The discovery of the Reliquary of Saint Trophimos at Chouhout Kassaba, the site of ancient Synnada in Phrygia, in 1907 provided an important piece of archaeological and epigraphic evidence for the study of early Christianity in Asia Minor. The small marble reliquary, fashioned in the form of an architectural sarcophagus, was discovered together with human cranial bones. Its subsequent study and publication by Gustave Mendel in his 1909 Catalogue des monuments grecs, romains et byzantins du Musée Impérial Ottoman de Brousse opened an important discussion concerning the date of the monument, the antiquity of the cult of the martyrs, and the degree to which Christian communities in Asia Minor expressed their faith during the period of Roman persecution.
The discovery itself belongs to the archaeological history of the late Ottoman Empire. The reliquary was not uncovered by a Western archaeological expedition, but first came to official attention through Ottoman administrative and museum channels. According to the official report preserved by Mendel, the reliquary was discovered by a local inhabitant of Chouhout, Chaban-oghlou Hallil, while searching for stones in a field near the town. Ottoman authorities subsequently dispatched Achmet tchélebi effendi, president of the local commission of public instruction and conservator of the local branch of the Imperial Museum, to investigate the discovery together with the mudir. The objects were then sent through the Ottoman museum system and entered the collections of the Museum at Bursa, where Mendel examined and published the reliquary in 1909.
The monument is associated with Saint Trophimos of Antioch in Pisidia, who, according to the literary tradition, suffered martyrdom at Synnada under the Emperor Probus (276–282), together with the martyrs Sabbatios and Dorymedon. The memory of these martyrs is preserved in the Acta Sanctorum and other Christian sources, while their names also appear in the ancient Nicomedian calendar, which formed the basis of the Hieronymian Martyrology. The discovery at Synnada is therefore particularly significant because it places a material monument bearing the name of Trophimos alongside the literary evidence for his martyrdom and subsequent veneration.
At the center of Mendel's discussion is a fundamental historical question: how visible was Christianity before the Peace of the Church established by Emperor Constantine? Earlier approaches to early Christian archaeology often associated explicit Christian monuments, inscriptions, and the public cult of martyrs primarily with the period following Constantine. The Synnada reliquary presents a more complicated picture. Its inscription explicitly identifies Trophimos as a martyr and contains a warning invoking divine judgment against anyone who might remove his bones: “Whoever should ever cast out these bones shall answer before God.” The question is therefore not simply whether Christians were present in Phrygia before Constantine, but how openly they commemorated their martyrs and what forms such commemoration could take during periods of persecution.
Mendel examines the possibility of a very early date through a combination of archaeological, paleographic, and epigraphic evidence. Although he treats the question with considerable caution, several features of the monument lead him toward a date in the late third or early fourth century. Henri Grégoire, whose important supplementary note Mendel included at the end of his discussion, goes further, stating that he inclines to date the inscription specifically to the third century. Their arguments are based not upon a single feature, but upon the cumulative evidence of the reliquary's form, the character of its inscription, its terminology, and the wider evidence for Christian communities in Phrygia during the age of persecution.
Several features of the monument are particularly significant.
1. The Absence of Explicit Christian Symbols. The reliquary bears no cross, Chi-Rho monogram, or other overt Christian symbol. For Mendel, this absence is significant when considered alongside the terminology of the inscription. Rather than employing the more explicit visual language that became increasingly common in Christian monuments after Christianity acquired public recognition, the reliquary retains a comparatively restrained funerary form.
2. The Warning Invoking God. The inscription on the lid concludes with the formula ἔσται αὐτῷ πρὸς τὸν Θεόν, “shall answer before God.” Mendel observes that this type of formula occurs with particular frequency in the inscriptions of Phrygia and considers it compatible with an earlier period. Grégoire likewise regarded the formula as significant. If the reliquary were instead a later monument created for a publicly established Christian basilica, the use of such a restrained and almost “timid” malediction would be more difficult to explain.
3. The Possibility of Partial Relics. The dimensions of the reliquary are too small to have contained an entire human skull, much less a complete body. Mendel therefore considers it plausible that the Christians of Synnada had obtained only a portion of the martyr's remains and preserved them in the small stone chest. He connects this possibility with the documented efforts of Roman authorities during periods of persecution to prevent Christians from recovering and venerating the bodies of martyrs. At the same time, Mendel carefully acknowledges that the official report of the discovery does not explicitly state that the cranial bones were found inside the reliquary. The connection is highly plausible in his view, but it remains an inference rather than something that can be established with absolute certainty.
