Coinage of the Tūlūnids

Grabar, Oleg
Numismatic Notes and Monographs
American Numismatic Society
New York
Worldcat Works




Open access edition funded by the National Endowment for the Humanities/Andrew W. Mellon Foundation Humanities Open Book Program.


Table of Contents




Some years ago the coin collection of Yacoub Artin Pasha, acquired by Robert C. H. Brock and presented by him to the University Museum in Philadelphia (hereafter U.M.), was transferred on indefinite loan to the American Numismatic Society. Dr. G. C. Miles has already devoted a monograph to Fāṭimid coinage and an article to Umayyad and ‘Abbāsid dinars,1 both of which were based in large part on the University Museum collection. Furthermore he has included some of the more outstanding specimens in his Rare Islamic Coins. 2

This study of the coins of the Ṭūlūnid dynasty in Egypt was similarly originally based on the collection of the University Museum (35 coins, all gold), to which were added the 18 specimens of the American Numismatic Society. There is, although it is now almost eighty years old, a corpus of Ṭūlūnid coins made by E. T. Rogers in 1877, which included at least 15 coins which were to become part of the Pennsylvania collection.3 Rogers knew of 125 coins which he attributed to the Ṭūlūnids, and divided them into 58 types, a type being defined on the basis of mint, date, and other characteristics. A thorough search through publications since Rogers’ time has led me to count almost 600 Ṭūlūnid coins known in all, falling into 97 types based on differences in metal, date, mint, and inscription.4 There is then some justification in assuming that a new publication of existing Ṭūlūnid coins is not out of place, since it will bring together documents which are only too often scattered in dozens of periodicals. Moreover, Rogers’ publication is purely analytical, giving only bare epigraphical and metrologic information. In fact one of the essential aims of numismatic studies is to provide the historian with documents which must be correlated with other sources. In Roman and Byzantine history, the works of A. Alföldi, C. H. V. Sutherland, M. Grant, and many others, and in Islamic history, the older works of Max van Berchem and recent studies by G. C. Miles, J. Walker, and others, have shown the extraordinary wealth of numismatic evidence for the understanding of civilization and history. In the specific case of the Ṭūlūnids a number of problems of great significance for the history of the ‘Abbāsid caliphate are posed and the coinage may perhaps be used to solve some of them.

The first part of this study will be devoted to a catalogue of the coins known to me which can be assigned without doubt to the Ṭūlūnids; that is, those coins bearing the name of a member of the Ṭūlūnid family.1 Only different types of coins will be described, but in each case the number of known coins will be indicated. They can all be found in the appended bibliography. The metrologic information (weight in grams, diameter in millimeters) has been, whenever possible, based on the coins of the ANS and of the University Museum in Philadelphia. Otherwise it has been copied from whatever publication has first dealt with a coin. Only one set of metro-logic characteristics is given whenever all coins whose weight and diameter are known are quite similar to each other in this respect. Any variation which may be of interest to the economic or political historian has been included, wherever it occurs. Since, on the whole, Ṭūlūnid coins follow the pattern of the classical 'Abbāsid type in appearance and in epigraphy, full description will be given only once. It is assumed that, unless specifically noted, each coin follows the pattern of the one preceding it. Whenever possible, a bibliographical reference has been given for each coin in one of the two standard and most complete catalogues (British Museum, Paris).1 Unpublished or rare issues are marked with an asterisk(*), illustrated specimens with a dagger (†).

The second part will include discussions of a series of problems posed either by the coinage itself or by the history of the dynasty, for which coins may provide an answer. A detailed history of the Ṭūlūnid dynasty has not been included, since it is easily available in the Encyclopedia of Islam,2 in G. Wiet’s contribution to Hanotaux’ Histoire de la Nation Egyptienne, and especially in Zaky Hassan’s Les Tulunides and Carl Becker’s superb evaluation of Egypt in Ṭūlūnid times in his Beiträge zur Geschichte Ägyptens.3 Most of these studies have in fact used numismatic evidence, especially for establishing dates. But the historical problem of the Ṭūlūnids is a much wider one.

It is often said that Aḥmad ibn Ṭūlūn became virtually independent, and his status is compared to that of the Ṭāhirids in Persia or to that of the Aghlabids in Ifrīqiyah.1 But it may be questioned whether the notion of independence is not in many ways ambiguous when applied in contemporary terms to what has been recently called “the Muslim City.”2

The intrinsic theoretical unity of the Muslim world, dār al-Islām, has often been emphasized.3 But it is equally well known that, in practice, the middle of the third century A. H. saw the first breaking up of the physical unity of the Muslim empire. Together with it, or somewhat later, there appeared also, at the head of the “politico-religious unity” characteristic of Islam,4 a breakdown of the unity of command with the multiplication of amīrs and later, sulṭāns, who first in fact and then also in right shared with the Commander of the Faithful the leadership, actually even the sovereignty, of the Islamic community. Islam underwent a practical breakup of its theoretical structure both in the center of the caliphate and in the provinces which made up the empire.5 A study of the Ṭūlūnids can only concern itself with the latter phenomenon, but the Egyptian dynasty is particularly important in this process of disintegration. Its early date is significant in the sense that it was the first true dynasty to develop within the Arab core of the Islamic empire and that, therefore, it created a precedent for all later developments. Furthermore, the personalities of the men who were involved in the Ṭūlūnid adventure (especially Aḥmad ibn Ṭūlūn and al-Muwaffaq) were such that through them one can get a glimpse of the ideology and of the motivations which led them to do and to say what they did and said.

The problem is not so much one of establishing facts as of determining in what ways the political morcellement of the empire was reconciled in men’s minds with the theoretical unity of Islam. In other words, how did the Ṭūlūnids (or any other third century A. H. dynasty) explain their own position in regard to the caliphate? The problem is not academic, for it is only through a study of each individual dynasty of the ninth, tenth, and eleventh centuries that one can properly understand the later formalization of a new Muslim concept of sovereignty and of a new political and social situation after the arrival of the Seljuqs, when mediaeval Islam was transformed.1 For the early period we do not possess a systematic treatise such as that of al-Māwardi or an analysis of the contemporary situation such as al-Ghazzāli’s. Nor are the individual stories found among chroniclers always sufficient to suggest a clear picture of the situation.2 The epigraphical material, however, spotty as it may be, and in particular coins, whose date and minting place are generally known, can be of great interest. On the one hand, this material, in most cases, is rigorously contemporary with the event with which it was connected; and, on the other hand, it is not a mere recording, but a definite expression of power and sovereignty.1 This study will attempt to determine, on the basis of coinage, inscriptions, and texts, the nature of the Ṭūlūnids’ power in Egypt in relation to the caliphate and in what ways their power was similar to or dissimilar from the contemporary sovereignty of Aghlabid and Ṭāhirid princes.2 Such an analysis is particularly justified in the case of the Ṭūlūnids, since, as will be shown, Aḥmad ibn Ṭūlūn for one was very careful throughout his life to have his acts sanctioned by religious authorities.3 It will be attempted to show that this was not a purely formal habit and that the life and utterances of ibn Ṭūlūn indicate a high degree of religious consciousness in all of his activities. But, even without this indication, it could be argued that his acts and decisions had in some way or other to be adapted to the politico-religious thinking of the time. Even if they were not always in accord with the general consensus of theological opinion, they must have been agreeable to a certain fraction of the religious community and, therefore, must have reflected a definite religious and ideological trend of the time. Thus, insofar as it is based on official epigraphical material, this study will not concern itself primarily with the political, economic, social, or military aspects of Ṭūlūnid history, which have been well analyzed by C. Becker, G. Wiet, Z. Hassan, and others (unless the coinage can lead us to new conclusions), but with the ideological or theoretical frame within which this history took place.

This evaluation of Ṭūlūnid coinage as a political and historical document falls under two headings: the coinage of Aḥmad ibn Ṭūlūn and the coinage of Khumārawayh ibn Aḥmad. In the case of the latter, only the coinage of the first few years of his rule is of any significance. Late Ṭūlūnid coinage is of little interest and serves only as a useful contemporary milestone for dates which are otherwise clearly established through texts.

End Notes

1 G. C. Miles, Fāṭimid Coins (New York City, 1951), and “Some Early Arab Dinars,” The American Numismatic Society Museum Notes, III (1948).
1 On the problem of defining a coin as belonging to this or that dynasty, see the remarks of G. C. Miles, in The Numismatic History of Rayy (New York City, 1938), p. 110.
1 S. Lane-Poole, Catalogue of Oriental Coins in the British Museum, 10 vols. (London, 1875–1889); H. Lavoix, Catalogue des Monnaies Musulmanes de la Bibliothèque Nationale: III Egypte et Syrie (Paris, 1896). These two catalogues will be abbreviated as B. M., and Paris respectively.
1 Wiet, p. 86; Becker, p. 150, where it is suggested that the similarity was not only one of situation, but also one of the means used to achieve the same results.
1 Cf., for instance, H. A. R. Gibb, “An Interpretation of Islamic History II,” The Muslim World, XLV (1955), pp. 124 ff.
1 On coinage as a prerogative of the prince see, among others, Tyan, pp. 480 ff; on inscriptions remarks are scattered throughout the numerous books and articles of Max van Berchem, in particular in his volumes of the Matériaux pour un Corpus Inscriptionum Arabicarum.
2 G. C. Miles, Rare Islamic Coins (New York City, 1950); hereafter RIC.
2 See also Zaky Hassan’s article “Aḥmad ibn Ṭūlūn,” in the new edition of the Encyclopedia of Islam.
2 L. Gardet, La Cité Musulmane (Paris, 1954).
2 This difficulty has already been pointed out by Tyan, pp. X-XI.
2 One should note that coinage or, in a more general way epigraphy, is not the only non-literary source which can be used to clarify the political and ideological relationship of various dynasties to the caliphate. Others are, for instance, clothes and arms. These are comparatively well known and accessible as far as the Mameluk period is concerned, after the studies of L. A. Mayer and D. S. Rice. But, so far as the earlier period is concerned, only a very careful combing of the sources could lead to the establishment of a consistent order among the numerous "robes of honor." Cf. some preliminary remarks in Tyan, pp. 488 ff., especially p. 496, where, of course, only caliphal dresses and insignia are mentioned. The essential point is that, when the literary sources are few or unreliable, and even when they are abundant, the mediaeval Islamic world has left us an enormous body of material which, following Sauvaget, one may call “archaeological,” and which supplements, when it does not actually supersede many a literary document. Cf. J. Sauvaget, Introduction à l'histoire de l'Orient Musulman (Paris, 1946), pp. 48 ff.
3 E. T. Rogers, The Coins of the Ṭūlūni Dynasty, in The International Numismata Orientalia, Part IV (London, 1877).
3 G. Wiet, L'Egypte Arabe, vol. IV of G. Hanotaux' Histoire de la Nation Egyptienne (Paris, 1937); Zaky M. Hassan, Les Tulunides (Paris, 1933); C. H. Becker, „Die Stellung der Tuluniden,” Beiträge zur Geschichte Ägyptens unter dem Islam (Strassburg, 1902–3). Among the sources I have been unable to consult al-Balawi, Sīrah Aḥmad ibn Ṭūlūn, ed. M. Kurd 'Ali (Damascus, 1358 A. H.).
3 Gardet, pp. 23 ff.
3 For example ibn Sa'īd, Fragmente aus dem Mughrib, ed. K. Vollers in Semitische Studien, I (Berlin, 1894), pp. 69–70; al-Kindi, Governors and Judges of Egypt, ed. Rh. Guest (London, 1912), p. 226. Hassan, Les Tulunides, p. 82, notices this tendency in ibn Ṭūlūn to look for a religious justification of his acts, but he does not explain it entirely.
4 This result would not have been possible without the remarkably careful bibliographical notations kept by Dr. Miles, whom I want to thank for the free use of his cards. It must also be added that most of these coins belong to European and American collections. Near Eastern collections are still too little known, but publications—even incomplete ones—such as that of the Khiḍr Ilyās treasure (N. al-Naqshabandi, "Kanz Khiḍr Ilyās," Sumer, X (1954), pp. 180ff.) show that a great number of numismatic documents are still untouched in the private and public collections throughout the Near East.
4 Ibid., p. 25.
5 On all these problems the latest study is that of E. Tyan, Institutions du Droit Public Musulman: I Le Califat (Paris, 1954), especially pp. 513ff.


