Post on 02-Jun-2018
8/10/2019 The 1958 dirham hoard from Tartu in Estonia / Thomas S. Noonan
1/25
THE
1958
F R O M .
T A R T U
IN E STON IA
T H O M S S . N O O N N
In 1958 a hoard of 41 dirhams was found in the
city
of Tartu
(Dorpat,
lur ev) in eastern Estonia. More detailed Informa tion abo u t the cir-
cumstances of the find is not available. The hoard was identified by
Aleksei Andreevich
Bykov,
former
head
of the numismatic
section
of
the
Hermitage,
and is
mnv preserved
in the
Tartu City Museum
Numis-
Inuentory
Book, vol. 5, TM 143,
N38, 1-41).
1
Only
two
brief
to the Tartu hoard have so far been published. The first
in E.
Tnisson s 1962 study
of
Estonian coin
and
treasure
of the ninth through thirteenth centuries. 2 Tnisson s descrip-
tion noted
only that
the hoard contained 41Arabic coins of the
eighth
through tenth centuries. The same
Information
was repeated in V. V.
Kropotkin s
enumeration
of
some
of the
recent
Kfic
coin
finds
from
the
Soviet Union. 8 Even this m inimal information about
the
hoard
was
Duringthecourseof a1975sabbaticalresearch tripto
obtain
further
information
on dirham hoards from theSoviet Union, I
was
able to visit T artu where the
officials
of the city museum granted meaccess to the unpublished account of the hoard con-
tained In their rccords. I should like to tnke this
opportunity
to thank th e C i t y
M useum of Tartu for its kindness and to express my apprecintlon to U ie f f i c e of
International Programs and the Putnam Dana
M cM il lan
Professorship Award
C om-
mittee, both
of the University of
M innesota,
for their generous financial support
of
my
sabbatical research
in the
Soviet Union.
I
should also like
to
thank M ichael
Bates of the American Numismatic Society for his critical evaluation of the
manu-
script and
various
suggestions for its improvement.
E. Tdnisson, Eesti aardeleiud
9.-13.
sajandist, Muistsed Kalmed
Ja
Aarded:
Arheoloogiline
Kogmik 2
ed. H.
M oora (Tallinn, 1962),
p.
199,
no. 80;
tables,
pp.244-45.
n V. V.Kropotkin,
Novye
nakhodki
sasanidskikh
i
kuficheskikh
monetv Vostoch-
noS
Evrope,
N 9
(1971),
p. 95, no.
204.
135
8/10/2019 The 1958 dirham hoard from Tartu in Estonia / Thomas S. Noonan
2/25
6
8/10/2019 The 1958 dirham hoard from Tartu in Estonia / Thomas S. Noonan
3/25
D I R H A M
H O A R D F R O M T A R T U 137
completely
accurate. The
Tartu hoard
is composed exclusively
tenth-century dirhams.
A ccording to
Bykov,
th e
Tartu hoard consisted
of the following
coins:
1.
SMNID
ISM IL B.
A ^ J M A D
1-2. al-Shsh,
286
H./.D. 899-900
3.
al-Shsh, 288/900-1
4. al-Shsh,
293/905-6
5.
al-Shsh, 294/906-7
A^MAD
B. ISM lL
6.
al-Shsh, 295/907-8,
7 .
al-Shsh, 299/911-12
NAR B. A^MAD
8.
al-Shsh, 308/920-21
9. al-Shsh, 316/928-29
318/930-31
al-Shsh,
320/932-33
3.
al-Shsh,
322/933-34
4.
al-Shsh, 323/934-35
5.
al-Shsh, 324(?)/935-36
6.
al-Shsh, 325/936-37
7.
al-Shsh, 331/942-43
8.
al-Shsh, 33(l)/942-43
9. Samar q and
317/929-30
0. Samarqand, 326/937-38
1 .
No mint, 309/921-22
NIJB. NAR
2. al-Shsh, 366/976-77*
3.
Samarqand, 340/951-52
4. Bukhr,
337/948-49
4 The
date should probably
be 336/947-48. Nh I b.
Nasr reigned
331-43. Nh
I
b.
M a ns rreigned from
365 to
387,
but his
father
was
M a ns r
not
Nasr.
8/10/2019 The 1958 dirham hoard from Tartu in Estonia / Thomas S. Noonan
4/25
138
T H O M A S
S .
N O O N A N
P
A B D A L - M A L I K
B .
25-26.
Samarqand, 343/954-55
MANUR
B. NH
27-28.
al-Shsh,
355/965-66
29.
al-Shsh,
357(?)/967-68
30. al-Shsh, 358/968-69
31-32.
al-Shsh, 359/969-70
33. al-Shsh, 366/976-775*
34.
Bukhr, 352/963-64
35. Rasht, 361/971-72
36. Nomint,
351/962-63
2. MUSFIRID OR SALLRID
I S M lL
A N D W A H S D N
37.
Barda'a, 3xx(?)6
3. 'ABBSID
AL-MUQTADIR
38.
Madinatal-Salm,
311/923-24
39.
al-Basra, (3)32/943-44
According to n m n y rcfcrencc works, M a n s r I dlcd in
365
and was succcc
b y h f s son
N f il i
II Ih n sa m cycar.
Ilowcvcr, d i r h a m s
of
M a n s O r
I f rom
al-Sh
wlth
the
ycar
30(5
arc rcporlcd from
o t h c r
K f i c hoards o f the
soulheastcrn Bn
and Russia. See, for example, R. R. Vasmer,
Ein im
D or fe Staryi Dedin
in
We
russland Gemachter Fund Kufischer
Mnzen
(Stockholm,
1929),
p. 17,nos.
