Macedonian Regal Coinage to 413 B.C.

End Notes

End Notes

For a century or more before the advent of the signed currency of Alexander, the whole Thraco-Macedonian coastal and hinterland area had produced a rich and diversified series of coins. These have been studied painstakingly by Svoronos in the article already cited, 1 but his assignments of the various types have not been generally accepted, nor are they all acceptable. Nevertheless, his collection and illustration of the varieties classed as "uncertain, probably Thraco-Macedonian" and homeless, lo, these many years, make it possible to survey the group in its approximate entirety. By considering the coins as a homogeneous group, one is able to see clearly a pattern in which three major religious or mythological types are associated with certain denominations, each of which is issued by one or more tribes. No tribe seems to have used all the types and all the denominations: some used several of each; some confined themselves to one. The three types are associated respectively with Dionysus, Ares, and Apollo-Hermes. 2

The types involving oxen, their driver, caduceus, and rose or other solar symbol 3 are those of Apollo-Hermes; the horse, alone, attended, being mounted or ridden, and the dog, possibly the lion, allude to Ares. Dionysiac types are more varied - nymphs, satyrs, centaurs and the vine. There are some additional types: Pegasus, Gorgoneion (see note 3), Herakles, and various animals, which do not fall into these categories. The pattern, reduced to its simplest terms, is this:

The fractions smaller than the stater continue these types in a more complex pattern, study of which would surely clarify much that is obscure in the arrangement of the coins. The reverses, with their various sorts of incuse and subsequent types, should also prove enlightening in the final attribution of the coins and the location of their issuing agents. Inscriptions seem to have been used first on the obverse and to have been considered originally as symbols; 4 later, they become so complete as to compare favorably with those on Hellenistic coins. The archaic form of some of the letters appearing on the tribal issues is retained on the earliest inscribed coins of Alexander.

The types most generally used (those of Apollo-Hermes, Ares, and Dionysus, as well as the somewhat less frequently occurring Herak- lean types) have a natural association with the area. Tradition and common consent localized the myth of Hermes and the cattle of the Sun in the north, 5 where Apollo was pre-eminently a sun-god. The Pangaean rose, the extraordinary beauty of which was mentioned as late as Pliny, 6 has the same applicability as a solar symbol as the canting type of Rhodes, the connection of which with Helios or Apollo is generally recognized. It is quite possible that the "dotted theta" was originally a stylized version of this rose. The rose, or an unidentifiable flowering plant, appears as a symbol with nearly every type and in nearly every denomination. As far as I am able to determine, the caduceus is used as a symbol only in Macedonia; e. g., as a brand on horses (see Plate II, 5); 7 in the hand of Hermes, it appears in other sectors. All signs point to Apollo as the more important of the two divinities: the legend itself, which makes Hermes' stay in the neighborhood brief and not particularly honorable; the types, which as frequently show the oxen drawing a cart as being led off by an attendant who is presumed to be Hermes even when he lacks the caduceus; and the more frequent use of various solar symbols.

Ares and Dionysus, who were always a bit embarrassing to the Greek Pantheon, were at home in the north. Ares is associated with the man-eating steeds of Diomedes of Thrace and with the supernatural horses of Rhoesus. Later, the "Thracian rider" or eques Romanus of many reliefs attests the relation. Recently, Jean Babelon 8 has attempted to give a name to this divine rider, but so much conjecture is necessary in any such essay that it is better simply to say that he is to be connected with the cult of Ares. Whether he be warrior or hunter is difficult to determine; his attire, on the coins, is that of a hunter while the use of the helmet as symbol or reverse type shows that the arts of war are somehow involved. In the catalogue of the Macedonian coins, where a rider is the type, I have used the term "warrior-hunter" because a decision is impossible. Later reliefs show the rider clad sometimes as a hunter and sometimes as a warrior. Dionysus spent his childhood on Mt. Pangaeon. Brought up by Nys(s)a, he rollicked there with all his merry crew- nymphs, satyrs, centaurs - and his worship was conducted by the Bessi. 9

The appearance of the hero Herakles as a type on the tribal issues is quite wide-spread and is explained by his deeds in the north. He captured the horses of Diomedes and then founded the city of Abdera, establishing games there in honor of his friend Abderus, who was killed while helping him secure the horses. It may be mentioned in passing that the later use of Herakles or his equipment as a type by the Macedonian kings seems to originate not so much in this natural association of the hero with the area as from the legend which traced the ancestry of the kings to Argos, whose hero Herakles was. 10

Of the animals which are used as types, the goat and the cow with suckling calf, while they are not so obviously appropriate to this area, have some justification. The goat is the canting type of Aegae, and may have a remote reference to Apollo, since in the legend of the coming of the Macedonian kings, the sunlight was their heritage and a goat marked their destination. 11 The cow r., with suckling calf, 12 although its significance is uncertain, can plausibly be associated, as a member of the genus bos, with the Apollo-Hermes group of types in this area.

The remaining types which appear in the pattern on page 44 the Gorgoneion and Pegasus, are apparently intrusive (but cf. note 3). They may legitimately be considered together on the basis of their blood relationship. Medusa, and Pegasus, are so closely connected with Corinth that there is every likelihood that their use as cointypes in the north arose from the early connection of Corinth with the north. (Cf. Chapter I.) Likewise, some significance may be attached to the fact that the Pegasus type is used on the tetradrachm, the denomination created to be used with the electrum, since the forepart of a Pegasus is one of the obverse types used on the eastern affiliates of the third group of electrum coins. The Gorgoneion appears on the electrum of the second group, 13 in a form stylistically earlier than that employed for coins in the Thraco-Macedonian area. Thus, their source may be either Corinth or Asia Minor.

The pattern of type distribution indicates some sort of monetary convention among the issuing agents. The survey made possible by Svoronos' collection of material reveals that obverse types were constant for certain denominations, whoever issued the coins, and that some tribes (e. g., the Orrheskians) issued inscribed coins in all three major types in some, but not all, of the denominations devoted to those types. From the extant coins there is no evidence that any one tribe used all the denominations above the stater as well all all the types. Some tribes apparently confined themselves to one type and one denomination, disregarding the fractional pieces. The preponderance of coins with Dionysiac types makes them, at first glance, seem the more important. However, the restriction of these types to the staters, and the use of the other types in the larger denominations, can well be taken to mean that the metal was apportioned in fairly equal quantities among the types characteristic of the three divinities, and that commercial or other considerations determined the amount to be used in various denominations. This hypothesis perhaps is over-complex; the assumption that these "rude and barbarous" tribes were capable of establishing and enforcing any such ingenious monetary convention as this may be difficult to accept. Let it be remembered that modern ignorance of these tribes does not prove their barbarous estate. All the evidence of the coins themselves, the variety of types and denominations, the vigor of the art and its stylistic development, indicate that the issuing agents have long been under-estimated because of absence of information about them. It is quite possible that detailed study of the coins would result in considerable modification of the suggestions made here. Such examinations as has been possible shows clearly the absence of hazard and the presence of design in their minting. 14

It does not lie within the scope of this study to investigate in detail the issuing agents of the mass of tribal coinage. However, two groups of coins which show a clear relation to the regal Macedonian issues must be examined in some detail. They will be seen to be the precursors of Alexander's coins and to furnish an explanation for the variety of his coin-types. They are the goat staters of Aegae and certain octadrachms and tetradrachms with types of Ares: horse and attendant, and horse and rider.

A. The Goat Staters, Plate I, 1-8.

The goat staters are quite generally assigned to Aegae. 15 Svoronos' assignment of them to the Derronians, 16 among whom the goat type has no recognizable meaning, is erroneous. The monogram on which he bases the attribution is that of Edessa, the older name for Aegae. 17 The fourth century coin of Lykkeios of Paionia 18 which has the inscription ΔEPPONAIOΣ indicates that that tribe is to be located at some distance from the Thermaic gulf; the succeeding monograms on the goat staters and have nothing to do with the Derronians, while a city monogram preceding a personal one is perfectly acceptable. The type is of purely local significance, a canting type which may have given rise to the fictitious oracles (see Chapter I). The quadripartite linear square on the reverse shows the development common to coins of the Thraco-Macedonian area: the earliest examples (Plate I, 1) are rude and irregular; the latest are precisely formed (Plate I, 5-8). The development is closely paralleled by that of the tetradrachms of Acanthus, 19 the different stages in each series being marked by similar symbols or monograms on their obverses.

The eight staters on Plate I show the various stages through which these coins developed. The earliest had no symbol or letter in the field. If one judges from the style of the reverse, the technique of the obverse, and the coins of Acanthus, the first added symbol was the flower, possibly the Pangaean rose. Only with this symbol is the goat represented as advancing left, with head reverted. The next symbol was the dotted theta, which may have been intended to represent a stylized flower or a solar symbol, or both. It was followed by a circular theta with a dot in the center, an obvious technical advance over the variety made by the drill alone. Then came the monogram and, finally, the first two letters of Alexander's name, in regular or reverse order. 20 On the coins of Acanthus in Desneux' catalogue, his order of nos. 1-92, includes successively those with no symbol, the flower, no symbol, the flower, dotted theta, circular theta, pecten shell, and finally A or AKAN in the field. This group of coins he dates before 480/79; the next series, with AKANΘION on the reverse, he dates after that time, on the analogy of the coins of Alexander., He places the earliest issue ca. 530 B.C., a date which is probably too low, although the paucity of specimens would seem to justify it were it not for the fact that the 92 die combinations include 88 anvil and 81 punch dies. In other words, there must be a great many missing die combinations.