The broader historical context makes this possibility particularly important. Mendel points to evidence that the veneration of martyrs was already well developed during the third century and that Christians sometimes went to considerable lengths to obtain and preserve the remains of those who had suffered for the faith. Roman authorities, for their part, sometimes attempted to prevent such veneration by destroying, concealing, or otherwise disposing of the bodies of martyrs. In this context, the preservation of even a portion of Trophimos's remains in a small reliquary would represent not merely an act of commemoration, but the preservation of a physical connection between a persecuted Christian community and one of its witnesses.
The evidence must nevertheless be distinguished from the later literary tradition. The reliquary provides significant material evidence for the memory of Trophimos as a martyr and for the presence of a Christian community at Synnada capable of preserving and commemorating his remains. It does not, by itself, establish every detail contained in the later Acts of Trophimos, Sabbatios, and Dorymedon. Nor can the surviving archaeological evidence determine with absolute certainty whether the cranial bones mentioned in the Ottoman discovery report were actually contained within the reliquary. What it does provide is a rare and tangible point of contact between the literary tradition and the archaeological record of early Christianity in Phrygia.
The later history of the monument adds another dimension to its importance. Its movement from the Ottoman Museum at Bursa to the Museum at Istanbul, followed by renewed scholarly examination in the twentieth century, demonstrates that the reliquary continued to attract attention long after Mendel's original publication. The 1934 Monumenta Asiae Minoris Antiqua expedition is particularly significant because its scholars encountered the object in Istanbul at a time when its authenticity had itself become a subject of discussion. Calder's defense of the monument shows that the Synnada reliquary was not simply accepted uncritically as an ancient Christian object, but continued to be evaluated against the growing body of archaeological and epigraphic evidence from Phrygia.
What follows is Gustave Mendel's detailed examination of this remarkable monument, together with the supplementary observations of Henri Grégoire. His study is valuable not only for its description of the reliquary and its inscription, but also for the window it provides into the methods by which early Christian archaeological evidence was interpreted at the beginning of the twentieth century. The monument itself, modest in appearance yet remarkable in historical implications, offers a glimpse into the ways in which Christians of Asia Minor preserved the memory of their martyrs through relics, funerary inscriptions, and sacred monuments during a period when the Church had not yet attained a publicly recognized position within the Roman Empire.
The Reliquary of Saint Trophimos
Musée de Brousse — Christian and Byzantine Monuments
102. (548) Reliquary of Saint Trophimos
Chouhout (Techifout)-Kassaba.
Given the importance and character of this monument, we reproduce here the official letter announcing its discovery and its dispatch to the Directorate of Public Education of the vilayet.
“Letter No. 221. From His Excellency the mutessarif of Kara-Hissar, dated May 30, 1323 (June 13, 1907). — The named Chaban-oghlou Halil, a resident of the Kalyn-Toungar quarter of the town of Chouhout, while searching for stones for washing in the field of Hadji Alaeddin, situated a quarter of an hour from the nahié of Chouhout, discovered an ancient casket with a lid and inscription, together with cranial bones. These objects were sent to us by the assistant to the mudir of Chouhout. Thereupon, we delegated, for the purpose of an investigation, Achmet tchelébi Effendi, president of the Public Education Commission and curator of the local branch of the Imperial Museum, who went to the site accompanied by the mudir. According to their report, they found two coins in the house of Hadji Halil and two others in that of Hadji Alaeddin. Nothing else was found. We are sending you these objects by post, etc., etc…”**
