"Sol Invictus, the Winter Solstice, and the Origins of Christmas", Mouseion, Number 47/3 (2003), 277-298 more

Mouseion, Series III. Vol. 3 (2003) 377-398 ©2003 Mouseion Sol Invictus. the Winter Solstice, and the Origins of Christmas STEVEN HIJMANS It is well known that we do not know the exact date of the birth of Jesus Christ. In most churches December 25 is celebrated as such, although the Armenian Church retains January 6 (Epiphany). But the fact is that the date was not recorded. Many early Christians, as Origen indicates in A.d. 245, actually opposed celebrating birthdays altogether, including Christ's, because it was deemed to be a pagan practice.1 All emphasis lay (and in the Orthodox church still lies) on the celebration of the passion and resurrection at Easter. In fact, the whole question of the exact date of Christ's birthday appears to have arisen only when Christian chronographers began writing their chronologies. Obviously the birthday of Christ had to be established in such chro- nologies, and numerous dates were proposed. In the late second century A.D. Clement of Alexandria refers to speculations (which he dismisses as superstition) which dated the birth on April 19 or May 20. Clement himself calculated it as November 17. 3 B.C.2 Other suggested dates in- cluded March 28 and April 2, but not December 25? None of the dates was influential or enjoyed any preference or official recognition. Their basis varied from learned calculations to pure guess-work. The De Pas- cha Computus, for instance, written in A.d. 243, argued that Creation began with the vernal equinox, i.e. March 25, and that the Sun, created on the fourth day, was therefore created on March 28. This obviously meant that Christ, the new "Sun of Righteousness," must have been 1 Origen, Comm. in Mt 10.22; cf. Kraabel (1982) 274. 2 Clement, Misc. 1.21; cf. Cullmann (1956) 22-23 n- 6. 3 March 28: see below: April 2: Hippolytos. Dan. 4.23. ms. A; Hippolytos cal- culated Christ's year of birth as A.D. 8 because that would make it 5500 years after creation as he calculated it—despite the fact that as a result Christ would have been crucified under Caligula rather than Tiberius. Cullmann (1956) 22 n. 5. Kraabel (1982) 274 (citing Cullmann), and the EECs.v. Christmas (206) all claim that as early as 221 Julius Africanus calculated the date as December 25 in his fragmentarily preserved Chronicle, but provide no reference. Dr. Martin Wallraff of the Rheinische Friedrich-Wilhelms-Universitat in Bonn, who is di- recting a project to produce the first critical edition of all preserved fragments of the corpus of Julius Africanus, has kindly informed me that he does not know of any such calculation by Africanus. 377 378 STEVEN HIJMANS born on March 28. To support these dates the author proclaimed ex- plicitly that he had been inspired ab ipso Deo.* It was only in the 330s, apparently, that December 25 was first promoted as a feast day cele- brating the birthday of Christ. Initially this happened only in Rome, but by the 380s it is attested as such in Asia Minor as well, and by the 430s in Egypt. Nonetheless, other churches, as we have seen, continued to maintain Epiphany—January 6—as the birthday of Christ, and do so to this day. It is usually suggested that establishing a feast day on the birthday of Christ became important as a result of doctrinal disputes concerning the human and divine natures of Christ. There had been numerous groups that argued for a strong distinction between the human and the divine. For example, in the second century the Basilideans taught that the divine Christ appeared on Epiphany to reside temporarily in the body of the human Christ. In their view, the date of birth of the human Jesus was of no interest, as he was only temporarily "host" to the divine Christ. Two centuries later the Manicheans went further, claiming that Jesus either was not born at all, or in any case did not take flesh of the Virgin Mary, but simply appeared among men—on Epiphany. One can easily imagine how the feast of Epiphany could be linked exclusively to the divine aspect of Christ, which was somehow "revealed" on January 6. If the church were to celebrate Epiphany only, but not the birth of Christ, that could be seen to emphasize that there was indeed a distinc- tion of importance between the two natures of Christ, human (birth, unimportant) and divine (Epiphany, important). Developing a feast for the birthday of Christ was a reaction to these views, counteracting such dichotomies by stressing the importance of the physical birth of Jesus.5 But why was the birth date set at December 25, rather than March 28, for instance, or one of those other dates previously proposed? In a recent article Francois Heim ([1999] 651) states that it is now universally accepted "que la date de Noel a ete fixee au 25 decembre pour opposer les festivites chretiennes aux festivites paiennes." The pagan festivities Heim is referring to are those celebrating the winter solstice, which in the Julian calendar was calculated as December 25th. According to the famous Calendar 0/354, 30 chariot races were held on this day to cele- 4 Cullmann (1956) 21-22. 5 Cf. Cullmann (1956) 30. Kraabel (1982) 276-278 discusses the lack of interest in Christ's birth displayed by the infancy gospels, such as the Protevangelium of James. He emphasizes the importance of a feast celebrating the birth of Christ (and thus his incarnation) to counter the views of docetism, of Arius, and later of Apollinaris and Nestorius. and suggests that it was no chance that the strongly anti-Arian Julius I (337-52) was bishop in Rome precisely at the time that Christmas was first celebrated there. SOL INVICTUS 379 brate the Natalis Invicti, that is, the birthday of Sol Invictus.6 This feast of Sol Invictus, then, was the festival that Heim claims the Church fa- thers wanted to displace with Christmas. And indeed, ever since Hermann Usener's studies of the feast of Christmas, the idea that De- cember 25 was chosen as Christ's birthday because of this important pagan festival has received wide acceptance.7 Usener referred to two sources in particular to support his conten- tion that December 25. the winter solstice, was consciously chosen be- cause of its pagan feast for the invincible sun. The first is a scholiast to the Syriac author Bar Salibi, whom Usener ([1905] 466) quotes as writ- ing: Die Ursache, weshalb die Vater das Fest des 6. Januar (Epiphanie) abanderten und auf den 25. Dezember verlegten, war folgende. Nach feierlichem Herkommen pflegten die Heiden am 25. Dec. das Geburts- fest des Sonnengotts (wortlicher: das Fest des Aufgangs der Sonne) zu feiern und zur Erhdhung der Festiichkeit Lichter anzuzunden. An die- sen festlichen Brauchen liessen sie auch das Christenvolk theilnehmen. Da nun die Lehrer der Kirche die Wahrnehmung machten, dass die Christen an diesem (heidnischen) Feste hiengen, kamen sie nach reifli- cher Erwagung zu dem Entschluss. an diesem Tage (25. Dec.) fortan das Fest des wahren Aufgangs (dh. Geburt), am 6. Januar aber das Fest der Erscheinung (Epiphanie) zu feiern. Und diesen Brauch haben sie bis zum heutigen Tage zugleich mit der Sitte, Lichter anzuzunden, festge- halten.8 We will return to this passage below (p. 396). The second passage comes from a homily referred to by Usener as the De solstitiis et aequi- noctiis, the full title of which is: De solstitia et aequinoctia conception^ et nativitatis Domini Nostri Iesu Christi et Iohannis Baptistae ("On the solstice and equinox of the conception and birth of Our Lord Jesus 6Stern(i953) 108. no,254,285;Salzman(1990) 149-153. 