162
W . Anderson, Der
M n z f u n d
von
Vaabina,
Sitzungsberichte
de r
Gelehrten
nischen Gesellschaftlpetatud Eesti Settsi Aasiaraamat
l (Tartu/Dorpat,
19
publ.1938),p. 52,nos. 276, 279.
6
Although listed
withthe
Smnid coins,
this
dirham most likely belongs
to
M usfirid Wahsdn
w ho ru l e d
from 330 to 355 and had ason,
Ism'il,
who. d
around
35 0
Encyclopaedia
o f
Islam
2nd
ed., s.v.
M usfir ids).
8/10/2019 The 1958 dirham hoard from Tartu in Estonia / Thomas S. Noonan
5/25
DlRHAM HOARD
FROM TARTU
139
4.
BUWAYHID
MU IZZ AL-DAWLA AND *IMD AL-DAWLA
Mint
illegible, (3)37/948-49
5. VOLGA BULGARS
MIK lL
B. JA FAR
.Madinat Bulghr, date illegible.
The
dynastic
distribution of the 41
dirhams
is
given
inTable l below.
s
we can
see,
the.hoard
was overwhelmingly
composed
of
Smnid
and the other foui dynasties
had
only
minimal representation.
TABLE l
Dynastic
Distribution of Dirhams in the Tartu Hoard
Dynasty Number of oins Percentageo f Hoard
Smnid
36
87.8
Abbsid 2 4.9
Musfirid l 2.4
Buwayhid l 2.4
Volga Bulgars
l 2.4
41
99.9
Theregional distribu tion
of the
mints
is
perha ps even more
significant
the dynaslic breakdown. Of
h e
38
d i rhamswhose
m i n t can be
34 89.5
percent)
are from mints located in
Transoxania-
Asia (al-Shsh,
26;
Samarqand,
5;
Bukhr,
2;
Rasht,
1).
Two
5.3
percent)
come
from
Irq Madinat
al-Salm and
al-Basra)
there is one
coin 2.6 percent)
from
Transcaucasia (Barda a)
and
e 2.6
percent)
from the m iddle Volga Madinat Bu lghr). In short,
struck in
Smnid mints formed
the
predominant
in the
Tartu
hoard.
8/10/2019 The 1958 dirham hoard from Tartu in Estonia / Thomas S. Noonan
6/25
140 T H O M S S.
N O O N N
The
oldest coin
in the
hoard
was a
Smnid dirham
of Ism l
Ahm ad struck in al-Shsh in 286/899-900. The
most
recent coin
a S mnid dirhamofM ansr b. Nh struck in al-Shsh in 366/976
The time span between the oldest and newest coins was thus 77ye
Table
2 gives a
breakdown
of the 39 dateable dirhams by decade.
figures show
that
the coins in the Tartu hoard are fairly evenly
tributed over
eight
decades. The hoard is prim arily composed
nei
of dirhams which
had
been
in
circulation
for a
long period befor
burial nor of recently-minted dirhams.
TABLE
2
Chronological
Distribution of Dirhams in the Tartu Hoard
Decade
890/91-899/900
900/1-909/10
910/11-919/20
920/21-929/30
930/31-939/40
940/41-949/50
950/51-959/60
960/61-969/70
970/71-979/80
Number of oins
2
4
1
5
8
6
3
8
2
Percentage of Hoard
5.1
10.3
2.6
12.8
20.5
15.4
7.7
20.5
5.1
39
100.0
The predominance of Smnid dirhams compels us to examine
distribution
by
ruler
in greater detail. The basic quan titative
can be
fond
in the
third
and
fourth
columnsof Table 3. The la
number of Smniddirhams about 3/8 was struck in the longreig
Nasr
b.
Alimad.
The
next largest group slightly over 1/4 com es
the reign of Mansr b. Nh during whose lifetimethe hoard was
sumably assembled. There are very few coinsfrom the relatively
s
reigns of A h m ad b. Ismail and f
A bd
al-Malik b. N h . In brief
quantitative breakdown by reign does not indicate any notewo
patterns.
8/10/2019 The 1958 dirham hoard from Tartu in Estonia / Thomas S. Noonan
7/25
TABLE 3
Distribution ofSmnid Coins in the Tartu Hoard by Ru
u r
and Years of Reign
Total Years Percentage of Total
Ni nber
of
Per
in
eign Span
(279-366) Coins
All S
Ism'il
b.
Ahmad
279-95
A h m a d
b.
Ism'il, 295-301
Nasr b. Ahmad 301-31
Nh
b.
Nasr,331-43
'Abd
al-Malik b. Nh,
343-50
Mansr
b.
Nh,
350-65
16
6
30
12
7
16
18.4
6.9
34.5
13.8
8.0
18.4
5
2
14
3
2
10
100.0
a
Mansr apparentlydiedin 365(see note5) but coinsdated366appear withhisname
a16-yearspanis usedfor the statistical purposes ofthis table.
8/10/2019 The 1958 dirham hoard from Tartu in Estonia / Thomas S. Noonan
8/25
142 T H O M A S S .
N O O N N
Columns one two,and
five
of Table 3 were used to determine
average number
of
Smnid coins
in the
hoard
per
year
for
each r
If th e
same numbcr
of coins was
issued each ycar
and
cach coin c
lated for
about
th e
same time,
w e would
expect
to find
fewer
per year
from
earlier reigns and
more
coins per year
from later
re
The
Tartu
hoard does not
conform
to this model. The largest nu
of
coins per year comes from
the
most recent
reign,
s
would be expe
but the coins peryear
f rom.
th e reigns of N h i and cA bd al-Malik
fewer than those
from the
much earlier reigns
of
Ism Il
and
Ahm
O fall the reigns beforethat of
Mansr,
Nasr s shows the greatest nu
of
coins per
year.