The marble is white; all four faces are worked in the same manner. There are several recent erosions on the lid, with traces of pick blows; the lid was sealed to the chest by two iron clamps, 0.07 m long, which must have been broken at the time of the discovery, although one of them was preserved and replaced in its mortises, as can be seen in the profile view. Total height, 0.25 m; height of the chest, 0.155 m; width at the lower molding, 0.20 m; at the upper molding, 0.195 m; at the inscription, 0.165 m; thickness at the lower molding, 0.195 m; at the upper molding, 0.175 m; width across the chest, 0.155 m; interior length of the chest, 0.135 m; interior width and depth, 0.11 m. The letters of the inscription are engraved between two incised lines, 0.014 m apart on the chest and 0.012 m apart on the lid; but the engraving is not very regular and is not the work of a highly practiced hand, and the letters sometimes fall below and sometimes extend beyond the alignment.¹
 |
| The Reliquary of Saint Trophimos, discovered at Chouhout (ancient Synnada) in 1907. Photographs reproduced from Gustave Mendel, Catalogue des monuments grecs, romains et byzantins du Musée Impérial Ottoman de Brousse (1909), figs. 46–47. |
The reliquary reproduces very exactly the form of a certain class of sarcophagi that were very widespread throughout Asia Minor; only the proportions have been modified, since the chest approaches a square and the whole is taller than it is wide. The chest, without any ornamentation, is enclosed between two thick moldings with a straight profile; the edge of the inner cavity is doubled by a small rounded molding that ensures the stability of the lid. The latter is in the form of a two-pitched roof, ornamented at the corners with quarter-sphere acroteria and, at the ends of the upper ridge, with hemispherical acroteria. As in the large sarcophagi, the lower face of the rim of the lid is cut in a bevel, except in the middle, where a “witness” has been left, which increases the bearing surface and protects the edge of the bevel when the lid is placed upon the chest. The underside of the lid is slightly hollowed; the lateral faces present a recessed tympanum, ornamented with a small circular medallion—missing on the left and mutilated on the right. The entire workmanship is executed with great care; in several places one can still see the incised lines that guided the mason’s chisel, but the whole is simple and heavy, without elegance. The inscription is engraved (1) on the chest; (2) on the lid.
Inscription on the Chest
“Ὠδε ἕνα Τροφίμου τοῦ μάρτυρος ὀστέα.”
“Here are the bones of the martyr Trophimos.”
Inscription on the Lid
“Τίς ἂν δὲ ταῦτα τὰ ὀστέα ἐκβάλη ποτέ ἔσται αὐτῷ πρὸς τὸν Θεόν.”
“But whoever should ever cast out these bones shall answer before God.”
In the penultimate line of inscription (2), the pick has spared only the Τ of the article τόν, but there is not enough room for the lapidary to engrave ON; it is probable that the Ν was omitted. The inscription presents no difficulty except in the form ἕνα, whose meaning is not doubtful, although I do not know how to explain it. Is it simply an error committed by the engraver who, having written the letters ἐν, thought by association of the word ἕνα? Is it a local form of ἐν? The grammarians must decide. On τίς—ὅστις, frequent in the inscriptions of Phrygia, cf. Dieterich, Untersuchungen, p. 200.
The Trophimos whose relics were contained in this small sarcophagus is the Trophimos of Antioch in Pisidia who suffered martyrdom at Synnada¹ under the Emperor Probus (276–282); cf. Acta ss. Trophimi, Sabbazii, Dorymedonis, in the Acta SS., September, VI, pp. 12–20; PG, ed. Migne, CXV, pp. 733–749; Mercati, Studi e testi, 5, Note di letteratura biblica e cristiana antica, XV, pp. 206–226, Un’apologia antica sotto forma di martirio; Harnack, Geschichte der altchristlichen Litteratur, II, 2, p. 481, note. The authenticity of the Acts, defended by Allard (Hist. des perséc., III, p. 279, 4) and Aubé (L’Église et l’État dans la deuxième moitié du IIIe siècle, p. 52 sq.), had been contested by Fr. Goerres (Jahrb. für protest. Theologie, XVI, 1890, p. 616 sq.). Mgr. Duchesne, to whom I had communicated photographs of the reliquary, was kind enough to reply to me on this subject:
“It is not possible to doubt the existence and the cult of the three martyrs of Synnada. However artificial their legend may be, their names appear in the Nicomedian calendar of the fourth century, which forms the basis of the Hieronymian Martyrology.”
The discovery at Synnada has thus come to confirm in a definitive manner the opinion of the learned historian.
As for the date, the same scholar declares himself inclined toward a date later than the fourth century. I feel my own incompetence too keenly to venture an opinion in the face of such high authority; I ask only permission to present a few observations of an archaeological and epigraphic nature.
From the archaeological point of view, the form of the reliquary can provide no indication; it is found as early as the Hellenistic period; it is extremely frequent in the Roman period, and it is still the form of the imperial sarcophagi preserved today near Saint Irene.
