George MacDonald.

The silver coinage of Crete, a metrological note online

. (page 1 of 4)
Online LibraryGeorge MacDonaldThe silver coinage of Crete, a metrological note → online text (page 1 of 4)
Font size
QR-code for this ebook

, JB -

a I

KS '> . " S




The Silver Coinage of


A Metrological Note


George Macdonald, C.B.

Fellow of the Academy

[From the Proceedings of the British Academy, Vol. IX]


Published for the British Academy

By Humphrey Milford, Oxford University Press

Amen Corner, E.G.

Price Four Shilling-'; net




By George Macdonald, C.B.


Communicated December 10, 1919.

A year or two ago I was able to acquire for the Hunterian
Museum an extremely interesting silver stater, struck at the Cretan
city of Itanus. It had formed part of a small collection belonging
to Sir Henry Yule, the editor of Marco Polo ; but there was nothing
to indicate how or when it had come into his possession. It appears
to be restruck, as are so many other Cretan coins. The style is
good, and points unmistakably to a date circa 400 b.c The descrip-
tion is as follows :

Helmeted head of Athena 1. [l]TANin[N] Eagle standing L,

looking backwards ; in field r.,
small Triton ; incuse square.

[Plate, Fig. 1]

The types, it will be seen, are not unfamiliar. What makes the
coin remarkable is its weight ; it turns the scale at 14*58 grammes
(225 grains), whereas the sixteen other known specimens of its
class range from a maximum of 11*74 grammes to a minimum of
10*18. The difference is so great as to be explicable only on the
hypothesis that the standard is not the same. No clue to the puzzle
could be found in the recognized authorities. A systematic survey
of Cretan coin-weights was accordingly undertaken, in the hope that
a clearer light would be thrown upon the matter. The results can
make no claim to finality. But they do seem to carry us some way
farther forward, and it may therefore be worth while putting them
on record. Professor Gardner's recent book affords convincing proof
of the importance of this obscure and difficult department of numis-
matic study.

From more than one point of view Crete offered special advantages
for a survey of the kind. The island is a limited and self-contained



area, and there was already in existence what may not unfairly be
called a corpus of its coins. 1 That is to say, the material was not
excessive in amount, and much of it was readily available. On the
other hand, its proper chronological arrangement will always be
exceptionally hard to determine. Unskilled engravers were employed
in the Cretan mints far more freely than in the mints of any other
part of the Hellenic world. Consequently, in applying the criterion
of style, it is not always possible to distinguish between degeneration,
strictly so called, and mere clumsiness in the copying of first-rate
models, such as is apparent in Plate, Fig. % when contrasted with
Plate, Fig. 1. The broad lines of development are, indeed, obvious
enough. It is in the handling of individual specimens that trouble
is apt to arise. Yet, for the right solution of a metrological problem,
reasonable precision of date may be hardly less vital than reasonable
precision of weight. At every step, too, progress is more or less
seriously hampered by the general limitation that besets all inquiries
into Greek coin -standards uncertainty as to what should be regarded
as the norm for each particular group of pieces. Even well-preserved
specimens struck from the same dies hardly ever agree exactly, and
sometimes the variation is considerable. The truth would appear to
be that. Greek mint-masters had perforce to acquiesce in rough-and-
ready methods of calculation. In default of elaborate mechanical
appliances, they probably put their chief trust in the acquired instinct
of their skilled workmen, who would be given a definite quantity
of metal with instructions to produce from it a specified number
of coins.

Current Views: Need for their Revision.

According to the hitherto accepted opinion two standards, and
two standards only, were employed in Crete prior to the introduction
of the Imperial coinage : the first was a debased form of the Aeginetic,
and the second was the Attic, which began to be used circa 300 b. c.
and which ultimately prevailed everywhere. The conclusions thus
summarized are those of the Historia Numorum, and they therefore
represent the outcome of a review which was broad and comprehensive
rather than detailed. Closer examination shows that considerable
modifications are needed. It reveals the fact that, if we are to get
at the truth, we must look not merely at the coins as a whole, but

1 J. N. Svoronos, Numismatique de la Crete ancienne (Macon, 1890). During
the past twenty or thirty years many additional examples have been published in
sale-catalogues and elsewhere. Due account was taken of these in the statistics
that were compiled for the purposes of this survey.