Of the coins which he assigns to the years 480-450 (or possibly a little later, although he is uncertain about an interruption of Acanthian coinage following the promulgation of the Athenian currency decree), there are 23 combinations of 22 anvil and 21 punch dies: again many die combinations must be missing. To be ultraconservative and take the number of extant dies as a close approximation of the total used during the thirty (or more) years, the resultant calculation suggests that the life of a die was a little over a year. On that basis, the dies for coins, nos. 1-92 in his catalogue should cover a period not less than 81 years (the number of surviving punch dies) and the beginning of the coinage of Acanthus should be placed ca. 560. The exact time of origin of the whole series of Thraco- Macedonian coins is uncertain, but for stylistic, metrological and typological reasons it should be placed early in the sixth century, if not at the end of the seventh. The survival of so many (92) die combinations for the early Acanthian coins and the parallelism between the progress of symbols on those coins and on the goat staters are indications that both series of coins began well before the middle of the sixth century. 21 Since we do not know the date of the accession of Amyntas, it cannot be said that he inaugurated the Macedonian coinage, or at what point he continued it if it had been begun before his time.

Turning to the latest of the goat staters, those with or in the field, we may conclude that they were struck after 498, the earliest probable date for the accession of Alexander. Possibly those with the monogram should likewise be assigned to that king. The whole series of coins, judged by published specimens, is rather small. It is questionable whether, on this account, one may argue that the monogram group is to be assigned to Alexander,even though there are so few surviving staters (and one of these plated!) with his own initials. The following list of goat staters and their fractions is not intended to be exhaustive; it contains only already published specimens in their relative chronological order. The larger proportion of surviving coins is in the monogram group, which may represent the issue of Amyntas. At all events, the indications both from the coins of Acanthus, where the floral symbol and the dotted or circular theta occur on so many surviving dies, and from the coins of Aegae, where nearly half the surviving dies are those of the monogram group, are that both series were begun before the middle of the sixth century. Beyond such an approximate date it is not possible to conjecture.

The fractional issues (Plate I, 9-12) of the same type as the staters, have most frequently two pellets which are surely marks of value, since the coins are diobols. One coin (Plate I, 11) has the dotted theta; there are a few others with the first two letters of Alexander's name, always retrograde (Plate I, 12). The following is not a complete catalogue, but it should serve to indicate that the stylistic development of the goat staters is a close parallel to that of the tetradrachms of Acanthus.

Staters

1. No symbol	Weber Coll., 1837	9.56
	Naville, 1922, 429 (Plate I, 1)	9.56
2. Flower	B.B. II, p. 166, 12 Plate I, 2)	9.57
	Hunter Coll., I, p. 267, 1	9.37
3. Dotted theta	Naville, 1927, 624	8.62
	Naville, 1925, 388 (Plate I, 3)	9.78
	Egger, 1914, 461	9.20
4. Circular theta	McClean Coll., II, 3098 (Plate I, 4)	9.15
	Weber Coll., 1839	9.18
	B.B. II, p. 166, 10	9.40
	BT I, 2, 1546	9.20
	Desneux Coll.
5. Monogram	Gillet (Jameson, 1836) (Plate I, 5)	9.27
	Locker-Lamson Coll., 152 (Plate I, 6)	9.31
	Berlin, AMNG III2, p. 19, 6	9.40
	Hirsch, 1933, 968	9.24
	BMC Mac., p. 37, 1	9.48
	Weber Coll., 1840	9.30
	B.B. II, p. 166, 11	9.30
	BT I, 2, 1540	9.41
6. Initials	BMC Mac. p. 37, 2 (Plate I, 7)	9.10
	B.B. II, p. 166, 13 (plated)	8.61
7. Initials	Berlin, AMNG III2, p. 19, 5 (Plate I, 8)	9.28

Diobols

1. Pellets	Naville, 1928, 416 (Plate I, 9)	.93
	May Coll. (Plate I, 10)	.87
	Delepierre 2323 (pierced)	.95
	Desneux Coll.
2. Dotted theta	Naville, 1921, 695 (Plate I, 11)	1.11
3. Initials	Berlin, AMNG III 2, p. 20,16 (Plate I, 12)	.90
	London, NC 1892, pp. 5-6	1.07

End Notes

15	Cf. Head, HN2, 198-199; Babelon, Traité, I, 2, 1095-1104. Gaebler, AMNG, III, 2, 18-20.
16	Svoronos, HPM, 6 and 7.
17	Cf. Babelon, Traité, I, 2, 1095 and Strabo, X, 1, 15.
18	Cf. Gaebler, AMNG, III, 2, pl. 37, 5.
19	Desneux, Les Tétradrachmes d'Acanthos, reprinted from Rev. Beige de Numis- matique, (1949), 5-122, with magnificent illustrations.
20	The order of the letters is of no assistance for purposes of dating the coins. The light tetrobols of Group I (nos. 38 and 39 in the catalogue) show the two letters retrograde.
21	I have made no attempt to assemble a complete corpus of the goat coins.

The illusion of great variety caused by a casual glance at the Alexandrian issues is dispelled by finding the roots of his coinage. The main type is that of Ares, similar to that used by other members of the Thraco-Macedonian monetary alliance; the caduceus (which is also used by Perdikkas) is the symbol of Apollo-Hermes (cf. note 28); the goat is the local canting type. Dionysiac types are not used, unless the very infrequent leaf as a symbol be a vine leaf (Plate IV, 15a). One new element is added: the heavy tetrobols have as their reverse type the forepart of a lion. One thinks immediately of the impressive lions on the coins of Acanthus, of the monumental lions of Amphipolis and Chaironeia, of coin types from Samos or Miletus, and of the various types of Herakles and the lion in the Thraco-Macedonian area. If Ionian or Aegean types were to influence coins in this area, they should have done so a century before. 31 The monumental lions are much too late. 32 The origin of the lion type at Acanthus 33 is as yet unexplained. We know from Herodotus (VII, 125) that lions were common enough and bold enough to attack the camels of Xerxes' supply train in Macedonia, but that is scarcely a reason for using them as a coin type. Perhaps the best case can be made out for a connection with Herakles, for Heraklean types were used in the north during the period of the tribal issues; the legendary association of that hero with the area provided sufficient reason for it, particularly in the case of Abdera and Thasos. As regards the Macedonians, the position of Herakles as ancestor of the Temenid house of Argos is a more valid source for the use of a type referring to him, since the locale of Heraklean exploits in the north was beyond the sphere of Macedonian influence. The king's efforts to be recognized as Greek might well include this allusion to Herakles on his coins. It should be observed that the types of Macedonian coins after 480/79 are exclusively regal; the ethnic is never used. The comparative anonymity of the lion is not relieved until the end of the fifth century, when Herakles' club, quiver and bow appear as a reverse type on a few of Perdikkas' fractional issues (Plate XI, f, with Herakles' head as obverse). In the fourth century, the lion is accompanied by a variety of Heraklean types and symbols.

Before a discussion of the coins, their arrangement, and their chronology, some observations may be made upon the technique and procedure of the Macedonian mint. These observations are culled from the various groups into which the coins have been divided and from the patterns of their die sequences. The die sequences for all groups are to be found at the end of this chapter.

Whatever the method of preparing the flans, it seems not to have changed throughout the fifth century. There is no apparent variation in the size of the coins; there are no residual protuberances on the edges of the coins, such as serve to indicate that the flans were cast in a series of moulds; there is likewise very little evidence of insufficient heating of the flans at the time of striking, for the edges of most of the coins are quite regular. The slight difference in the sizes of the flans for the light and the heavy tetrobols indicates that mint officials were able to determine quite accurately the amount of metal necessary for coins weighing within a gram of one another. Where the coins in a comparatively few cases (see pp. 39-40) show very irregular weights, I should prefer to attribute them to carelessness in the selection of flans, rather than to inability to prepare them properly.

The dies and their size and form may be reconstructed from the coins. In Group I (Plates III-V), both dies are smaller than their flans and considerable care has been taken to center them in most cases. With the beginning of Group II, both anvil and punch dies increase in size, except those for the octadrachms (Plates VI-IX). Here, the border of the obverse and the corners of the incuse on the reverse are frequently off flan. There is evidence (see below Chapter VI) that in Group II other than Macedonian die-cutters worked in the mint. It is impossible to do more than suggest that these outsiders were responsible for the change. Whatever its cause, once made the charge was permanent. There is an apparent exception in the case of the heavy tetrobols: these were not struck in Group II, and the few specimens which remain in Group III do not show much increase over those of Group I. However, the heavy tetrobols of Group IV, struck by Perdikkas, show an increase in the size of both dies (see Plates XIII-XV). 1

Macedonian coins exhibit a phenomenon which, so far as I can ascertain, is not characteristic of other ancient coins and which is to be explained partly by the design and construction of the dies and partly by the size of the dies in relation to the flans. Because of the pressure applied directly to the punch in the ancient method of striking coins, in most issues more punch than anvil dies are necessary. The opposite is true of these Macedonian coins: there are more signs of die damage on the obverse than on the reverse, most often beginning at the edge of the die. The obverse dies are in much higher relief than the reverses, in each denomination; the punch itself was considerably larger than the die contained in it and flat, rather than beveled away from the die. 2 The actual size of the punch cannot now be discerned, for its edge is not visible on any of the coins catalogued here. In some cases, one can make out the edge of the anvil die, which seems to be very little larger than the border of the design. (See Plate III, 2, left side, and Plate VI, 48a, upper left.) These are both octadrachms, which were the only coins not to show an increase in the size of the anvil die. In the other denominations, the anvil die is so much larger than the flan that part of the design is left off. These technical details indicate that the pressure of the blow from the hammer would fall not only on the rather shallow die, but rather on the whole large area of the punch. In the case of the obverse dies, some of the pressure would fall on the die itself, outside the edges of the slightly smaller flan. Since the higher relief of the die rendered it the more fragile originally, direct pressure without the flan to act as cushion would make it suffer damage. This is not to imply that no punch die shows signs of damage; some do so, and that fact has been noted in the catalogue. Several punch dies, notably T P 17, (cf. Plate IX, 62b and 63a) and H P 26e (cf. Plate XIV, 216-218) show signs of an original miscutting and of use with more than one anvil die without either repair or damage. The example on Plate IX, of a goat type reverse, which is in the highest relief of any reverse type, is a good illustration of the salutary effect of the pressure distribution over a punch larger than its die by which the life of these punch dies is prolonged. The following list of known dies shows how constant is the larger number of anvil than punch dies.

When the figures are given by denominations rather than by groups, there is a slight change.