Rather
than
the increase in number of coins
year
we
would anticipate with each successive reign,
we
encount
decline under
Nh and
Abd al-Malik.
Their
reigns
are thus
un
represented
in the hoard while that of Nasr is
overrepresented.
T
figures may
reflect
lower
production
in
some
or all
Smnid m
during
the reigns of Nh and cAbd al-Malik. There appears to
been increased production in some or all Smnid mints under N
The
information
on
coins
per
year
is
therefore more revealing
the
basic data
on the
number
of
coins from
the
reign
of
each r
T A B L E
4
Distribution of Dirhams in the Tartu Hoard by Mint
n
al-Shsh
Samarqand
Unknown
Bukhr
Rasht
Barda
c
a
Madinat
al-Salm
al-Basra
M a di na tBulghr
Number
Percenlage of
oard
26
5
3
2
1
1
1
1
63.4
12.2
7.3
4.9
2.4
2.4
2.4
2.4
2.4
41
99.8
8/10/2019 The 1958 dirham hoard from Tartu in Estonia / Thomas S. Noonan
9/25
DlRHAM
HOARD FROM TARTU
43
The
mint distribution
of the 41
dirhams given
in
Table
4
presents
interesting
phenomenon. Thevastm ajority of the coins in the hoa rd,
w as
slruck
in
al-Shsh.
The
onlyo ther m intwi th
a
significant
of the Tartu dirhams was Samarqand. Furthermore, Table
that
the
al-Shsh
dirhams
are
well
represented
among
e coins from most of the Smnid rulers, and composed 72.2 percent
all
Smnid
dirhams in the hoard. In this connection, it is striking
the percentages for the two reigns with the largest concentrations
are
78.6 percen t Nasr
b.
Ahmad)
and
70.0 percen t M ansr
Nh). Aside from the period
331-50
H.,w hich is poorly represented
the hoard, al-Shsh dirhams predominate among both older and
Smnid dirhams.
TABLE 5
Distribution
of
al-Shsh
Dirhams
in the
Tartu H oard
by
Ruler
Percentage
ul r
ll b. Ahmad
b.
Ism il
b. Ahmad
b. Nasr
bd
al-Malik b. f c
b.
Nfr
Total
Dirhams
5
2
14
3
2
10
Number al~Shsh
5
2
11
1
0
7
al Shsh
100.0
100.0
78.6
33.3
0.0
70.0
The weak representation of al-Shsh dirhams from 331-50 H. is
when
we
examine
the
number
of
al-Shsh dirhams
per
year
each reign Table
6). It is
interesting
to
note
that the
precipitous
in the
number
of
al-Shsh dirhams
per
year
and the marked
rise
the
number
of
dirhams
per
year
from
other mints comes
during the
of Nh and
Abd al-Malik,
a
period whose
coins, s we
have
are noticeably
imderrepresented
in the hoard
s
a whole. These
raise the possibility that, while the overall production of S-
mints m ay have declined in the period
331-50
H., the decline
relatively
more severe in al-Shsh than in other Smnid mints.
8/10/2019 The 1958 dirham hoard from Tartu in Estonia / Thomas S. Noonan
10/25
14 4
T H O M A S
S .
N O O N N
T A B L E 6
Number
of
Smnid Dirhams
Per
Year from
Mints
in the Tartu Hoard Given by Reigns
ul r
al Shsh CoinslYear
Olher
Mints Coinsl
Ism Il
b.
Ahmad
A ljma d b. Ism ll
Nasr b.
Ah.mad
Nh. b. Nasr
Abdal-Malik b. Nb.
M a ns r b . Nh.
.31
.33
.37
.08
.00
.44
.00
.00
.10
.17
.29
.19
In
summary,
the
Tartu
hoard
was composed
overwhelmmgl
Smnid
dirhams
from
Transoxania which had been struck ove
prior eight decades. There was no notable concentration of
either
new or
very old dirhams. Among
th e
Smnid dirhams,
those
fro
Shsh were pr edom ina nt. Pcrhaps
th e
most striking
featurc of the
h
was the disproportionate number of dirhams per year from
se
Smnid reigns
and the sharp
fluctuation
in the
number
of al-S
dirhams per year for various reigns. These variations can
per
best
be
explained
by
changes
in the
production
of the
Smnid
m
during
the
time under consideration.
A swe have seen, Central Asian dirhams composed
almost
90pe
of the coins in the Tartu hoard. It
might
appear from these fi
thatthehoard s awhole originated in Transoxania. Such a suppos
is
strengthened
whenweconsider
that
the most recent coins (nos
35) are all
Smnid dirhams from Central Asia. Under normal condit
local coins would presumably occupy
a
ma jor place among
the
recent dirhams
in a
hoard.