7 Usener (1889) and (1905): Cullmann (1953); Heim (1999) 651-652 (with refs.). Kraabel ([1982] 277-278) lists three possible reasons: actual belief in the date, synchronicity, and Realpolitik. He suggests that insofar as Realpolitik played a role, it was not merely the replacement of a Solar festival which was at stake, but more generally the construction of a uniform, coherent Christian calendar. 8 "The reason why the Fathers changed the feast of January 6 (Epiphany) and moved it to December 25 was as follows: according to festive custom the pagans were wont to celebrate the birthday of Sol (literally: the feast of the rising of the sun) on December 25, and to heighten the festivities they lit lights. They also let the Christians participate in these festivities. Now because the doctors of the church saw that the Christians were attached to this (pagan) feast, they decided after long deliberation to celebrate henceforth the feast of the True Arising (i.e. birth) on this day (i.e. December 25), but on January 6 to celebrate the feast of the Appearance (Epiphany). And to this day they have maintained this custom, and with it the custom to light lights." 38o STEVEN HIJMANS Christ and John the Baptist").9 Towards the end. we read about the winter solstice: sed et invicti natalem apellant. quis utique tarn invictus nisi dominus noster qui mortem subactam devicit? vel quod dicant solis esse natalem ipse est sol iustitiae de quo malachias propheta dixit: orietur vobis ti- mentibus nomen ipsius sol iustitiae et sanitas est in pennis eius.'0 According to Usener, these two passages indicate with "surprising openness" how Christmas was created by church politics to replace a dangerous pagan feast of great significance. As Heim has indicated, most scholars since have followed his line, and scholars have invariably linked this supplanting of the cult of the sun to the pervasive use of the sun as metaphor for Christ in early Christian literature, in which he is often referred to as "Sol verus." "Sol Justitiae." etc." Nonetheless. Usener's position raises serious questions. There can be no doubt that the church fathers who elected to celebrate December 25 as the day of Christ's birth were fully aware of the significance that that day had in the cult of Sol Invictus. The question is whether they chose it because of or despite that pagan significance. If we ignore, for a mo- ment, the two sources quoted by Usener—both will prove to be more problematic and less conclusive than they may appear at first glance—then one must wonder whether the latter is not the more likely. If we accept that the decision to celebrate Christ's birth with a separate feast was primarily a post-Nicaean move designed to emphasize the incarnation of Christ—a move taken in the context of the struggle against the Arians and other, comparable groups—that would make it primarily an intra-faith (Christian versus Christian) polemical move, rather than an inter-faith (Christian versus pagan) one, with the initia- tive (and therefore the risk) on the side of those breaking with tradition. 9 Henceforth referred to in this article as De Solstitiis. There is no critical edition available of the homily, but Botte (1932) gives the full text, based on his collation of a limited number of manuscripts. Usener translates: "Aber man nennt den Tag auch Geburtsfest des Invic- tus. Ja wer ist denn so unbesiegbar ausser unserem Herrn. der den Tod sieg- reich unterworfen hat? Und wenn man sagt, es sei der Geburtstag der Sonne, nun er selbst ist die Sonne der Gerechtigkeit, von dem der Prophet Malachias gesagt hat: Aufgehn wird euch. wenn ihr seinen Namen furchtet, die Sonne der Gerechtigkeit, und Heil ist in ihren Schwingen'." ["But they also call this <day> the birthday of the invincible one (Invictus). But who then is as invincible as our lord who defeated the death he suffered? And if they say that this is the birth- day of the sun, well He Himself is the Sun of Justice ..."]: cf. Botte (1932) 105, lines 434-39- " jEECs.v. "Christmas": iMCs.v. "Geburtstag," sp. 239; cf. jRACs.v. Epipha- nie; Heim (1999) 651, with refs.; cf. now also Wallraff (2001). SOL INVICTUS 381 i.e., the anti-Arians. They would need strong arguments to defend their move against the inevitable opposition, and one can easily imagine the problems they would face if their main reason for choosing December 25 was the fact that it was the date of a dangerous pagan feast. Of course, it remains equally possible that the move was initially a relatively insignificant, local development in Rome, which was not di- rected explicitly against the Arians or other Christian groups, and which for a significant time had little impact beyond the city. Seen from the Christian perspective, the latter scenario would leave more oppor- tunity for the move to have been directed explicitly against a popular pagan feast or practice in Rome, in casu the celebrations in honour of Sol Invictus. However, whichever scenario one considers intrinsically more likely from a Christian perspective on the choice, it is obvious that the nature of the pagan feast in honour of Sol which is recorded for De- cember 25 will have played some role in the considerations, and it is to this feast that we must now turn. From Usener (1905) to Heim (1999), all scholars who argue that Christmas was instituted to counteract the pagan feast of December 25 in honour of Sol strongly emphasize the pagan nature of that feast and the great importance and popularity of Sol Invictus in late antiquity. Little evidence is offered for the former contention beyond reference to the 30 chariot races in honour of Sol recorded for December 25 in the Calendar of 354. The latter contention, however, is generally supported by extensive discussions on the nature and importance of Sol Invictus. Heim dedicates over half his article to a substantial review of the cult of Sol Invictus in the Roman Empire (640-645), and the continuing influ- ence of the sun god after Constantine (645-651)." Unfortunately this review and others like it are seriously flawed, both in their interpreta- tion of the Roman sun god in general and specifically in their treatment of the sun in Late Antiquity as a typical—and important—pagan god. Indeed, I would argue that they misunderstand or misinterpret funda- mental aspects of Roman polytheism and its interaction with Christian- ity which are directly relevant to the problem at hand. Rather than analyze and refute the views of all relevant scholars, I will limit myself to Heim (1999) as the most recent, but also quite typi- cal, example. In his article, Heim essentially repeats the position, first proposed in the nineteenth century, that Sol Invictus was an oriental newcomer to Rome. According to this view, Rome had had no sun god since the disappearance of the traditional, but unimportant Sol Indiges 12 The rest of his article (651-666) is devoted to fourth-century and fifth- century homilies in which reference is made to the winter solstice—December 25, birth of Christ—and the summer solstice. 