that the animal signs of the zodiac too had such a function before they were identified with constellations. If they had, our zodiac would represent a selection from a third set of lunar symbols, the other two being those of the London Papyrus and of the Orphics. But of course this is just a conjecture which cannot be proved.

¹⁰² PGM. II, p. 39Google Scholar (l. 899 ff.); CCAG. VII, 177 ff.Google Scholar; Reitzenstein 257, 2; Boll, , Aus der Offenbarung Johannis 36Google Scholar.

¹⁰³ Hippol., Elench. V, 14Google Scholar; Boll, , Sphaera 309 f.Google Scholar; Reitzenstein 257; on the twelve hours as twelve luminous kings see Reitzenstein, , Hell. Mysterienrel.³ 227Google Scholar.

¹⁰⁴ On other sets and on the Roman tradition see Boll, , Sphaera 472 ff.Google Scholar; Weinreich, , Myth. Lex. VI, 764 ff.Google Scholar; 823 ff.; on representation Weinreich 826; Ferri, , Bull. comun. LXVII (1939, Bull. Mus. Imp.), 39 f.Google Scholar

¹⁰⁵ Fr. 8 B.-G. = Hierosolym, Cosmas., Patr. Gr. XXXVIII, 461Google Scholar; cf. Bidez-Cumont, , Les Mages hellénisés I, 175 ffGoogle Scholar. The codex Clarkianus 12 (cf. Gaisford, , Catal. … MSS. qui a cel. E. D. Clarke comparati in Bib. Bodl. adservantur (1812)Google Scholar: not used by B.-C.) contains the following readings: Τιθύς; Κοῦρος (no doubt correct: Καιρός Vat., B.–C.: Κόρος Boll); Λύμη (also Vat., perhaps correct: Λοιμός Boll, B.-C.); Φόρη (also Vat.: Κόρη B.-C).

¹⁰⁶ Cf. Boll, , RE VII, 2557 f.Google Scholar; Nock, , AJA L (1946), 164Google Scholar n. 88.

¹⁰⁷ Cf. Boll-Bezold, , Sternglaube³ 58Google Scholar; 148. Incidentally, this distribution is based on the normal sequence of the signs, beginning however with Aquarius instead of Aries and ending with Capricorn so that Cancer and Leo, the single domiciles of the Moon and Sun respectively, occupy the centre. The other planets receive the next signs as their domiciles, one on the right, the other on the left cf this centre. The order of the planets too is correct, except that the Sun is moved into the centre of the two halves from its place between Mars and Iuppiter. Now if we identify Aquarius with January, we observe that March-Aries is under Mars, April-Taurus under Venus (on Aprilis ∼ Aphrodite see Walde-Hofmann I, 847), May-Gemini under Mercury (= son of Maia), June-Cancer under the Moon (cf. Iuno Lucina). Coincidence?

¹⁰⁸ Cf. Chavannes, loc. cit. 51 ff.; Halévy, , T'oung Pao N.S. VII (1906) 270 ff.Google Scholar; Pelliot, ibid. XXVI (1929) 204 ff.; Coedes, ibid. XXXI (1935) 315 ff.

¹⁰⁹ See above n. 94.

¹¹⁰ Cf. Boll, , Sphaera 476 ff.Google Scholar; Weinreich 819; 822.

¹¹¹ Cf. Boll, , Sphaera 336Google Scholar, 2; contra Sethe loc. cit.. On the Babylonian origin of this doctrine see Van der Waerden, B. L., Journ. Near Eastern St., VIII (1949) 22 f.Google Scholar

¹¹² Cf. Sen. NQ. II, 41, 2; Aug. CD. IV, 23; Mart. Cap. IX, 914; Weinreich 818 ff.

¹¹³ Sethe, , Nachr. Gött. Ges. (1920) 98Google Scholar, 3; cf. also Ioh. Gaz. I, 314 ff.; on representations of the twelve hours in Dendera and Edfu, Graeco-Roman period, see Daressy, , Annales du Service des Antiquités de l'Égypte XVII (1917) 197 ff.Google Scholar

¹¹⁴ Gunkel, , Schöpfung u. Chaos 302 ff.Google Scholar; id. Zum religionsgesch. Verständnis d. NT. 42 f. But his further explanation, that the twenty-four δικασταί and πρεσβύτεροι represented the twenty-four mansions of the Moon (this number was at one time presumed by Hommel, , Zeitschr. d. Deutschen Morgenl. Ges. XLV (1891) 599 f.Google Scholar) was rightly, I think, rejected by Boll 35, 6. The Persians too had a set of twenty-four gods (Plut. Is. 47) but it is not certain that they belong to this context, see Bousset's Commentary 247; Bidez-Cumont, op. cit. II, 76 n. 16.

¹¹⁵ This view must have been held more widely in the ancient world because a magical text, published by Reitzenstein, , Poimandres 301Google Scholar, contains a list (sic) πρεσβυτέρων.

¹¹⁶ It is thought possible by Bousset (op. cit.) and Clemen, , Religionsgesch. Erklärung d. NT.² 375 f.Google Scholar; rejected by Charles, Lohmeyer and others quoted in the voluminous monograph by Michl, J., Die 24 Ältesten in d. Apokalypse d. hl. Johannes (1938) 133 ff.Google Scholar

¹¹⁷ Scott, , Hermetica III, 371Google Scholar; Bidez, , Mélanges Capart (1935) 53 fGoogle Scholar. It is worth adding that the decans went through the same transformations as did the other symbols of time. There was a set of thirty-six animals (an enlarged version of the twenty-eight); there were many lists attributing to them secret names, Egyptian and other demons; their personifications are also found in monuments and in illuminated manuscripts: see Gundel, , Dekane u. Dekansternbilder (1936) 223 ff.Google Scholar