If this small monument really dates from the end of the fourth or the fifth century, is there not reason to be surprised that it bears no external sign of Christianity, neither a cross nor a monogram? On the tombstone of Saints Perpetua and Felicity, recently discovered at Carthage, each line of the epitaph is preceded by a cross, and Father Delattre attributes it to the beginning of the fourth century, approximately a hundred years after the martyrdom, which occurred in the year 203 (cf. C.-R. Acad. Inscr., 1907, pp. 176 and 193–195). Less significant perhaps, but nevertheless worthy of note, is the use of the word μάρτυς and the absence of the word ἅγιος, which seems to indicate a distinction between the two terms that would no longer be explicable in the fifth century; at that time one would rather expect the formula τοῦ ἁγίου μάρτυρος, which occurs, for example, in the dedication of the Martyrium of Saint Christopher, erected in 450–452 (BCH, II, p. 289). On the meaning of the two words, cf. Hauck, Realencykl. f. prot. Theol. und Kirche, vol. I, s.v. Acta martyrum et acta sanctorum; on the use of the word ἅγιος, Harnack, Mission und Ausbreitung², I, pp. 339–340.
Taken as a whole, the inscription clearly presents the “veiled type” of the South Phrygian inscriptions of the third and beginning of the fourth century. The formula ἔσται αὐτῷ πρὸς τὸν Θεόν, localized almost exclusively in this region, seems to belong properly to the third century, and only one example is cited that might be from the fourth (cf. Ramsay, CB, II, p. 514 sq.; Cumont, MAH, XV, 1895, p. 252). A conspectus of the inscriptions in which it occurs is given in Monumenta ecclesiae liturgica, ed. and under the direction of F. Cabrol and H. Leclercq, I, Reliq. liturg., sect. prima, Reliq. epigr., p. 12, no. 2798. It was not yet known in Synnada, but it occurs in Antioch of Pisidia, the homeland of Trophimos.
The characters of the inscription would not contradict this date. It does not contain the form Ϛ, which is found from the middle of the third century (SBB, 1881, p. 448), and even exceptionally at the end of the first (BCH, VII, 1883, p. 453, l. 14, right-hand column; inscription of Sebaste in Phrygia); on coins, it is already found on a coin of Caracalla (Rev. numismatique, 1883, pl. IX, no. 3). Nor, on the other hand, does it contain the form Α that appears only around the middle of the fourth century. The Ρ has the classical form, without elongation of the vertical stroke. The only objection would be the ligature by which the group στ is rendered in line 4 of inscription (2): it already resembles the Byzantine form, but differs from it in that each of its two elements retains its individual character, and one can affirm that it is even older than the earliest paleographic form in which the curve of the σ continues directly into the horizontal bar of the τ (cf. Gardthausen, Gr. palaeogr., pp. 207 and 265). It may be compared with ligatures such as ϟ = τσ, which are found in pagan inscriptions of the third or even the second century (e.g. AM, XXIX, 1904, p. 303; BCH, XXV, 1901, p. 11, no. 147).
In any event, it would perhaps be hazardous to attach too much importance to this isolated sign, considering that as early as the second century there are ligatures that are genuine monograms (e.g. AM, L. l., p. 313, fig. 34). The profile of the molding indicates, with considerable probability, this period or a nearby one, and letters placed one within another are found in the inscription of Museum No. 423, whose beautiful, deep, and regular engraving makes it possible to attribute it to the Antonine period (BCH, L. l., p. 18, no. 148 bis). The monogram is reproduced typographically in a lamentable manner, but I had made a good copy of it, and the fragment is certainly from the same period and probably from the same hand as No. 147, bearing the name of the legate C. Julius Scapula. The inscription gives me, it seems, the impression of a date earlier than CIG 8607, which dates from around 330 (cf. Gardthausen, Gr. Palaeogr., pl. I, col. 1), and it presents, if not in the engraving, at least in the form of the letters, certain analogies with the series of inscriptions from Altyn-tach, which are generally placed at the end of the third or beginning of the fourth century, particularly No. 49. On examining it closely, one sees that it is engraved carefully, by a hand that is diligent but lacks experience; one could reasonably suppose that it is the work, not of a professional, but of a member of the Christian brotherhood of Synnada, who executed it from a reasonably accurate copy—with no iotacism—and approximately reproduced its paleographic form at this point.