1 Æ 258 Miṣr 3 Rogers, nos. 1 and 2 Obv. لا اله الا
الله وحده
لا شريك له
Rev. لله
Obv. margin: ornament.
Rev. margin: بسم الله ضرب هذا
الفلس بمصر سنه ثمان وخمسين و مائتين
†2 Æ 25[9?] Miṣr 19 2.24 4 RIC, nos. 392–3 For both 1 and 2 see commentary pp. 30 ff.
†3 image 266 Miṣr 23.1 4.22 13 ANS, UM
Rogers, nos. 9–10
B.M., II, no. 218
Obv. — — —
المفوّض الى الله
Rev. — — —
المعتمد على الله
احمد بن طولون
Usual margins: Qur’ān, XXX, 3–4, and IX, 33; mint and date.
4 image 267 Miṣr 22.1 4.12 20 UM (2)
B.M., II, no. 219
5 image 267 Miṣr 24 2.45 2 Paris, III, no. 13
Rogers, no. 17
6 image 267 al-Rāfiqah 21 3.70 4 Rogers, no. 16
Paris, III, no. 1
†*7 image 267 al-Rāfiqah 25 2.91 1 Yale University
8 image 268 Miṣr 23
14 ANS
B.M., IX, no. 219b
*9 image 268 al-Rāfiqah 1 Jerusalem, Flagellation No description given.
10 image 268 al-Rāfiqah 3.30 1 Paris, III, no. 2 Note the low weight, below pp. 58–9.
11 image 268 al-Rāfiqah 23.2 3.81 5 Rogers, nos. 18–19 Rev. — — —
احمد بن طولون
لؤ لؤ
†12 image 269 Miṣr 22.9 4.05 12 UM
Rogers, nos. 20–21
B.M., II, no. 220
No لؤ لؤ
*13 image 269 Dimishq 1 American University of Beirut I have not seen this coin. Information G. C. Miles.
†14 image 270 Miṣr 22 4.15 27 ANS, UM
Rogers, nos. 22–24
B.M., IX, no. 220b
The weight of these coins varies between 4. and 4.25. See p. 68.
15 image 270 al-Rāfiqah 3.50 1 Rogers, no. 26
(Paris, III, no. 3)
Note the low weight.
16 image 270 Dimishq 21 2.60 1 Paris, III, no. 12
17 Æ bef.
1 Zambaur, Contrib. I, pp. 74ff. No mint. No date.
Rev. احمد بن
18 Æ bef.
5 Miles, Tarsus, p. 301. Obv. — — —
المعتمد على الله
Rev. احمد بن
On rev. traces of marginal legend including المؤمنين.1
19 image 270 al-Rāfiqah 3.58 1 Paris, III, no. 19 Obv. — — —
المفوّض الى الله
Rev. — — —
المعتمد على الله
خمارويه بن احمد
†20 image 271 Miṣr 22.1 4.13 12 UM
Rogers, no. 27
21 image 272 Miṣr 21.5
19 UM
Rogers, nos. 30–31
B.M., IX, no. 220t
*22 image 272 al-Rāfiqah 2.58 2 Markoff, add., p. 928, no. 7c
R.N. (1935), p. 34
Not described.
Note the extraordinary weight.
23 image 272 Dimishq 25 2.80 1 Paris, III, no. 32 Rev. pellet.
†24 image 273 Miṣr 21.9
27 UM
Rogers, nos. 35–37
B.M., II, no. 221
Rogers, no. 35, weighs only 3.95.
25 image 273 al-Rāfiqah 22.8 4.10
7 Rogers, nos. 32–33
B.M., II, no. 222
Paris, III, no. 20
All coins, except Rogers no. 32 (B.M., II, no. 222) and Khedivial, no. 910, whose weight is known, weigh around 3.50.
26 image 274 Miṣr 22.9
19 UM (2)
Rogers, nos. 39–41
B.M., IX, no. 222c
†*27 image 274 al-Rāfiqah 23.5 3.96 2 ANS The second coin (R.N., 1935, p. 35) weighs even less (3.62).
28 image 274 Ḥimṣ 3.60 1 Rogers, no. 42
(Paris, III, no. 17)
Note weight.
29 image 275 Miṣr 22.8 3.95 7 UM
Rogers, no. 44
Paris, III, no. 27
Square piercing.
The Paris coin weighs 4.30.
30 image 275 al-Rāfiqah 3.50 3 Rogers, no. 45
Paris, III, no. 27
Note weight.
31 image 275 Dimishq 2.90 1 Paris, III, no. 35
32 image 276 Miṣr 22.8 4.05 8 UM
Rogers, nos. 47–48
Paris, III, no. 28
33 image 276 al-Rāfiqah 20
4 Rogers, no. 49
Rogers, no. 50
Paris, III, no. 22
Rogers, no. 49 is the same as B.M., II, no. 23.
Note the very low weight.
34 image 276 Dimishq 23.7 2.55 1 Rogers, no. 52
35 image 276 Ḥarrān 25.4 4.16 2 Rogers, no. 51
Khedivial, no. 914
The Cairo coin (apparently the same as in Sumer, III, p. 277) is remarkable for its small size (19.).
36 image 276 Anṭākiyah 4.10 1 Paris, III, no. 15 image
Obv. — — —
Rev. — — —
37 image 277 Miṣr 22.5
25 ANS
Rogers, nos. 53–55
Paris, III, no. 29
No marks.
†*38 image 277 al-Rāfiqah 19 4.60 1 ANS Note the unusual weight and size.
39 image 277 Dimishq 23 4.15 3 Rogers, no. 56
B.M., II, no. 224
40 image 277 Filasṭīn 3.20 2 Paris, III, no. 23 Note weight.
*41 image 277 Anṭākiyah 1 Private collection (Casablanca) Communication G. C. Miles.
42 image 278 Miṣr 22.5 4.10 21 UM
B.M., IX, no. 224c
Rogers, nos. 57–8
43 image 278 al-Rāfiqah 24 3.60 3 Rogers, nos. 59–60
44 image 278 al-Rāfiqah 2 Rogers, no. 61 Rev. — — —
†*45 image 278 al-Rāfiqah 23 3.70 1 ANS Floral tail to the nūn of bn.
*46 image 278 Ḥimṣ 1 Constantinople, II, no. 798 No marks.
47 image 278 Filasṭīn 4.20 4 Paris, III, no. 24
48 image 278 Anṭākiyah 25.1 3.54 1 Rogers, no. 62 Obv.: ornamental tails on, و and ي of المفوّض الى الله; pellet between them.
Rev.: to the side of area: الملك
†*49 image 278 Anṭākiyah 23 3.21 1 ANS Obv.: normal inscription, but different style of epigraphy.
Rev.: — — —
50 image 2XX Dimishq 25.1 2.82 1 Rogers, no. 63 Since the reverse bears the name of al-Mu'tamid and the obverse that of al-Mufawwaḍ, this coin could not be later than the beginning of 279.
*51 Æ 1 Soret, RNB (1854), p. 18 Obv. — — —
المعتمد على الله
Rev. — — —
احمد بن
Cf. below p. 77
†*52 image 279 Miṣr 22.3
2 UM
Obv. — — —
المعتضد بالله
Rev. — — —
المعتمد على الله
خمارويه بن احمد
53 image 279 Anṭākiyah 4.35 1 Paris, III, no. 16 Note weight.
54 image 279 Dimishq 21.8 4.14 1 Iraq Museum, no. 3878 (Sumer, III, p. 277).
55 image 279 Miṣr 4.10 141 Rogers, nos. 64–65
Paris, III, no. 31
Obv.: — — —
No name.
Rev.: — — —
المعتضد بالله
خمارويه بن احمد
56 image 279 al-Rāfiqah 1 Rogers, no. 66 This coin was published by Soret (RNB, 1856, p. 132) as having on the obverse خمارويه بن احمد and on the reverse المعتضد بالله. Rogers corrected it to fit with the common type 55.
†*57 image 279 Ḥalab 19 3.25 1 H. W. Glidden coll.
58 image 279 Anṭākiyah 4.30
2 Rogers, nos. 67–68 Cf. 53. Rogers, no. 67 is supposed to be in Paris, but does not agree with Paris, III, no. 16.
59 Æ 279 31 Miles, Tarsus, p. 302
لا اله الا
الله محمد
رسول الله
مما امر به
الامير محمد
بن موسى
Cf. below p. 33.
60 image 280 Miṣr 3.80 1 Paris, III, no. 32 Obv. — — —
Rev. — — —
61 image 280 Miṣr 4.20 1 Rogers, no. 69 Obv. — — —
Rev.: no sign.
†62 image 280 Miṣr 22.9 4.09 8 UM
Rogers, nos. 70–72
B.M., IX, no. 224h
Obv. — —
Rev. — — —
63 image 281 Miṣr 21.9 4.02 10 UM
Rogers, nos. 73–75
B.M., IX, no. 224k
64 image 281 Dimishq 4.40 1 Paris, III, no. 18
65 image 281 Filasṭīn 1 Z. für N., XI, p. 64 Not described.
66 image 281 Ḥalab 21.9 4.12 4 Rogers, no. 76 Obv.: no pellet. Rev. — — —
67 image 282 Miṣr 22 4.08 15 UM
Rogers, nos. 77–78
B.M., II, no. 225
Like 62. Note that the mint in Rogers no. 78 is not certain. Cf. Tiesenhausen in RNB, vol. XXXI, p. 360.
68 image 282 Miṣr 3. 2 Rogers, no. 79
Paris, III, no. 36
No pellet, no. ح.
†*69 image 282 Filasṭīn 20 4.03 1 ANS
†70 image 283 Miṣr 22.7 4.22 21 UM
Rogers, nos. 80–85
B.M., II, no. 226
Rev. — — —
جيش بن خمارويه
Note that one coin, Khedivial, no. 924, has the very low weight of 3.18.
71 image 283 Miṣr 20 3.90
3 Rogers, no. 86
Khedivial, no. 925
Rev. — — —
هرون بن خمارويه
Note the comparatively low weight of these Miṣr coins.
†72 image 284 Miṣr 22.2
12 UM
Rogers, nos. 87–91
B.M., II, no. 227
No pellet, no ح.
73 image 284 Dimishq 2 B.M., IX, no. 227b No weight given. Four rings.
74 image 285 Miṣr 21
18 UM
Rogers, nos. 92–95
Paris, III, no. 40
No rings. As 72.
The ANS weight is exceptional.
All other weights known to me are below 3.80.
†*75 image 285 Ḥimṣ 21.8 4.07 2 ANS
76 image 285 Filasṭīn 22 4.29 4 UM
Rogers, no. 96–97
B.M., II, no. 228
The coin from the Rogers collection weighs only 3.75.
77 image 285 Ḥalab 22.6 3.97 1 Rogers, no. 97
78 image 286 Miṣr 21 3.73 7 UM
B.M., IX, no. 228c
79 image 286 Miṣr 4. 1 Rogers, no. 98
(Paris, III, no. 41)
Rev. — — —
†80 image 287 Miṣr 21.9
4 UM
Rogers, no. 102
Nothing below area on rev. or obv.
†81 image 287 Miṣr 22 4.14 251 UM
Rogers, nos. 99–101
(Paris, III, no. 92)
Obv. — — —
Rev. — — —
82 image 287 Dimishq 2 Johnston, N.C., p. 266 Like 73.
*83 image 287 Filasṭīn 1 Markoff, p. 928, no. 13b No description given.
84 image 288 Miṣr 21
23 UM
Rogers, nos. 103–7
B.M., II, no. 229
Like 81.
85 image 289 Miṣr 23.4 4.08 1 B.M., IX, no. 229c image on obverse only.
86 image 289 Miṣr 4.15 1 Paris, III, no. 45 Rev. — — —
المكتفى بالله
هرون بن خمارويه
No image on obverse.
87 image 289 Miṣr 22
202 UM
Rogers, nos. 108–11
Paris, III, no. 44
Like preceding, but with image on both sides.
88 image 289 Miṣr 1 Rogers, no. 112 Rev. — — —
†*89 image 28X Ḥalab 21.1 4.03 1 UM Bears the name of Harūn.
No signs.
†90 image 290 Miṣr 21.9 4.02 12 UM
Rogers, nos. 114–17
B.M., IX, no. 229e
Like 87.
91 image 290 Miṣr 22 1 Constantinople, no. 799 image on obv. only.
92 image 290 Dimishq 2 Zambaur, Contrib. I, no. 71 No signs.
93 image 290 Filasṭīn 22.5 3.43 1 Rogers, no. 113
Khedivial, no. 933
Note low weight.
†94 image 291 Miṣr 21.9 3.82 23 UM
Rogers, nos. 118–23
B.M., II, no. 230
Obv. — — —
No sign on rev. The UM coin is lower in weight than most of the other coins whose weight is known (4.10–4.20).
95 image 291 Miṣr 1 Lane-Poole, Johnston, p. 57 No description given.
†*96 image 291 Filasṭīn 22 3.63 2 UM Pierced. Note that there are already ‘Abbāsid coins struck in Filasṭīn in 291 (Zambaur in N.Z., 1922, p. 9). Note also the weight.
97 image 292 Miṣr 23.4 3.88 1 B.M., IX, no. 230c Like 94. Note that there are already ‘Abbāsid coins struck in Miṣr in 292 (Rogers, nos. 124–5). The B.M. coin belongs to what must have been the last Ṭūlūnid issue.

End Notes

1 There is little doubt that nos. 17 and 18 belong to the same series. They are separated here only because in detail there occur a few minor discrepancies between the Tarsus coins and the specimen published by Zambaur. Cf. below pp. 30ff.
1 Are included here 8 coins from the Khiḍr Iliyās treasure, some of which could conceivably belong to type 52.
1 Undescribed coins have been included here, although a number could presumably belong to the preceding type.
2 Some of these coins, which are not described, could equally well belong to any of the four 289 types.


I. Aḥmad ibn Ṭūlūn and early Ṭūlūnid coinage

According to the mediaeval chroniclers, Aḥmad ibn Ṭūlūn, after some thirty-five years spent either as a student at Tarsus or as an officer in Samarra—where his life was largely devoid of the usual more or less criminal intrigues which characterized the ‘Abbāsid court after the death of al-Mutawakkil—was given his first appointment in Egypt in 254 A.H./868 A.D. Al-Ṭabari simply says that in that year Bayākbāk, or Bākbāk, the apanagist of Egypt and Aḥmad’s stepfather, “entrusted ('aqada) Aḥmad ibn Ṭūlūn with the governorship of Miṣr.”1 Maqrīzi and ibn Sa'īd, who repeat the contemporary account of ibn al-Dāyah and whose reliability as far as Egypt under ibn Ṭūlūn is concerned is greater than al-Ṭabari’s, are more specific. According to them, Aḥmad ibn Ṭūlūn was appointed as the agent of Bayākbāk in Egypt (fī khilāfatihi 'ala Miṣr) and in fact his appointment was only over the main part of the province, al-qaṣbah, to the exclusion of the regions which depended from it, such as Alexandria, etc.2 For sixteen years thereafter, until his death in 270 A.H./884 A.D., Aḥmad ibn Ṭūlūn remained in Egypt. But the nature of his power there did not remain the same throughout these years.

The second stage in Aḥmad ibn Ṭūlūn’s rule over Egypt began in 257 A.H., when Yārjūkh, the new apanagist of Egypt and Aḥmad’s father-in-law, gave him control over Alexandria and Barqah.1 But the major new event occurred in 258. In that year Aḥmad received control over the finances of Egypt, after the removal of his rival ibn al-Mudabbir, and also the governorship (wilāyah) of the Syrian marshes.2 And ibn Sa'īd adds that the letter of appointment which was brought to ibn Ṭūlūn by a personal servant, khādim, of the caliph al-Mu'tamid, was witnessed by two respected judges, one from Wāsit, the other one from the marshes.3 The fact itself and the station of the people involved would further enhance the suggestion made above that, in all his dealings with the caliphate, Aḥmad ibn Ṭūlūn was always careful to receive the approbation of the religious leaders of Islam.

In the same year, 258/871–2 Yārjūkh died and Ja'far ibn al-Mu'tamid, later called al-Mufawwaḍ ila Allāh, became heir to the caliphate.1 Some sources assert that it is soon after that, in 259, that Aḥmad ibn Ṭūlūn became independent, while others have even believed him to have been so as early as 258. But the literary evidence does not give any information to suggest any difference in Aḥmad’s status between 258 and 259; and the two late texts (ibn Khaldūn and ibn Taghribirdi) which do speak of independence are contradicted either by earlier documents or by their own further accounts.2

At the same time, 258 is the date which is usually taken to have witnessed the first appearance of Ṭūlūnid coinage. In 1854 and 1856 F. Soret published two copper coins, dated in 258, struck in Miṣr, and bearing in the lower part of the reverse the sign |◊||, which was interpreted by Soret to mean Aḥmad.3 The reading is, of course, unacceptable, but these two coins have been taken by Rogers as nos. 1 and 2 of his classification of Ṭūlūnid coins. In 1904, E. de Zambaur published a coin with the same sign at the bottom of the reverse, but without any trace of mint or date. Over the sign appears the name Aḥmad ibn Ṭūlūn, while on the obverse the first part of the profession of faith is followed by the name of the caliph al-Mu'tamid ‘ala Allāh.1 But the two Soret coins did not bear the name of the Commander of the Faithful. Zambaur assumed this sign to be a date, 262, but in a second article2 he seems to have accepted Nützel’s suggestion that it is a mere ornament. In 1950, George C. Miles published two similar coins bearing the probable date of 259,3 while the collection of the American Numismatic Society contains a number of other coppers of the same type. Five specimens of a similar variety were discovered by the American excavators at Tarsus.4 The question is whether these coins all belong to one group and whether they should all be considered as Ṭūlūnid, bearing in mind that the problem is further complicated by the fact that, with an exception to be mentioned presently, no other copper coin is known until the rule of Khumārawayh, twelve years later.