T he few
coins
from
Transcaucasia
and
found in the Tartu hoard can be
attributed
to the circulation of
dirhams
from
the
more westerly provinces
of
Islam within Transoxa
At the
saine
time, we must aeknowkdge that the presence of a di
issued
by the Bulgar state along the middle Volga River raises ce
Problems concerning
the
Central Asian origins
of the
hoard.
exclude s
unlikely the
possibility
that the
Volga Bulgar dirham
obtained in Transoxania, it appears
that
this dirham wasadded to
Tartu
hoard
somewhere
in eastern Europe. Oneobvious explana
8/10/2019 The 1958 dirham hoard from Tartu in Estonia / Thomas S. Noonan
11/25
DlRHAM
H O A R D
FROM
TARTU 45
that the
Tartu
hoard was
brought
from
Central Asia
to
Estonia
via
e
Volga Bulgar
lands where a local dirham
might easily have been
Alternatively,the
hoard
could
h ave been
formed in the
Volga
lands,
at someother
interm ediary point ,
or in
Es tonia. Tenth-
dirham
hoards
from
Russia
and
adioining
areas were ovenvhelm -
com posed
of
Smniddirhams.7
In
addit ion,
Volga Bulgar dirham s
e found in many Russian and
Estonian
hoards of thetenth Century
several separate finds
of
tenth-century VolgaBulgar
d i rham s
have
recordedinthesesameregions.8 Inbrief,the
Tartu
hoardmay well
beenform ed somewherein eastern Europefrom the dirhamsavail-
there.
The above discussion clearly demonstrates that even if the Tartu
originated
in
Central Asia,
it was not
brought
to
Estonia unaltered.
e
Tartu hoard
is not
tmique
in
this respect.
Volga
Bulgar dirhams
addedtom any tenth-century hoards
from
eastern Europe
while
the
from 970
also
saw the
appearance
of
wes tern E uro pea n deniers
Russ ian
and
Estonian
hoards .
Byzantine coins, too,
are
found
eastern European hoards
of the
tenth
and
eleventh centuries.9
short, dirhams were
not
necessarily brought
directly from
Central
to
Estonia
and then immediately
buried there.
The
mixture
of
Bulgar, western European,
and
Byzantine coins with
d irhams
Transoxania in eastern European hoardsof thetenthand eleventh
seemsto
showthat
dirhams circulated,at leastto a limited
in
eastern
Europe, and
that
it was duringthe course of such
that
coins struckindif ferent partsofEurope were m ixed with
from Central As ia. Sucha conclusion agreeswith
r written
sources, which strongly suggest that Islamic merchants
et
their
eastern
European
counterparts in
such interm ediary places
R. R. Vasmer, Ob izdanii novoi
topografii
nakhodok kuficheskikh
monet
v
Evrope, Izvesitia
Akademii
Nauk SSSR, Otdelenie
obshchestvennykh
6-7 (1933), p. 478; V. L.
lanin,
Denezhno-vesovye sistemy russkogo sred-
Domongol skii period (Moscow, 1956),
pp. 118-40.
V. V.Kropotkin, Torgovye sviaziVolzhskoiBolgariiv X v. po
numizmatiches-
dannym, revnie slaviane
i ikh sosedi,MIA
176
Moscow, 1970),
pp.146-50.
9V. V.
Kropotkin,
Klady vizantiiskikh
monet na territor SSSR Mos cow, 1962)
d
Novye
nakhodld vizantiiskikh monet na territorii SSSR, V
V
26 (1965),
166-89.
8/10/2019 The 1958 dirham hoard from Tartu in Estonia / Thomas S. Noonan
12/25
8/10/2019 The 1958 dirham hoard from Tartu in Estonia / Thomas S. Noonan
13/25
D l R H A M H O R D
F R O M T R T U
147
T A B L E
7
Four Estonian
CoinHoardsofthePeriod
966/67-986/87
a
Location
of
Find,
of Most
Recenl
Coin Percentage
(Number
of Coins)
D i rhams
e ts a k l a , 970(51) 49.0
rtu, 976-77 (41) 100.0
980
(128)
96.9
985(335) 100.0
Percentage
W .
European
Denier s
51.0
0.0
0.0
0.0
Percentage
Other
Coins
0.0
0.0
3.1
0.0
a
The
K u m n a
(Metsakla)
hoard
is
discussed
in A.
Friedenthal,
D er M n z fu n d
n Kumna, Beitrge zur Kunde
Estlands
18
(Tallinn,
1932), pp. 5-13; the firra-
hoard in V.V. Kropotkin,
Novye
n a k h o d k i
sasanidskikh,
p. 95, no. 205;
e
Essemgi(smae)hoard
in H.
Frank, D ie
baltisch-arabischen
Fundmnzen,
Gebiete der Geschichte Liv-, Est- und
Kurlands
Gesellschaft
Geschichte
und Altertumskundeder
OstseeprovinzenRusslands,
18
(Riga, 1908),
393, no. 50.
T heTartu hoard isinteresting
because chronologically
it
fol lows
after
the
Ku m na (M etsakla) hoard,
the
f irst Estonian
which contained wcstern European coins. Table 7
indicates
two qu i te different
types
of
hoards existed
in E sL o n i a
d u r i n g
th e
years
under discussion. T he Kum na (M etsakla) hoard \vith its
equal K f i c an d western European components represents the
hoards which
w ereto
characterizethis
arcafor the
next
Ce ntury
d a half . The
Tartu
hoard, however , be longs to
Ihe sccond
Lype,lhat
,those overwhelmingly or entirely composed of
eiLher Kfic
orw estern
For the two
decades
in
which
we are interested, the
group w as represented by predominantly or exclusively Kfic
T he Tartu hoard, l ike those
from
Erra-Liiva an d Essemgi
did not contain a single w estern European coin . It w as a
hoard made up primarily o f dirhams rather than a new
mixed hoard i nc l u d i ng western R u r o p c a n deniers. O n e
m i g h t
conjeclurc t h a t
the
T a r t u h o a r d
w as
buried soou a f l o r
its
d i r h a n i s
Estonia, beforethey had the chance to
c i r cu l a te
au d
become
xed w ith western European and Byzant ine coins.