It may be recalled here, in order to forestall a fairly serious objection, that the cult of martyrs began as early as the time of the persecutions; by the third century it was highly developed; in the time of Diocletian it was already giving rise to certain abuses; the faithful spared neither money nor effort to secure possession of a corpse and, when the entire body was unavailable, a part of the body or at least an object that had belonged to the Saint. This is demonstrated by the very measures taken by the authorities: the proconsul of Smyrna forbade the Christians to recover the remains of Polycarp; at Lyon, the bodies of the martyrs were burned and their ashes thrown into the river so that no relics might be made of them; the same things occurred during the persecution of Diocletian.¹ If these facts are admitted—and, as far as I can judge, the texts leave little room for doubt—one may likewise admit that the faithful of Synnada, having recovered under these circumstances a portion of the skull of Trophimos, enclosed it in this reliquary.²
M. H. Grégoire has kindly communicated to me the interesting note which he authorizes me to publish here and which very aptly completes what precedes; may he kindly accept once again my affectionate thanks.
“I saw, at the Musée de Brousse, the curious monument of Trophimos, and I am inclined to believe that the inscription is indeed from the third century. Formerly, one would undoubtedly have thought that an epigraphic text attesting to the cult of a martyr must necessarily postdate the peace of the Church; but the inscriptions recently published prove that in Phrygia, above all, certain populations, as early as the third century, professed the Christian faith with extraordinary boldness; the stelae bearing the formula Χριστιανοὶ Χρηστιανοί (see above), of which M. Cumont knew only three examples in 1895, have multiplied; several bishops of the third century are named in the inscriptions; the Christian symbol par excellence, the simple cross, occurs before the chrism and the monogrammatic cross, on the fronts of the stelae (Studies, pp. 37, 84, 217); in Phrygia and Isauria, the persecutions preceding that of Diocletian scarcely interrupted the religious peace enjoyed by the Christian communities; (one must not forget, in this connection, that the persecution of Probus which affected Trophimos was the least extensive, the least bloody, and the shortest of all).
Under these circumstances, it would be astonishing if there were no epigraphic mention of martyrs; it would be astonishing if these Christians, who no longer concealed either their faith or their hierarchy, had always refrained from indicating on their tombstones their status as confessors of the faith. From these regions of Asia Minor, we possess at least two funerary inscriptions in which ἀθλοφόροι μάρτυρες appear.³ Eugène, Marcellus, Alexander, Macedonius, and Nonna are among them. This group of martyrs is named in inscription CIG 9266 (Hierapolis); M. Ramsay was the first to recognize the meaning, elsewhere very clear, of the words τὸ τῆς ζωῆς μέρος (cf. CB, II, p. 730); he has given excellent reasons for believing that these five confessors suffered under Decius; apart from the expression ἠσυχάζειν τὸ τῆς ζωῆς μέρος, the wording of the inscription scarcely departs from the usual funerary formula: ἔστησα τὴν εἰσ[ήτην] μνήμην Εὐγένει καὶ Μαρκέλλῳ καὶ Ἀλεξάνδρῳ καὶ Μακεδονίῳ καὶ Νόννῃ τοῖς γλυκυτάτοις τέκνοις … ὅς ἂν δὲ προσεψῇ … ἄρρωστα τέκνα ζήσωσι.”
Here now is a text in which the name martyr is written out in full:
“Νούμμος καὶ Οὐαλέριος ἐχώσαμεν Παῦλον τὸν μάρτυραν μνήμης χάριν.”
(BCH, X, 1886, p. 509; Studies, p. 62; cf. Analecta Bollandiana, XXVI, 1907, p. 464.)
This inscription is from Derbé in Lycaonia. Regarding the verb ἐχώσαμεν, Miss Margaret Ramsay, the latest editor, says: “The formula with the verb ἐχώσαμεν is common in the epitaphs of Lycaonia about the third and fourth centuries, and it is natural to infer that this stone was, like the rest of the Dorla monuments, a gravestone and not a mere honorary inscription.” In fact, it is not this banal formula, which has nothing specifically Christian about it, that one would have engraved on the tomb of a martyr after the peace of the Church.