A first point to be made is to compare the evidence of the fals with that of the other coins from Miṣr. Egyptian gold and silver issues are known for almost the whole period between 255 and 267.5 All these issues, except one, are purely ‘Abbāsid, without any mention of Aḥmad ibn Ṭūlūn. The 265 coin in the ANS, the only dirhem of the whole group, is of particular importance, since it suggests that the two undescribed ones of 264 were also purely ‘Abbāsid. But, among these coins, one issue poses a problem and it is noteworthy that it is the 258 issue, i.e., the issue which is contemporary with the major group of problematic copper coins. Five of these coins are known to me: one in the Bibliothèque Nationale,1 two in the University Museum collection,2 one in the National Library (ex-Khedivial) in Cairo,3 and one in a group of coins published by Rogers in 1875.4 All these coins, the only ones known for Miṣr in 258 (with the exception discussed in note 3), have on the reverse, under the name of the caliph, at the place where one generally finds the name of the governor, a word ( ?) which so far has not been identified (PLATE III). But there is no doubt that it cannot be Aḥmad ibn Ṭūlūn. In other words for the year 258 (and perhaps 259 also as far as the copper coins are concerned) we deal with a body of gold and copper coins with a strange sign on the reverse and no obvious Ṭūlūnid name.

The second remark which can be made is based on a comparatively recent archaeological discovery. Among the coins found by the Tarsus excavators, we meet not only with the five previously mentioned coins with the name of Aḥmad ibn Ṭūlūn, but also with 27 specimens of a copper type which bears on the reverse the name of MuṬammad ibn Mūsa.1 Similar coins were already known in Paris 2 and by Soret.3 G. C. Miles has argued that Muḥammad ibn Mūsa was in fact a nephew of ibn Ṭūlūn, who, under Khumārawayh, governed Tarsus for a short time in 279/892–3,4 and that this whole group of coins should be assigned to the thughūr al-sha'miyah, which, from 258, were under the theoretical control of the Ṭūlūnids. This argument is most plausible and the typological similarity of the Zambaur coin to this later group added to the facts that it is epigraphically notably different from the 258 coins and that a number of specimens of the same type were found at Tarsus would indeed suggest that all the coppers bearing the name of Aḥmad ibn Ṭūlūn should then be connected with the thughūr. It is, of course, impossible to determine the date of these coins with any certainty. But since the dated Miṣr specimens do not bear any proper name, and in particular not the name of ibn Ṭūlūn, it is doubtful whether the Zambaur and Tarsus coins should be considered as contemporary with them. A date after 265, that is after Aḥmad’s triumphal expedition through Syria, is perhaps more likely,5 inasmuch as this date would correspond with the first issues of coins in Miṣr with the name of the Ṭūlūnid prince.

In other words the copper coins bearing the name of a Ṭūlūnid should be attributed to the Syrian marshes and are probably to be dated after 265, while the coins struck in Miṣr in 258 and 259 do not have any Ṭūlūnid name. The only point of contact between the two groups of coins is that on both the sign |◊|| occurs and the question which arises is, of course, the origin of that sign. Chronologically the Egyptian samples should be considered as the first to have used this sign, but did the sign originate in Egypt? To my knowledge there is no similar decorative motive in Islamic coinage, either before or after the Ṭūlūnid period. A heraldic sign or some symbol of Central Asian Turkish origin seems very unlikely for so early a period. But, if one considers that this type of copper coin was most common in the thughūr area, it may be suggested that the sign was an imitation of one of the marks of value, monograms, letters, or countermarks on Byzantine coins.1 Although such adaptations of non-Islamic, particularly Byzantine, types are well known at various times throughout Islamic numismatic history,2 this explanation cannot be, for the time being, more than a simple suggestion. It may perhaps be strengthened somewhat by the fact that Aḥmad ibn Ṭūlūn spent many years in Tarsus, studying with the many learned men who were there and participating in the constant Holy War which was waged at the frontier.3 It is because of this attachment to Tarsus and to the idea of the war against the Byzantines that he received the financial and political control of thughūr before acquiring Syria. In the little-known early ghāzi milieu of Tarsus, Byzantine objects and coins must have been quite common; and it is possible that attempts were made then to copy or transform Byzantine symbols, just as the latter certainly were taken over a few centuries later when the Danishmend and Ortoqid cultures developed in another area, but in the same relationship to the Christian civilization of Anatolia. It is quite possible that ibn Ṭūlūn had seen some such Byzantine type, and then adapted it to Egyptian copper coins, since there is no reason to doubt his genuine concern for the jihād against the infidels, which he wanted to undertake as an amīr rather than as an individual soldier of fortune.1

At the same time it must be emphasized that, whether one considers the motive of the diamond and three bars as an imitation from Byzantium or as a mere decoration, the copper coinage on which it appears in no way reflects the position of Aḥmad ibn Ṭūlūn vis-à-vis the caliphate. If it were not for the Zambaur and Tarsus coins which bear the sign and the name of the prince, there would not be any justification in considering the coins as Ṭūlūnid. Even then, these copper coins, which can by analogy be associated with the person of ibn Ṭūlūn, can in no way be taken as evidence of his office, much less so of his independence from the caliphate. In a strict sense, whatever power Aḥmad ibn Ṭūlūn may have acquired in 258, it was not considered by him, or by the caliphate, as a type of appointment which permitted him to add his own name to that of the caliph on the coinage of the province of which he was vice-governor and financial director. In that sense his position was definitely different and lower than that of the numerous governors who, before him, had put their names below the names of the caliph on the coins issued in their capitals.1

This still formally subservient position of Aḥmad ibn Ṭūlūn is further confirmed by the fact that the known ṭirāz inscriptions from Egypt dated in 260, 262, and 2632 do not mention the name of the then governor of Egypt. Yet the letter of appointment which was sent by al-Mu'tamid in 258 did mention ṭirāz together with finances, although the text is admittedly quite vague.3 A third series of more or less contemporary documents is provided by a weight and a stamp recently published by M. Jungfleisch.4 The stamp is dated 260 and mentions neither the name of the caliph nor of the Ṭūlūnid. The weight is dated 259 and does bear the name of Aḥmad ibn Ṭūlūn, called there mawla amīr al-mu’minīn.5 The formula used on these two objects, as has been pointed out by M. Jungfleisch, is interesting in that it used the full expression bism Allāh al-raḥmān al-raḥīm, which is only known on two objects from the times of al-Wāthiq and al-Mutawakkil and which had been abandoned by ibn Ṭūlūn’s immediate predecessors. These two objects seem to confirm the evidence of the coins: in 259 we meet with the name of the Ṭūlūnid on an official object, just as he is indirectly suggested on the copper coins, but, in both cases, we are dealing with a practice which was common to many governors who have never been suspected of trying to become independent. One may call this "routine" procedure, except for the fact that the usage of a less typical formula on glass (a formula, it may be noted, which is characterized by the completeness of its divine invocation) and of a mysterious sign on coinage does indeed suggest that we are dealing with a rather extraordinary man. At the same time, the glass inscription, like the gold coins between 259 and 265 and the ṭirāz formulas which belong to the usual ‘Abbāsid type, does not mention the name of the governor.1

Thus, in dealing with this group of Egyptian documents between the years 254 and 265, we are led to the conclusion that there is only one piece of evidence which definitely and unequivocally mentions the name of Aḥmad ibn Ṭūlūn: the 259 weight. The coincidence of a curious sign on a group of coins of 258 and 259 with the appointment, known through texts, of ibn Ṭūlūn to the control of the finances and of the administration of Egypt and its similarity to the sign on a coin from a different area suggests, without proving, that the copper coins could be considered as Ṭūlūnid. But in none of these cases do we meet with any indication that the de facto ruler of Egypt considered himself independent from the caliphate. If it had not been for what followed, these documents, or the corresponding literary texts, would not have warranted any extensive study. As it is, however, when connected with the texts, they do offer a great deal of interest in illustrating two important points in any explanation of ibn Ṭūlūn’s career: his concern for the absolute legality of his acts towards the caliphate,1 and a strange interest in the activities of the Byzantino-Arab frontier.2 His actions seem meant to appear as conforming to the dictates of Islamic legal and political theory. It must be recalled that the first part of the third century A.H. was one of the most flourishing ones in the development of Islamic civilization, but one that is especially characterized by the activities of the great traditionists and codifiers of Islamic law. Practically alone among the Turkish officers, whose scandals fill the political chronicle of Iraq, Aḥmad ibn Ṭūlūn takes some part in the religious life of the time and withdraws to the frontier with its ribāts and its ghāzis without taking his family along.3 It would not seem consequent then to suggest that his life of earnest retirement was followed by a sudden and intense involvement in intrigues which would ultimately make him independent, and put him in a position contrary to the very nature of Islamic political theory. That he did intrigue to achieve his aims is amply proved by the texts: it was the only means to achieve any result. But to attribute to him an aim to match his means is not to do justice to his past; and such an interpretation is certainly not justified by the contemporary and official documents of the time—as opposed to the later texts, which, influenced as they were by the later developments of Islamic history, tended to simplify earlier events. And it will be attempted to show that these very same religious and legal concerns directed ibn Ṭūlūn’s later actions.

We must, however, first turn to one unsolved problem of the coinage of the time, the problem of the five gold coins of 258. These, it will be remembered, are perfectly normal ‘Abbāsid coins, except for the fact that on the fifth line of the reverse area we meet with a word which has not, so far, been understood [Plate III]. It has been read as Najrān or Baḥrayn, the latter being properly discarded by G. C. Miles,1 since Miṣr appears as a mint on other coins of the same type, and since this would have been a most unusual place for the mint name at the time. The position of the word would almost require that it be a name, presumably the name of the governor responsible for the mint. But the possible names of Yārjūkh, the absentee lord of Egypt in 258, ibn Ṭūlūn, or Ṭa'laj or Ṭughj, who briefly replaced Aḥmad ibn Ṭūlūn while the latter went to Alexandria, do not fit the epigraphical data; and I have been unable to find a possible explanation for this word. Could it be a metrological term? Or are we dealing here with some symbol?

The first evidence we possess for a change in ibn Ṭūlūn’s position occurs in 265. From this year we have two inscriptions. One is on a textile.2 It is of the usual type, except for the fact that the praise of the caliph is followed by the words: "this is what the amīr has [ordered] to be executed in the public ṭirāz, in Miṣr, in the year 265." The name of the amīr is not given, but there is little doubt that it is ibn Ṭūlūn. The other two ṭirāz inscriptions known during his life-time and after 265 are both of the type of the 265 one, not of earlier ones.3 The second inscription of 265 is the well-known one celebrating the foundation of the mosque of ibn Ṭūlūn.4 In it we meet again with the title of amīr and with the expression mawla amīr al-mu'minīn,5 but the importance of the inscription as a document consists essentially in the emphasis given in it to religious themes not usually found in construction inscriptions.1 Aside from the usual Qur’ānic quotations (II, 256; IX, 18; XXIV, 36–38), we meet with quotations (XLVIII, 29; III, 106) which are less common on building inscriptions and which emphasize the duty of the Muslim against the infidel,2 thus pointing once more to the importance given by ibn Ṭūlūn to the religious motivations of his acts and to the Holy War. Expressions such as tasniyah al-dīn and 'ulfah al-mu'minīn within the inscription itself give it an exhortative quality which was common enough on mosque inscriptions but whose emphasis relates it fairly closely to the later Seljuq and Ayyūbid inscriptions, at a time when the main task of the rulers was to restore the spiritual unity of the Islamic community. These two inscriptions are then followed in 266 by the gold coins struck in Miṣr with the name of Aḥmad ibn Ṭūlūn under the name of the caliph. It is, therefore, only from that date that one could conceivably speak of Aḥmad ibn Ṭūlūn as officially independent from the caliphate. But, even then, two questions are raised by this innovation. First, why 266 and not any other year after 258? And, second, what exactly can be meant by independence in the third century A. H., bearing in mind of course that we are not dealing so much with the de facto situation as with the de jure one? Is this the same kind of independence as is evidenced in Persia and North Africa at the same time?

The reason for raising the question of the year 266 is that, so far as one can gather from the texts, there was no new appointment given to Aḥmad ibn Ṭūlūn in that year. One would have to assume that more or less arbitrarily, in that year, Aḥmad suddenly decided to give an official confirmation to his actual rule over Egypt. However, the events of the preceding three years, that is since 263, together with the events which followed 266, may suggest a more complex explanation for the introduction of the Ṭūlūnid’s name on the coinage of Egypt, and one that would be more consistent with his political and legal ideas.

In 258, according to Ṭabari, al-Mu‘tamid had given his brother al-Muwaffaq the provinces of Diyār Muḍar,1 Kinnasrīn, and al-‘Awāsim.2 This tradition is definitely suspect, since it is contradicted by the existence of two coins from al-Rāfiqah (in the Diyār Muḍar) dated in 259 and bearing the name of Ja‘far.3 According to ibn-Sa‘īd, the whole empire was divided into two parts, the western part being farmed out to Ja'far al-Mufawwaḍ, al-Mu'tamid’s son and first in line of succession.4 No date is given by ibn-Sa'īd, although a later Egyptian tradition (Taghribirdi) gives the date 256;5 but the same event is related by Ṭabari for the year 261.6 The coinage of Egypt has the name of Ja'far as early as 258,7 and, since the name appears already in 256,8 one might assume that the division was effected as soon as al-Mu'tamid became caliph. On the other hand, the 256 coin is from Baghdad which, according to the texts of the division, belonged to the realm controlled by al-Muwaffaq. Taghribirdi’s date is thus certainly wrong, but whether this means that 261 should be accepted as the official date is not entirely certain. The numismatic and ṭirāz evidence does confirm the fact of the division,4 but to my knowledge there is no epigraphical material between the years 256 and 261 which would give us the exact date of the division. In using the name Ja'far on the coins the mintmasters could have referred to the heir apparent to the throne. But at the same time it should be noted that we do not have any textual reference to al-Mu'tamid’s heir apparent until 261, in Ṭabari’s text, from which it appears that this is the date at which the name al-Mufawwaḍ was given to him. Ṭabari’s statement in this respect is confirmed by the coins, and the date 261 seems the most likely one for the division of the empire.2 Why then does Ja'far’s name appear on coins before 261? It has been suggested by Becker that it is because, at the death of Yārjūkh in 258, Ja’far was probably made successor to all of the latter’s functions.3 But while this is possible,4 it would not explain the existence of coins as early as 256 with the name of Ja'far. One would have to dissociate in Ṭabari's text the two facts of the division of the empire and of the succession to al-Mu‘tamid. Ja'far was already heir apparent in 256. In 261 he acquired a laqab and was associated with al-Muwaffaq in the rule of the empire, both decisions being strengthened by the sending of the agreement to Makkah, where it was put in the Ka'bah, following probably the precedent of the earlier agreement between al-Amīn and al-Ma’mūn.1 As far as Egypt is concerned the years 258–261 are characterized by the fact that there was no proper authority replacing Yārjūkh. Ibn Ṭūlūn did not receive a new appointment nor did he accomplish any action which would indicate that he considered himself independent, although, of course, there was no one to prevent him from doing what he wanted.