From this chart, it will be seen that there are fewer octadrachm obverse dies (the only ones which did not increase in size) and only two more tetradrachm obverse than reverse dies. Since the extant die combinations in these denominations are so few (22 octadrachms and 28 tetradrachms) the evidence from the smaller denominations, of which there are many more die combinations, is more valuable.

The pattern of die sequences makes it possible to draw some conclusions about Macedonian minting practices. In every group, in one or more denominations, three obverse dies are used with one reverse. 3 In Group I, this arrangement occurs in the heavy tetrobols; in Group II, in the octobols and both series of light tetrobols; in Group III, in the tetradrachms; in Group IV, in both the light and heavy tetrobols. Without the appearance of this procedure in the case of the tetradrachms in Group III, it would not be possible to generalize; the manner of producing small denominations might well differ from that for large coins. The evidence, however, allows one to conclude that regular mint activity consisted in using three obverse dies, all probably set in one anvil, 4 with one punch die. The fact that the octadrachms do not survive in sufficient numbers to show this procedure cannot be taken as a sign that it was not used in all denominations; so few octadrachms have survived that it is clear that we know only a few of their dies. 5 It is worth noting that the coins from Olynthus which immediately precede the League coinage, as well as the League tetradrachms and tetrobols, also show this use of one punch with three anvil dies.

Elaborations of this pattern occur at two points in the series of coins, once under Alexander, in Group II, where there are other signs of increased mint activity and once under Perdikkas, in Group IV, when the mint output is quadrupled. In Group II, the A series of light tetrobols (Plate VIII, 76-95) shows the use of one punch die with five, and another with six obverse dies. In both cases, some of the obverses are used with other punch dies; probably production was doubled by the use of twice the normal number of obverse dies. In Group IV, among the heavy tetrobols, there is a confusing pattern of seven punch dies in use at once, with at least eighteen obverses. Additional charts following the diagrams of die sequences at the end of this chapter show their ramifications. They are discussed in greater detail in Chapter VIII.

The use of three punch dies with one obverse does not appear before Group II, where it occurs once in the octadrachms, nos. 49-51, Plate VI (the obverse die shows hard use) and twice in the H series of light tetrobols, nos. 99-101 and 103-105, Plate IX. In Group III, it occurs once in the heavy tetrobols, nos. 118-120; and in Group IV, once in the first series of light tetrobols, nos. 138-140, Plate XI, and seven times in the heavy tetrobols, nos. 176-178, 186-188, 189-191, 204-206, 221-226, 228-230, Plates XIII-XV. This is not the normal replacement of worn dies, as would be the case if one obverse die were used with two reverses, or vice versa. The absence of this pattern from Group I may be fortuitous, but its use may rather be associated with the need for increased mint output. Group II shows various signs of abnormal mint activity. In Group III, the striking of heavy tetrobols is recommenced after a lapse of about fifteen years; a need may have been felt for more than normal quantities of them. In the light tetrobols of Group IV, the pattern appears at a time when that denomination was the sole product of the Macedonian mint. Of the seven instances of it among the heavy tetrobols of the same group, three are near its inception, when that denomination is struck again after a period of about twelve years; four are during the period when the mint was issuing great numbers of these coins. In the last series of Perdikkas' coins, there is no sign of it. Given the normal mint activity of three obverse dies with one punch, production would have been tripled when three punch dies were used simultaneously; presumably, that would indicate the use of nine obverses, although there is as yet no concrete evidence for it. The heavy striking of light tetrobols in Group II, when five or six obverse dies were used with one punch, and of heavy tetrobols in Group IV, when seven punches were used at once with at least eighteen obverse dies, points to very extraordinary demands for coinage at those times. In each case, the political situation will be seen to afford a reason.

GROUP I

Octadrachms

Tetradrachms (helmet reverse type)

Tetradrachms (goat's head reverse type)

Heavy Tetrobols

Light Tetrobols (goat's head reverse type)

Light Tetrobols (helmet reverse type)

GROUP II

Octadrachms

Tetradrachms

Octobols

Light Tetrobols, A series

Light Tetrobols, H series

GROUP III

Octadrachms

Tetradrachms

Heavy Tetrobols

Light Tetrobols

GROUP IV LIGHT TETROBOLS*

* The numbers in parentheses above the line refer to the Series within the group.

GROUP IV HEAVY TETROBOLS

Details of die sequence in Group IV heavy tetrobols, Series 2, nos. 195-233.

Beginning with no. 195 there is a period of abnormal mint activity, marked by a change of obverse type. The walking horse is supplanted by a prancing horse. Eight A dies of the walking horse type are used with seven P dies which are also used with ten prancing horse A dies. These latter A dies are also used with six other P dies, which are then used with other A dies of the prancing horse type. The dies involved have been numbered and lettered:

P 24, followed by a letter (a-g), is a die used with both styles of horse.

A 26, followed by a letter (a-h), is a walking horse A die used with a P 24.

A 27, followed by a letter (a-j), is a prancing horse A die used with a P 24, and in most cases with a P 25 (plus letter, a-h).

A 28, followed by a letter (a-g) is the next series of prancing horse A dies, used with P 25 (a-h) and with P 26 (a-b).

The combinations with P 24 must be regarded as simultaneous, with an abrupt abandonment of the walking horse dies, since none of those dies (A 26 +letter) is worn. With the use of A 27 (a-j), more than the seven P dies used at the time of the shift are involved. The following charts show the interlocking dies at this point.

PATTERN OF DIE SEQUENCES, SERIES 2, HEAVY TETROBOLS

There is no reason for questioning the commonly accepted date, 1 480/79, for the beginning of Alexander's epigraphically or otherwise identifiable coinage. Whether the coins appeared first after Salamis or after Plataea, it is impossible to determine. The sophisticated complexity of Alexander's issues can only be the result of a carefully thought-out plan; his use of types and denominations shows clearly that neither the early staters of Aegae, nor his own octadrachms, tetradrachms and fractional issues in the monetary alliance (see Chapter III) were adequate for his purposes. In this first national coinage, three aims were embodied: (1) it was to be readily exchangeable with the most influential currencies of the Aegean area (see Chapter II); (2) it was to be readily recognized as Macedonian; (3) it was to be unmistakably a regal issue.

The coins themselves fall naturally into three groups, as an examination of the octadrachms on Plates III, VI, and X will make clear. The smaller coins follow the style of the larger ones. Since there are no magistrates' symbols, or other means of ascertaining dates for these groups, the dates given below are not absolute in all cases. Historical lacunae make it necessary to rely on the coins themselves for information. The date of the end of Group I has been determined by the content of Group II, whose commencement is shown clearly by the coins to be 476/5. The date of the end of Group III is, of course, coincidental with the death of Alexander. This is usually given 2 as 454, but the coins indicate that he ruled some years longer, probably to 451. These then are the suggested dates for the coins of Alexander I:

Group I: 480/79-477/6

Group II: 476/5-ca. 460

Group III: ca. 460-451

Group I consists of four major denominations, two large and two small, in addition to fractional pieces whose weights are difficult to classify. The stater (9.82), with the obverse canting type of a goat, is no longer in use; its weight was unsuited to that of other currencies in the Aegean area. The coins struck are octadrachms, with a derived tetrobol, and tetradrachms, with a derived tetrobol. The weight norms for these denominations are:

The various ways in which such coins could be integrated into other currencies have been discussed in Chapter II. The passing of the stater, the original one-hundredth division of the Babylonian common mina on which all the Thraco-Macedonian tribal coinage was based, has been the source of the difficulty of classifying the weights of the fifth century coinages in the area. It had served its term of usefulness and was no longer to have a part in ancient commerce.

Before discussing the types and their implications, it must be pointed out that the mass condemnation of the tetradrachms of this and subsequent groups by Gaebler 3 cannot be endorsed. It is unnecessary to refute his arguments one by one. His misunderstanding of their weight has already been mentioned. His pronouncements on style and technique are unjustified on comparison of these coins with others in Group I, as well as with those tetradrachms immediately preceding it (Plate II, 7-9); his remarks on their color are not corroborated by such coins as I have been able to see. His theory that the reverse type of two of the light tetrobols (nos. 34 and 35, Plate V) provided a modern forger with models is a misinterpretation. The same Macedonian(s) did the work on all of these coins. A great many of these coins were collected by Rhousopoulos, 4 some in a hoard or hoards, under circumstances which leave their authenticity unquestionable. In the catalogue these coins have been indicated by the insertion of Rhousopoulos' name as their source. Likewise the coins condemned by Gaebler have been indicated in the catalogue. No. 22 (a and b) lists two possible forgeries, one known to Gaebler; the defects of these coins have been discussed at their catalogue entry. Since Gaebler uses the same type of argument, equally unwisely, in his "exposure" of tetradrachms in Groups II and III, no further reference to his conclusions is necessary.

While the denominations and the main outlines of the typological content of the coins had been determined upon at the time that the coins of Group I appeared, the actual type distribution among the denominations betrays some indecision at first. Combinations are tried out and discarded for others which, at the end of Group I, become constant for the denominations and remain unchanged thereafter. The metrological relation of the heavy tetrobols to the octadrachms and of the light tetrobols to the tetradrachms is clearly marked by stylistic details. The obverse type of all denominations is from the Ares-type of the monetary alliance: horse, horse and rider, horse and attendant. In Group I all obverse borders are dotted; subsequently, the dotted circle is retained on the octadrachms and heavy tetrobols, while on the other two denominations it gives way to a linear circle. These latter show on both obverse and reverse an uncertainty of orientation which persists even on the coins of Perdikkas; the others do not.

The reverses show plainly the conscious association of types with denominations. The octadrachms always have the king's name around a four-part linear square in the incuse. The heavy tetrobols, which introduced the lion type, 5 attempted to carry the inscription of the octadrachms also. Fortunately, this attempt was soon abandoned (two dies are known Plate V, nos. 23-26), for of all the poorly executed lions of Group I, these are the worst. The design is small; the die- cutters were working with an unfamiliar type; the space available for the inscription was microscopic. The letters themselves are illegible on the first die (no. 23) and almost as bad on the second. The tetradrachms and light tetrobols, although they have different obverse types (the obverse of the tetradrachms is the same as that of the octadrachms and heavy tetrobols), share an uncertainty as to reverse type and both show a tendency to replace the dotted square by the linear, which is uniformly used on both denominations in the later groups.