382 STEVEN HIJMANS in the first century A.D. It was, we are told, the fourteen year old Ela- gabalus, coming to Rome from Syria in A.D. 219, who was the first to impose this oriental cult of Sol Invictus on Rome as its supreme deity. After Elagabalus' death in A.D. 222, the cult of Sol Invictus was discred- ited until it was reintroduced by Aurelian in A.D. 274, and subsequently the sun-god remained a major deity in the Roman pantheon until well into the reign of Constantine (ca. A.D. 32o).'3 Faced by the rapid rise of Christianity, the cult of Sol declined swiftly in the fourth century A.D., although Julian (361-363) was a strong supporter of the cult, as shown by his Hymn to Helios. As I have argued in extenso elsewhere, this synopsis of the history and development of the cult of Sol in Rome and the Roman Empire is simply untenable in light of the extensive archaeological, iconographic. epigraphic, numismatic, and literary evidence that documents the con- tinuous presence in Rome of the sun-god in some form or another from as far back in history as we can trace Roman religion at all.14 Literary sources place Sol (and Luna) among the gods introduced by Tatius'5: Sol was already portrayed on one of the earliest coins minted by Rome16 and continued to appear on Roman state coins for the next 500 years—sporadically until the late second century A.D., and almost an- nually, often with multiple issues in one year, from the reign of Corn- modus until the reign of Constantine17; in art the iconography of the 13 Heim ({1999] 643) does not mention the Republican Sol Indiges. but does refer in passing to the solar aspects of the Augustan Apollo. Otherwise, he fol- lows the standard paradigm, although he suggests that during the quarter- century of the Tetrarchy the cult of Sol was eclipsed, but not stifled, by the cults of Hercules and Jupiter. For a comparable synopsis of the cult of Sol. cf. Salzman (1990) 149-153 (with refs.). 14 Hijmans (1996). 15 Varro, L. 5,74. 16 An uncia minted between 217 and 215 B.C. (LJMCs.v. Helios/Sol 18; Gobi [1978] nr. 1419). Various comparable coins were minted outside Rome in this period, and there is some discussion whether this coin was minted by Rome itself, or elsewhere under Roman jurisdiction. Cf. Thomson (1961) 2.229-231; Crawford (1974) 150,39/4. 17 This presence of Sol on Roman Imperial coins is itself sufficient evidence of the fact that Sol was not deemed to be "foreign" or Syrian. The Romans freely venerated deities of non-Roman origin, of course, and many gained a significant degree of popularity (Isis and Sarapis, Jupiter Dolichenus. and Mithras are numbered among the 14 deities most mentioned in Latin inscrip- tions; cf. MacMullen [1981] 5-7). However, these non-Roman deities were rarely—if ever—depicted on official, imperial Roman coins. Sol, by contrast, is depicted on well over 1500 issues. For the Republic, cf. the issues (mainly de- narii) of M. Aburius, ca. 124-103 B.C. (Griiber [1910] 1.146 nos. 995-997); Cornelius Sisenna, ca. 118-100 B.C. (LIMCs.v. Helios/Sol 398); Aulus Manlius, SOLINVICTUS 383 Roman Sol was firmly established well before the destruction of Pom- peii18 and remained esentially unchanged in subsequent centuries'9; likewise in inscriptions Sol occurs regularly through the centuries, without any break or hiatus, nor with any indication of a radical change in his nature.™ ca. 115-107 B.C. (Gobi [1978] nr. 3339. incorrectly stating that this is the oldest Roman coin of Sol); M.L. Aquila, probably 108 B.C. (Mattingly {1960] 61, pi. XIV nr. 18); Lucretius Trio, 76 B.C. (Gobi {1978] 3443); C. Coelius Caldus, middle of the first century B.C. (Gobi [1978] 3833); Cordius Rufus, 46 B.C. (quinarius; LIMCs.v. Helios/Sol 20); Valerius Acisculus, 45 B.C. (Gruber [1910] 1.4110, p. 536); Mark Antony, 42 B.C. (Gruber [1910] 2.398, nr. 60; LIMCs.v. Helios/Sol 21); Mussidius Longus. 39 B.C. (Gruber [1910] 1.578, nr. 4250); P. Clodius, 42-38 B.C. (aureus, denarius; Gruber [1910] 1.585, nr. 4284-4289). The issues under the empire are far more numerous, and can only be summarized here without ref- erences (for the following survey, cf. Hijmans [1989] 1.16-34 and 2.140-170, with refs.; Hijmans [1996] 135-137): under Augustus, cf. a range of coins of L. Aquil- lius Floras, ca. 14 B.C. (LIMC s.v. Helios/Sol 24): Vespasian minted a rare but famous coin with Sol, and both he and his successors Titus and Domitian issued coins of Aeternitas holding busts of Sol and Luna (see n. 38 below); Trajan and Hadrian minted a number of coins with Sol, and there are also some second- century medallions depicting Sol; from Commodus and especially Septimius Severus onwards the number of imperial coins with Sol increased significantly, and in the third century the only emperors not to mint coins depicting Sol were the four who reigned between 249 and 253 (Trajanus Decius, Herennius Etrus- cus, Trebonianus Gallus, and Volusianus). Clearly, the image of Sol was well established on Roman coins long before the arrival of Elagabalus and never disappeared after his death. Certainly it is impossible to maintain that his cult was somehow reintroduced by Aurelian, given the fact that he has a very sig- nificant presence in the coinage of, inter alia, Gordian III (238-242, annual emis- sions), Gallienus (253-268, probably annually, often multiple emissions per year), and especially Claudius Gothicus (268-270, a surprisingly high number of emissions). The number of emissions under Aurelian is high, but by no means unprecedented. Following his reign, the number of issues is quite uneven; high under Florianus (276), Probus (276-282), Numerianus (183-284), Maximinus Daia (305-313) and to a lesser extent Galerius Maximianus (293-311), but insignificant under Tacitus (275-6), Carinus (282), Carus (282-3), Diocletian (284-305), Maxi- mianus Herculius (286-305), Constantius Chlorus (293-306) and Severus II (305-307). Constantine, finally, minted very significant numbers of coins until about 317, after which the number of emissions declines drastically until, ca. 323, the last Roman coin of Sol was minted. 18 Cf., e.g., LIMCs.v. Helios/Sol 90. 