The situation was regularized in 261 when Aḥmad ibn Ṭūlūn’s area of influence, Egypt and the thughūr, fell to the lot of Ja'far and, while al-Muwaffaq had his hands full in the murderous Zanj war, Egypt, with its wealth and army, was the only remaining force in the empire. The events which followed are well known and can be found in all reference books. Al-Muwaffaq asked for money, ṭirāz, slaves, and horses. But at the same time al-Mu'tamid, the caliph, sent a secret message to the effect that al-Muwaffaq’s messenger was but a spy sent over to plot against ibn Ṭūlūn. The latter kept the messenger under surveillance, but he did make an important contribution for the Zanj war, which, let us note it again, was witnessed officially by various Egyptian military and religious leaders. The sum was thought insufficient by al-Muwaffaq and the insulting manner in which his envoy was received led him to write an injurious letter to Ibn Ṭūlūn and to remove him from his position. He even levied an army under the command of Mūsa ibn Būghā to move against Ibn Ṭūlūn. Ibn Ṭūlūn also started to build up his defenses, but at the same time he wrote a letter to al-Muwaffaq, the text of which has been preserved by Ibn Sa'īd and Maqrīzi, and which is one of the most important documents to explain all further actions of the Egyptian prince, especially if we keep in mind the all-important fact that the caliph himself had advised Ibn Ṭūlūn against sending help to al-Muwaffaq. The main point of the letter is legal: "he (al-Muwaffaq) has no power over my office; he did not invest me with it... The division [of the empire] was between him and the amīr Ja'far. The oath of allegiance was sworn to them provided they keep their faith ... [If not,] the community (al-ummah) [may] break from him and from [their] oath to him,"1 al-Muwaffaq did try to use force, but his army stopped at Raqqah and it soon disintegrated. The only result was that Ibn Ṭūlūn lost his official control over the thughūr, but here again the reason given by Ibn Sa'īd is interesting.2 It is that the thughūr should be held by "somebody who would be there, who would participate in the ghazw with its people, while Aḥmad Ibn Ṭūlūn had sent somebody who was not doing anything about it." If one recalls the attachment of Ibn Ṭūlūn to the frontier area, his next move is perfectly understandable. Disregarding his dismissal, which he probably believed had been forced on al-Mu'tamid by his brother, and profiting from the fact that Amajūr, the governor of Syria, had died in 264, he invaded Syria and received the allegiance of the local prefects. But here again, while the possession of Syria is a long-standing economic and strategic necessity for whoever rules Egypt, it should be noted that formally Aḥmad Ibn Ṭūlūn was only asking for the passage of his army to Tarsus, from where he was planning to start the war against the Christians.1 Although the Tarsus expedition was only half successful, he intended to stay there;2 but the news arrived of his son's revolt in 265. Returning to Egypt in 266, he gathered forces against his son; and in that same year the first coins bearing his name appeared in Fusṭāṭ, although no new investment was bestowed on him, nor had he taken any new title for himself.

It has been argued3 that at that time Ibn Ṭūlūn finally felt that he had reached the pinnacle of his power. But it must be admitted that Ibn Ṭūlūn had been just as powerful and secure two years earlier when Mūsa ibn Būghā’s army collapsed in Iraq. Furthermore, the example of his own lieutenant Lū'lū', who, in 268, issued coins with his own name in Rāfiqah, shows that it was not necessary to wait until one’s power was fully established to start using the privileges of the sikkah. Aḥmad’s action must, I believe, be explained in terms of his opinion on what was happening in Iraq at the time. His earlier letter had shown that he believed that al-Muwaffaq had usurped a power which did not belong to him and his relations with al-Mu'tamid led him to believe that the caliph felt the same way.

Around 266, Aḥmad went a step further and assumed that the caliph was no longer free of his movements, that he was virtually the prisoner of his brother. The events which followed have been well summarized by Hassan in his history of the Ṭūlūnids:1 al-Mu'tamid’s unsuccessful flight from Iraq; the attempt to establish the caliphate in Egypt; Aḥmad’s dismissal by al-Muwaffaq and his replacement by Isḥāq ibn Kundāj, the governor of Mosul; his unsuccessful expedition to occupy Mecca; and finally the meeting at Damascus in 269, when the judges and religious leaders of Syria and Egypt were gathered to hear about the tragic state of the caliph and to exclude al-Muwaffaq from the succession to the throne. The meeting in Damascus is the crucial act in ibn Ṭūlūn’s activities. He did not try to use religious authority to foster the aims of his own ambition. Rather he tried to persuade the religious authorities of the truth and validity of his own inner conviction that the Commander of the Faithful was a prisoner and that he had to be delivered.2 The meeting of 269 was certainly not called on the spur of the moment, but had been brought about by a long series of events which are not always very clear in the sources.1 It may be suggested that the appearance of the name of ibn Ṭūlūn on the coins of Egypt in 266 can provide us with the date at which the vice-governor of Egypt and director of its finances felt that, since his contacts with the caliphate through proper channels were broken and since the caliph could no longer voice proper authority, he could participate in the official right of sikkah. In fact it was his duty to do so.

For this whole development illustrates yet another significant point. In the eleventh-century al-Aḥkam al-Sulṭāniyah of al-Māwardi, there is a passage which discusses the case when the freedom of the imām is impaired.2 Two possibilities exist: either a subordinate of the imām takes more power than that to which he is entitled, or the imām is the prisoner of a victorious enemy, a polytheist or a rebellious subject. In the first case the procedure is for the imām to look for an ally or helper who will put an end to the domination of the usurper; in the second case, the whole community (ummah) must try to liberate him, through arms or through ransom. It could be argued that, in the case of the relationship between Aḥmad ibn Ṭūlūn and the caliphate, an evolution occurred which corresponds to the two possibilities outlined by al-Māwardi. First, the caliph, feeling that al-Muwaffaq took on too much power,3 appealed to ibn Ṭūlūn for help.

The second step appears to have been taken by Ibn Ṭūlūn independently from the caliphate, when he was led to believe—or tried to make others believe1—that the caliph was a prisoner of his brother. In the first act of the drama, at the time of his letter to al-Muwaffaq, Aḥmad was still considering himself as bound to the agreement which made al-Muwaffaq second in succession to the caliphate. It is only when he thought it incumbent upon himself, as the head of the only sizeable force in Islam outside of al-Muwaffaq’s, to move against the latter—or shortly before that—that he felt free to strike coins with his own name below that of the caliph and with the name of the heir apparent and lord of the West, al-Mufawwaḍ, on the obverse. Through his appeal to the ijmā' of the religious leaders of the area under his control, through his usage of a procedure which later became part of the textbooks on Islamic law, Aḥmad Ibn Ṭūlūn also shows that his acts corresponded without doubt to a legal thought which must have been formulated somewhere in his time and which he probably learned during his years at Tarsus. In this sense, it is interesting to compare him to his opponent in Egypt itself, Bakkār ibn Qutaybah, the chief qāḍi of Egypt,2 who refused to accept the Damascus verdict on the grounds that there was insufficient proof of the plight of al-Mu'tamid and that al-Muwaffaq had not, therefore, forfeited his position.3 Bakkār certainly believed that the preservation of the faith was in a way independent from political contingencies.4 Thus, next to ibn Ṭūlūn’s comparative radicalism, there seems to appear in Bakkār that principle of ḍarūriyah (necessity) which will be so much more strongly emphasized by al-Māwardi and al-Ghazzāli, several centuries later.1 It would be interesting to know to what extent and in what ways these two attitudes, this "ambivalence" so characteristic of any living culture,2 corresponded to any well-defined position of the religious and legal groups of Islam in the ninth century.

Another question raised at this stage is that of the origins of ibn Ṭūlūn’s attitude. Does his legal framework reflect that of the early ghāzi of Islam? Too little is known about the organization and the life of the frontier in the ninth and tenth centuries A.D., but the previously quoted statement of ibn Hawqal 3 suggests that there was an intensive intellectual life there together with the constant warfare.4 We may meet in early Islam with a frontier psychology which is perhaps better known in the case of the later Danishmends, but which may at that time have been characterized by strict obedience to religious and legal principles.5 The literary evidence is not very clear on the question of ibn Ṭūlūn’s religious affiliation. According to Taghribirdi and the Egyptian tradition he was a ḥanīfite,1 like Bakkār, and this is accepted by Hassan.2 But another possibility is suggested by ibn al-Athīr,3 who, although late, was fairly well informed on Ṭūlūnid matters.4 His description of Ibn Ṭūlūn mentions the latter not only as a pious man full of concern for religious problems, but also as a staunch shāfi'ite. The question cannot be settled without a more thorough attempt at comparing ibn Ṭūlūn’s actions and opinions with the lessons of the two schools, but could it be perhaps that, in his somewhat puritan approach, Ibn Ṭūlūn had been influenced by certain shāfi'ite methods of thought?5

If these hypotheses are acceptable, one can see how the coinage of Aḥmad Ibn Ṭūlūn serves to qualify the notion of the independence of the first dynasty of Egypt. We must now examine the problem from another direction and attempt to compare the expression of sovereignty of Aḥmad Ibn Ṭūlūn with that of the other dynasties of the century. The Ṭāhirids and the Aghlabids, although both dynasties started considerably earlier than the Ṭūlūnids and although the collapse of the Tāhirids actually corresponded to the rise of the Ṭūlūnids, are most likely to be fruitful as comparative subjects, since the establishment of the literally contemporary Ṣaffārids was the result of a conquest, while Ibrāhīm ibn al-Aghlab and Ṭāhir ibn al-Ḥusayn, just as Aḥmad Ibn Ṭūlūn, were appointed to their offices.

The 184/800 appointment of Ibrāhīm as governor (amīr)6 of Ifrīqiyah was truly "la solution avantageuse d’une situation difficile qui n’avait que trop duré."1 The "difficult situation" was not the ambition of one leader, but the chaos of North African politics in the late eighth century. Ibrāhīm ibn al-Aghlab appeared to be the only leader who could unite all parties, and Hārūn al-Rashīd empowered him to do so. From then until the end of the dynasty we do not have much evidence of any continuous political relationship between the various Aghlabid rulers and the caliphs. In particular there is no indication that each Aghlabid prince was invested anew by the Commander of the Faithful.2 At the same time we have no evidence that there was any break between Baghdad and Qayrawān. In one case where it is related that Ziyādah Allāh, while drunk, wrote a challenging letter to al-Ma’mūn, it is added that he immediately tried to overtake the messenger, and, failing to do so, that he sent a second letter in much more submissive terms. al-Ma’mūn, we are told, did not pay any attention to the first letter.3 Furthermore, when al-Mu‘taḍid ordered Ibrāhīm ibn Aḥmad to resign in favor of his son, he did so; but, instead of appearing before the caliph to justify himself, he preferred to go and die in the Holy War against the Christians.4 Certain coins also were marked as being for the caliphate,5 and it seems6 that there was a continuous tribute sent from North Africa to Iraq. Similarly the spiritual prestige of the caliph was in no way impaired.1 And yet, on the coins of the Aghlabids, with the exception of the coins of Ibrāhīm I, there is no mention of the name of the caliph.2 For the first quarter of the third century this may not be too surprising, since, in a number of cases, even gold and silver coins struck in areas directly controlled by Baghdad did not give the name of the ruling Commander of the Faithful.3 But from the reign of al-Mu'taṣim, and certainly from the time of al-Wāthiq, the name of the caliph always occupies the lower part of the reverse area, with, more often than not, the name of the heir apparent on the obverse. This practice was not followed by the Aghlabids. Did they simply continue the traditional type? Or were they indicating that, while vassals of the caliphate as an institution, they did not necessarily feel subjected to any specific caliph and that they considered themselves as politically and financially independent? The second explanation is certainly the more likely one and corresponds quite clearly to the reality of the time.4 In other words, in the case of the Aghlabids, we deal with a dynasty, whose first representative was legally appointed by the caliph and which perpetuated itself without interfering in the affairs of the caliphate and without being touched by it. It is, as has been justly seen by Marçais, the perfect application of the first kind of amirate according to Māwardi,5 but it went much further in its expression of independence than any previous dynastic group (always with the exception of Umayyad Spain), for it completely ignored the caliphal right of the sikkah, and only preserved the higher one of the khuṭbah.

At the other end of the empire the major power until the middle of the third century A. H. was that of the Ṭāhirids. They are generally considered, and in many ways justifiably so, in relation to the later so-called Persian dynasties, but the curious point about them is that, regardless of their significance as the first materialization of a Persian renaissance, from the caliphate’s point of view they fulfilled in the East the same function mutatis mutandis as that of the Aghlabids in the West. They did not assume the rule of Khurāsān themselves, but it became apparent to the caliphs that the troubled situation in northeast Persia could only be resolved through the appointment there of Ṭāhir.1 After his death his descendants were all regularly appointed by the caliphs until in 259 Muḥammad ibn Ṭāhir was forcibly removed by the first Ṣaffārid.2 But it is doubtful whether one can properly speak of a Ṭāhirid dynasty in Khurāsān, since many Ṭāhirids held high office in other parts of the empire, combining these at times with their own governorship in Khurāsān. The annals of 'Abbāsid governors in the first half of the third century show us many examples of sons succeeding fathers at the helm of various provinces3 and the fact that five generations of Ṭāhirids followed each other as rulers of Khurāsān shows simply that from the point of view of the caliphate they fulfilled their purpose without endangering the unity of the caliphate or creating undue trouble.1 The literary evidence is here fairly well confirmed by the numismatic. On the one hand the existence of copper coins with debased Sasanian busts2 shows the concern of some Ṭāhirids for the Sasanian heritage still alive in east Persia. But at the same time their coinage was not much different from that of many a governor. The name of the caliph is always mentioned and the few exceptions that are known are called "revolutionary" by G. C. Miles.3 Since they were struck by Ṭāhir ibn al-Ḥusayn in 206, they should probably be connected with his tentative separation from the caliphate mentioned in a previous note.4 In general, however, to consider Ṭāhirid coins separately as the expression of a new independent dynasty would lead us logically to consider any coin bearing the name of an officially appointed governor as the expression of the same independence. This does not mean denying the importance of the Ṭāhirids in the growth of independent movements in Persia. It merely suggests that, from the point of view of the official expression of sovereignty, there is no justification in calling the Ṭāhirids an independent dynasty. The whole situation was, of course, to be altered in Persia with the advent of the Ṣaffārids and Sāmānids, just as it had been altered in the West with the conquering dynasties of the Umayyads of Spain and of the Idrīsids of Morocco.

Thus in comparing the Ṭūlūnids with the other two "dynasties" of appointed amīrs of the third century A. H., we can see that there was a definite difference between what happened in Ifrīqiyah, Miṣr, and Khurāsān. In North Africa, the Aghlabids, officially appointed by the caliphs before taking things into their own hands, soon dropped the name of the caliph from their coins but maintained the principle of their appurtenance to the Islamic koiné through regular gifts and through sporadic announcements of their activities. The caliphs, probably largely ignorant of what was happening in their far-flung province, now and then showed some interest and some initiative in Aghlabid affairs. The Aghlabids, therefore, come as close to being independent as is possible within the framework of sunni Islam. The Ṭāhirids of Khurāsān would really be at the other extreme. This "iranisierte Familie arabischen Ursprungs"1 belongs in fact to the category of hereditary top civil and military servants of the state who have taken to heart the interests of one part of the empire. Many examples could be found of similar occurrences in other civilizations and at other times. They were identified with a specific cultural area of the Islamic world, but, with a unique and short-lived exception, they did not sever their allegiance to the caliphate, either in fact or in their coinage.

The case of the Ṭūlūnids, and especially of Aḥmad Ibn Ṭūlūn, belongs somewhere between these two extremes. Aḥmad was officially appointed in Egypt, but only in a junior capacity. Through his abilities and his intrigues, to which must be added the Zanj war that immobilized the whole strength of Iraq, he rose in power. Official investments were only given to him piecemeal and his assumption of the title of amīr together with the striking of coins bearing his own name occurred only when, rightly or wrongly, he felt that it was his legal and religious duty to save the caliphate, and that he was the only one who could do so. Such questions as his personal ambitions and the justification of his assumptions do not matter on this score,1 since in numismatic and epigraphical analyses we are dealing essentially with the formal rationale of man’s actions; however, it should be noted that Aḥmad’s constant concern with religious forms, the Holy War and religious justification of his actions could not have been very hypocritical, at a time when power politics of the roughest kind were shaking the Muslim world.2 His coinage retained all the official notations expected in the realm controlled by the caliphs. Its importance, therefore, resides not so much in what appeared on it as in the timing of its appearance. Thus, while on the one hand the constant focusing of his attention on the affairs of the caliphate brings him closer to the situation of the Ṭāhirids (especially the earlier ones), his more or less self-imposed isolation from the actual power in Iraq led him to a degree of independence which belonged rather to the Aghlabid type. The ambiguity of the Ṭūlūnid phenomenon in Egypt will appear most strikingly in the time of Aḥmad’s son and successor, Khumārawayh.