These reverse types reveal two different elements in Alexander's coinage. The goats, used first on the light tetrobols and then on the tetradrachms, were taken, in abbreviated form, from the staters; no inscription was necessary on a coin bearing this type. The helmets, first on the tetradrachms and finally on the light tetrobols, are new to Macedonian coins, but cannot be considered a new type since many of the tribal issues had a helmet as symbol or reverse type. 6 It is possible that this adoption of the helmet as a reverse type is proof that Alexander conquered at least one of the tribes which he has been credited with subduing and whose coinage he supplanted. The Orrheskians, among others, used the helmet as a reverse type; their staters have already been included among the Ares types having some similarity to the early Macedonian coins (Chapter III and Plate II, 13). Nothing is known of their location. 7 They may have been a tribe incorporated into Alexander's kingdom, but of sufficient size and importance to have their tribal integrity preserved by the use of one of their types. The helmet continued as a reverse type through the coinage of Perdikkas on the light tetrobols. The goat disappeared finally at the end of Alexander's reign when Perdikkas discontinued the denomination bearing it.

Symbols are used only occasionally and with no evident purpose. The octadrachms show one anvil die with a leaping hound in front of the horse (Plate III, 5). This is the anvil die which appears to be by the same hand as one of the anepigraphic series (Plate II, 6). One hound, however, is not quite assurance that the rider should be called a hunter rather than a warrior. No. 4, an octadrachm, has a huge frog below the horse. It is without precedent and is never repeated. Its significance is unknown; though there are a few coins, of Aeginetan weight, attributed to an indefinite Aegean island, 8 which have as type a swimming frog, it seems impossible to connect them with this coin. The same anvil die which has the hound also has a caduceus on the horse's rump. This symbol is a reference to the Apollo-Hermes type and had already appeared on some of the earlier, tetradrachms, also as a brand. It keeps recurring in various positions throughout the reigns of Alexander (as a reverse symbol on tetradrachms and small issues) and of Perdikkas (as a reverse symbol on the heavy tetrobols). One tetradrachm reverse die (no. 15) has a leaf, also used on some fractions, below the goat's chin. It has more resemblance to an ivy leaf than to a vine leaf, and it cannot be associated with the flower which appears on the early goat staters and later on Perdikkas' heavy tetrobols. During Alexander's reign, the small coins have no symbols; in Group I, the first two punch dies for the heavy tetrobols and one anvil die for the light tetrobols have inscriptions or monograms. Types alone were considered sufficient for these pieces, as well as for the fractions. It is not until the time of Perdikkas, when there were no large denominations, that the small coins carried symbols. As Group I remains, it cannot be said that the symbols have any major significance; they afford no help in arranging or dating the coins.

Even a cursory glance at Plates II, IV, V, shows the homogeneity of Group I, and, in the case of the large denominations, their immediacy of succession upon the octadrachms and tetradrachms of the anonymous coins attributed to the Macedonians. 9 The small pieces are the stylistic counterparts of the larger ones. A glance at Plates VI-IX shows the marked difference in style of Group II. The arrangement of the coins in the various denominations within the Group is dependent upon factors that are susceptible of criticism; therefore, a justification of the methods employed to arrive at the present catalogue order is necessary.

A. The octadrachms (Plate III) furnish no clue from the obverse to the order of their striking. The one die link (the A die of nos. 1 and 2) is of no value, since it may very well be that no. 2 is only another specimen of no. 1. The condition of the reverse is so poor that, while the legible portion is almost exactly similar to the reverse of no. 1, the condition of the rest makes it impossible to say that the die is the same. On the other hand, the letter forms and their arrangement on the P dies show changes which furnish some basis for their sequence. P 1 (No. 1) has for A and for P; both of these forms occur on tribal issues. P 3 10 (no. 3) has for A, but P is regular; the A is tilted across the corner like the A on some Acanthus coins. 11 P 4 (no. 4) has both A and P in regular form, but the initial A is tilted in the corner. P 5 (no. 5) has both A's and the Δ tilted in the corners (again, similarly to some Acanthus coins) and some granulation within the four-part square. P 6 (no. 6) has a fine regular inscription with no tilted letters; there is a trace of granulation within the square. It is logical to conclude that the order given here is close to that in which they were struck. The obverses are very little different from those on the anonymous coins; the die-cutters worked in accustomed forms. The letters were something new and the coins show some uncertainty until P 6 is reached. Here, the letter forms are as good as those in the next Group.

B. The heavy tetrobols (Plate V) are the only coins in the group which arrange themselves with something approaching finality. As with the octadrachms, the evidence lies in the reverses rather than in the obverses, since the latter show little stylistic progress. The six punch dies show increasing ability on the part of the artists to master this new type, after the pathetic attempts to include the inscription. Die sequences, of which there are three, are marked by wear and the break in HP 6, which occurs after its first use in no. 31 and is greater in no. 33 than in no. 32. The order of this denomination can be regarded as secure.

C. The tetradrachms and the light tetrobols (Plates IV and V) are susceptible of two different arrangements: the one actually used in the catalogue and another that has been ignored because of the ambiguity of the evidence of the light tetrobols. The difficulty is caused by the duplication of reverse types in the two denominations; each uses both the helmeted head and the goat's head. The arrangement that has not been used would involve the simultaneous use, in both series, of these two reverse types. Stylistically, this is quite possible; TP 1, TP 2, TP 3 (nos. 8-10) and TP 7 (no. 14) seem equally early, likewise TA 5 (nos. 12 and 13) and TA 12 (no. 21) are close in style. The fact that there are six TP dies with the helmet and seven and one half 12 with the goat's head, taken in conjunction with the appearance of six punch dies each for the octadrachms and heavy tetrobols, hints that a figure near that number is the total for each denomination of Group I. Thus, there would be two series of tetradrachms, each equalling in volume the other denominations. Unfortunately, the lack of corroboration for this from the light tetrobols casts light on its validity. There, only one LP die (no. 34) has a goat's head, while all seven others have the helmet. Both types seem contemporaneous on both denominations.

It is attractive to suppose that LP 1 is the sole survivor of a group of six or seven punch dies of that type, particularly in view of the seven helmet type LP dies. If that could be assumed, then there would have been two issues of both tetradrachms and light tetrobols in Group I. But such speculation is too dangerous; one must confine himself to the extant material. Therefore, the catalogue arrangement has been based on the fact that subsequent Groups show as standard types the rider r., and goat's forepart for the tetradrachms and the horse and helmet for the light tetrobols. The orientation of the riders on the obverse of the tetradrachms affords some evidence that the catalogue arrangement is correct: the obverses of the helmet series all face right; the earlier of the obverses of the goat's head series all face left. The last one (no. 21) which has the best goat's head, has the rider of the obverse facing right, a position which is unchanged thereafter. The catalogue order is admittedly unsatisfactory, since there is no conclusive evidence to justify either the arrangement used or the one discarded.

The artistic skill of the makers of the dies of Group I has already been mentioned casually. It deserves some special attention, in view of the signs of vigor and ability of the artists. Their mistakes are as revealing as their successes. There is no need to reproach them further for the first sorry lion reverses on the heavy tetrobols. One difficulty they never succeeded in overcoming (this is true even in Group III) 13 appears on those obverses where the rider is carrying his spears on the outside (that is, in his right hand if he is facing right, and vice versa). All their skill in differentiating planes and indicating depth was not sufficient to make the spears appear separate from the body of the horse; they curve around it in most unconvincing fashion. (See Plate III, 1, 2, 4 and 5, and Plate IV, 11). A touch of realism on those riders facing left appears in their costume. The chlamys normally was fastened at the right shoulder, thus leaving the right arm free for action; the artist remembered, in showing the other side, that the chlamys covered his left arm completely and he indicated the folds. On the dies where the rider is facing left, the spears are held in the right hand, which is free. This touch of realism raises the question how such a rider managed his reins (held in the left hand) and how the riders facing right, who hold the reins in the right hand, managed to keep both spears in the closely wrapped left hand. These artists never resort to the device used elsewhere (and in Macedonia on the coins of Archelaos) of having the chlamys hang free from the shoulders to enable the wearer to use both arms.

All the artists, and it is possible to distinguish the work of individuals, indulged in freedom in details: the angle of the horse's foreleg, the shape and size of the petasus, the length of the chlamys, and the pattern of its folds, are individual matters. One trait they had in common was the decorative use of the drill. The matter of the dotted borders has already been remarked. Dots are used for the exergual line, for the horse's tail, and sometimes for the edge of the lion's or goat's body. Some dies, notably the obverse of No. 5 (Plate III) and of nos. 16 and 17 (Plate IV), show an incomplete masking of the original use of the drill along the upper part of the outer hind leg of the horse. In most cases, however, traces of the drill are blotted out unless they are to serve as decoration.

The helmets present peculiarities of detail. The form of the crest is different from that on most other coins; in fact, many helmets on the tribal issues have no crest at all. Such helmets as have been recovered from excavations in Thrace and Macedonia 14 do not have crests. Their material is of course so much more perishable than that of the helmets that their non-survival is not surprising; it is in some cases impossible to tell whether the helmets recovered had a device to which a crest like that of the Macedonian could be attached. At any rate, the Macedonians had a crested helmet; it is not safe to call it either Corinthian or Attic. The shape of the bowl itself has its counterpart on sixth century vases, 15 but the shape of the crest and its close attachment to the bowl can be paralleled only by some of the helmets used as a reverse type or as a symbol on some of the tribal issues. 16 These helmets, as well as those of Group II, have a nose piece as well as cheek and neck pieces. The artists in all cases inserted a dot for the eye, although there is no other evidence that a helmeted head, rather than simply a helmet, was the type. There is no bit of the neck visible, either below or in the space between neck and cheek pieces. Since it is hardly conceivable that an eye shield was intended, the appearance of the dot for the eye may be considered a bit of archaic attention to detail such as is often evident on early vases.