Likewise, from the phrase τίς ἂν ἐκβάλῃ …, e.g. in the inscription of Synnada, I believe one must conclude that the monument is earlier than the end of the persecutions. If, in fact, the small sarcophagus was, in the ritual sense of the word, a reliquary made specifically for the basilica of the Saint, after the institution of a public cult in his honor, the hypothesis of the violation of the bones would become almost monstrous, and the old “crypto-Christian” formula, ἔσται αὐτῷ πρὸς τὸν Θεόν, would appear, in the face of such an offense, a very timid curse. There is, it is true, a difference between the monuments of Hierapolis and Derbé and that of Synnada: whereas those are stelae or tombstones, this one is a small casket containing only a portion of the bones of the deceased.⁴ This is undoubtedly because the Christian community of Synnada did not obtain the entire body of the martyr from the authorities. “Rome did not war against the dead; and the remains of the martyrs were allowed to be buried by their friends”: this statement by M. Ramsay (CB, II, p. 730) is too absolute, even for the earliest persecutions, as M. Mendel has shown.⁵ — On the antiquity of the Church of Synnada, attested since 218, cf. Harnack, Mission und Ausbreitung², II, pp. 181, 182, 185.
Endnotes1. Cf. nos. 47, 48, 49.
2. This letter does not mention fragment No. 27, which is said, moreover, to have been found at the same place.
3. On the location of Synnada at Chouhout-Cassaba, cf. Perrot, RA, 1876, I, p. 190 sq.; Ramsay, AM, VII, p. 132; Koerte, ibid., XXII, pp. 5 and 28. — M. Ramsay writes (CB, II, p. 716): “In the Synnada district we find no Christian inscription before Constantine…”; and it is clear that an assertion of this kind, accurate at the moment when it was formulated but based on data necessarily incomplete, is always subject to revision and cannot be taken into consideration in the discussion that follows.
4. I quote here almost verbatim Lucius, Die Anfänge des Heiligenkults in der christlichen Kirche, p. 72 sq., where the references will be found; cf. also Linsenmayer, Die Bekämpfung des Christentums durch den römischen Staat, ch. IV, especially p. 47 sq.
5. The dimensions of the chest do not permit us to suppose that an entire skull could have been deposited in it; one must observe, moreover, that the terms of the report do not formally state that the “cranial bones” were found in the reliquary, although, strictly speaking, this hypothesis is all the more plausible because no other bones are mentioned, nor the tomb in which they would have been enclosed.
6. I leave aside the epitaph of Bishop Akylas (Zemné), Studies, pp. 125–126 and 201, because I am by no means certain that this bishop actually suffered martyrdom, as MM. Ramsay and Anderson believe. — Cf. a very ancient inscription in the Christian Museum of Athens: Pargoire, Viz. Vrem., X, p. 636: Μαρτύριον Ἰωάννου, Λουκᾶ, Ἀνδρέου, Λεωνίδα [καὶ …] μαρτυρησάντων πρὸ πέντε καλανδῶν Ἰανουαρίου (Keserli, Thessaly). These martyrs, like those of Hierapolis and Paul of Derbé, are unknown to the martyrologies and synaxaria.
7. The same verb has a very different meaning in the fifth-century Paphlagonian inscription published by M. E. Legrand, BCH, XXI, 1897, p. 101: Ἀγαλασμὸς μάρτυρος ἀθλοφόρου ὅλον κοσμήσατο μνηόν.
8. The inscription alone could not prove this, for it is a Jewish epitaph—there can be no question here either of a reliquary or of a martyr—and it reads: ἔνθα κεῖται ὀστᾶ τοῦ σώφρονος Παύλου διακόνου (CIG, 7270).
9. Neither the inscription of Hierapolis, nor that of Derbé, nor that of Synnada proves that the martyrs whom they mention were already then the object of a cult. But the stone of Derbé appears to have been placed later in the basilica of the Saint (cf. Analecta Bollandiana, XXVI, 1907, p. 464); and such a transfer is infinitely plausible in the case of the relics of Trophimos.
Source: Gustave Mendel, “Catalogue des monuments grecs, romains et byzantins du Musée Impérial Ottoman de Brousse,” Bulletin de Correspondance Hellénique, vol. 33 (1909), pp. 342–348. Translation from the French and Introduction provided by John Sanidopoulos.