A last point must be made about the coinage of Aḥmad ibn Ṭūlūn. Aside from Miṣr, only two mints are known, al-Rāfiqah and Dimishq.1 Dimishq appears only twice, in 269 and in 270, and, although generalizations cannot be made on the basis of unique coins, it may be suggested that the appearance of the Damascus mint must be related to the meeting of 269 and to its consequent full assertion of ibn Ṭūlūn’s power. As to the al-Rāfiqah mint, of which we have many examples for 267, 268, and 270 (one coin), there is no doubt that the choice of the easternmost outpost of ibn Ṭūlūn’s possessions must be connected with his claim on the person of the caliph and with his opposition to al-Muwaffaq. The year 267 was perhaps the one in which the flight of al-Mu‘tamid was planned and the sending to Diyār Muḍar of Lu’lu’, ibn Ṭūlūn’s most important lieutenant,2 may have been connected with these plans. Lu’lu’'s importance and comparative independence from ibn Ṭūlūn is shown by the addition of his name to the coins of 268. In spite of Lu’lu''s defection and of the failure of al-Mu‘tamid’s escape,3 and although at that time ibn Ṭūlūn’s control over a part of the Jazīrah was not secure, the striking of al-Rāfiqah coins was continued in 269 and 270. Ibn Ṭūlūn was still intent on asserting his power as close as possible to the caliphate. And perhaps the more political or strategic character of the al-Rāfiqah coins could find a further argument in the fact that with one known exception4 all the coins whose weight is published are definitely lighter and therefore more debased than the usual Miṣr dinars.1 As far as the latter are concerned, not only were they kept at a consistent weight and gold standard, but it was part of the policy of Ibn Ṭūlūn to maintain a high standard of currency. In a story transmitted by Ibn Sa'īd, it is said that for a king the purity of the gold used in his dīnārs means the purity of his life and afterlife.2 It is added that the best dīnārs are of the Mu'taṣimi and Sindi types, and that Ibn Ṭūlūn did his best to reach the latter’s level.3

End Notes

1 Ṭabari, Annales, ed. M. de Goeje and others (Leyden, 1879–1901), III, 1697.
1 Ibn Sa'īd, p. 11; al-Kindi, p. 216.
1 Cf. the discussion in Hassan, pp. 51–2. Some sources claim that in that year Ja'far was made overlord of all the western provinces. Cf. below pp. 42 ff.
1 E. de Zambaur, “Contributions à la Numismatique Orientale I,” Numismatische Zeitschrift, XXXVI (1904), pp. 74ff.
1 Paris, I, no. 1020.
1 Miles, Tarsus, p. 302. In this catalogue, no. 59.
1 W. Wroth, Catalogue of the Imperial Byzantine Coins in the British Museum (London, 1908), II, pp. 661–3 for a list of such marks; pls. XLV, 11–12, LI, 4–7 for marks of the ninth century, which could have been copied and simplified by the Muslims.
1 Ibn al-Athīr, VI, pp. 14–15. For more on the Holy War and ibn Ṭūlūn, see below.
1 Cf., for instance, B.M., I, index of names and passim. There is a basic difference between the coins which bear the name of a governor after mimmā amara and those which have merely a name under the mention of the caliph. In the first case we are dealing with the expression of a direct executive order or of an official prerogative. The second type reflects rather the actual importance of the governor of a certain province, without necessarily implying the same specific prerogative of striking coins.
1 To deduce from that that "c’est bien en 260 H. qu’il faut dater la première indépendance de l’Egypte en période d’Islam" (Jungfleisch, p. 8), seems to me farfetched. These were years when ibn Ṭūlūn’s hands were full with the pacification of Egypt and there is no valid indication that there occurred any change in the status he had in 258.
1 Cf. also ibn Sa'īd's description, p. 16, of the reasons for which he demanded the control over finances; Hassan, p. 51.
1 RIC, nos. 154–5.
1 The financial-religious allusions found in this inscription have been fully discussed by G. Wiet, MCIA I Egypte, vol. 2 in MIFAO, vol. 52 (Cairo, 1930), pp. 80–1.
1 Not of Misr, as Hassan writes, p. 52. The error, which consists in the omission of a dot, seems to have already been made by some editors of ibn al-Athīr, V, p. 365.
1 While in the east coins and ṭirāz inscriptions mention al-Muwaffaq (B.M., nos. 352, 355, 356; Répertoire, no. 753), in the west the name of Ja'far, later under his laqab al-Mufawwaḍ ila Allāh, is usual (B.M., I, nos., 353, 358, 366, 374).
1 Ibn Sa'īd, p. 19; Ṭabari, III, p. 1890; Maqrīzi, II, p. 178.
1 Ibn Sa'īd, pp. 21 ff., especially p. 22; Maqrīzi, II, p. 179. This letter, which is more clearly preserved in Maqrīzi than in Ibn Sa'īd, has been well paraphrased in Hassan, pp. 60–61.
1 Ibn Sa'īd, p. 55; Kindi, p. 219. In view of the preceding evidence I would be less inclined than Hassan, p. 65, or Becker, p. 178, to be overly suspicious of ibn Ṭūlūn's purpose in moving into Syria. The story related by Nuwayri in Hassan, pp. 66–7 and whose text is given in Taco Roorda, Abul Abbasi Amedis .... Vita et Res Gestae (Leyden, 1825), p. 81, shows quite clearly that, although the reception given to Ibn Ṭūlūn in Tarsus was far from being friendly, he used a trick which must have been heartbreaking for a general (let his forces retreat without fighting in front of the inhabitants of Tarsus) in order to demonstrate to the Byzantines the great strength of the city. Cf. also Mas'ūdi, Les Prairies d'Or, ed. C. A. Barbier de Meynard and A. Pavet de Courteille (Paris, 1861–77), VIII, p. 67, where the word ghāzw is specifically used and where it is added that ibn Ṭūlūn’s army was followed by volunteers (muṭawwa'ah) from Egypt and Syria. Such expeditions were constantly taking place on a smaller scale, sometimes led by ibn Ṭūlūn's lieutenants, Ṭabari, III, 1952. All texts on Tarsus are conveniently gathered in H. Grégoire et M. Canard, La Dynastie Macédonienne, in A. A. Vasiliev, Byzance et les Arabes, II (Bruxelles, 1950), pp. 4–23. But it is only rarely that one can witness an expedition as large as the one set up by Ibn Ṭūlūn.
1 Ibid., pp. 81–4.
1 Note that ibn Ḥajar al-Asqalāni seems to admit that al-Mu'tamid had sent a letter dismissing al-Muwaffaq from the succession, al-Kindi, p. 512. This statement, which is not confirmed by any other text, is one of many which tends to confuse the exact sequence of events in the five years before 269.
1 The point whether the first or the second of these hypotheses is the correct one is not germane to this study, since it proposes essentially to give the theoretical framework within which Ibn Ṭūlūn acted. However, in the argument over whether Aḥmad or al-Muwaffaq begun the quarrel between the two men, Hassan appear to me to be right (p. 88) as against Becker (pp. 177–8), in believing that the Iraqi was responsible.
1 Gardet, Cité, pp. 178ff.
1 Taghribirdi, III, p. 3.
1 G. Marçais, La Berbérie et l'Orient au Moyen Age (Paris, 1946), p. 59; see, in general, pp. 57–63 for a lucid statement of the respective positions of the Aghlabids and of the caliphate.
1 See the examples given by Marçais and in particular the curious 250 inscription in the mosque of Tunis, where the name of the caliph is mentioned, but not that of the Aghlabid; cf., lately, G. Marçais, L’Architecture Musulmane d'Occident (Paris, 1954), p. 7.
1 Cf. ibn al-Athīr, V, pp. 196–7; Ṭabari, III, p. 1054; B. Spuler, Iran in frūh-islamischer Zeit (Wiesbaden, 1952), p. 59; and pp. 320–1 for a discussion of the Ṭāhirids’ relationship to the caliphate and reliance on the older Sasanian tradition.
1 The only exception to this would be the curious and sudden dropping of the name of the caliph in the khuṭbah of 207 by Ṭāhir ibn al-Ḥusayn. His death, which followed the event quite closely, did not permit him to pursue whatever course he had in mind. Cf. al-Ya‘qūbi, II, p. 556; Ṭabari, III, p. 1064; ibn al-Athīr, V, p. 204. In view of what followed the importance of this event seems to me to have been overestimated by V. Bartold, Turkestan down to the Mongol Conquests (London, 1928), p. 208, and by many others.
1 Spuler, p. 59.
1 The whole psychological problem of Aḥmad Ibn Ṭūlūn has been admirably analyzed by Becker, and, with minor exceptions, the coinage confirms his analysis, but it may be wondered whether his diagnosis of a pure power struggle between al-Muwaffaq and ibn Ṭūlūn is not oblivious of the tremendous power of religious motivations in the mediaeval world.
1 In a forthcoming work G. C. Miles attributes, on epigraphical grounds, a series of copper coins from Ashmūn to the Ṭūlūnids.
2 Ibn Sa'īd, p. 7. Maqrīzi, Khiṭaṭ (Cairo, 1270 A.H.), I, 314; in G. Wiet’s edition, in Mémoires de l’Institut Français d’Archéologie Orientale, vol. 53 (Cairo, 1927), pp. 144ff.; translation by P. Casanova, in MIFAO, vol. II (Cairo, 1906), p. 208. On the exact meaning of the word qaṣbah, cf. Becker, p. 160, “nicht die Hauptstadt, sondern das Hauptland im Gegensatz zu den See- und Grenzprovinzen." On the whole system of lieutenants in the provinces cf. Makrīzi, Wiet ed., pp. 145–6.
2 Ibn Sa'īd, pp. 16–17; al-Kindi, p. 217. There is a curious discrepancy between al-Kindi’s statement that Aḥmad had received the governorship of the marshes and ibn al-Dāyah’s that it was only the financial control. Al-Kindi’s statement appears more convincing, since it is followed by a relation of the difficulties found by ibn Ṭūlūn in getting a representative in Tarsus, whose functions were to be political. It should also be noted that ibn al-Dāyah was more interested in anecdotes and that his facts (or is it ibn Sa'īd's relation of them?) are not always too precise; cf. Becker, p. 152. Whatever happened, whether Aḥmad’s control was purely political or financial, or both, it is an odd assignment if one considers that he did not control Syria at the time. It can only be properly understood if one admits that the war against the infidels was one of the foremost aims in ibn Ṭūlūn’s mind, as is suggested by a number of sources, ibn al-Athīr, al-Kāmil (Cairo, 1353 A.H.), VI, pp. 14–5 and ibn al-Maḥāsin Yūsuf ibn Taghribirdi, al-Nujūm al-Zāhirah (Cairo, 1929), III, p. 5.
2 Ibn Khaldūn, 'Ibar (Cairo, 1284), IV, 298. See also ibn al-Athīr, V, p. 367 and Taghribirdi, III, p. 6, where ibn Ṭūlūn is made independent at the death of Bākbāk, which is contradicted by all the other sources and by Taghribirdi himself, who writes, p. 7, that ibn Ṭūlūn did not control the finances as yet. Rogers, p. 6, claims that, at the death of Yārjūkh, ibn Ṭūlūn became independent by succeeding to all of the latter’s functions. This is not brought out by the sources.
2 "Contributions à la Numismatique Orientale II,” ibid., XXXVII (1905), pp. 194–5.
2 RIC, nos. 154–5.
2 Paris, I, no. 1666.
2 Cf. especially the Umayyad period and later, the Urtuqid and Seljuq periods in Asia Minor. It may be noted that some of the coins found by the Tarsus expedition show as early as in the end of the third century A. H. features derived from local traditions, Miles, Tarsus, pp. 310–11.
2 Répertoire, II, nos. 646, 656, 667.
2 See below.
2 Répertoire, no. 685.
2 For a similar idea expressed through Qur’ānic quotations in the Umayyad period, cf. my forthcoming analysis of the Dome of the Rock in Ars Orientalis, vol. III.
2 Ṭabari, III, pp. 1859–60.
2 Note the existence of a 261 coin with the name of Ja'far, B.M., I, no. 374, which probably belongs to an early issue of that year. Cf. also Paris, I, nos. 261–2.
2 Ibn Sa'īd, p. 24.
2 Kindi, p. 220.
2 It is on religious grounds that he refused to follow the advice of those who did not deem it wise to fight against al-Muwaffaq, Ibn Sa'īd, p. 70.
2 al-Māwardi, al-Aḥkam al-Sultāniyah, ed. Maḥmūd Aly Ṣubayḥ (Cairo, no date), p. 18; tr. E. Fagnan (Algiers, 1915), pp. 38–41. Cf. also ibn Khaldūn, Prolegomènes, tr. de Slane (Paris, 1863–68), I, p. 394. Al-Māwardi’s statement is interpreted in a way similar to ours, but referring to a later period, by H. A. R. Gibb, "Al-Māwardi’s theory of the khilāfah," Islamic Culture, XI (1937), pp. 297 ff.
2 Cf. the texts and the commentary in Hassan, pp. 87–8 and 260ff.
2 Ibid., p. 29.
2 Hassan, p. 222.
2 Cf., however, the use of the word aqarra "to confirm," in the cases of al-Ma’mūn and al-'Amīn with respect to Ibrahīm and Abdallāh, in ibn Idhāri, pp. 94 and 95. The word does not occur for any later prince.
2 For instance, B.M., II, nos. 188–217.
2 Ṭabari, III, pp. 1065, 1102, 1339, 1505–6, 1881–2.
2 B.M., II, no. 240.
2 Ibn Ṭūlūn was not the only one who used religious arguments. In a panegyric written for the caliph al-Mu‘taḍid a number of years later, the Ṭūlūnid is called "a second Pharaoh of Egypt, who defied God and obeyed Satan." C. Lang, "Mu'taḍid als Prinz und Regent, ein historisches Heldengedicht von ibn al-Mu‘tazz," Zeitschrift der Deutschen Morgenländischen Gesellschaft, XL and XLI (1886–7), verses 35–36.
2 Ibn Sa’īd, pp. 67–8.
3 It may be noted in passing that the latter was Sāliḥ ibn Aḥmad ibn Ḥanbal, the well-known son of the founder of the fourth major Sunni school of theology. Cf. a few words on him in H. Laoust, art. “Aḥmad ibn Ḥanbal,” in the new edition of the Encyclopedia of Islam.
3 F. Soret, ‘‘Lettre à M. Lelewel ...,” Revue de Numismatique Beige, X (1854), p. 20; and “Lettre à M. le Conseiller d’Etat de Dorn,” ibid., XII (1856), p. 132. Pl. II, fig. 11 (a drawing), in the latter of these works, shows the name Aḥmad more clearly than on the other coin. I believe, however, that we are dealing with the very same motive and that the difference is due to the state of the coin or to a defective die. Cf. also Paris, III, no. 14, where the same mistake is repeated.
3 RIC, nos. 392–3.
3 Khedivial, no. 619. This is actually a problematic coin. The mint has been obliterated. The date 261 seems to be certain. It would be tempting to attribute this coin to Miṣr, since it is only on Egyptian coins that we meet with the odd signs described below. On the other hand, the numerous coins known between 258 and 261 do not possess the sign, and one might wonder why it suddenly reappeared. Properly speaking this coin does not belong to the group of 258 coins. It is simply mentioned here, because it may eventually be of use in finding the solution to the problem of this whole group of coins.
3 In RNB, 1854, P. 22.
3 Ibn Hawqal, Kitāb, ṣūrah al-arḍ, ed. J. H. Kramers in Bibliotheca Geographorum Arabicorum, vol. II (2nd ed., Leyden, 1939), pp. 183–4. Ibn al-Jawzi, al-Muntaẓam (Hayderbad, 1357), vol. V, part 2, pp. 71–74, also emphasizes that Aḥmad was interested in religious matters and that he went to Tarsus in order to fight the infidels.
3 Ibn Sa'īd, p. 16.
3 Hassan, pp. 30–31.
3 Ibid., nos. 699 and 702.
3 Tiesenhausen, no. 1998; H. Nützel, Katalog der Orientalischen Münzen I (Berlin, 1898), no. 1553.
3 Becker, pp. 162–3. A tradition existed, Ya'qūbi, II, 624, to the effect that it is the son of al-Muwaffaq, the future caliph al-Mu'taḍid, who took over Yārjūkh’s appointment. Becker has shown that this is most unlikely, but the error shows that, even to contemporaries, the events concerning the succession of al-Mu'tamid and the legal rule of the empire were highly confusing.
3 Hassan, p. 77.
3 Note the usage of the word ghalaba by Mas'ūdi when talking about al-Muwaffaq, VIII, p. 67.
3 Kindi, p. 226; cf. also pp. 512ff., where ibn Ḥajar’s stories are given.
3 Cf. above p. 34, note 3. See also Ṭabari, III, 1930–1, for another personage who settled in a ribāṭ; ibid., III, 2193, for a curious system of electing chieftains; also Isṭakhri in Bibl. Geogr. Arab., I (Leyden, 1927), p. 55.
3 Ibn al-Athīr, VI, p. 56 (under year 270).
3 al-Nuwayri in Ibn Khaldūn, Histoire des Berbères, tr. de Slane and Casanova, I (Paris, 1925), p. 413.
3 Ibid., I, nos. 255, 256, 267, among many other examples.
3 E. de Zambaur, Manuel de Généalogie (Hanovre, 1927), passim.
3 G. C. Miles, The Numismatic History of Rayy (New York City, 1938), nos. 108 B and C, and commentary pp. 110–111. See also the note of Lane-Poole in Khedivial, p. 80, where it is pointed out that a great number of coins were issued in places which were definitely under Ṭāhirid control without mentioning the name of the Ṭāhirids. On this subject see also E. de Zambaur, "Contributions II," Numismatische Zeitschrift, XXXVII (1905), pp. 119ff., and especially pp. 125ft.
3 Tabari, III, 2037–8.
4 Miles, Tarsus, pp. 297 ff.
4 E. T. Rogers, “Notice on the dinars of the Abbaside dynasty,” Journal of the Royal Asiatic Society, n.s., VII (1875), p. 283.
4 Hassan, p. 122.
4 M. Jungfleisch, "Un poids et une estampille en verre," Bulletin de l'Institut d’Egypte, XXX (1949), pp. 4ff.
4 Ibid., no. 682.
4 Ibn Sa'īd, p. 19. For another statement without date, see al-Ḥalabi in F. Wüstenfeld, Die Statthalter von Aegypten zur Zeit der Chalifen (Göttingen, 1875), III, p. 58.
4 Ṭabari, III, p. 1873, mentions the fact that Ja'far attended Yārjūkh’s funeral. His presence there may perhaps be a sign of his relation to the deceased or to his former positions.
4 See the story given by Hassan, pp. 87–8.
4 Note that as important a religious leader as the son of inbn Ḥanbal was qāḍi in the thughūr al-sha’miyah; cf. above p. 29, note 3. Note also that this appears to be the period when certain themes of the "frontier legend" were established; see R. Goossens, "Autour de Digénis Akritas", Byzantion, VII (1932), and esp. M. Canard, "Delhemma," Byzantion, X (1935), pp. 283–300.
4 Cf. below.
4 Ibid., pp. 431–2.
4 Cf. the words of Nuwayri in ibn Khaldūn, pp. 397–8.
4 Cf. above, note 1.
4 Sumer, III, p. 279.
5 255: RIC, no. 150; 256: no known coin; 257: RIC, nos. 152–3; 258: Paris, I, no. 1020 and RIC, nos. 154–5 (on these see below pp. 38–9); 259: Rogers, no. 4, Khedivial, no. 616, Markoff, p. 345, no. 1; 260: Rogers, nos. 5 and 6, UM, etc....; 261: Khedivial, no. 619 (there is a problem connected with this coin, cf. below p. 32); 262: no known coin; 263: Rogers, no. 7 and 8, Markoff, p. 345, no. 2, Khedivial, no. 618; 264: Markoff, add., p. 928, no. 2a, Z. für N. (A. Erman, „Die im Jahre 1882... erworbene... Münzen,” Zeitschrift für Numismatik, XI, 1884), p. 64 (neither of these coins is described); 265: RIC, no. 291.
5 There had been a Ṭūlūnid governor in Tarsus before 266, al-Kindi, p. 217, and, considering the shaky nature of Aḥmad's control over the thughūr, the name of the governor was more likely to have been used before 266 than ibn Ṭūlūn’s.
5 On this formula see below p. 39 note 5.
5 The question here is to know whether, at this time, the expression mawla amīr al-mu'minīn is the statement of a certain personal relation between the two men or whether it is an official title. In a series of passages Max van Berchem has brilliantly shown that all titles in amīr al-mu'minīn imply a certain share in the sovereign power of the caliph, a "partage du pouvoir." See, for instance, Matériaux pour un Corpus Inscriptionum Arabicarum I Egypte, in Mémoires de la Mission Archéologique Française au Caire, vol. XIX (Cairo, 1903), pp. 81 ff.; or "Eine arabische Inschrift aus dem Ostjordanlandes," Zeitschrift des Deutschen Palästina Vereins, XVI (1893). In his commentary on the inscription of the mosque of ibn Ṭūlūn, MCIA, p. 29, van Berchem assumed that already at that time the expression was a title with the same implication as that found on later inscriptions. At the same time it can be open to doubt whether the formula, when used after the names of governors and financial administrators throughout the Umayyad and ‘Abbāsid periods on stamps and weights, did mean much more than a personal relationship to the caliph. Cf. G. C. Miles, Early Arabic Glass Weights and Stamps (New York City, 1948), index under mawla amīr al-mu'minīn. It seems to me that in the third century A.H. most titles were not yet formalized and that, in the case of Aḥmad ibn Ṭūlūn one should still consider the formula as the expression of a personal relationship between the caliph and his subordinate, perhaps a delegation of authority, but certainly not a sharing of it, since the formula was used too often after the name of comparatively minor officials.The same problem can be raised in the case of the word amīr, which refers either to a function or to a title. A. Mez, The Renaissance of Islam, Eng. tr. (London, 1937), pp. 81 ff., defines the amīr as the military commander, to whom was joined an 'āmil, whose responsibility it was to deal with financial problems. While perhaps true in the tenth century, this is not so in the case of ibn Ṭūlūn. It is clear from al-Ya‘qūbi, Historiae, ed. T. Houtsma (Leyden, 1883), II, p. 620, that he was the 'āmil of Miṣr, while ibn al-Mudabbir was controlling the finances. In the examples used by Mez amīr refers really to a function. Similarly, when in his letter to al-Muwaffaq (see below p. 45) ibn Ṭūlūn refers to the latter as amīr, he no doubt meant the function of the caliph's brother as commander of the army, since the latter was only appointed as wāli al-mashriq or wāli al-'ahd (Ṭābari, III, 1890). Similarly ibn Ṭūlūn had been appointed as wāli or 'āmil, not amīr. And yet, when on the 259 glass or on the mosque inscription we meet with the expression al-amīr Aḥmad ibn Ṭūlūn, there is little doubt that a rank or a title are meant as much as a function. There seems to be a distinction, in the third century A.H., between titles which were given and titles which were assumed. Either Aḥmad ibn Ṭūlūn liked to be known essentially as a military leader, or we are dealing here with the first steps of a formalization of titles which will be more completely achieved in the following centuries, but the whole question of the development of titles would warrant a more complete study. Cf. just recently A. A. Duri, arts. 'Āmil and Amīr in new ed. of the Encyclopedia of Islam.
5 Taghribirdi, III, p. 24. But ibid., p. 33, he repeats the same story for the year 261.
5 That the Holy War was given some consideration in the center of the empire, even when the caliphate was not at its highest, is shown by Ṭabari, III, 1481–85. On the jihād as a state institution rather than an individual one, see M. Khadduri, War and Peace in the Law of Islam (Baltimore, 1955), pp. 60ff.
5 See J. Schacht, The Origins of Muhammadan Jurisprudence (Oxford, 1950), pp. 283 ff., 317–8, and passim.
5 J. Farrugia di Candia, "Monnaies Aghlabites du Musée du Bardo," Revue Tunisienne, vol. VI (1935), p. 272; also vol. VII (1936), p. 179.
5 Marçais, Berbèrie, loc. cit.; Māwardi, tr. Fagnan, pp. 59ff. Here again the question can be raised whether in the third century A.H. we are already dealing with a formalized legal theory of the amirate or whether cases like that of the Aghlabids prompted the formulation of the theory.
6 Ṭabari, III, p. 1890.
6 It would seem that the word amīr refers to a function here rather than to a title, ibn Idhāri, Kitāb al-Bayān, ed. G. S. Colin and E. Levi-Provençal, vol. I (Leyden, 1948), p. 92. Cf. also above p. 39, note 5.
6 Marçais, pp. 60 and 62.
7 Rogers, no. 3.
8 B.M., I, no. 358.