Since the period of Group I is precisely defined, between 480/79 and 477/6, the extant dies and their combinations furnish indications of length of use of the dies. It will be remembered that in Chapter IV evidence was cited for concluding that normal procedure in the Macedonian mint involved the use of three anvil dies with onepunch. While many dies are lacking from Group I, particularly of the octadrachms, helmet series tetradrachms, and goat series light tetrobols, the survival rate of die combinations and specimens in the heavy tetrobols, the goat series tetradrachms, and the helmet series light tetrobols, makes it possible to draw some conclusions about the original number of dies in use and of the number of dies normally used per year in the Macedonian mint.

The numbers six, seven, and seven and a half (the half indicates that a die was used first without a symbol and later recut with a symbol: possibly the number should be eight) for P dies and seven, nine, and eleven for A dies lead one to believe that nearly all the punch dies survive. For a period of four years, the rate of two punch dies would demand eight P dies for Group I; any other calculations would not account for the proportion of die combinations to extant specimens in Group I.

If three A dies were used with each P die, the maximum number of A dies for the four-year period would be twenty-four; this would be the case if no A die was used with more than one P die. The die sequence charts (pp. 69-70) show, however, that this was not the case. Every series except the heavy tetrobols shows one A die link. Since three A dies were in use at once with one P die, the minimum number of A dies for the four-year period would be twelve; eleven of the number survive in the heavy tetrobol series, which is the only series not to show A die links. Therefore, the number must be fifteen, eighteen or twenty-one A dies for the four-year period. Fifteen would not be sufficient to compensate for the lack of A die links in the surviving heavy tetrobol combinations; twenty-one is an unnecessarily large figure - possible but larger than survivals from this Group or any other warrant. It seems reasonable to posit eighteen A dies for this four year period, making the life of the three A dies used at one time about two-thirds of a year.

The two series of tetradrachms present as great a problem in this connection as in the matter of their arrangement. There are enough surviving P dies in each to lead to the conclusion that both series were being struck at the same time, and the total number of A dies is only one more than in the heavy tetrobols. Unfortunately, the only A die links (see p. 70) are within each series, not between them. Here, the question of wear on the dies has some bearing. Nos. 12 and 13 use the same A die. In the first specimen of no. 12 (Plate IV), the beginning of a break at the right foreleg is barely visible; in the second (wt. 10.80) there appears to be repair on the coin itself, a cutting away to sharpen the outline of the leg to such a degree that the weight of the coin has been affected. In no. 13 (not illustrated), there are three specimens, showing an increase in the size of the break at the forefoot and another at the tail. The P die used in no. 12 shows no sign of deterioration in either specimen; therefore, the P die used in no. 13 was not a replacement for a worn die. Much the same facts appear from an examination of nos. 16 and 17a (Plate IV). There the anvil die shows no more wear in one than in the other; the order of striking could not be determined if the P die of no. 17 were not that of no. 16 with the caduceus added. In the five specimens of no. 17, the P die shows progressive wear at the edges of the incuse and linear square and the sharpness of the outline of the A die is reduced after no. 17a. It is conceivable that the six (?) missing anvil dies for the tetradrachms would provide links between the two series to show that they were struck simultaneously, for a year or two at least.

The light tetrobols have only one surviving goat's head reverse die, which makes it rash to believe that the two series in this denomination ran concurrently, even though they may have done so in the tetradrachms. Until more dies turn up, the arrangement of the two denominations must be left uncertain; there is no conclusive evidence pointing in one direction. At least it can be said that in the helmet series of light tetrobols and in the heavy tetrobols, a total of eight P dies and eighteen A dies was used for the four years; the same is probably true for the tetradrachms and may be true for the octadrachms, although here few anvil dies survive.

One other factor must be taken into consideration: the relation of the weights of the coins to those of communities with which Alexander came in contact. In the Thraco-Macedonian area, the weights of his coins were the same as those of other issuing agents. Greece proper, with which he sought to unite himself during the Persian Wars and with which he hoped to cooperate subsequently, had no coinage with which the octadrachms (wt. 29.46) were readily exchangeable. It may be that a smaller number of coins of this denomination was struck. The tetradrachms exchanged with Attic tetradrachms at the rate of four to three (13.08 × 4 = 52.32 and 17.44 × 3 = 52.32), a heavy tetrobol and a tetradrachm together equalled a Chian-Rhodian tetradrachm (15.52) and the light tetrobol was exactly equivalent to an Attic tetrobol. Therefore, the numbers of coins struck in all but the largest denomination would be consequent on a far larger trade potential than would the octadrachms.

There are a number of small coins of various recognizably Macedonian types which are of the same fabric and style as the catalogued pieces in Group I and must therefore be considered a part of that Group. The main varieties are illustrated on Plate V, a-e. All but one have the anonymous four-part linear square in the incuse on the reverse; one has a goat's head right. The first four might be considered as belonging in the period of the uninscribed Macedonian coinage discussed in Chapter III, and, indeed, there is no certainty as to their date, except that they cannot be later than Group I. All obverses show a horse, its forepart, or its head in a dotted circular border. The small size of these coins makes it almost impossible to say that they are or are not stylistically contemporaneous with Group I, since we have already noticed the close similarity in style between the uninscribed earlier octadrachms and tetradrachms and those of Group I. Plate V, a and e, are particularly nondescript.

It is well-nigh impossible to give a name to the majority of these fractions, or to assign them to the standard of the octadrachms or the tetradrachms, for two reasons: (1) The extant weights have no correspondence with the types, and (2) the original difference in the norms is in itself very slight. Given ancient irregularities of striking, one cannot distinguish denominations. The norms are as follows:

The weights of the coins illustrated on Plate V are: a, .46; b, 2.22; c, 1.03; d, 1.05; e, .85. It will be noticed that b could be a light tetrobol, or even a heavy tetrobol, for the list appended at the end of Chapter II shows some specimens of both denominations which weigh less than Plate V, b.

The types do not afford a criterion for determining denominations. The difference in weight between b and c - more than a gram - although both coins have the same type, makes it difficult to believe that they were intended to represent the same value. It might be conjectured that b is a light tetrobol of the period before Group I, but the style of its reverse is finer than most of the coins of that time. The forepart of a horse on the obverse of d contradicts its weight, by which it should be the same denomination as c, possibly the diobol on the tetradrachm standard, but such a conclusion leaves out of account the difference in type. Plate V, d and e, both with the forepart of a horse, should be fractions of something - of a light tetrobol and a light triobol? But here again is a duplication of obverse type of coins of two different denominations. The reverse type on e, the forepart of a goat, is stylistically contemporaneous with the tetradrachms (see Plate IV) and one light tetrobol (Plate V, no. 34) of Group I. Plate V, a, wt. .46, is obviously an obol, but with the difference in the norms only .07, it is impossible to say which obol this coin is intended to be.

There are other coins, as yet unassigned, which have the forepart of a horse on the obverse and a kneeling goat on the reverse (Gaebler, AMNG III, 2, pl. XXVII, nos. 2-5) and one (pl. XXVII, no. 6) with the same type of goat on the obverse and horse's forepart on the reverse. The weights range from 1.00 to 1.78. These coins may be Macedonian, struck ca. 492 when the Macedonians under Alexander joined the tribal alliance against the Persians. It is as difficult to determine their denominations as to explain the reversal of obverse and reverse types.

The knowledge that we possess of the kingdom of Macedonia under Alexander after the defeat of the Persians is very scanty, actually not so much knowledge as inference. The events of the "Pentecontaetia" as recorded by Thucydides 1 do not include any action by Alexander, but, since the locale of the earliest action of the Delian Confederacy, the reduction of Eion by Cimon, is close to Macedonia, we may be sure that Alexander was concerned in it. Remembering the king's words to the Athenian generals when he came secretly to them on the eve of the battle of Plataea, 2 we can imagine the eager anticipation with which Alexander awaited the arrival of the Athenian fleet. Plutarch 3 is the only ancient author to mention Alexander in this connection; he says that one of the charges against Cimon in 461 was that he had been bribed by Alexander not to seize Macedonia when he drove the Persians out of Eion. A nice bit of calumny which arose out of events some ten years later than the time to which it was assigned.

The coins given here as belonging to Group II show first the alacrity of Alexander to assist the Athenians and later his determined opposition to them when in 466/5 they made their first attempt to found a city on the site of the later Amphipolis. The changes in denomination, type and style of the coins show that Group II began as an issue designed (a) to exchange readily with Athenian currency, (b) to rally the moribund, if not defunct, tribal alliance to work with the forces of the Confederacy, and (c) to provide an earnest of Alexander's membership in the Hellenic race. A little less than ten years after these changes were made, they were modified in a way that reflects the disillusionment of Alexander about the aims of Athens.

a. The change in denomination was a simple one: for the heavy tetrobols (2.45) was substituted a unique issue of octobols (4.36). This substitution made the exchange between Attic and Macedonian currency a simple affair, for the octobol was the equivalent of an Attic drachma, 4 the light tetrobol was in weight an Attic triobol, and the tetradrachm of 13.09, of the worth of three Attic drachms. The octadrachms, which were not readily exchangeable with Attic currency, were nevertheless retained, even though their fraction, the heavy tetrobol was intermitted. The reason, I think, is not far to seek. Alexander needed to make purchases in the north, possibly of supplies for the Athenian fleet, for which he would need the large denomination in the currency of the region. All the smaller coins struck were, of course, a part of the Thraco-Macedonian system, 5 but they were also Attic in weight. As we do not know the rate of pay at that time for Athenian troops, 6 we cannot assume that the light tetrobols (an unusually heavy issue) and the octobols were designed to help Cimon meet his payroll. It is obvious, however, that these small pieces, as well as the considerable variety of fractional issues (Plate IX, a-e) would facilitate trade by the troops with Macedonian χαπήλοι.

b. The changes in type, which are in a measure consequent on the change in denomination, are significant of Alexander's interest in the earlier tribal alliance, which had turned military in 492. 7