8/10/2019 The 1958 dirham hoard from Tartu in Estonia / Thomas S. Noonan
14/25
8/10/2019 The 1958 dirham hoard from Tartu in Estonia / Thomas S. Noonan
15/25
D I R H A M
H O A R D F R O M T A R T U 149
T A B L E 8
Non-Western
E u r o p e a n Dynasties Represented in Estonian K f ic
Hoards of the Period 966/67-986/87
Tart
Kumna
Erra Liiva
sm e
ara
Kawast
Smnid
x x
Abbsid x x
M usfirid
x
uwayhid
x x
Volga Bulgar
x
yzantine
India
U m a y y a d
Imitations
X
X X
X X
X
X
X
X
X
X
X
X
X
six hoards contained Abbsid and Buwayhid dirhams. The 1958
Tartu hoard therefore appears to be typical ofEstonian hoards of this
period in three of its dynastic components. On the other hand, Volga
Bulgar
dirhams were found in only one other hoard while none of the
other five hoards had
M usfirid
coins. The presence of these dirhams
in
the
Tartu hoard probably reflects
the
great diversity
of
coins
fo u n d ,
in
small numbers,
in
many
of the
hoards
of
this
era. Incomplete in-
formation
about most of the other Estonian hoards makes any attempt
to reach meaningful comparisons of
their
dynastic components impos-
sible.
In sum, the types of
K f ic
coins
fo u n d
in theTartu hoard do not
appear
to
diverge from those found
in
other contemporaneous Estonian
hoards.
This does
not mean
that
the
Tartu hoard
was
necessarily typical
f
th e dirhams found in Estonian hoards of the same period.
Ithad
been
my
intention
to
compare
the
Tartu hoard with
Latvian
from
the 20-year era under discussion in order to contrast the
artu dirhams with hoards
from the
region directly south
of
Estonia.
H owever, l
have been unable
to
identify
any
Latvian hoards
of
this
period.
On the other hand, there are several
dirham
hoards of this
Urne
from
the
regions
of
Russia
adjoining
Estonia,
that is, the
Novgorod
an d
Pskov
lands.
I have also included another
Kf ic
hoard from
north-
western Russia which
was
published
in
fll.
The
abundant data
on
these Russian hoards enable us to pursue our comparative analysis
the
Tartu hoard
in
depth.
8/10/2019 The 1958 dirham hoard from Tartu in Estonia / Thomas S. Noonan
16/25
S -
i i
< t
lf
?
l
- g
S
S
en
V
w
0
S
?
ts5
0
?
r.
g1
c&
M
S;
"
O
^
g
8.
0
tn
X *
o
.
o
M-
rf
CO
p
cn
p
cn
M-
bt
W
e?
^
0
v
CO
o
CD
oo
< i
co
o o
o
CO
IN3
b
cn
p
co
M .
cn
p
co
$
,
g
c o
r
5
pr
CO
v i
_O5
"^3
I
co
bo
vl
~4
*
M.
CO
y >
C D S*
C n 3
5 ^ i
p >
i g
3 g
C D9
"h? ^
P. C A
'S
9
7
,
5
(
4
cn
< s
co
io
M >
C n p
b io
v j
8/10/2019 The 1958 dirham hoard from Tartu in Estonia / Thomas S. Noonan
17/25
DlRH M
HO RD FROM T RTU 5
Table9containsthe
data
on thedynastic distribution inpercentages
of
six contemporaneous hoards
from
northwestern Russia. The
most
recent
coin from
cach of these
hoards
dal es to the
970s.
A ll sixRussian
hoards are composed
predom inantly
of Smnid dirhams the
numbers
ranging
from
76.1 percent to 86.9 percent. In additio n all six hoards
contain some Abbsid coins
and at least one
Hamdnid
and
Ziyrid
dirham. Imitation
Smnid coins
and Buwayhid
dirhams
are found
in five
hoards. Byzantine coins
and
dirham s issued
by the
Volga Bu lgars
and the Amirs of
Andarba were
in
three
of the six
Russian hoards.
A few coins from six other dynasties were contained in one or two of
the hoards. In general the larger the hoard the more diverse its dy-
nastic composition.
The
1956 Nerev hoard
the
largest
in our
group
contained 12
diff erent
types of coins. The Novaia
MeFnitsa
and Velikie
Luki hoards which have
less than
60 dirhams
each
contain only six
different types
of coins. Regardless of size howev er the percentage
of
Smnid coins remais fairly constant.
The percentage of Smnid dirhams
(87.8)
in the
Tartu
hoard is
comparableto that in the
northwest Russian hoards;
it is
slightly higher
than
that found in f ive of the
Russian hoards
but
less than
one
percent
more
than the
figure from
the 1953 Nerev hoard . It is
therefore hard
to
point
to any
significant
difference
between
the
Tartu
and
Russiau
hoards
with
respect
to
their overwhe.hning
Sfimfinid composilion.
O n
the
other
hand there is
some
difference in
their percentages
of
non-S5mnid
dirhams.
TheTartu
hoard
did not
contain
the
Ham dnid
or
Ziyrid coins
found
in all the Russian hoa rds or the Imitation dir-
hams
found in
five
of the
Russian hoards. Volga Bulgar dirhams which
are
in the
Tartu hoard
are
found
in
only half
th e
Russian hoards
while
none of the Russian hoards contains a Musfirid d i rham which was
ound
in the
Tartu hoard.