2. The coinage of Khumārawayh ibn Aḥmad and late Ṭūlūnid coinage

A discussion of the coinage of Khumārawayh ibn Aḥmad Ibn Ṭūlūn may best be attempted from two different points of view. First, there is an historical problem, to the solution of which the numismatic evidence may be of some help. Second, there is a problem posed by the coins themselves in the state in which they have been preserved. It must be borne in mind, however, that, except for the earliest period of Khumārawayh’s rule, the history of the Ṭūlūnid regime in Egypt after the death of Aḥmad is far less interesting and far less eventful than the few years which saw the establishment of the dynasty in Fusṭāṭ, and the coinage does not pose any major problem.

Khumārawayh ibn Aḥmad succeeded his father without any difficulty, at the age of twenty.1 The army and the civil government of Egypt accepted him as their leader.2 But his position was not confirmed by the caliphate, in whose eyes Isḥāq ibn Kundāj was the legal governor of Egypt. Therefore Khumārawayh’s first task was to be recognized by the supreme power of Islam. There is no evidence to suggest that he tried in any way to maintain the positive moral and legal position of his father. Neither his character nor his training had prepared him for such an attitude. His sole aim was to stay in power. But this did not mean that he would meekly accept the dictates of al-Muwaffaq. After a series of military campaigns (on which more will be said below), in 273 a political and perhaps financial arrangement was reached with al-Muwaffaq. The latter was again mentioned in the khuṭbah, which suggests that Khumārawayh had maintained for three years the situation which existed at the time of Aḥmad’s death, including the cursing of al-Muwaffaq.3 The political agreement between Egypt and the caliphate is important in that it closed the chapter opened ten years earlier with Aḥmad’s refusal to send more money to al-Muwaffaq than was due to him. The politico-religious attempt of Aḥmad’s had failed, since al-Muwaffaq was still in power. But, at the same time, the so-to-speak illegitimate succession of Khumārawayh to his father’s position was recognized by the caliphate, largely because the military expedition sent to crush him had failed. No information is given on the specific financial arrangements made at that time, although there probably was some agreement on that matter; but the point is established that Khumārawayh and his successors (wulduhu) receive the governorship (wilāyah) of Egypt and of the Syrias (al-sha'māt, meaning Palestine, Syria proper, the thughūr, and, almost certainly, the right bank of the Euphrates) for thirty years.

The importance of this agreement and of the events which preceded it consists in that it emphasizes two points: the absolute necessity for a working arrangement with the caliphate;1 and the fact that the caliphate neither recognized the independence of the Ṭūlūnids nor accepted without reservations the fact of a dynasty. By setting a time limit al-Muwaffaq acknowledged the existence of the Ṭūlūnids,2 but refused to farm Egypt out to them permanently and thereby admit the theory of dynastic succession outside the caliphate in the Muslim community. This agreement, no doubt, represents a step toward independence when compared with the regular appointments of Ṭāhirid governors, but it is not yet on a par with the situation of later centuries, when the dynastic sense increased and the relationship to the caliphate became more tenuous. In 279, after another series of military actions, in which al-Muwaffaq may not have been directly involved, a new agreement was reached with al-Mu‘taḍid, the new caliph and former commander of the armies which had attacked Khumārawayh in 271. Actually this new agreement confirmed and clarified the first one. The area to be controlled by the Ṭūlūnids was defined as extending from the Euphrates to Barqah; administrative and judiciary autonomy was accorded to them; and the tribute was fixed permanently.1 This agreement was maintained until 286, when the Ṭūlūnid Hārūn ibn Khumārawayh lost his northern possessions, was compelled to increase the tribute, and finally seems to have been forced to accept caliphal supervision in his administration.2 In 292 Egypt became again a province directly administered by the caliphs.

The coinage of Khumārawayh and of his successors confirms this state of affairs. From 270 on until the end of the dynasty it reflects quite faithfully the situation of the caliphate. Al-Mu‘tamid and al-Mufawwaḍ continue to appear on coins. al-Muwaffaq does not, and it should be presumed, although for obvious tactical reasons the question does not appear to have been raised during the discussions, that the Egyptians still maintained the fiction of the division of the empire. al-Muwaffaq died in Ṣafar 278 and was succeeded by his son al-Mu'taḍid in his position of second in the line of succession.3 Exactly a year later al-Mu‘taḍid edged out his cousin al-Mufawwaḍ and became the only official wāli al-'ahd.1 In Rajab of the same year 279 al-Mu‘tamid died and al-Mu'taḍid became caliph.2 Two coins struck in Miṣr and preserved in the University Museum and American Numismatic Society collections (no. 52) belong to the early part of 279, since they have al-Mu'taḍid as heir apparent, but a number of coins of the same year already have him as caliph. They must have been struck shortly after he became caliph and confirm the fact that it was customary to strike coins as soon as a new man attained supreme power. The same thing occurred at the death of al-Mu'taḍid in Rabī' II 289.3

Another body of epigraphical evidence parallels that of the coins and can help in defining the relationship between the Egyptian dynasty and the caliphate. It consists of the ṭirāz inscriptions, of which a comparatively large number survives from the period under consideration. Two inscriptions are particularly interesting since they are dated in 272, that is before the agreement was reached which recognized Khumārawayh as governor of Egypt.4 One was made in Tinnīs, and contains not only the name of the amīr Khumārawayh, mawla amīr al-mu’minīn, but also that of the caliph and of al-Mufawwaḍ, curiously enough here simply called Ja'far. This inscription confirms the evidence of both coins and texts, showing that Khumārawayh preserved the situation as it was at the time of Aḥmad’s death in spite of the illegality of his situation. The other ṭirāz inscription, made this time in Alexandria,5 bears the name of the caliph only. In other words, during the time when Khumārawayh was actually fighting the forces of the caliph, textiles were officially made in Egypt for the caliphate without mention of the Egyptian ruler’s name.