I. The type of the octadrachms is once again the horse and attendant of the anepigraphic octadrachms issued in the late sixth (?) and early fifth centuries. At least one Bisaltian octadrachm of the same type (Plate II, 3, from the Hunterian Collection) must be dated by its style to this same time. 8 The retention of Alexander's name on the reverse shows that he was not merely a participant in the alliance as he had been earlier, but its leader. 2. The tetradrachms, whose type shows no change beyond the substitution of the goat's forepart for its head, have the initial of Alexander's name added to the obverse, at the exergual line, half-way through the series as it now remains. 3. The type of the octobols, obverse and reverse, is the same as that of the octadrachms. This is a major change, for the octobols are metrically related to the tetradrachms, not to the octadrachms. It will be remembered that in Group I (cf. pp. 87-88) the types tended to emphasize the metrological variants. Here, the variants are linked by type. 4. The light tetrobols, while retaining the type which became fixed before the end of Group I (horse-helmet), form two separate series marked by divergencies of style. One series has the letter A above or below the horse on ten of its anvil dies. Five of the other six anvil dies are linked by punch dies with the monogram dies; the sixth shows by its style that it belongs in the A series (A 19, no. 86). The other series has the letter H on five of its anvil dies; the remaining three are linked to them by punch dies. The horse itself on the H series, is not a Macedonian horse (see p. 114). The combination of the "foreign" horse and the letter H, which is hard to connect with Macedonia, 9 are additional indications of the more than national significance of this group. The intermission of the heavy tetrobols leaves their reverse type, the forepart of a lion, unused in Group II. Since this is the only type used by Alexander which did not have a place in the tribal issues, 10 its absence increases the "international" purpose of the Group. The strictly national, or regal, nature of Group I has thus been altered in two respects: (1) exchange with Athens was facilitated and, (2) emphasis was placed on those types which had more than national meaning.

c. A comparison of Plates III, IV and V with Plates VI, VII, VIII, and IX will show clearly the great improvement in style in Group II. Since the group follows immediately upon Group I, the change must be due to a fresh artistic impulse for the origin of which it is natural to look to the south. The greater skill in the execution of the dies is more readily apparent on the octadrachms and tetradrachms. The smaller denominations do not offer quite the scope for beautiful work; the changes, while marked, do not result in such splendid improvement as on the larger pieces. While the interpretation of stylistic change, or variation, is very subjective, 11 it is possible to identify the work of southern artists on some dies. Since the reverses of the octadrachms offer an opportunity for little more than fine letter cutting, the obverses only need be considered. No. 45 (Plate VI) is probably the best and has been put first for that reason. The proportions of horse and attendant are well calculated; their poses are relaxed yet alert; the head (of the attendant), hair and petasus, as well as the upper part of the chlamys, show a skill which is entirely lacking on the octadrachms of Group I. The horse is sleek and subtly modelled, very similar in its restraint and anatomical simplicity to the handsome and spirited bronze horse in the Metropolitan Museum, dated between 490 and 480. 12 Details of mane, hooves and muscles show a close relationship. On the coin, the composition is perfectly adjusted to the available space. Unfortunately, the only specimen is marred by a deep slash at the left side, but the beauty of the piece is still unmistakable. The other octadrachms, except no. 46 show the influence of this one, which surely was made by a Greek, probably an Athenian die-cutter. The smoothness of the musculature, the simplicity of modelling, and the composition of nos. 47-57 (Plates VI and VII) are witness to the ease and speed with which Alexander's die-cutters learned new methods. It is chiefly in the stiffness of the heads of horse and attendant, in the awkwardness of the bunched folds of the chlamys at the neck, and in the proportions of the petasus that these anvil dies fall short of the first one.

The obverses of thetetradrachms (Plate IX) show an improvement over those of Group I (Plate IV) comparable to that of the octadrachms, but there is no anvil die demonstrably superior to the others. As will be seen below, one of the octobol dies enjoys this distinction, which fact leads to the conclusion that an Athenian (or, at least, a Greek) artist made one large and one small die as a model for Macedonian workmen. Every coin in Group II shows this diecutter's influence, for all the dies are marked by a smoothness and "polish" which actually enhances the vitality already evident in the heavy, rather clumsy, horses and riders of Group I. Since Group II follows immediately upon Group I, the improvement cannot be con- sidered as having its source in Macedonia, particularly because of the rather slight advance of the coins of Group I over those of the period just before it, when the Macedonian coins were struck; compare Plates II and III and contrast them with Plates VI, VII and IX. The reverses of the tetradrachms, with the forepart of a kneeling goat, are much more elegant than those of Group I, although, as in the case of the obverses, there is no one die of marked superiority. There is every possibility that one will some day be found, to show that a southern artist made a model for this type also. TP 17, used with two different obverses (nos. 62 and 63, Plate IX) was originally miscut at the left side, but seems to have suffered no subsequent damage in use. There are five specimens of no. 62, among which none shows the A die badly worn; it would appear that TA 17 and TA 18 were in simultaneous use and that the miscut TP 17 was used because of the necessity of heavy striking in this series; production could not be held up while a better die was being cut.

An examination of the light tetrobols shows an advance in style over those of Group I similar to that of the other series, but as in the case of the tetradrachms, there is no anvil or punch die (with the possible exception of LP 22) which shows clearly that it was made by a southern artist. The reverses show a marked variety in the shape of the helmets, which retain the nose piece shown in Group I. In general, it may be said that the chief change is in the elongation of the cheek piece, although one P die in the earlier Group, that of nos. 39 and 40 (facing 1.) anticipates this lengthening. Likewise, in most cases, the bowl is more sharply differentiated from the remainder of the helmet. The helmets themselves retain the close-fitting crests which seem to be exclusively northern; the articulation of the members may possibly represent (at least in some cases) an interpretation of the northern helmet by southern artists, or the form of the helmet itself may have undergone some modification at this time. One LP die, that used on the last coins of the H series (no. 107, Plate IX) deserves special mention. Here, a head, with the neck visible below, is shown in a helmet with a most peculiar crest, curving below the base of the neck. The die is of singular beauty and has no affinities among the reverses of the other light tetrobols. It does not recur in subsequent groups. This may have been the one die where a southern artist used a type of helmet, fanciful or otherwise, which pleased him as an artist, instead of modifying the regular type.

While the horses of the A series of light tetrobols call for no comment beyond the mention above of their advance in style over those of Group I, those of the H series are in a special category. (See Plate IX). They are a slim, wiry breed of horse, not at all like the stocky or rangy horse which is customary on Macedonian coins, with the sole possible exception of No. 44 in Group I (Plate V). Even here the similarity is not great. When in motion, and motion beyond the tentative raising of a forefoot is not characteristic of the horses on Alexander's coins, these horses spring forward in a manner that cannot be described as either prancing or galloping, since all four legs are angularly bent, the foreknees almost touching the chin. On the much later tetrobols of Perdikkas, when the regular Macedonian horses start prancing (Plates XII, 163, 164, 166-9, XIV, passim, and XV), then galloping (Plate XIII, 170-175), they do so in the manner one expects of a horse; these are quite different. Although the action is more violent, the horses themselves bear a close resemblance to those on the earlier staters of the Ichnaians, Orrheskians and Tyntenians (Plate II, 10, 11, 13), in pose and proportions. These coins have already been discussed in Chapter III. Their archaic style is not in evidence on these light tetrobols; the advance in technique is comparable to that of the remainder of Group II. The horses are so different from those on the A series of the same denomination that the one unlinked, anepigraphic, obverse die (that of no. 86, Plate VIII) automatically takes its place in the A series because of the style of the horse. 13

The style of the octobols (Plate VII) is very like that of the octadrachms. Nos. 66-68 may well be from a die cut by the artist of the octadrachm, no. 45. There is the same easy pose of both animal and man, the same slight forward tilt of the man's head, and the same soft folds of the chlamys at the neck. Nos. 69-72 are more like the octadrachms nos. 47-57, showing a smoothness of modelling, but a certain stiffness of pose. The A dies of the last three octobols, nos. 73-75, present a peculiarity which must be taken in connection with the two octobols illustrated on Plate VI, a and b, and the octadrachm no. 46. The latter two octobols have their obverses from the same die; one reverse has the Bisaltian ethnic, the other has the name of Mosses. 14

This group of A dies has the lower part of the attendant's chlamys visible below the body of the horse. All three (SA 7, SA 8, and that of the two on Plate VI, a and b, have various slight differences: that of those on Plate VI has a small attendant, with a rather long-bodied horse; the composition is not so well fitted to the field. That of no. 73 (SA 7) is the only die either of the octobols or the octadrachms where the attendant is taller than the horse, although the composition as a whole is better than that of the preceding die. The die of nos. 74 and 75 (SA 8) is the best and was, I believe, made by a foreign (southern) artist. It has the feature that was imitated on the octobols just mentioned, the chlamys hanging below the body of the horse, but it has a refinement which was not appreciated by the imitators. The horse is standing perfectly still; on the small compass of the octobol, this is a neater pose than the one suitable to the octadrachm, where the horse has his left forefoot raised. There is some question whether the arrangement of the octobols is justifiable, since the dies with the long chlamys, one possibly by a Greek, are put last. My confidence in the rapid response of Macedonian die-cutters to outside influence leads me to prefer this arrangement on the ground that the native artists would soon have seen effectiveness of the standing horse and would have made their dies accordingly. Since there is no wear on any of the linking P dies, the order could be changed, but there is no tangible, incontrovertible proof for either arrangement.

The octadrachm, no. 46, which is out of place anywhere in its series, looks like an attempt by a die-cutter who had more courage than skill to reproduce this type variant, for it is closely modelled after the die of nos. 74 and 75. It stands out among the octadrachms and has no compeers among the tetradrachms; its oddness is further emphasized by the fact that the spears are reversed. There is not the slightest reason for condemning the piece, which for all its awkwardness shows the vigor and originality always associated with northern artists, whatever their degree of skill. 15 The arrangement of the octadrachms whereby this coin is placed second in the series may be questioned in view of the fact that the octobols to which it is similar are placed at the end of their series. It has been done because with the exception of no. 45 and it (no. 46), the octadrachms are the work of the Macedonians, among which a little sequence dependent on the few linking dies is possible. Actually, it may be almost contempo- raneous with no. 53, the last before the waning moon symbol is added. The chart of die sequences at the end of Chapter IV shows how little of this series remains; consequently, any arrangement depends on intangibles. In this case, it seemed better to isolate no. 45, which is surely not by a Macedonian, and no. 46, which, although probably native, is quite unlike the others. If more coins come to light, a different order may be possible.