In the
area
of
non-Smnid coins
the
chief
and
perhaps, sole similarity between the Tartu and Russian hoards
lies in
their
small percentages of Abbsid and Buwayhid issues.
It is
difficult
to determine if the differences in percentages of non-
Smnid coins in the Tartu and northwest Russian hoards have any
significance.
As we
have seen
the number of
dynasties
represented in
hoard depends to a largeextent, on its size. The absenceofv rious
in the
Tartu
hoard canthus be linked to the fact
that
it is
smaller
than any of the
northwest
Russian
hoards. Furthermore,
the
8/10/2019 The 1958 dirham hoard from Tartu in Estonia / Thomas S. Noonan
18/25
152 T H O M S S.
N O O N N
fragmentary data on contemporaneous Estonian hoards
prevents
a
definitive conclusion
about which non-Smnid dynasties were
rep
scntcd in Estonian
hoards
of this cra. Even more importnnt, it se
safe to assumethat
most
ofthe
non-Smfmid
dirhamswere broug ht
i
eastern Europe
from
Central Asia along
with
Samnid dirhams. T
presenceorabseneeofvarious non-Sm nid dirham s
in
eastern Europ
hoards of
this
time would, therefore, depend
greatly
on the vagaries
monetary cireulation in Central
Asia.
It can therefore beargued
th
the
small non-Smnid component
of the
northwest Russian
and
Ta
hoards reflects, more
than anything
eise the non-Smnid dirha
available in Transoxania at the timethese
hoards
were being
form
The
presence
of one or a few
non-Smnid dirhams
in one
easte
European hoard and
their
absenee in
another may
be the
result
chance. Finally I am not aware of any reason why Hamdnid
Ziyrid dirhams, which were accepted
in
northwest Russia, should
deliberately excluded from Estonia.
All
this discussion suggests th
we
shouldnot
attach great
imp ortanceto the
differences
in the dynas
distributionof the non-Smnid dirhams between the
Tartu
and nor
west Russian
hoards.
TABLE
10
Regional Distribution of S ix
Dirham
Hoards
from
Northwest Russi
(Expressed
in
Percentages)
Central Asia Near
East
Elsewhe
1953 Nerev
Novaia Mel'nitsa
1956 Nerev
Velikie Luki
Erilovo
Staryi Dedin
87.2
81.3
85.5
79.3
81.4
76.1
8.8
17.0
10.8
20.7
15.0
6.5
4 0
1.7
3.7
0.0
3.6
16.9
Table 10shows the regional
dislribuo n
of the
coins
in the sixnort
west Russian hoards. As
could
be expected, the abundanceofS amn
dirhams has resulted in a decided Central Asian dominance. Coi
from the more western regions of Islam, here designated Near Eas
8/10/2019 The 1958 dirham hoard from Tartu in Estonia / Thomas S. Noonan
19/25
DlRHAM
H OA RD
FROM
TARTU
53
rsimplicity,comprised nly 6.5percent to20.7 percentof
the
dirhams
the
hoards.
Coins f rom other regions imitation dirhams,
Bulgaria
the.
Volga,
B y z n n t i u m , wcstern Europe vary frain 0.0 percent to
In theTartu
hoard,
89.5 percentof thedirhams originated
Transoxanian
mints,
7.9
percent
were
f rom
the
Near
East,
and 2.6
camef romelsewhere. Inother words,the
Tartu
hoard contained
higher
percentage
of Central Asian diihams than did the
Russian hoards. We may recall, in
this
respect, that the
of Smnid dirhams in the Tartu hoard was higher
than
for all six of the
northwest Russian hoards. Given
the
absence
other
contemporaneous
Estonian hoards, it is hard to say what
any, should
be
attached
to the
somewhatgreater promi-
of
Central
Asian/Smnid
dirhams
in the
Tartu
hoard.
TABLE 11
Chronological
Distribution
of the
Dirhams
f rom
FiveHoards
f rom
Northwest Russia Expressed in Percentages)
I
j
8/10/2019 The 1958 dirham hoard from Tartu in Estonia / Thomas S. Noonan
20/25
154 T H O M S S. N O O N N
The chronological distribution of the dirhams
from five
northw
Russian
hoards is shown in
Table
11. The most
interesting conclus
to bc
drawn
from U ns
table concerns
th e
similarities
and differen
within the
Russian hoards.
All
contain
a
small percentage
of co
struck before
890; in
four
of the hoards
this
is less
than
one perc
Similarly, none of the hoards contain a significant percentage ofco
issued before
920
and,
in four of the fivecases,
before 930.
At thesa
time, the three somewhat earlier hoards seem to have a chronolog
distribution
pattern
distinct
from that of the two slightly
laterhoar
In each of the first threehoards, there is a significant concentration
dirhams of the 960s which ranges from 1/3 to over 2/5 of the en
hoard.
The
issues from
the twolater
hoards
are
distributed much
m
evenly
over
the
period
from
the
930s
to the
970s.
The
Tartu
hoard
d
not fit either of these general patterns and also seemsnot to have
same overall chronological distribution sthe Russian hoards. First
contains none
of the
pre-890 dirhams present
in
small quantities
all the R ussian hoards. S econd,a significantly
greater
percentage o
twice) of the Tartu hoard
than
is true for the Russian hoards
da
from
the 890s and 900s; the percentage
from
the 920s and 930s in
Tartu hoard isalsogreater. Inshort, the Tartu hoard containsa sig
ficantly
larger percentage
of
earlier coins
than
do the
five
R ussian
hoar
Between 1/2 and 5 /8 of the dirham s in the northw est R ussian
hoa
were minted after 949/50,
while
only 1/3 of theTartucoins weremin
during
thecomparable period.