19 For an overview cf. LIMCs.v. Helios/Sol and Hijmans (1996) 127-134, 137-138, listing numerous examples from the first and second centuries A.D. On the significance of the raised right hand, an element of Sol's iconography from the late second century onwards and often identified by scholars as an "orien- tal" gesture indicating Sol's non-Roman origins (e.g. L'Orange [1935] 93-94), cf. Hijmans (1996) 124-125 (with further refs. in n. 65). M Hijmans (1996) 123-127. On the epithet invictus, of which the earliest 384 STEVEN HIJMANS It is important to note that in the nineteenth century, when the para- digm of a third-century A.D. oriental origin for Sol Invictus was estab- lished, this evidence was ignored, not as a result of poor or superficial scholarship, but because of a racistly tainted ideological bias against so- lar cults in general and, in particular, against its postulated corollary: ruler worship." Both were deemed decadent and therefore incompati- ble with the Roman psyche, making it both convenient and attractive to relegate their origin to the Semitic (and therefore decadent and infe- rior) Orient.» It is this ideological bias which originally produced the paradigm of an oriental origin for Sol Invictus which makes it especially unfortunate that it continues to be repeated uncritically in studies and handbooks today.23 It is clear, then, that ve must beware not to take established views on the origin and nature of Sol Invictus at face value. Furthermore, it must be stressed, pace Usener, that December 25 was neither a longstanding nor an especially important official feast day of Sol. It is mentioned only in the Calendar of 354 and as far as I can tell the suggestion that it was established by Aurelian cannot be proven. In fact, there is no firm evi- dence that this feast of Sol on December 25 antedates the feast of Christmas at all.24 The traditional feast days of Sol, as recorded in the firmly dated occurrence with Sol is on an inscription from A.D. 158, cf. also Weinstock (1957). 21 The radiate crown of the Roman emperors was understood as the link be- tween the two. Bergmann (1998) has recently reiterated the position that the imperial radiate crown was indeed a solar symbol with no real-life counterpart in antiquity, an opinion that I contest (Hijmans, forthcoming). 22 For an extensive analysis of this process, cf. Hijmans (1996) 115-126. 23 Cf., e.g.. the OCT? s.v. Sol. 24 On G. Wissowa's suggestion that the festival was instituted by Aurelian. cf. Salzman (1990) 151 n. 106; Heim (1999) 643 with refs. There is no explicit evi- dence stating that the feast of December 25 was instituted by Aurelian. In fact the Calendar of 354 is our only conclusive evidence for an official feast day in honour of Sol on that day. Other sources certainly suggest that the winter sol- stice was marked as more than a purely astronomical event, referring to it as the natalis of Sol. In the Calendar of Antiochus there is a discrepancy of three days between the actual solstice (marked for December 22) and the birthday of the sun, marked for December 25; that the latter was nonetheless linked to the solstice is suggested by the addition au£et cpcoc (Boll [1910I 16, 40-44). Just how complicated matters could become is shown by another calendar (CCAG 9.1.128-137; Weinstock [1948]), that marks the solstice on December 20, has nothing for December 25, and April 12 as iivycoua 'HXtou and November 22 HAioSuciot. Weinstock explains that these dates are of astrological significance, and their exact meaning need not concern us. For the important point for our purposes here is that none of this is evidence for the celebration of an official, traditional feast day for Sol on Deceember 25. In fact, none of the other feast- SOLINVICTUS 385 early imperial fasti, were August 8, August 9. August 28, and December 11.25 Of these, only August 28 is still mentioned in the Calendar of 354. along with October 19 and October 22, the latter being the most impor- tant, judging by the 36 chariot races with which it was celebrated.26 There is direct evidence that in A.D. 274 Aurelian instituted quadrennial agones in honour of Sol, which were held from October 19-22, and cul- minated in the unprecedented 36 chariot races recorded in the Calendar of 354 for October 22.17 However, while our sources are quite clear on the novelty of these quadrennial agones, they do not indicate whether they were an expansion of a pre-existing annual festival in honour of Sol or were something totally new. In the context of the nineteenth- century paradigm that Aurelian was (re)introducing the whole cult of Sol, scholars have automatically assumed the latter, but in fact there is no evidence that this was the case. On the contrary, the evidence for a continuous veneration of Sol in Rome, outlined above, postulates that we assume some continuity of festivals as well. One must wonder, therefore, whether Aurelian ignored the existing festivals when he in- stituted the agones, or expanded one of them. There is certainly circum- stantial evidence for the latter, because although the early imperial fasti do not indicate a festival specifically in honour of Sol for October 19, they do record an armilustrium for that day, which was held in the Cir- cus Maximus. The circus, of course, was closely linked to Sol, and prior to Aurelian the Circus Maximus was the site of the most important days for Sol had any connection with cosmic, astronomical, or astrological dates. This is true not only for the traditional feast days but also for the feast days in October, which sources do link directly to the Aurelian cult innovations (see below). It is therefore unwarranted to take for granted that the actual festi- val celebrating the Natalis Invicti with 30 chariot races, mentioned in the Calen- dar of 354, was itself older than the Christian feast honoring Christ's birthday on that day. On the evidence currently available we cannot exclude the possibil- ity that, for instance, the 30 chariot races held in honor of Sol on December 25 were instituted in reaction to the Christian claim of December 25 as the birthday of Christ. This is purely conjectural, of course, but by no means unlikely. As Bowersock has argued ([1990] 26-27, 44-53), the extent to which late pagan fes- tivals copied, incorporated, or responded to Christian practices, elements, and dates deserves far more attention than it has received 25 It is not clear whether the references to August 8 and 9 are to one festival or two (Latte [1960] 231-232). Cf. Degrassi (1963) 493 (August 8, 9), 503 (August 28), 535-536 (December 11), all dependent upon the early imperial Fasti- inscriptions. Comparable sources for the period between the last fasti (middle of the first century A.D.) and the Calendar of 354 are lacking. Cf. Hijmans (1996) 117 nn. 15.16 (with refs.). 26 Scullard (1981) 203; Stern (1953) 110; Salzman (1990) 149-153. 27 Salzman (1990) 150 (with refs.). 