to some extent also at the issues of individual cities. During the
fifth and fourth centuries, more especially, there was far less homo-
geneity than is usually supposed. Once this has been realized, the
problem assumes a new aspect. Meanwhile it can be most con-
veniently approached through the quotation of Head's ipsissima
verba 1 :

4 Down to this time [i. e. about the end of the fourth or the
beginning of the third century B.C.] the weight standard employed
throughout the island had been the Aeginetic, or more properly
a debased form of the Aeginetic approaching in weight to the
Persic standard which prevailed along the south coasts of Asia
Minor and in Cyprus. After the age of Alexander, whose coinage
has left but slight traces in Crete (although the absence of Cretan
coins in the third century suggests the inference that the currency
of the island was at this time Alexandrine), the Attic standard
creeps in and replaces the older Aeginetic. In the second century
a general revival of the coinage takes place, at first on the pattern
of the new Athenian tetradrachms, which afterwards give place to
local Cretan types. This coinage continues sporadically until the
conquest of Crete by Q. Caecilius Metellus in b.c. 67 when auto-
nomous issues for the most part appear to have been put an end to,
until, in the time of the Empire (Augustus to Trajan), a new
Romano-Cretan coinage makes its appearance.'

Prevalence of the Aeginetic Standard Proper : Two Norms

There can be no doubt as to the Aeginetic standard having been
employed in Crete from the earliest period of coinage there, or in
other words from circa 480 b.c onwards. But it is hardly justifiable
to describe the Cretan form of it as 'debased 1 , since that term,
as applied to money, conveys a suggestion of deliberate intention.
The fall in weight can be accounted for in a perfectly natural way.
Wherever a coin-standard spread from one region to another by the
method of gradual infiltration, some reduction during the process was
virtually inevitable. The phenomenon recurs again and again, and
it can be very simply explained. We have seen that in Greek mints the
weighing of blanks was not carried out with the scrupulous nicety
that modern invention has made possible. Under Gresham's Law
it would doubtless be the heavier specimens that would find their way
abroad. But, in the course of their travels, even these would be
brought below the normal through the wear and tear of circulation.
It would thus be 'specimens reduced by a natural and inevitable
process that would serve as a model when the standard came to

1 H. N.\ pp. 383 f.



be introduced elsewhere. The more numerous the intermediate
stages, the lower would be the level that was finally touched. The
Aeginetic standard had, of course, to travel southwards through
the Archipelago before it could take root in Crete, and we catch an
instructive glimpse of it upon its journey. The norm for the silver
stater of Aegina in its original home may be put at 12-60 grammes,
although that maximum is rarely attained and still more rarely
exceeded. By the time it had reached the Cyclades in its progress
southwards, it had lost fully half a gramme. Imhoof-Blumer says of
the archaic coins of that island group, which are somewhat older
than the earliest of the Cretan issues, that ' the weight of the best
preserved staters is twelve grammes on the average, and is in point
of fact more frequently below that figure than above it V Assuming
the same ratio of decline to be maintained, we should expect to find
the Cretan norm in the neighbourhood of 11*40 grammes. What
are the facts ?

If we were to take Crete as a whole, disregarding for the moment
the differences between individual cities and making no allowance for
the way in which the condition of specimens may vary, 2 we might fix
as a rough average for the three chief denominations

Staters or Didrachms .... 11 grammes.

Drachms 5*40 grammes.

Hemidrachms 2*60 grammes.

These, it must be remembered, are averages only. If the figures
relating to the staters could be set out in the form of a graph, the
curve might rise as high as 12*25 on one side of the line and might
sink as low as 9*40 on the other, 3 an excess over 12 being, how-
ever, of more common occurrence than a fall below 10. The other
denominations, if similarly treated, would show a similar oscillation,
although the distance between the extremes would naturally be less.
Tetrobols were occasionally struck, 4 and it is important that these

1 Griech. Miinzen, p. 13.

2 However desirable it might be to allow for this factor, it is not possible to do
so, since the c sources ' seldom trouble to note the condition of the pieces that
they register.

3 Even these limits are exceeded in a few exceptional cases. Thus Svoronos
(op. cit., p. 160, No. 25) records a stater of Gortyna, now in Berlin, which is said
to weigh 12*95 grammes. A misprint is, of course, possible, and at present the
statement cannot be verified. At the other extreme is the stater of Itanus
which was published in Num. Chron., 1913, p. 384. It more than
7"68 grammes, but is admittedly c badly worn, hence its low weight'.