Finally,
the
Tartu
hoardlacksthe
nota
concentration
of
coins from
the
960s
found in the
first three Russ
hoards,and itsdirhams werenot s evenly distributed overthe deca
from the 930s
s
was true for the
latter
two hoards.
The considerable differences in chronological distribution
betwe
the
Tartu
and Russian hoards are also apparent when we
compare
distribu tion of the Smnid dirham s in each. Table 1 2 shows t
distribution fo r three of the northwest Russian hoards. A sma
perccntage of the Smnid dirhams in the Russian hoards
comes f
th e
reigns
of
Ismail
b. Ahmad, Alimad b. Isml l, and
Nasr
b.
Aljm
and a greater percentage from the reigns ofNl.i b. Nasr,
Abd
al-Ma
b. Nh,, and Mansr b.
Nfc.
We
can
summarize
this
discussion
by
emphasizing
the
importa
chronological
differences between the
Tartu
and Russian hoards. F
8/10/2019 The 1958 dirham hoard from Tartu in Estonia / Thomas S. Noonan
21/25
D I R H A M
H O A R D
F R O M
T A R T U
155
T A B L E 12
D istribution of
Smanid
D irhams by R u l e r in Three
Iloards
from No rthwest R ussia (E xpressed
in Percentages)
1953
Nereu
1 9 5 6 N e r ev
Staryi
Dedin
203 ) 3 5 5 ) 1 5 3 )
b.
A h m a d
or A h m a d
b. Ismail
b.
A h m a d
N h
h b.
NasrJ
h
or
Abd
al-M alik
d
al-M alik
b. Nh- ^
d al-M alik or M ansr
ansr
b.
Nh
h b. M ansr
2.0
0.5
1.5
23.6
0.0
16.2
1.5
15.3
1.0
38.4
0.0
0.6
0.3
1.4
21.1
5.1
20.3
0.8
10.7
0.0
39.7
0.0
7.2
0.0
3.3
31.4
0.0
15.0
0.0
9.1
0.0
32.7
1.3
unknown reason,
the Tartu
hoard contained
a
larger percentage
older coins than the R ussian h oard s and the new dirham s in the
did not follow either of the two Russian patterns for the
more
recent
dirhams.
While the
vast
predominance
Smanid dirhams created some general chronological similarities
the
Tartu
and Russian hoards,
more
detailed analysis has de-
the existenceof
distinct
diff
erences between
the
ages
of the
found
in the
Tartu hoard
and the
hoards
o f
northwestern R ussia.
T A B L E 13
Num ber of Smanid Coins/Year for Each Ruler in T hree H oards
from
Northwest R ussia
1953 Nerev 1 9 5 6 Nereu Staryi De din
Il b. Al.imad
b. Ism il
b . A h m a d
r b. Nasr
bd
al-M alik b. Nh
ansr b. Nh
0.25
0.19
1.60
2.75
4.43
4.88
0.13
0.83
2.50
6.00
5.43
8.81
0.69
0.83
1.60
1.92
2.00
3.13
8/10/2019 The 1958 dirham hoard from Tartu in Estonia / Thomas S. Noonan
22/25
156
T H O M A S
S.
N O O N A N
The chronological differences between the
northwest Russian
Tartu
hoards
also become apparent when we measure
the numb
Smnid dirhams per
year
in the
Russian hoards.
The
figures
for
such
hoards, givenbyreigns, are
found
in Table 13. In allthree
ho
the
number
of
coins
per
year
fo r
each reign steadily increases
approach
the
approximate date
of
burial.
The
only exceptions,
rather
small
onesat that,involve the data for Nh and
rA bd
al-M
in the
1956 Nerev
hoard and for Ismail an d
Ahmad
in the 1953 N
hoard. In
general,
the
Russian hoards
conform to our
ideal
mod
an
increasing number
of
coins
per
year
for
each successive
ruler.
theTartu
hoard,
it
will
be recalled, deviated
noticeably
from this
tern for the
reigns
of Nh and cAbd
al-Malik. Regardless
of the
fl
ations
in the
production
of
Smnid mints,
our
original expectatio
more
coins
per
year
for
each successive ruler seems
to be
validated
we
cannot explainth edivergent
figures
for theTartu hoardby
varia
in
mint
Output in
Transoxania.
We shall
return
to this
question
TABLE 14
Mint
Distribution
of
Dirhams
in
Three Hoards
from
Northwest
R
Expressed
in
Percentages)
1953 Nerev 1956 Nereu Stariji
1 5 4 342) 166
al-Shsh
Samarqand
Bukhr
Others
31.8
37.7
16.2
14.3
31.6
31.0
12.9
24.5
39.2
32.5
10.8
1 7 5
Table
14
gives
the
distribution
of the
dirhams
in
three Russian
ho
by
mint. Al-Shsh
and
Samarqand were clearly
the two
most-re
sented
mints. Bukhr,
the
third
most-represented
mint,
was far
be
The
mint distribution
of the
Tartu hoard
differs distinctly from
Russian pattern.