386 STEVEN HIJMANS temple dedicated to him.28 Given the location, it is at least conceivable, perhaps even likely, that some celebration of Sol was associated with or incorporated in this armilustrium. This may well have evolved into a festival dedicated primarily to Sol at some time between the extant fasti and the reign of Aurelian. The Julio-Claudian fasti, then, give direct evidence for significant festivals of Sol on August 8,9,28, and December 11, and may give a con- text for a festival on October 19. Other sources indicate that the Sol- festival(s) of October 19 and 22, mentioned as such only in the Calendar of 354, were linked to or instituted by Aurelian as quadrennial agones. However, there is no evidence whatsoever that the 30 chariot races held on December 25 were instituted by Aurelian, far less that this celebra- tion in honour of Sol had older roots as a public festival. One must con- clude that in the early fourth century A.D. anyone surveying the festivi- ties in honour of Sol would identify the period from October 19 to October 22 as far more important than December 25, and the festival of August 28 as far older. It is true, of course, that being the winter sol- stice^—the natalis invicti—December 25 was the most logical of the sun's feasts to serve as birthday of Christ, if that was the way the "selection process" went. But this leads us to a different, important consideration. None of the traditional religious feast days of Sol were connected in any way with a specific astronomical date—such as one of the solstices or equinoxes—except the late feast of the winter solstice on December 25. Yet we know from the abundance of iconographic and other evidence that astronomical dates and phenomena such as the solstices and equi- noxes played an important role in Roman society from at least the first century B.C. onwards. One need but think of the numerous depictions of seasons, zodiacs, calendars and planetary gods representing the days of the week, already quite prominent in Pompeian art in the early first century A.D., and popular throughout the Roman imperial period. The parapegmata (peg-calendars) discovered throughout the empire fur- ther attest to the pervasive influence of such concepts as the seven-day week,29 with each day represented by one of the seven planets (Sol, Luna, Mars, Mercury, Jupiter, Venus, and Saturn).30 This evidence is suggestive, because it indicates that there was great interest in the as- tronomical, astrological and calendrical aspects of the sun, sol, as heav- 28 Degrassi (1963) 532-534- 29 On the Republican origin of the seven-day week in Rome, cf. Brind'amour (1983) 256-268 (with refs.). 30 Hijmans (1996) 142-144; LIMCs.v. Helios/Sol, passim; LIMCs.v. Planetae, passim. Cf. Meyboom (1978); Sadurska (1979); J. Riipke, A/eue Paulyq s.v. parapegma, 322; Gundel (1992). SOL INVICTUS 387 enly body, without these being central to the cult calendar of the Roman Sun God, Sol Invictus. This apparent dichotomy between astronomy and cult in the Roman polytheist attitude towards the sun deserves closer attention. It touches upon the central, indeed to my mind the fatal flaw in the contention that Christmas was instituted on December 25 to counteract a pagan feast. Much of this is connected with the traditional understanding of Roman polytheism in general that underlies the specific interpretations given to Sol Invictus and his cult. It is clear both from the language used and from the interpretation given to the nature and history of the cult of Sol in late antiquity that scholars have viewed it in essentially the same terms as Christianity. They present the cult as a henotheistic, almost monotheistic, religion which viewed Sol as the supreme deity and strove hard to attain uni- versal acceptance within the Roman Empire. After Heliogabalus' initial failure to do so, solar religion was succesfully introduced from the ori- ent under Aurelian, resulting in the acceptance of Sol as the most im- portant Roman god for the next half century until the reign of Constantine. Under the impetus of Christianity the cult subsequently rapidly disappeared, although it enjoyed a short revival under Julian. In every respect this picture is very unrealistic. To begin with, there is the sociologically problematic notion of a god, with a suspect past (under Elagabalus, 218-222), being "established" at a stroke of the pen as supreme deity by one emperor (Aurelian, 270-275). Could Sol really have been accepted by "the Empire" (whatever that may mean) to such an extent that his cult became sufficiently widespread and firmly rooted to be a serious competitor for Christianity? After all, this cult is said to have retained its position of dominance for a mere fifty years (until Constantine, 313-337), only to disappear again as rapidly as it surfaced! To my mind this is as blatant an example as any I know of the subordi- nation of social reality and socially realistic potentiality to an abstract academic juggling of perceived "facts" or "data." As an analysis of the role and influence of solar cult, this approach lacks even the most basic respect for internal logic and cohesion. Think about it: there is no solar cult in Rome, we are told, until Aurelian reinstates a previously dis- credited cult in A.D. 274. Roman polytheism is by then so fickle3' that Romans immediately allow this oriental god to supplant older, Roman deities such as Jupiter Optimus Maximus as the "most important" (whatever that may mean in a polytheistic context) god of the empire. Yet those same fickle Romans who in 274 readily downgraded the old 31 Halsberghe ([1972] 42) calls Roman polytheism of the second century A.D. an "undermined and weakened body. " 388 STEVEN HIJMANS and hallowed Roman cults in favor of a new and oriental one, in the 330s so tenaciously maintain their ("old"? "new"?) solar festivals that church leaders felt obliged to neutralize the threat of those ("en- trenched"?) practices by imposing Christian feast days on the pagan ones. This will not work. It is troubling enough that such an unrealistic understanding of Roman Solar cult could be developed in total disre- gard of the evidence for Sol's longstanding but minor role in Roman paganism, discussed above. But let me stress that I am not, at this point, taking issue with the perceived facts, which I have already refuted ear- lier, but with the lack of any realistic underlying understanding of the nature of Roman polytheism in support of this interpretation of any such facts. It is the very understanding of Roman polytheism which is at issue here, and not merely the actual data concerning the chronology and precise nature of the cult of Sol Invictus. If the misinterpretation of the available data obliges us to reexamine the nature and chronology of Roman solar cult in antiquity, the flawed and unrealistic notion of the nature of Roman polytheism which underlies the discredited paradigm should alert us to the fact that we must question even the most funda- mental assumptions concerning the nature and religious role of this god and heavenly body in Roman religion. Of course, the suggestion is no longer accepted that paganism was somehow so moribund in late antiquity that the radical introduction of new cults was relatively straightforward.32 But to state that paganism was vigorous in the third century A.D. and beyond still does little to define more clearly what it is that we mean by that word, first coined by Christians not to define but to demean the traditional Roman relig- ions) which they opposed. It is actually extremely difficult, if not im- possible, to define Roman "religion" because, as Beard et al. have put it ([1998] x): one thing does seem clear enough: that many of our familiar categories for thinking about religion and religious experience simply cannot be applied here. ... Even the idea of "personal belief" (to us, a self-evident part of religious experience) provides a strikingly inappropriate model for understanding the religious experience of ... Rome. This is particularly important when we face the problem of analyzing 33 Lane Fox (1986); Bowersock (1990); Fowden (1993) 37-60; Beard era/. (1998) 364-388. 