4 I have noted an unmistakable group of four at Cnossus, with an average weight
of 3'91 grammes. Two of these are in the Hunter Collection (Catalogue, ii.
p. 175, Nos. 15 f.), and the latter of them is figured here (Plate, Fig. 7).


should not be mistaken for abnormally light drachms or abnormally
heavy hemidrachms. Where, as at Cnossus, there are drachms and
hemidrachms belonging to the same issue, the distinction is quite
clear. Smaller denominations, such as obols, hemiobols, and tri-
hemiobols, are not infrequent, but in the present state of our
knowledge an examination of the weights of these would not be
worth the trouble it would cost.

Based though they are on the coinage of the island as a whole, the
averages given above are sufficiently high to make it practically
certain that the reduction which the Aeginetic standard underwent
in Crete was the outcome of purely general causes. At the same
time it is not altogether satisfactory to be confronted with a stater
of 11 grammes where one of about 11-40 was anticipated. It there-
fore becomes necessary to see whether a better result can be got by
considering the issues of individual cities. Even a superficial scrutiny
brings marked differences to light. Thus, out of 211 staters of
Phaestus, of whose weights I have been able to secure a record,
there are only 17, or 8.05 per cent., that fall below 11 grammes.
At Gortyna with 198 staters the corresponding percentage is 10*1.
At Cnossus, on the other hand, with 128 it rises as high as 41*5.
A contrast so startling affords prima facie grounds for suspecting the
presence of two distinct norms. And, if we probe the matter farther,
we shall find that the suspicion is more than justified. Here a brief
digression on procedure is required.

The difficulty of determining the norm in any particular case has
already been alluded to. The obvious plan is to calculate the
average weight of all the members of the group in question. Indeed,
where only a few coins are available, there is nothing else to be done.
At the best, however, the results are apt to be of doubtful value ;
two or three abnormal pieces may disturb the balance of the whole.
After a good deal of experiment I am convinced that by far the most
trustworthy method is one originally suggested by M. E. Babelon, 1
and first turned to practical account by Mr. G. F. Hill 2 the con-
struction, with the aid of squared paper, of what may be termed a
table of frequency \ Each column of squares is regarded as repre-
senting some small fraction of the unit, a twentieth (say) of a gramme,
and a series of such columns is marked accordingly, beginning with
the heaviest coin in the group and continuing in regular descent until
the lightest has been reached. Within the limits thus staked off
the weights of the various specimens are then indicated by dots

1 Traite, i. pp. 577 f., note 4.

2 Num. Chron., 1906, pp. 342 f.


placed inside the blank squares, one to a square and each in its
appropriate column. The outcome is a complete conspectus of the
weights of the group. While abnormal examples are revealed as
abnormal and are not allowed to mar the general effect, the approxi-
mate position of the norm stands out conspicuously in the shape of
a solid block of dots, very often crowned by a more or less prominent
projection. 1

As was hinted above, it is essential to the success of the method
just described that the coins available for the construction of the
tables should be reasonably numerous. Consequently, although an
attempt was made to apply it to the Aeginetic staters of all of the
Cretan towns, there were only four cases in which a perfectly certain
and definite result was obtained. These were Gortyna, Phaestus,
Cnossus, and Cydonia. At the two former towns the norm lay
between 11-20 and 11-85 grammes, being slightly heavier at Gortyna
than at Phaestus. At the two latter it lay between 10-65 and 11-30,
being rather lighter at Cydonia than at Cnossus. We shall not be far
wrong if we fix it at 11-55 for Gortyna and Phaestus (with a tendency
to sink lower at Gortyna and to rise higher at Phaestus), at 11-10 for
Cnossus, and at 10-95 for Cydonia. 2 The drachms of the same four
towns, when tested in the same manner, bear witness to a similar
difference of norm. Further, the evidence from the remaining cities
of Crete, incomplete as it is, is useful for confirmation ; in the great
majority of instances the lighter norm would appear to have been
followed. It will be observed that at Gortyna and Phaestus the
reduction in the weight of the Aeginetic stater, as compared with its
weight in the Cyclades, is even less than we were prepared for. We
may conclude that these were the two earliest of the Cretan towns to
strike their own money, and that it was from them that the practice
spread to the other parts of the island. This gives a new significance
to the archaic staters with the quaint legends through which the
types are made to proclaim themselves ' the stamp of Gortyna '
(Plate, Fig. 3) and ' the stamp of the Phaestians ' (Plate, Fig. 4). 3