Before considering such meager evidence as there is for the duration of Group II, the relation between the peculiarities of the group and the events occurring in the north should be pointed out. At the beginning of this chapter, reference was made to Cimon's attack on Eion in 476/5; 16 there is no hint in Thucydides' abbreviated "Pentecontaetia" that Alexander was consulted or even informed in advance of Greek plans. The clue pointing in such a direction is the A die of the octadrachm no. 45. Since this obviously influenced the makers of the other dies, it was placed first; it is hard to account for the presence of an Athenian (or other Greek) artist in Macedonia at this time if Alexander was not transacting business with the Greeks and privy in their plans. So far as we know, the two regions were not in close contact with one another so early in the century; it looks as though the plan for the change in denominations and the importation of a southern artist were the result of some sort of planned cooperation between Alexander and the Greeks. In any case, whether these conclusions be justifiable or not - and if they are not, we have to assume that Alexander quite independently decided to be ready to work with the Athenians when they arrived - the coins themselves probably did not appear before Cimon and the fleet were in northern waters. Therefore, it is safe to date the beginning of the group in 476/5. This date (and, consequently, that of the end of Group I, 477/6) and that of the beginning of Group I in 480/79 are the only dates in the coinage of Alexander which can be fixed with any degree of assurance. The date of the end of Group II is hard to define and both beginning and closing date for Group III are somewhat arbitrary, as will be seen in Chapter VII.

From the coins, it seems that Alexander was reviving some portion, at least, of the old inter-tribal cooperation of the sixth century, which in the years of the Persian advance, notably 492 and following, had become a military alliance - probably only for defensive purpose. He succeeded in influencing the Bisaltians, the tribe over which Mosses ruled, the tribe represented by the H and the "foreign" horse, and possibly even the Edonians, to unite. These latter, who struck coins using the Apollo-Hermes types, lived a little above Eion on the Strymon (Thuc. I, 100); hence they would be concerned with Athenian action so near home. The probability of including the Edonians in the "alliance" is enhanced by the form of the letters on some of their octadrachms. These coins bear on the obverse the variety of the Apollo-Hermes type which shows two oxen and an attendant, who is presumably Hermes; 17 the type is thus similar to the form of the Ares type used always by the Bisaltians, with the ethnic on the obverse, used without an inscription earlier by the Macedonians (see Chapter III), and only here with Alexander's name. The form 18 of the A in the inscription on the reverse ΓEΤΑΣ BΑΣΙΛEΟ[Σ] HΔΟΝEΟΝ is the one which displaced the shortly after 479 on Macedonian coins and on various others in that area, and which is used on the coins of Acanthus, first inscribed at this time. 19 The earlier coins of the Edonians likewise have the form . It seems fair to conclude that Getas was a ruler of the Edonians contemporary with Alexander, who took part in the renewal of the old alliance revived to give assistance to the Delian League when its contingents moved north.

Some ten years after the liberation of Eion, the Athenians returned to the north to defeat the Thasians who had differed with them about the ownership of the mainland and the mines. Thucydides 20 states that at the same time the Athenians sent out 10,000 colonists from Athens and the allies to found a colony at Ennea Hodoi, on the Strymon, in Edonian territory. This action is generally dated in 465, 21 despite the statement of Diodorus that it took place directly after the release of Eion. Thucydides concludes his brief account of the first attempt at colonizing Amphipolis with the statement that the Athenians, after conquering at Ennea Hodoi, went thence inland and were utterly defeated by all the Thracians at Edonian Drabescus. There is no doubt that this abortive colony at Ennea Hodoi was as unpleasing to Alexander as to the Edonians. Of course, his realm did not yet extend beyond the Strymon, and we may be sure that any plans he may have had for such an extension were not known to Getas, but for other Greeks to seize land so close to him was as bad as having Persians in the vicinity. He would have felt doubly insulted, for he claimed Greek ancestry and was accepted as a Greek at the great sanctuaries of Delphi and Olympia.

It is of some significance that Thucydides, for all his brevity, mentions that the land was Edonian, but does not mention the king, and says that all the Thracians defeated the Athenians. In the first chapter, it was pointed out that the term "Thracian" and "Macedonian" were used rather indiscriminately by both Herodotus and Thucydides; thus the wording does not exclude Macedonians from participation in the conflict. This is the action of the revived tribal alliance, no longer cooperating with the Delian League, but opposed to its members. That Alexander was the prime mover is to be seen from the coins. The latest of his octadrachms (OA 13, 14, 15) carry in the upper left field the unusual symbol of the waning moon. This I take to be Alexander's way of making known his sentiments to the Athenians, who were at Thasos conducting a siege for two or three years after Drabescus. It was the symbol they had used to mark their defeat of the Persians, interlopers into their world. 22 Alexander used their own symbol to mark their defeat as interlopers into his world. 23

A curious aftermath of the two ventures of Cimon in the north comes out in Plutarch, 24 where the story of the charge of bribery is told. Plutarch states that when Cimon was accused of treason in 461, one of the charges against him was that he had taken a bribe not to invade Macedonia when he freed Eion. It can be conjectured that the Athenians, who felt keenly the defeat at Drabescus, willfully or otherwise confused the monetary assistance given Cimon by Alexander and the tribes in 476/5, and presumably for some time thereafter, with the continued issue of alliance coins and the defeat at Drabescus. It is in a sense the confirmation of the hypothesis that Alexander led the coalition in 465, since there is no hint in the ancient authorities that the plans of the Athenians in 476 included the annexation of Macedonian territory.

The terminal date for Group II is not readily apparent. I believe that the symbol on the octadrachms is to be dated after 465, but how much longer the issue lasted is quite indeterminable. The number of known dies, die combinations and specimens, and the die sequences cast no light on the matter. The die combinations are as follows:

The evidence from this chart and from the die sequences (Chapter IV) shows how few remain from the two large denominations. Using as a criterion the proportion of three A dies to one P die (see Chapter IV), one can easily judge how large is the number of missing A dies. There is no clue to the number of missing P dies. If we add to this scanty die survival the circumstance that among the octadrachms one of the three links is that where one A die is used with three P dies (nos. 49-51), a sign of increased mint activity, it becomes clear that we have only fragmentary remains of a heavy issue. The fact that there are only two more specimens than die combinations is another indication that many dies or their combinations are lost. The rate of survival of the tetradrachms is somewhat better: there are eight combinations and eighteen specimens, but no significant links.

The smaller denominations are better represented. The octobols are entirely linked, the links including two occurrences of one P die with three A dies. The rate of twenty-four surviving specimens for ten combinations is quite high. In the two instances where one A die is used with two P dies (nos. 66, 67, and 74, 75) none of the punch dies is worn, which may well mean that they were used simultaneously. The H series of light tetrobols is almost as unrevealing as the two large denominations: an equal number of both dies and only three more specimens than combinations. However, the die links, with two instances of one A die used with three P dies (nos. 99-101 and 103-105), provide the series with almost complete linkage and furnish the same evidence of heavy striking as was indicated in the octadrachms. The A series of heavy tetrobols shows the clearest signs of heavy minting, and it may be that this issue, which provided convenient Attic change and which was stamped as Alexander's, was more heavily struck than any of the others. Six P dies and eighteen A dies, twenty combinations with twenty-nine specimens, one P die used with five, and another used with six A dies, all indicate a large issue. The incongruity of the evidence is too great for any assumption that all denominations were struck in such quantities as the A series of light tetrobols, and the remains make it difficult to judge the length of the period during which Group II was struck.

Possibly some such conclusion as this might be reached. In 476/5 the issue was begun with octadrachms (without the moon), tetradrachms (probably without the A), octobols, and both series of light tetrobols, the A series in larger numbers than the other. 25 At this time also the Bisaltian octobols and those of Mosses were struck. 26 About the time of the Thasian revolt (466/5), and the coming of the 10,000 settlers, perhaps shortly before that time if Alexander heard about Athenian plans for Ennea Hodoi, the issue of octobols ceased and the letter A was added to the tetradrachms. The Bisaltian octobols and those of Mosses probably ceased with those of Alexander, since the need for that denomination had apparently never been very great in the north. The light tetrobols, even though they were the equivalent of the Attic triobol, had always been popular in the area, to judge from remains, and continued to be struck. After Drabescus (465), the waning moon was added to the octadrachms. The terminal date cannot be reached from consideration of such evidence as we now possess. The conjectured date of 460 is merely a suggestion, based partly on the small size of Group III. It is quite possible that Alexander struck no coins at all for some years following the waning moon octadrachms. When the Athenians retired from the north, after the subjection of the Thasians, the country may have been quiet and the currency already in circulation sufficient for local needs.