33 Beard et al. are referring in particular to early Roman religion, but given that many hold that "belief" did not become a central focus of religion until the Reformation or the Enlightenment, I feel justified in quoting this caveaf in rela- tion to late antique paganism as well. Cf. Rothaus (2000) 4 (with refs.). SOLINVICTUS 389 the interaction and competition between "paganism" and "Christian- ity," as in the case of Christmas. Rothaus ([2000] 4-7) has succinctly and clearly summarized the issues which lie at the root of the problem. In the traditional dichotomy pagan-Christian, the people of late antiquity are essentially divided into a defined set (those who worship Christ) and a non-set (those who do not worship Christ). The latter are unde- fined, as they share no common characteristic and their "set" therefore has no boundary or defined content.34 Of course, the dichotomy pagan- Christian has long been recognized as false, but it still clearly influences much of our current understanding of paganism. In practice, what many scholars have done is to approach individual pagan cults as inter- changeable with paganism, treating those cults—singly or as a group—as the "opposites" of Christianity. In this way they provide con- tent to the otherwise indefinable set "pagans." Opposite "those who worship Christ" they set "those who worship Sol and/or Zeus and/or Isis etc." The Christocentricity of this approach is immediately appar- ent, and its effects clearly visible. As we discussed above, the Sol of the nineteenth-century oriental paradigm is implicitly understood as a god with a self-contained cult and institutions, rooted in the orient and transposed from there by believers (priests like Elagabalus, and other adherents) to Rome and other parts of the empire whose inhabi- tants—through force, decree, or proselytization—were subsequently "solarized" and came to revere this new god as their supreme deity within a "fluid" polytheist context. Conceptualized in that way, the dif- ference between the cult of Sol and Christianity lies in what they be- lieved, not in how they perceived (and celebrated or experienced) re- ligion. Such a Christocentric approach can only obscure, not elucidate, the nature of Roman polytheism. Many have argued that we must move away from a focus on its supposed teachings and beliefs (p. 388 above) to emphasize practice and ritual.35 That is to say: we must divest our understanding of Roman religion of its (early modern. Christianizing) restrictions if we wish to grasp the full scope and effect of polytheism within Roman society.36 As a god, Sol, in everyday Roman polytheist 34 The former, of course, are also undefined, as one can hardly speak of "those who worship Christ" as one clearly distinct group. Rothaus's point re- mains valid, however, not because the set actually existed, but because the set ("Christians") has been perceived to have existed—in clear opposition to "pa- gans"—by countless scholars of late antique polytheism and early Christianity. 35 Price (1984); Rothaus (2000) 4-7. 36 Cf. MacMullen (1981): Price (1984); Lane Fox (1986); Versnel (1990); Feeney (1998); Beard ef ai. (1998). 390 STEVEN HIJMANS consciousness, would then have been defined primarily by (public and private) practice and ritual (whether daily, annual or quadrennial)—the realm of religion, one might venture. Speculations concerning the na- ture of Sol, and his place and role in the cosmos of deities and human- kind, by contrast, were restricted to a tiny elite and deemed more the realm of philosophy, perhaps. It should come as no surprise that our written sources concerning Sol—in any event both scarce and late—are more or less restricted to the philosophical musings of the elite, but shed little if any light on the broader practice and ritual surrounding Sol.37 More informative are material sources—often directly reflecting ritual action—but they pose their own interpretative challenges. Nonetheless, in the following I will argue that material evidence for Sol and his cult presents a very differ- ent picture of the nature of this deity in Roman polytheism, especially if we abandon Christocentric preconceptions concerning polytheist cult and deity in general. I have already suggested that there was a certain duality or ambigu- ity in the nature of Sol in Roman religion. In many respects, the differ- ence between the Sol whose major feast is celebrated October 22 and the Sol who is the planetary symbol for Sunday, for instance, is the same as the difference, in modern editions of Latin texts, between Sol with a capital letter and sol without. On the one hand there is a deity, with a cult, temples, festivals and rituals, and on the other there is a heavenly body which rises in the morning in the east and sets in the evening in west on fixed times and according to a never-changing pattern. This ambiguity in the nature of Sol—and Luna—is not shared to an equal extent by any of the other Roman gods and has not been sufficiently ac- knowledged by scholars dealing with the origins of Christmas Day. In Roman art the manner in which the iconographic image of Sol is used clearly reflects and to some extent defines this ambiguity. On the one hand, there are images which depict Sol as a god among gods, ap- parently on an equal footing with his peers. There were presumably cult-statues in temples (although none have been preserved), on occa- sion statuettes in lararia (rare), depictions of Sol on coins (sometimes claimed to reflect cult statues), a small number of reliefs depict- ing—inter alia—a cult statue of Sol, etc. (fig. 1). However, indisputable representations of Sol as a god among gods were rare. Far more com- mon were representations of Sol as a cosmic and temporal symbol. On coins minted by Vespasian, Titus, and Domitian, we find Aeternitas standing and holding small busts of Sol and Luna in her hands.38 Parallel Fauth (1995)- LIMCs.v. Aeternitas, 2-4. SOL INVICTUS 391 Fig. 1. Two men and a child bringing a sacrifice on an altar in front of a statue of Sol. Despite the destruction of the head, Sol is readily recognizable because of the characteristic iconography (naked but for a chlamys, right hand raised, traces of a radiate nimbus immediately above his shoulders) which is standard for him. Reign