The existence, after 273, of well over thirty published ṭirāz inscriptions from Egypt during the Ṭūlūnid period can help in providing an explanation for this anomaly. Of these inscriptions all those made in Tinnīs, except one, and the only one known from Damietta1 bear the name of the Ṭūlūnid ruler, while all those made in Alexandria or Miṣr bear only the name of the caliph.2 It will be recalled that Aḥmad Ibn Ṭūlūn had received the right to deal with ṭirāz matters.3 It may be assumed that Khumārawayh maintained the practice established by his father and that it was accepted by the caliphate after 273 or 279, although there is no mention of it in the texts. But what is important to us is that at no time4 did the Ṭūlūnids control all the textile factories of Egypt. Serjeant, in his study of texts on Islamic textiles, has suggested that Tinnīs and Damietta were centers where Christians from Egypt worked and where a Coptic type of textile was made.5 If his conclusions are correct, the Ṭūlūnids only had power over the purely local, Coptic factories; but throughout their history the cloth manufacture in the two largest cities of Egypt, Fusṭāṭ6 and Alexandria, was not their direct concern but that of the far away caliphs. The implication of this state of affairs in the organization of textile manufacturing, a state concern almost on the same level as the striking of coins,1 does not concern us here,2 but these examples are important in illustrating once more the curious dichotomy of mediaeval Islamic civilization, not only a dichotomy between practice and theory which has often been discussed,3 but a dichotomy between levels of legal authority. With textiles, just as with coins, it was essential to preserve the fiction of an effective caliphal control. And, thus, just as in Islamic art a series of unifying factors tie together the tremendous variety of themes and ideas which found their way into mediaeval Near Eastern culture,4 the more or less self-sufficient political centers (outside of the remote areas of North Africa and Spain) felt bound to accept symbols of a politico-religious authority which at times was not capable of raising an army to defend itself.

There is yet another historical problem where the coins can be of help. Between 270 and 279, as has been mentioned, Khumārawayh was involved in a series of military operations whose purpose was to consolidate his position in Syria. These operations are described in the texts, but since they were numerous and as they involved always the same people, many a chronicler appears to have been confused about their order and their exact dates. In the case of one mint the remaining coins can be used to determine the succession of the campaigns. The case is that of the coins struck in al-Rāfiqah. They are known for every year between 270 and 279. After 279 they still appear, but they are of no importance to us here since they are purely ‘Abbāsid,1 reflecting probably the fact that, as was previously mentioned, the 279 agreement limited Khumārawayh’s possessions to the areas west of the Euphrates and presumably did not include the Diyār Muḍar in its totality, if at all. It is the period between 270 and 279 which interests us here and it is the most eventful one in terms of military expeditions. Among the coins the name of Khumārawayh appears in 270, 273, 274, 275, 276, 277, 278, and 279. But, in 270, 271, 272, 274, and 275, coins were also struck in al-Rāfiqah which do not bear the Ṭūlūnid’s name and should therefore be considered as purely ‘Abbāsid.2 The question is whether this body of coins can help us determine with some degree of certitude the exact succession of events.

It will be recalled that in 269 Isḥāq ibn Kundāj had been appointed by al-Muwaffaq as governor of all of ibn Ṭūlūn’s possessions.1 This man, whose lack of military ability (I believe that he lost just about every battle in which he was involved during the period under consideration) was only matched by his pugnacious interference wherever he thought he might gain some advantage, will become the main actor in the subsequent fighting. At the time of ibn Ṭūlūn’s death he was in Mosul. The news of ibn Ṭūlūn’s death traveled fast and, while one of the first acts of the new ruler of Egypt was to strike coins with his own name in both Miṣr and al-Rāfiqah, ibn Kundāj moved towards Syria, underestimating, as ibn al-Athīr specifically says, the power of Aḥmad’s son.2 There is no doubt that he went first to al-Rāfiqah, since there is a purely ‘Abbāsid coin from al-Rāfiqah in 270 and since Ṭabari states that in 270 a fight occurred there between ibn Kundāj and the Ṭūlūnid representative.3

In the meantime, ibn Kundāj had got in touch with the caliphate, which sent an army under abu al-‘Abbās Aḥmad ibn al-Muwaffaq, the future al-Mu'taḍid. This army moved along the Euphrates, then crossed over to Qinnasrīn, and, finally, went south to Palestine where, between Jerusalem and Ramleh, it encountered the main Ṭūlūnid force at the tragi-comic battle of al-Ṭawāḥīn.4 Al-Kindi puts the battle in Ṣafar 271, but this sequence of events poses a problem. According to the Egyptian chroniclers (al-Kindi, al-Maqrīzi, Taghribirdi), Aḥmad died on the tenth of Dhū al-qa'dah 270. Could it be that during the three months which elapsed between the day of his death and the battle of al-Ṭawāḥīn all this could have occurred: the news had traveled as far as Mosul; coins had been struck at al-Rāfiqah both by the new ruler of Egypt and by the invader; an army had been raised in Baghdad and Mosul and had reached southern Palestine? One should take into consideration here a tradition transmitted by Ṭabari alone1 to the effect that the news of Aḥmad’s death had reached Baghdad as early as Sha'bān 270. Tabari puts the battle of al-Ṭawāhīn in Shawwāl 271, which would give us some fourteen months between the two events. The dates of the Egyptian historians are definitely unacceptable. And, although one could conceivably adopt Tabari’s date for the death of Ibn Ṭūlūn and the Egyptian date for the battle, it is generally apparent that the chronology of events transmitted by Iraqi historians is more trustworthy than that given by the Egyptians. A difficulty would be that no coins from Miṣr are known for the year 270 with the name of Khumārawayh, an odd thing if four months had elapsed between the death of Aḥmad and the end of the year. Yet such coins may still be found and may even be included among the 27 specimens of our type 14, many of which are undescribed, inasmuch as there is a Rāfiqah coin struck in 270 with the prince’s name. One should, I think, adopt Ṭabari’s dates for both events.

That in 271 we have only an ‘Abbāsid coin from al-Rāfiqah is perfectly natural, since, although the Ṭūlūnids won the battle, they only pursued the enemy as far as Damascus which was properly and officially re-occupied. The events of 272 and 273 are more complex and the chronicles are highly confusing. We know practically nothing of what Khumārawayh was doing during most of the year 272. He appears to have stayed in Fusṭāṭ. He left in Dhū al-qa'dah for Syria, where he had his governor in Damascus murdered for insubordination in Muḥarram 273.2 At the same time a quarrel took place between ibn Kundāj and ibn abi al-Sāj, another roving ‘Abbāsid governor, half official, half adventurer, who was governor of the Euphrates area south of al-Rāfiqah.1 Ibn abi al-Sāj called Khumārawayh for help, subjecting himself and the area under his control to the Ṭūlūnid. Together they moved against ibn Kundāj, whom they defeated near Raqqah in Jumāda I 273, and pursued as far as Samarra.2 Ibn abi al-Sāj obtained from Khumārawayh all that had been conquered from ibn Kundāj. The expedition was followed by the agreement of 273, actually reached through the intermediary of ibn Kundāj. Territorially the agreement probably acknowledged the status quo, and thus can be explained the purely Ṭūlūnid coins of 273.3 A problem arises, however, about the existence of two Ṭūlūnid coins of 272 struck in al-Rāfiqah. Only two possible explanations for their existence can be found. One is that some local governor acknowledged Khumārawayh for a while without its being recorded in the literature. It could not have been ibn Kundāj and, according to ibn al-Athīr, ibn abi al-Sāj, who had been governor at Raqqah, only controlled Qinnasrīn at that time. But one may wonder whether ibn al-Athīr is correct in writing that ibn abi al-Sāj was in Qinnasrīn. In 269, according to Ṭabari and ibn al-Athīr, he was in Raḥbah, on the Euphrates below Raqqah, and had gone as far north as Qarqisiyā (at the mouth of the Khābur and the Euphrates),4 whence he had dislodged ibn Ṣafwān, a Ṭūlūnid governor.5 There is, so far as I have been able to discover, no information about his whereabouts between 269 and 273. But, when Ṭabari mentions the battle that took place between him and ibn Kundāj in 273,1 he places it at Raqqah, which would be perfectly logical if the man had still been in the middle Euphrates area. Ṭabari says nothing about his allegiance to Khumārawayh, but the 272 coin would indicate that he had already passed into the Ṭūlūnid camp. Thus, while ibn al-Athīr’s sequence of events is in all probability the correct one, his dating and his account of the areas in which the events occurred is confused. The 272 coin and the information given by Ṭabari permit a more secure understanding of what happened, where, and when.

A second series of military operations took place between 274 and 276. There were in fact two major operations.2 First ibn Kundāj, in 274, started a new offensive towards Syria. Khumārawayh opposed him brilliantly, crossed the Euphrates, and defeated him. Ibn Kundāj gave up all his possessions to the Ṭūlūnid and Khumārawayh returned to Egypt. Immediately afterwards ibn abi al-Sāj revolted. Khumārawayh defeated him alone near Damascus (in Muḥarram 275), then pursued him to the Jazīrah. There Khumārawayh received help from ibn Kundāj, who was, however, defeated. But Khumārawayh himself was victorious over ibn abi al-Sāj; and the latter, who had been in constant touch with al-Muwaffaq, withdrew to Baghdad, where he received a new appointment.3 These events are confirmed by the numismatic evidence, which definitely favors ibn al-Athīr’s account over that of the Egyptian chronicles. In 274 there appears a series of ‘Abbāsid issues, which correspond to ibn abi al-Sāj’s revolt. Since there is also a 275 ‘Abbāsid coin, it follows that ibn abi al-Sāj remained in Raqqah during the early part of that year, but the Ṭūlūnid issues of the same year illustrate Khumārawayh’s victory. From 275 until 279 only Ṭūlūnid coins are known from al-Rāfiqah. After that date, willingly or not, Khumārawayh lost control of all areas east of the Euphrates.1

While the coins provide an answer for a number of historical problems, in themselves they pose a major one to which historical texts should provide an answer. One of the striking features of Khumārawayh’s coinage is the sudden multiplication of the number of mints. While under Aḥmad only three are known (Miṣr, al-Rāfiqah, Dimishq), under his son five new ones appear (Ḥims, Filasṭīn, Ḥalab, Ḥarrān, and Anṭākiyah), and none of the older ones is abandoned. Furthermore these mints do not, so far as our present evidence goes, appear from the very beginning of his rule, but only from the year 276, with the exception of Ḥimṣ, which occurs already in 274. It is admittedly true that a numismatic discussion can never wisely be based on negative evidence, unless one deals with hoards or large numbers of coins, and it may very well be that future discoveries will complete or change the picture. I do not believe, however, that future discoveries will alter the picture too much, since it is between 273 and 276 that Khumārawayh had established himself firmly as ruler of Egypt and Syria, and it is only then that he was free to administer his provinces without interference.

The questions that are raised by the introduction of the new mints are essentially two: what led to the specific choice of mints? What was the function of coins struck in these mints in the economy and life of Ṭūlūnid controlled lands? The questions are really too vast to be dealt with entirely at this juncture, inasmuch as contemporary texts are of little help. Furthermore these are not problems which can be easily solved through the case of the Ṭūlūnids alone, but ones which should be started with studies of the Umayyad and early ‘Abbāsid periods in order to show the precedents which might have led to the situation under the Ṭūlūnids. Only a few remarks can be made here which may contribute towards a better understanding of the whole problem.

First it can be asked whether the Ṭūlūnid mints were deliberately located so that every province (jund) should have its own. Do the coins show something about the administration of the land? Miṣr, Filaṣtīn, Ḥimṣ, Ḥalab, and Anṭākiyah are all in different junūd. 1 In the cases of Dimishq and Ḥimṣ, both the province and the capital bear the same name. In Palestine Filasṭīn was used to mean the capital Ramlah,2 while Miṣr is almost always identified with Fusṭāt.3

In the two cases of Ḥalab and Anṭākiyah, we are dealing with the capitals of the provinces of Qinnasrīn and al-‘Awāsim.1 No city other than the capital is known as a mint in any of these provinces. This points to a remarkable organization of the minting processes in Ṭūlūnid times. Instead of using a large number of mint names often with comparatively little concern for administrative units, as was often the case in earlier times, we meet now with what appears to be a conscious pattern of unity. This pattern will be followed in the ‘Abbāsid period after the fall of the Ṭūlūnids and under the Ikhshīdids, although in Fāṭimid times again the number of mints will increase, often with little relation to the organization of the state.2 The only mint which appears immediately after the Ṭūlūnids in the area under their control is that of Ṭabariyah, capital of the province of Urdunn, and, even then, a new discovery may easily fill in this gap, so that there is no need to assume that al-Urdunn was not a Ṭūlūnid province with its own administration. An interesting fact about these mints is that in at least two cases the Ṭūlūnid initiative seems to have acted as a stimulant. Ḥalab and Anṭākiyah were extremely rare mints before the Ṭūlūnid period,3 but Ḥalab became quite common in later years.4 while Anṭākiyah remained as an Ikshīdid mint, but was soon after taken by the Byzantines.

A last question to be raised with regard to the administrative significance of these mints is why, in the cases of Anṭākiyah and Ḥalab in particular, was the name of the town chosen rather than the name of the province, since the evidence definitely indicates that the mints were chosen because they represented the provinces. The problem is important, for an answer to it would lead to an eventual understanding of the nature of the process of localization of gold minting in mediaeval Islam. Were the towns chosen indiscriminately? Were the capitals of the provinces chosen because the coins were actually struck there? Is there a difference between coins struck in "royal" foundations and those struck in administrative capitals?1 While a valid answer could only be given after a comparative analysis of mints on a large scale, one may venture the suggestion that in Ṭūlūnid times, the mints were restricted to provinces, but that each province tended to become characterized by its chief city (whether the reasons for determining the chief city be political or economic). Filasṭīn and Miṣr appear to be exceptions, but even there the official identification of the two terms with the cities of Ramlah and Fusṭāt seems likely, and in the case of Miṣr, this identification would probably be confirmed by the evidence of the ṭirāz.