It is, of course, possible though hardly probable that the
Aeginetic standard reached the northern side of Crete by a different
and more circuitous route, the longer journey involving a greater loss

1 For a practical example, see E. S. G. Robinson in Num. Chron., 1915 , p. 261.

2 This, too, although there has been excluded from the table the specially
light set of Cydonian staters which will be dealt with presently.

3 It is interesting to find that one of these archaic staters, which is uninscribed
and may, therefore, belong to either city, is restruck upon a stater of Siphnos :
it weighs 11*76 grammes. See Imhoof-Blumer, Antike griechische Miinzen, p. 13
and cf. supra, p. 4.


than would have been incurred if the direct line from the Cyclades
had been followed, as it apparently was to Gortyna and Phaestus.
One thing, however, is clear. We may dismiss the idea that the
prevalence of the Persic standard on the southern coasts of Asia
Minor and in Cyprus had any influence in bringing about the reduc-
tion we have been discussing. The Cretan staters, which not
infrequently approach or even exceed 12 grammes, can hardly have
been interchangeable by tale with staters of the Persic standard, for
which a maximum of 11*50 may safely be assumed. And, unless
a system of interchange was secured, no practical purpose would have
been served by the reduction. A workable agio could have been
arrived at without it. Moreover, although it is notorious that
restriking was a very common custom in the mints of Crete, I have
not met with a single instance of a coin of Persic weight being
selected for such treatment, whereas the choice of foreign coins of
Aeginetic weight has frequently been noted. 1

Incidentally, the testimony of restruck coins throws an interesting
light on the official attitude towards the two distinct norms whose
existence within the island was revealed by our tables of frequency.
The difference between them would seem to have been ignored.
Both at Gortyna and at Phaestus the mint-masters used staters of
Cnossus as flans on which to impress their own types, 2 thus affording
another proof of that neglect of scrupulous exactitude which has
already been mentioned as characteristic of Greek moneyers. The
bulk of the pieces struck on the Aeginetic standard in Crete must
have been looked upon as readily interchangeable, although each city
continued, as a rule, to maintain at its own mint the particular norm
which it had adopted when it first began to issue coins. Here and
there exceptions may perhaps be detected. At Praesus, for example,
the table of frequency, inconclusive though it be owing to lack of
numbers, seems to suggest that the older staters were minted upon
the heavier norm and their successors upon the lighter one. At
Lyttus, again, something of the same sort may have happened,

1 Thus we have staters (Svoronos, op. tit., p. 159, No. 8, and p. 160, No. 25),
drachms {ibid., p. 66, No. 5), and hemidrachms {ibid., p. 163, No. 42) of Aegina
itself; drachms {ibid., p. 261, No. 41) and hemidrachms {ibid., p. 279, No. 17)
of Larissa in Thessaly ; and hemidrachms of Argos {ibid., p. 270, No. 7, and p. 279,
Nos. 16 f.). As to staters of Cyrene {ibid., p. 165, Nos. 65 f., p. 262, No. 53, and
p. 263, No. 59) and drachms of the same city {ibid., p. 278, No. 11) see Robinson
in Num. Chron., 1915, p. 263.

2 e.g. Svoronos, op. tit, p. 162, No. 36, and p. 255, No. 2. The latter of
these weighs only 10*15 grammes, showing that it was not simply the heavier
pieces that were chosen.


although there, in the absence of any well-defined change of types,
nothing short of a careful examination of the whole of the actual
coins would enable one to say whether chronology and metrology
were in agreement as to a line of demarcation. But at none of the
four cities for which reasonably full statistics were procurable does
the table give a clear indication of a transition from one norm to the
other. 1 Tradition, entrenched behind convention, did not deem it
needful to discriminate.