of M. Aureiius or Caracalla. Naples, Mus. Naz. 6678. Photo: author. issues depict Aeternitas holding a Phoenix. These coins can serve as an introduction to the most common type of representation of Sol, namely as a symbol or attribute, together with Luna, in a wide range of con- texts. Sol and Luna are usually depicted as small busts flanking or framing some scene, but sometimes the pair is represented as full- length figures (very rare) or charioteers of their quadriga and biga re- spectively (figs. 2-4). That they do not participate in the main scene but Fig. 2. Dolichenus triangle, fragment, with busts of Sol (right, nine rays, whip) and Luna (left) below a bust of Jupiter Dolichenus. From Heddemheim, Wies- baden, Mus. Inv. 6776. Photo: Mus. Wiesbaden. SOL INV1CTUS 393 Fig. 3. Mithras slaying the bull. Upper corners: Sol in quadriga (left) and Luna in biga (right). Relief, sandstone, from Osterburken; Karlsruhe, Bad. Landes- museum C 118. Photograph: author, of the copy in the Umesmuseum at Aalen. Fig. 4. Statue of Aphrodite of Aphrodisias, detail: mutilated busts of Sol (left) and Luna (right). Aphrodisias, Museum, second century A.D. Photo: author. 394 STEVEN HIJMANS have a symbolic function in relation to the scene to which they are ap- pended is invariably clear from their position (peripheral), relative size (small), iconography (bust), or a combination of these factors. As the coins of Vespasian indicate, Sol and Luna in this symbolic role are attributes defining or guaranteeing eternity. They are, of course, the two most visible cosmic bodies whose appearances follow clear, immu- table patterns, but as the parallel use of the phoenix as symbol of Aeter- nitas implies, the symbolism goes beyond that. The sun and the moon are in a constant flux—the moon either waxing towards or waning from full moon, and the sun never rising and setting at the same time two days running; from the winter solstice to the summer solstice the days lengthen and the nights grow shorter, and then for the next six months the pattern is reversed. Sol and Luna, then, not only symbolize Eternity but also explain how cosmic stability can be sustained despite temporal fluctuations. This symbolic use of Sol and Luna was common in a wide range of religious contexts in the Roman Empire: we find it on Jupiter Doli- chenus triangles, Mithraic reliefs, Danubian rider plaques, the epen- dytes of Aphrodite of Aphrodisias, Jupiter columns, etc.39 Nor was this symbolic use limited to Sol and Luna as a pair. Comparable Danubian rider plaques depict Sol alone, and there are other examples of Sol alone as a symbol as well. The temporal and cosmic symbolism of Sol in such cases is exceptionally clear on a Mithraic bronze relief from Brigetio.4" Framing the main scene of the tauroctony we have Sol and Luna, the four seasons, and the seven planetary gods of the week, all clear cosmic symbols providing a framework for the main scene, rather than any actors in it. In all these images Sol is clearly not depicted as a god to be celebrated with feast days and chariot-races, but as a cosmic body whose presence contributes to the framing and defining of the nature of the main scene. That this was the case should not surprise us. The cosmic reality and astronomical predictability of Sol was and is incontestable:—not a mat- ter of religion but a matter of fact. It was to those incontestable aspects of the sun and moon to which Sol and Luna or Sol alone refer, directly as well as metaphorically, in the contexts we have just discussed. It is also those aspects of Sol which are important in, for instance, the mosaic floors of the late antique synagogues of Hammath Tiberias, Beth Alpha, and elsewhere. In the large, central panel of the floor of the main nave we find Sol in the traditional Roman iconography on a frontal chariot within a zodiac and with the four seasons in the corners. Here too Sol is 39 LIMC s.v. Helios/Sol, passim. 40 LMCs.v. Helios/Sol. 288. SOL IN VICTUS 395 the anthropomorphic representation of the visible "planet," and as such is a cosmic symbol rather than a polytheistic god, further emphasized by the zodiac around him and the four seasons accompanying him. This is the most plausible interpretation—given the problems which would arise if we were to treat these as images of the Roman polytheistic sun god, forcing us to explain why he dominated the floors of no fewer than four synagogues—and any lingering doubts about that interpretation were removed by the recent discovery of yet another mosaic floor of this type, in which the head of Sol was replaced by a radiant disc.4' The adoption of this formerly polytheist symbolism is not limited to Judaism. In early Christian art we find the same use of the busts of the sun and the moon, flanking crucifixion scenes as well as other major events in the life of Christ, such as his baptism.42 In short, the most pro- lific imagery related to Sol (and Luna) treats the sun not as a god but as symbol, and despite the polytheistic religious source of the anthropo- morph iconography employed, the actual cosmic-symbolic connotations of this imagery were so well understood that it could be employed without significant variations by pagans, Jews, and Christians alike. Does this help us account for the choice of December 25 as Christmas day? I think it does. It shows that the sun played an ambiguous role in the Roman world both as a cosmic body and as a god. Christians could deal with Sol. whose cosmic nature and reality were undeniable and whose potential for cosmic symbolism was inspiring, without necessar- ily dealing with the pagan god Sol Invictus. The homilies dealing with this topic support this view, for while they routinely refer to the winter solstice and the like, they rarely mention (or attack) pagan feasts on De- cember 25. In fact, one of the earliest such homilies—if, that is, it has been dated correctly to the early fourth century A.D.—is the De solsti- tiis. quoted by Usener, and it makes quite clear that we should see the date of December 25 not in isolation but as part of a cosmic-symbolic system. The author argues that the birthday of John the Baptist was ex- actly six months before that of Jesus, so that he was born on the summer solstice. This means that John was conceived in the autumn—on the equinox to be exact—while Jesus was conceived on the vernal equinox. 41 The anthropomorphic Sol was depicted in the synagogues of Hammath Tiberias (ca. A.D. 300), Husaifah (fifth century), Beth Alpha (sixth century) and Na'aran (sixth century); cf. LIMC s.v. Helios/Sol 292. For the recently discov- ered mosaic at Sepphoris, where the anthropomorphic Sol has been replaced by a disc but, strangely, the quadriga of Sol has been maintained, see Weiss and Netzer (1996). esp. 26-29. 42 Compare an intaglio in the British Museum, depicting Christ on the cross, with the small busts of Sol and Luna above (inv. EG 561). On this symbolism in early Christian art, see Deonna (1946) and (1947), with numerous examples. 