It will be noted that two Ṭūlūnid mints have been omitted from this discussion: al-Rāfiqah and Ḥarrān. As far as al-Rāfiqah is concerned, I hope to have established that, since the province was never held by the Ṭūlūnids long enough to be truly administered by them, the coins issued there had a political rather than administrative significance.2 The appearance in the year 276 only of the mint of Ḥarrān is definitely puzzling. It is in the Diyār Muḍar, but much farther north than the usual scenes of the fighting between Khumārawayh and ibn Kundāj or ibn abi al-Sāj. Furthermore, Khumārawayh was back in Fusṭāṭ in Jumāda I of that year and, so far as I have been able to ascertain, Ḥarrān is never mentioned by the chroniclers as being part of the areas belonging to Khumārawayh, although an army of Aḥmad Ibn Ṭūlūn did reach Ḥarrān.1 The most likely explanation for these coins could be the following. In 275, when ibn abi al-Sāj left Raqqah and moved first to Mosul and then to Baghdad, ibn Kundāj remained in the Diyār Rabī'ah and the Diyār Muḍar.2 At that time he was presumably still an ally of Khumārawayh and these coins could be a sign of his allegiance to the Ṭūlūnid, struck where he had established himself at the time, that is, at Ḥarrān. This allegiance was probably very short-lived, since in later times we do not hear any more of any relations between the two men, nor are there any coins which would suggest one, although in 280 ibn Kundāj was still active in the same areas, involved this time in ‘Abbāsid politics.3

These coins from Ḥarrān are interesting in yet another respect. With the exception of an Arab Byzantine copper4 they are the first known coins from that mint and here again the mint was fairly active in the years which followed.5 What seems to have been an accidental event of Ṭūlūnid coinage had significant repercussions. But the importance of these coins, as far as the Ṭūlūnids are concerned, is, just as in the case of al-Rāfiqah, purely political. Their aim was propagandistic not economic or administrative; and in that sense they no doubt fulfilled an essential function of coins throughout the middle ages, if not even in the modern period.

A last remark should be made about Ṭūlūnid coins as a whole. Almost always—the few exceptions are indicated in the catalogue—the coins other than those bearing Miṣr as a mint are lower in weight by a rather considerable margin. In view of the importance of the sikkah as a gubernatorial and princely prerogative, it is probable that this custom was imposed from Egypt (it may be worth noting that the Ḥarrān coins certainly struck outside the regular channels are of higher weight than the rest of the provincial coins). It is all the more remarkable because the Miṣr coins are almost throughout1 of a high and consistent weight, thanks to the wealth of the country and to the high standards imposed by Ibn Ṭūlūn.2 Was this an attempt to assert the high value of the currency struck in Egypt? Does it imply that the coins bearing provincial mints were not necessarily struck there?3 Can it be connected with attempts to protect or to enhance the buying power of the inhabitants of Egypt? These questions are directed to the economic historian and cannot be answered before more complete data are available from other times and places.

The marks found on Ṭūlūnid coins should be connected either with economic matters or with the identification of certain issues. The general problem of marks is, however, far from being solved and we lack sufficient evidence to draw any general conclusions. An interesting point, however, is that these marks, which are of the most common type (‘ayn, hā, ḥā, tā-mim, bā) do not occur at the beginning of the dynasty. The earliest example is on the 276 coin from Anṭākiyah and it should be noted that these signs begin on provincial mints, although by no means all of them, and that they do not become common on Miṣr coins before 280. Here again we seem to meet with a difference between provincial and central mints, whose significance is not very clear and which cannot, I believe, be explained through Ṭūlūnid coins alone.

The only individual coin which may still deserve comment is the one numbered 51 in the catalogue. It was published over a hundred years ago by Soret;1 by the inclusion of the name al-Mu'tamid it is datable before 278, and it also bears the name of Khumārawayh. On the reverse this small copper bears a name which has been read by Soret as Aḥmad ibn Ya‘qūb. Soret suggested first that this may have been an Aḥmad ibn Ya'qūb who had for a while governed Sicily for the Aghlabids and who would have gone to the Orient at a certain time. For some reason Soret assumes that he ruled in 268. This hypothesis has to be discarded since both ibn al-Athīr and ibn Idhāri mention that the man, who had indeed been governor of Sicily in 257, was killed in 258.2 A second suggestion made by Soret was that Aḥmad ibn Ya'qūb belonged to a heretical group called al-Ya‘qūbiyah, which had its center not far from Mosul, at a place called Marj, and which in 273 fought against ibn abi al-Sāj.1 I have been unable to find anything out about the leaders of this sect at that time, but I wonder whether Defrémery, and following him, Soret, did not confuse a small heretical group2 with the Christian sect of the Jacobites called al-Ya'qūbiyah by Arabic authors, which was certainly represented in the area of Mosul.3 There is little doubt that they did not strike coins as a group and, while it is not impossible to imagine that Ṭūlūnid coins were struck in the area of Mosul, the Soret coins cannot be taken as evidence for it. I have not been able to discover another man by the name of Aḥmad ibn Ya'qūb who could conceivably have been involved in Ṭūlūnid politics. Just as in the case of the other copper coins struck under the Ṭūlūnid regime this one should perhaps be considered as originating from the thughūr and Aḥmad ibn Ya‘qūb may have been one of the many commanders who at one time or another took the field against the Byzantines.

End Notes

1 However in so far as I have been able to gather in the published catalogues and at the American Numismatic Society, the provincial 'Abbāsid coins from that period are similarly lower in weight than average; see below.
1 On Khumārawayh, see the article in the Encyclopedia of Islam, Hassan, pp. 107–133, and the chronicles. Becker, pp. 182ff., shows as usual a remarkable insight into the period. In general, however, we are far less informed on Khumārawayh than on his father and, as has been noted by Hassan, p. 116, the great compilers of world annals such as Ṭabari and ibn al-Athīr are often clearer in their exposition of Ṭūlūnid activities than the Egyptian writers.
1 On this point see the remarks made by Becker, p. 184, about the apparent necessity for Turkish leaders throughout Muslim history to be fully accepted within the theoretical framework of the caliphate. Cf. Taghribirdi, III, p. 4, where Ibn Ṭūlūn is shown as hating other Turks very strongly. Note also that most of his sons received names from pre-Islamic Arab traditions. There is here again an interesting example of adaptation to and adoption of a culture by a second generation of men who came from an alien group.
1 Kindi, p. 240; for the tribute see Hassan, p. 118.
1 Ibid., III, p. 2131.
1 Ibid., nos. 757, 767, 774, 785, 788, 805, 813, 814, 818, 825, 847; the exception is no. 769.
1 See Serjeant’s first chapter, Ars Islamica, IX, pp. 60ff.
1 281: Khedivial, no. 623; 283: Paris, I, no. 1043; 291: Tiesenhausen, no. 2182.
1 Ṭabari, III, p. 2048.
1 Ṭabari, III, p. 2104.
1 For his earlier career see Ṭabari’s index, esp. III, pp. 1941–2 and 2025ff. For his situation in 269, see ibn al-Athīr, VI, p. 50. In general see article "Sādjids" in Encyclopedia of Islam.
1 Ṭabari, III, p. 2112.
1 Two coins have been published which present something of a puzzle. J. Allan, "Unpublished Coins of the Caliphate," Numismatic Chronicle, vol. XIX (1919), p. 187, mantioned a B.M. coin dated in 284 and minted in Damascus without the name of Aḥmad Ibn Ṭūlūn. Dr. J. Walker kindly informed me that this is a misprint and that the coin is dated in 294. The second coin, in the Palestine Archaeological Museum, was published by G. H. Nassar, "The Arabic Mints in Palestine and Transjordan," Quarterly of the Department of Antiquities in Palestine, vol. XIII (1948). It was struck in 277 in Filaṣtin and is described as 'Abbāsid, not Ṭūlūnid. I do not know of any historical event which would have justified an ‘Abbāsid issue at that date in Palestine. The sources (Kindi, pp. 239–240; Taghribirdi, III, p. 52) do say that in that year Khumārawayh left for Syria on some "necessary business," but they do not say what. He returned to Egypt shortly thereafter. Was there some rebellion in Palestine? Or has the coin been incorrectly read?
1 All the texts dealing with the organization of Syria at that time are conveniently gathered in G. LeStrange, Palestine under the Moslems, (London, 1890), pp. 24ff.
1 The town of Qinnasrīn was at that time in decadence and Aleppo was in fact the capital of the jund.
1 For instance al-Mu'izziyah al-Qāhirah as opposed to Miṣr in the Fāṭimid period.
1 Ibn al-Athīr, VI, p. 18.
1 Exceptions would be provided by some of the later issues (283, 285, 286, 291). By that time both the political and the economic situation of this regime had considerably deteriorated.
1 Soret in RNB (1854), p. 19.
1 Ibn al-Athir, VI, p. 61; M. Defrémery, "Sur la famile des Sadjides," Journal Asiatique, 4ème série, vol. IX (1847), p. 434.
2 Ibn Sa'īd, p. 33.
2 Ibn Sa'īd, p. 74.
2 Note that Ṭabari and others, when talking about Egyptian army at that time, do not say jaysh Khumārawayh, but jaysh ibn Ṭūlūn.
2 Taghribirdi, III, p. 118; Hassan, pp. 141–2.
2 Ibid., III, p. 2133.
2 Ibid., nos. 736, 758, 762, 768, 778, 779, 780, 784, 793, 824, 827, 828, 837, 838.
2 A distinction may be emphasized here between the ṭiraz al-khāssah and the tirāz al-'āmmah, Wiet, Exposition, p. 7. See also A. Grohmann, article "Ṭirāz," in Encyclopedia of Islam (note in particular all the additions found in the supplement), where a number of texts are given about the various Tinnīs factories.
2 270: Porter in Numismatic Chronicle (1921), p. 323; 271: Casanova, no. 639; 272: B.M., IX, no. 352n and U.M.; 274: Rogers, no. 38, Paris, I, no. 998, U.M.; 275: Tiesenhausen, no. 2876. The 274 coins present a curious problem. They bear on the reverse the name of Aḥmad ibn al-Muwaffaq, the future al-Mu‘taḍiḍ, while the obverse has the name of al-Mufawwaḍ. The oddity of the coin consists in the fact that in the month of shawwāl 274 Aḥmad quarrelled with his father and was jailed (Ṭabari, III, 2115). The coin may have been struck in the early part of the year, but, so far as I have been able to determine, Aḥmad had no official position at the time. That he was at the head of a political coterie at the time is made quite clear from the various stories told about him by Ṭabari (cf. index) and it appears that the activities of that coterie led to his arrest. Was this coin the result of some manoever of that political group? Did ibn abi al-Sāj belong to it?
2 Ibn al-Athīr, VI, p. 56.
2 Kindi, p. 236; Taghribirdi, III, p. 51.
2 Kindi, p. 236; Ṭabari, III, p. 2112; but the most coherent account is that of ibn al-Athīr, VI, p. 61.
2 I am following here ibn al-Athīr, VI, pp. 62–64. Kindi, p. 238 and Taghribirdi, III, p. 56 seem to have confused the two operations.
2 Note the statement of Naṣir-i Khusraw, Sefar-nāmeh, ed. and tr.n C. Scheffer (Paris, 1881), p. 65 about Ramleh that "this city is known in Syria and the Maghrib under the name of Filasṭīn." Compare the modern usage of ash-shām for Damascus.
2 G. C. Miles, Fāṭimid Coins (New York City, 1951), pp. 50–51.
2 Above p. 58.
2 Ibn al-Athīr, VI, p. 64. Already in 273 he had withdrawn at one point to Mardīn, ibid., p. 61, and it does seem that he had considerable support in this whole area of the Jazīrah.
2 Cf. above p. 59.
2 Ibn al-Athīr, V, p. 364; ibn Idhāri, ed. Colin and Levi-Provençal, I, p. 115.
2 ‘Abd al-Qāhir al-Baghdādi, Moslem Schisms and Sects, tr. K. C. Seelye (New York City, 1920), pp. 45–6.
3 Ibid., p. 34. The story occurs also in Maqrīzi, Wiet’s ed. in vol. 30 of the MIFAO (Cairo, 1911), pp. 181–2, and in al-Balawi, Sīrah Aḥmad ibn Ṭūlūn, ed. Muḥammad Kurd 'Ali (Damascus, 1358 A.H.), p. 196. The latter book was unfortunately unavailable to me and I owe the reference to A. S. Ehrenkreutz, 'The fiscal administration of Egypt in the Middle Ages," Bulletin of the School of Oriental and African Studies, vol. XVI (1954), p. 510, note 3. There seem to be slight variations in these accounts and Maqrīzi, in particular, relates that Aḥmad's coins became in turn noted for their excellency and even their superiority over Mu'taṣimi and Sindi coins. They were called Aḥmadiyah. On all these terms cf. Sauvaire’s "Matériaux....," Journal Asiatique, 7ème série, vol. XV (1880), pp. 271–2. A Sindi coin was of the type developed by al-Sindi, who took over the mint after the fall of the Barmakids, while a Mu'taṣimi was obviously according to the standard of the coins struck by the caliph by the same name. See Miles, Rayy, pp. 84, 115, 117 for these personages, but I do not know of any evidence showing that in their own times their coins were especially noted for their excellency.
3 Kindi, p. 237; Maqrīzi, I, p. 321.
3 Ṭabari, III, p. 2123.
3 Ibid., III, pp. 2206–7.
3 Above p. 36.
3 Cf., most recently, J. Schacht, Esquisse d'une Histoire du Droit Musulman (Paris, 1953), pp. 70ff.
3 Ṭabari, III, pp. 2104–5.
3 ‘Abbāsid coins of that year, if found, would have to belong to the early part of the year.
3 The information in Maqrīzi, I, p. 321 and Taghribirdi, III, p. 52 is entirely erroneous as far as the dates of all these events are concerned.
3 See above p. 64, note 6.
3 Ḥalab: 146 (Tiesenhausen, no. 747), 153 (ibid., no. 825), 138 (RIC, no. 371), 164 (Tiesenhausen, no. 961), all but one coppers. Anṭakiyah: 138 (RIC, no. 220, where the final statement in the commentary should be amended).
3 Ṭabari, III, p. 2137.
3 There are several pieces of evidence suggesting that provincial coinage, in particular that of al-Rāfiqah, was not struck in the provincial city itself, but either in Baghdad-Samarra or in Fusṭāṭ. On the other hand, it must be pointed out that a glance at the illustrations (see, for instance, pl. I, nos. 27, 45) shows that there are notable epigraphical differences between al-Rāfiqah and Miṣr coins. It would seem a priori unlikely that the mint masters in the capitals would also devise a special epigraphy for provincial coins. This phenomenon should perhaps be connected with the information given by W. Barthold, Turkestan down to the Mongol Invasion (London, 1928), pp. 204ff., that in Khwarizm debased coinage was introduced to prevent the circulation of local coinage beyond the frontiers of the province.
3 Yāqūt, II, p. 689.
4 Répertoire, no. 731.
4 The statement of ibn al-Athīr, VI, p. 50, quoted by R. B. Serjeant, "Material for a History of Islamic Textiles," Ars Islamica IX (1942), p. 72, that one of the reasons for the cursing of Ibn Ṭūlūn was that he had dropped the name of al-Muwaffaq from the ṭirāz is rather curious, since the name of the heir apparent occurs only occasionally and that of the "heir in second" is not known.
4 R. Ettinghausen, "Interaction and Integration in Islamic Art," in Unity and Variety in Muslim Civilization, ed. G. von Grünebaum (Chicago, 1955). passim. Most of the studies in this book emphasize the same theme from different angles.
4 Ṭabari, III, pp. 2106–7; Kindi, p. 235; Hassan, pp. 110–112.
4 Ṭabari, III, p. 2049; ibn al-Athīr, VI, p. 50.
4 I have only checked through the Ikhshīdid period.
4 RIC, p. 17.
5 Répertoire, no. 732.
5 Serjeant in Ars Islamica, XIII-XIV, pp. 88 and 91ff.
5 Hassan, p. 80.
5 The next coin is ‘Abbāsid in 284, RIC, no. 160.
6 Miṣr is taken here to mean Fusṭāṭ. Serjeant assumed it to be so for textiles and Lane-Poole for coins. For a different view cf. G. Wiet, L’Exposition Persane de 1931 (Cairo, 1933), p. 6.