Sporadic Appearance of a Debased Form of the Aeginetic


The general truth, then, remains as stated. The statement, how-
ever, was expressly limited to * the bulk ' of the pieces struck on the
Aeginetic standard in Crete. So far as I am aware, it has not
hitherto been recognized that at one time there was employed, in
a particular corner of the island, a variety of the Aeginetic standard
which may quite legitimately be called * debased ', and whose staters
can hardly have been interchangeable with those of either of the
ordinary norms. The reason for the debasement is quite obscure.
But the evidence for it seems indisputable, and accordingly the coins
concerned have up to this point been left out of account. They form
a class by themselves, and the fact that this has not previously been
noticed is undoubtedly responsible for much of the ill repute that has
come to attach to the Cretan form of the Aeginetic standard. Their
exceptionally low weight has produced a false impression by distorting
the perspective. Now that they have been segregated, we can deal
with them separately. In doing so it will be best to begin at the
fountain head, which would appear to have been Cydonia.

The most numerous, though not the oldest, group of Cydonian
staters have on the obverse the head of a nymph or Maenad, r.,
wreathed with vine-leaves and grapes, and on the reverse a figure
of a naked archer, probably the hero Kydon, stringing his bow
(Plate, Fig. 8). Occasionally a dog with raised forepaw stands facing
the archer. My list contains twenty- six examples of this series which
are unquestionably minted on the lighter of the two usual Cretan
norms ; they range between 11-35 and 9-53 grammes, and yield an
average of 10-87 grammes, all save the lowest being well over 10.
There is, however, a unique stater with similar types, which is distin-
guished from the others by the presence of a symbol above the

1 Some of the earliest coins of Cnossus are, however, heavy enough to suggest
that they may reflect the norm used throughout at Phaestus and Gortyna.


dog, and which weighs no more than 9*25 grammes (Plate, Fig. 9). 1
Had it stood alone, the abnormal weight might readily have been
attributed to accident. It is only when we examine the next group
that its significance becomes apparent.

It will be found that the latter pieces have the head on the
obverse turned to 1. and wreathed with vine. The reverse has the
archer Kydon as before, and sometimes he is accompanied by his dog,
always with a symbol in the field above. But here the presence of
the dog has a real meaning, as will be clear from the following list of
all the examples whose weights I have been able to ascertain :

Coins without dog or symbol. Coins with dog and symbol.

11-83 9-72

11-52 9-51

11-41 9-50

11-38 9-49

11-30 9-44

11-20 9-43

11-04 9-42

11 9-40

10-98 9-40

10-85 9-27

10-80 9-26

10.48 9-23

9-95 9-14


Examination reveals a difference in the obverse also ; on the lighter
series the bunches of grapes in the nymphs hair have been replaced
by ivy leaves. And the evidence of the contemporary drachms, so
far as it goes, is to the same effect, for the average weight of 11
examples on which there is no dog is 5-37 grammes, giving a stater of
10-74, whereas the solitary drachm on which the dog appears, weighs
no more than 4-65 grammes, 2 and would therefore give a stater of only
9-30. The gap between the two classes is thus about 1^ grammes, too
considerable to be the result of an accident, particularly as it is clean-
cut, with little or no gradation on either side and practically no
difference in the comparative preservation of the specimens. The
first explanation that suggests itself is that the lighter pieces are later,
and that their appearance indicates the substitution of one standard
for another. Plausible as it may seem, this theory must be dismissed

1 Svoronos, op. tit., p. 99, No. 2. The coin subsequently passed into the
Bunbury Collection (Sale-Catal. ,T*t. i. Lot 1162), and is now in the British Museum.

2 Ibid.,?. 101, No. 17.



as untenable. The symbol above the head of the dog establishes a
direct connexion with the unique stater mentioned above (Plate, Fig. 9),
and so proves that the earlier group, too, had a light series as well as
a heavy one. Unfortunately the condition of the obverse is such that
it is not possible to determine definitely whether there was a corre-
sponding difference in the head-dress of the nymph. It will be
interesting to have this point settled should a better- preserved
specimen come to light.

Moreover, it can be shown that, on the one hand, the drop
was not permanent, and that, on the other, it was not due to a
temporary lapse at this particular mint. We find that the same
division recurs in the succeeding series of Cydonian staters. The
beautiful pieces bearing on the obverse a female head wreathed with
ivy, and on the reverse a hound suckling an infant, yield when
weighed two very similar groups, which are, however, indistinguishable

1 3 4

Online LibraryGeorge MacDonaldThe silver coinage of Crete, a metrological note → online text (page 1 of 4)