Al-Shsh dirhams are over 60percent of the
T
dirhams. Samarqand dirhamsconstitute
approximately 1/5 the num
of
al-Shsh dirhams
in the
Tartu hoard
while the
number
o f
Samarq
8/10/2019 The 1958 dirham hoard from Tartu in Estonia / Thomas S. Noonan
23/25
DiRHAM
HOARD FROM TARTU
57
al Shsh dirhams
from the
Russian hoards
is almost the
same.
e
percentage
of
Bukhr coins
in the
Tartu hoard
is
also
much less
2
to 1/3) thanthat in the Russian hoards. While al-Shsh dirhams
the
vast majority
of the
Tartu hoard issues, they constitute
a sizeable minority of the
coins
in the
Russian hoards.
of the mint distribution between theTartuand northwest
hoards thus reveals a significant difference: al-Shsh dirhams
twice s prominent in the Tartu hoard s in the Russian hoards
the relative number of Samarqand and Bukhr dirhams is con-
smallerin theTartu hoard. Itis not immediately clearwhy
mint
distribution of the di rhams in the Tartu hoard should be so
that f ound in the
Russian hoards.
e
first
part
of
this
study
sought
to
determine
the
more important
o f the
dirham hoard found
a tTartu in
1958.
Identifica-
by the comparison a n d
contrast
e Tartu hoard
with
contemporaneous di rham ho ards from Estonia
adjoining region s. This co mp ara tive an alysis could not be fully
because oflimited data on the five other known Estonian
hoards o f this pcriod. The absence o fco ntemporaneous La tvian
ruled
out the comparison ofthese hoards with the Tartu hoard .
was available on several hoards of the 970s
f ound
in the
o f
Russia which borders
on
Estonia.
The
comparison
of the
Tartu
with these Russian hoards
showed
that there were some broad
which arose from the fact that both were overwhelmingly
o f
Smnid d irhamsfrom C entral Asia. Even m o re interesting,
was the
discovery
of
several significant differences between
Tartu an d no rthwest Russian hoards. The Smnid dirhams from
Tartu
hoard were generally
older and
al-Shsh dirhams were much
prominent .
The
number
o f
Smnid coins
per
year steadily in
dur ing each reign in the Russia n ho ard s, unlike the
pattern
o f
Tartu hoard . The
Tartu
hoard therefore possessed several key
ics which distinguished it from
Kflf ic
hoardsfound in nearby
lands.
e
might
well
have expected the Tartu and Russian hoards to be
similar.
After
all,
the
d i rhams found
in
Estonian hoards
pre
camefrom the same coin stock s the d irham s buried in
north
Russia.
The
only
difference
between
the two
would seem
to be
8/10/2019 The 1958 dirham hoard from Tartu in Estonia / Thomas S. Noonan
24/25
158 T H O M A S S. N O O N A N
the route by which
they moved
north.
Some dirhams moved
fro
upper Volga region to the north by various rivers toward Nov
whilc otlicr
di rhnms
wcnt; north from thc
uppcr
Volgn vin
Lake
and the Narva
River
alongthe
Russian-Estonian
frontier. The
hoard, however, raises
the
possibility
that
the
differences betwee
tonian and northwest Russian hoards during the period966/67-98
went far beyond alternate routes. T he absence of detailed da
contemporaneous Estonian hoards makes it impossible to deter
if
theTartu hoard was typical or
anomalous. Furthermore,
the
size
of the
Tartu hoard could meanthat relatively
few
Smnid
may have distorted
the chronological and
mint
distributions.
N
theless, one possible explanation for the differences between the
and Russian hoards should be noted.
Written
evidence suggests
strongly
thatdirhams
did not
come directly
from
Transoxania
to
ea
Europe but were brought from
Bulgaria
on the Volga or, at an e
period,
from
Khazaria by Rs/Rus' merchants. T he existence of
intermediaries
also explains howByzantine, Volga Bulgar, and we
European coins weremixed with the dirhams sent from Transox
While Volga Bulgar dirhams were most
likely
added in Bulgar
the Volga, neither Byzantine coinsof the
tenth-eleventh
centuries1
very many western European coins ever reached Volga Bulgar
la
These coins were most probably mixed with dirhams
at
some Ru
center
where
coins of various
types
and origin could be found.
also logical to assume that the Rs/Rus merchants would retu
the
Russian centers wherethey
had
obtained
the slaves an d
fr
to Islamic merchants in the Volga Bulgar and Khazar market
thereforc
seems
imlikely
that
dirham
hoards went directly
from
#ir iu on lli Volga t.o K s l . o n n . T he
dirh. 'itns
in
Iho. T u r i n
hoard
suinahly cwno
Uirongh
a
Hnssian
center
such ns Novgorod or
P
T he Tartu hoard could well have been composedor altered
from
selected by a
m erchant
in Russiawho was visiting Estonia. T he c
Kropotkin, Klady vizantiiskikh
monet
maps 7-8.
16 See the map accompanying V. M.Polin, Topografiia nakhodok zapadno
peiskikh monetX-XIIIvv. na territorii
drevneiRusi, Trudy
G rmitzah 9
pp.106-94
8/10/2019 The 1958 dirham hoard from Tartu in Estonia / Thomas S. Noonan
25/25
DlRHAM HOARD FROM TARTU 59
coins under such circumstances
may
have beenarbitrary
but it
could
have
involved
a
conscious decision about
the
type
of
dirhams
to
includod In nn y c v c M i l
th e
e x i v S l e n e e of
addi t ionnl
inlortncdiary
and further handling seems the bestpreseut explanation for the
between
the
artu
and
northwest
Russian hoards Monetary
may have been
limited
but it does appear to have had some
in a few cases The discovery of further Estonian
dirham
hoards
the 970s should enable us to resolve some of the problems presented
the 1958 artu hoard