End Notes

1	The relevant chapters are 96-101 in Book I. Briefly, these chapters include the following facts. (The dates given are either those generally accepted or those of Gomme, A Commentary on Thucydides, Book I, pp. 361—413. He has made a thorough study of Thucydides' purpose and achievement in this section of Book I and discusses the problems of chronology resulting (a) from his brevity, (b) from the very different purpose of Plutarch in his Cimon and Pericles, and (c) from the unreliability of other chroniclers who have included mention of the interval between the Persian and the Peloponnesian Wars in their chronographies). 1. The first collection of tribute from members of the Delian League in 476. 2. The expulsion of the Persian garrison at Eion by Cimon in 476/5. (Herodotus, VII, 107, gives the story of the heroism of the Persian commander Boges; Diodorus, XI, 69, 1, states that the Athenians settled cleruchs there immediately after the expulsion of the Persians, although Thucydides does not mention it.) 3. The revolt of the Thasians in 466/5. 4. The Athenian punitive expedition of the same year and the dispatch of 10,000 settlers of the Athenians and their allies to Ennea Hodoi. 5. The defeat of the Athenians at Drabescus in 465. 6. The subsequent siege of Thasos, its request for Spartan aid in 464, the Spartan promise and failure of fulfillment because of the earthquake and the revolt of the Helots in the same year. 7. The final capitulation of the Thasians in 463/2, their relinquishment of claims to the mainland and their inclusion in the tribute lists.
2	Cf. Herod. IX, 44-45, and Chapter I.
3	Cf. Plutarch, Cimon, 14, and Pericles, 10.
4	See Chapter II, note 14.
5	The octobol does not seem to have been struck so frequently as other small coins, but it was in use from at least the middle of the sixth century.
6	Information on that subject is extremely scanty at any period. We know that it was raised to three obols a day under Pericles, or a drachma if the troops found their own food. At how early a period pay for troops began is uncertain, but perhaps the levying of tribute from members of the Delian Confederacy was occasioned not only by the need for supplies, but also for the pay of troops who might in a sense be considered as mercenaries. There are some references in Thucydides to rates of pay during the Peloponnesian War. The matter is discussed in CAH, V, 23-32, but for the period immediately following the Persian Wars there is no actual information.
7	Cf. Chapter III, pp. 57-59.
8	Cf. also Svoronos, HPM, Pl. XI, 6 (London) which has the form A in the inscription. There is no justification for the oft-repeated statement that "Alexander subdued the Bisaltians and adopted their coin-types". Cf. Babelon, Traité I, 2, p. 1037. The latest occurrence of the idea is in Edson, CP, XLII (1947), 88-105, "Notes on the Thracian 'Phoros'," 95, note 56.
9	I have no explanation for this letter; it might refer to Eion or to the Edonians. Neither is very plausible, since the Edonians used the Apollo-Hermes types and Eion used a water (?) bird on the few small coins generally attributed to it. The "foreign" horse on these coins, similar to the horse on coins of tribes or cities in the area, is not associated with any tribal group whose name begins with that letter.
10	Cf. Chapter III for a discussion of the origin of the lion type on Macedonian coins.
11	Cf. C.H.V. Sutherland, "What Is Meant by 'Style' in Coinage?" ANS Museum Notes, IV, 1-12.
12	Richter, Animals in Greek Art, figs. 62, 63.
13	For a similar horse on a gem in Boston, see Richter, Animals in Greek Art, fig. 60.
14	Bisaltian octobols of this style are rather rare. I have been able to find the one illustrated, from Naples, another in Klagenfurt (AMNG, Pl. XII, 11), and a third in Cambridge ( McClean Coll. II, 3105, Pl. 113, 8). The coin of Mosses illustrated on Plate VI is in Brussels. Gaebler condemns all the Bisaltian octobols of this style: the one in Klagenfurt he lists as a forgery without explanation; the other two he dismisses (Falsch. I, 195 ff.) as copies of the Mosses coin and claims that the reverses as well came from that coin, the forger having left the O of the Mosses inscription in that of the Bisaltian. I have not been able to examine the coins, but his arguments lack force. The Bisaltians struck octobols earlier (see Plate II) one of which (AMNG, Pl. XII, 8, from Berlin) has the monogram on the horse's rump. None seems to have been struck beyond the period of the revived tribal alliance in 492 except these listed here. The coins of Mosses, however, show a continuous development which sets them apart from the Bisaltian coins. For this reason, as well as because of the fact that it seems most unrealistic to assume that a ruler issued only small coins in his own name while issuing both large and small in the name of his people, the supposition that Mosses was a ruler of the Bisaltians is unwarranted. He may well have been one of those tribal chieftains mentioned by Thucydides (II, 99) who retained a measure of autonomy while paying tribute to Alexander. A number of his octobols are illustrated in AMNG, III, 2, Pl. XXVII, nos. 37-42.
15	Cf. Brunn, "Paianios und die nordgriechische Kunst," Sitzb. Bayer. Akad ., 1876,1, 1, 315. While it is perhaps erroneous to postulate a "school" in this area, nevertheless the coins show a homogeneous style. For other Greek artists working in the north, cf. the remarks of Casson, Macedonia, Thrace and Illyria, pp. 232-233 apropos of the coins of Thasos after 463. Significantly or not, these coins were found in the Tigris river in 1818.
16	See Note I.
17	With , Babelon, Traité, I, 2, 1458; pl. 45, 7.
18	With A BMC Mac., pr 144, 1 and 2.
19	Desneux, Les Tétradrachmes d'Acanthos, pl. XIV, 90-92.
20	This passage in Thucydides (1, 100), and another where reference is made to the first colonizing of the site in describing the settlement of Amphipolis in 437 (IV, 102) have been the subject of considerable discussion because of the ambiguity of the antecedent of αύτοί (1, 100) and οἳ (IV, 102). Who actually suffered defeat at the hands of the Thracians at Drabescus - the settlers, or Athenian troops from the scene of the siege of Thasos? Meritt-Wade-Gery-McGregor, The Athenian Tribute Lists, III, pp. 106-110, explain the passages sanely, referring αύτοί in I, 100, and οἳ in IV, 102, to the Athenians, after the defeat of whom the abortive colony at Ennea Hodoi was withdrawn. Following Isocrates (8, 86) both Walker, CAH, V, 58, and Gomme, Commentary on Thucydides, 1, 297, took it to be the settlers. That the military action, as contrasted to the attempted settlement, was undertaken by the military arm is, of course, only logical: Drabescus was at some distance from Ennea Hodoi; the settlers would have no occasion to stray so far inland. Further, Pausanias, I, 29, 4 states that a list of casualties on a stele was the first monument set up by the Athenians for military personnel, in 465/4. This stele is probably IG I 2 , 928, (cf. also ATL, III, 259) and confirms the conclusion that the defeat at Drabescus was suffered by Athenian troops, not settlers from Athens and the other allies.
21	Cf. Note 1.
22	Cf. Seltman, Athens, its History and Coinage, Chapter XIV, and Greek Coins, 91-92.
23	It may be mentioned that Walker, CAH, V, 58-59, without adducing any evidence from the coins, poses the question: "Can it be that Alexander had a hand in bringing about this disastrous defeat of the Athenians?".
24	Plutarch, Cimon, 14, and Pericles, 10.
25	The two series of light tetrobols are unconnected by P dies. It is impossible to determine whether they were contemporaneous or successive one upon the other. There is no linking die between them, although the style of the helmet on the reverse of both series is quite similar. It seems best to consider them roughly contemporaneous, with the H series perhaps beginning slightly later, when the tribe represented by that series was brought into the alliance; the two series would subsequently be struck simultaneously.
26	Possibly the octobols with Dionysiac types on the obverse and BEΡΓΑΙΟΥ around a four-part linear incuse square on the reverse (HPM, pl. X, 32; BMC Thrace, p. 205,1) were struck at the same time, by the people of the town Berga, on the upper Strymon. This place would be vitally affected by an Athenian colony at the mouth of the Strymon. It was tributary to Athens following 454, but was absent from the tribute lists during some of the crucial years before and after the founding of Amphipolis. Cf. Chapter VIII.

The fractional issues assigned to Group II are stylistically associated with that Group. Coin a on Plate IX has as obverse type a horse very like those on the" A" series of light tetrobols such as Plate VIII, nos. 85-88. The reverse is a four-part linear square in the incuse, with an upright caduceus in the center. None of the other fractions in this group or in Group I has the linear square within the incuse, although the incuse is quadrilaterally divided. The weight of the coin illustrated is 1.99 and others of the same type weigh about the same. This probably should be classed as a triobol on the heavy standard (norm 1.83), but it is to be noticed that many of the light tetrobols in all groups weigh no more than this. In Group II, the only coins in the catalogue which are on the heavy standard are the octadrachms. It is rather surprising to find any small coins at all on that standard during this period.

Coin b, which has a standing horse r., with a leaf above similar to that on the left on the reverse of the tetradrachm no. 15 (Plate IV), wt. .93, likewise appears to be on the heavy standard, a trihemiobol (norm .91), but since the light norm is .81 for the trihemiobol and 1.08 for the diobol, there can be no certainty about the denomination of this piece. Coin c, also a standing horse r., on the obverse, wt. 1.01, is probably the same denomination as b. It will be remarked that the coins here described show the same lack of association of a single type with a single denomination as was the case with the fractions of the first group.

Coin d is the lightest of the fractions in this group, wt. .53, and is certainly an obol, although one cannot specify the norm. This is an interesting type of fraction, the head of a man wearing a petasus. It is probably not intended as portraiture, but rather as shorthand for the warrior-hunter or the attendant of the horse on the larger coins. The occurrence of a head as type on a coin is certainly no rarity, but if its use here has a fractional value, as does the forepart of a horse on other fractions, it is unique as far as I know. The work is neat, although the dotted border is rather heavy, and the four-part incuse on the reverse is as anonymous as ever. Another similar coin, which must lie outside the Macedonian regal issues, should be mentioned. Unassigned, it has a head, wearing a petasus, and to judge from illustrations, it seems to be stylistically quite close to this coin, even in the matter of the rather coarsely dotted circle. The reverse, however, has an elaboration of diagonals and pellets within the four parts of the linear square which is quite foreign to the Macedonian mint. It is best illustrated in G. AMNG, III, 2, pl. 26, no. 17, wt. .43. It is conceivable, although incapable of proof, that this is a fractional piece issued by one of the tribes associated with the Macedonians first to aid the Athenians, then to hinder them. The source for the reverse type is unknown, although it could be taken as an elaboration of the reverse on the electrum pieces i and j on Plate I. So far as I know, it is not used again.

The fifth fractional type on Plate IX, e, wt. 1.05, perhaps belongs in Group I, since the helmet on the reverse is very like those in that group. It is a light diobol or trihemiobol and may be contemporaneous with Plate V, e. The obverse type is the same and both reverses, helmet and goat's head, had been associated in Group I with coins of the light standard, the tetradrachms and light tetrobols. Because the execution of the coin, Plate IX, e, appears better than that of Plate V, e, it has been assigned to Group II.