396 STEVEN HIJMANS Having built up and argued this cosmic symbolic system throughout his homily, the author at the very end, in the aside referred to at the begin- ning of this article, deals with the problem that the winter solstice also happens to be a feast day of Sol Invictus. That Christ was born on such a feast day can be no problem, he argues, as Christ is the true "sun of jus- tice" anyway. In fact, apart from the scholiast to Bar Salibi (p. 379 above), we find no indication that the feast day of Sol Invictus played any role in the choice of December 25, while we do find countless refer- ences to the winter solstice and cosmic considerations in numerous homilies by many church leaders (Heim [1999] passim). So let us return to the scholiast. Closer consideration shows that Usener's quotation of the passage is misleading. Not only does Usener translate "birthday of the sun god" (polytheistic religious) where the Syriac merely states "feast day of the rising of the sun" (cosmic sym- bolic), more importantly he presents the scholiast's passage as an objec- tive statement of fact even though it is actually a hostile one. For imme- diately preceding the section quoted by Usener, the scholiast states: The Lord was born in the month of January, on the day on which we celebrate Epiphany; for the andents celebrated the birth and epiphany on the same day, because He was born and baptized on the same day. Likewise the Armenians celebrate both feasts on the same day. One should also mention here the Doctors of the Church who speak of both feasts together. Only then does the scholiast continue with the words quoted by Usener: "The reason why the Church fathers transposed this feast from January 6 to December 25 is, it is said, the following ..." (Usener omits the "it is said"). Clearly, the scholiast is orthodox, and he rejects the date of De- cember 25. One such late, and obviously hostile, source has little value in this case. I would argue that the evidence is clear. In the Roman Empire, the nature of Sol was ambiguous. As a heavenly body, the sun was often used—together with Luna—as a cosmic symbol or allegory for eternity. The astronomical reality of the sun and the moon precluded such sym- bolism from being exclusively pagan, and the evidence of the De solsti- tiis, as well as the numerous passages from a wide range of homilies collected by Heim (1999), suggests that it was readily adopted by Chris- tians. It is cosmic symbolism of this type which inspired the Church leadership in Rome to elect the winter solstice, December 25, as the birthday of Christ, and the summer solstice, June 25, as that of John the Baptist, supplemented by the equinoxes as their respective dates of con- ception. While they were aware that pagans celebrated a festival in honour of Sol Invictus on that day, this did not concern them, and it cer- SOL INVICTUS tainly did not play any role in their choice of date for Christmas. Department of History and Classics University of Alberta Edmonton, AB T6G2H4 References Beard. M.. North. J., and Price, S.R.F. 1998. Religions of Rome I. A History. Cambridge. Bergmann, M. 1998. Die Strahlen der Herrscher. Theomorphes Herrscherbild und politische Symbolik im Hellenismus und in der romischen Kaiserzeit. Mainz. Boll, F. 1910. Griechische Kalender 1. Das Kalendarium des Antiochos. Sitzungsberichte der Heidelberger Akademie der Wissenschaften, Philoso- phisch-historische Klasse, 16. Botte. B. 1932. Les origines de la noel et de Vepiphanie: latude historique. Lou- vain. Brind'Amour, P. 1983. he calendrier romain: Recherches chronologiques. Col- lection d'etudes anciennes de l'Universit6 d'Ottawa 2. Ottawa. Crawford. M.H. 1974. Roman Republican Coinage, 2 vols. Cambridge. Cullmann. O. 1956. "The origin of Christmas." in A.J.B. Higgins. ed. The Early Church. London. Deonna, W. 1947. "Les crucifix de la vallee de Saas (Valais): Sol et Luna, histoire d'un theme iconographique (premier article)," Revue de 1'Histoire des Re- ligions 132:5-47. _. 1948. "Les crucifix de la vallee de Saas (Valais): Sol et Luna, histoire d'un theme iconographique (deuxieme article)," Revue de l'Histoire des Religions I33-49-i02. EEC = E. Ferguson, ed. Encyclopedia of Early Christianity. New York/London 1990. Fowden, G. 1993. Empire to Commonwealth: Consequences of Monotheism in Late Antiquity. Princeton. Gobi, R. 1978. Antike Numismatik, 2 vols. Munich. Gruber, H.A. 1910. Coins of the Roman Republic in the British Museum, 3 vols. London. Gundel, H.-G. 1992. Zodiakos. Mainz. Halsberghe, G. 1972. The Cult of Sol Invictus. Etudes preliminaries de Religions Orientales 23. Leiden. Heim, F. 1999. "Solstice d'hiver, solstice d'ete dans la predication chretienne du Ve siecle. Le dialogue des evfiques avec le paganisme, de Zenon de Verone a saint Leon," Latomus 58: 640-660. Hijmans, S.E. 1989. Sol Invictus. Een iconografische studie. Diss. Groningen. _. 1996. "The Sun which did not rise in the East: The cult of Sol Invictus in the light of non-literary evidence," Bulletin AntiekeBeschaving 7v. 115-150. _. 2000. "Language, metaphor, and the semiotics of Roman art," Bulletin 398 STEVEN HI)MANS Antieke Beschaving 75:151-168. _. forthcoming. "Reality and symbol: The radiate crown of Roman emper- ors." Acta of the XVJth International Congress of Classical Archaeology. Boston 2003. Latte, K. i960. RomischeReligionsgeschichte. Munich. LIMC= Lexicon Iconographicum Mythologiae Classicae. Zurich 1981-. L'Orange. H.-P. 1935. "Sol Invictus Imperator. Ein Beitrag zur Apotheose," SO 14:86-114. MacMullen, R. 1981. Paganism in the Roman Empire. New Haven/London. Meyboom. P.G.P. 1978. "Un monument 6nigmatique 'Dusari sacrum' a Pouz- zoles." in Hommages a Maarten J. Vermaseren. Leiden: 782-790. Price, S.R.F. 1984. Rituals and Power: The Roman Imperial Cult in Asia Minor. Cambridge. RAC= Reallexicon fiir Antike und Christentum. Sachworterbuch zur Ausein- andersetzung des Christentums mit der antiken Welt Leipzig 1950-. Rothaus, R.M. 2000. Corinth: The First City of Greece. An Urban History of Late Antique Cult and Religion. Leiden. Sadurska. A. 1979. "Rzymskie Kalendarze manipulowane oraz ich uwarunkow- anie historyczne." Archeologia 30:69-86. Salzman, M.R. 1990. On Roman Time: The Codex-calendar of 354 and the Rhythms of Urban Life in Late Antiquity. Berkeley. Scullard. H.H. 1981. Festivals and Ceremonies of the Roman Republic. London. Stern, H. 1953. Le calendrier de 354. Etude sur son texte et ses illustrations. Paris. Thomson, R. 1961. Early Roman Coinage, vols. 2 and 3. Copenhagen. Usener, H. 1889. Das Weihnachtsfest. Bonn. _. 1905. "Sol Invictus." RhM6o: 465-491. _. 1911. ReligionsgeschichtUche Untersuchungen. Bonn. Wallraff. M. 2001. Christus Verus Sol. Sonnenverehrung und Christentum in der Spatantike. Jahrbuch fiir Antike und Christentum. Erganzungsband 32. Minister. Weinstock, S. 1948. "A new Greek calendar and festivals of the Sun." JRS 38: 37-42. _. 1957. "Victor and Invictus," HTR 50:211-247. Weiss. Z. and Netzer, E. 1996. Promise and Redemption: A Synagogue Mosaic from Sepphoris. Jerusalem.