Skip to main content

Full text of "The Arabic Original of the Report of R. Nathan Hababli"

See other formats


Early Journal Content on JSTOR, Free to Anyone in the World 

This article is one of nearly 500,000 scholarly works digitized and made freely available to everyone in 
the world by JSTOR. 

Known as the Early Journal Content, this set of works include research articles, news, letters, and other 
writings published in more than 200 of the oldest leading academic journals. The works date from the 
mid-seventeenth to the early twentieth centuries. 

We encourage people to read and share the Early Journal Content openly and to tell others that this 
resource exists. People may post this content online or redistribute in any way for non-commercial 

Read more about Early Journal Content at 
journal-content . 

JSTOR is a digital library of academic journals, books, and primary source objects. JSTOR helps people 
discover, use, and build upon a wide range of content through a powerful research and teaching 
platform, and preserves this content for future generations. JSTOR is part of ITHAKA, a not-for-profit 
organization that also includes Ithaka S+R and Portico. For more information about JSTOR, please 



My friend Dr. Alexander Marx, who also has rendered 
me valuable assistance in the preparation of this article, 
some time since drew my attention to a fragment now in the 
library of the Jewish Theological Seminary in New York, 
belonging to the Genizah Collection of Taylor-Schechter 
(T-S. Loan 48) 1 . The fragment consists of one leaf, paper, size 
12x7, written on both sides in Hebrew characters. The 
handwriting is oriental with a turn to cursive. The language 
is Arabic. A part of the upper margin, unfortunately that 
containing the beginning of the fragment, is torn off. In 
several places the MS. has suffered, but the words can still 
be deciphered. In the upper left-hand corner we still read 
the words ♦ . . ffivi ban rb>yJ? s/,, b, and under it my ijn npao. 
On the lower margin a few words are still visible, written — 
like those of the upper margin — in a different, very careless 
hand. The ink is very much faded. I believe I can 
discern a group of letters — Wiuaint yn&DK . N — which I can- 
not explain, but I feel no certainty even about the letters. 
At any rate, both these and the words in the upper margin 
have scarcely anything to do with the text. Upon 
examining the contents of the MS. I found that it was 
the Arabic original of a part of the well-known report 
of R. Nathan Hababli on the Academies in Babylon 
(Neubauer, Mediaeval Jewish Chronicles, II, 78 f., beginning 

1 I am indebted to Professor Sohechter for the permission to use 
the MS. 
3 Possibly 'id. Apparently an abbreviation for bvd. 


with 78, 20 and ending with 79 penultima, with the words 
•tnora nts>jny msn). 

Up to this time the Hebrew report has been considered 
the original. But in my opinion the original character of 
the Arabic cannot be doubted. 

1. It is to be presumed that a native of Bagdad, such as 
E. Nathan apparently was, used Arabic as the language of 
his narratives. 

2. The Arabic is written in a very pure style, and reads 
most fluently. 

3. A number of additional words and phrases, some of 
essential character (see later), are to be found in the Arabic, 
which would be quite impossible in a translation. 

4. The Hebrew, though in general fluent and free from 
the distressing Arabisms which mar the philosophical style 
of Hebrew translations, betrays in several places the 
influence of the Arabic. Confining ourselves to the part 
to which our MS. offers the text, we find so obvious an 
Arabism as in Neubauer, 79, 8, below, * tfwa jnu ,,,'ti D1K 
= w_jj,*Xl or j)+i\\ (the word is torn off in the MS.) 
followed by v>. The strange expression Dtwn by 'h uroi 
(79, 19), "they should do homage to him" = «J^ IjJAoj 
(verso, 1. 12), is explicable only if we remember that «wc 
means also " to write." DlSDD by 1DJJ xbv 2 (78, 8, below) 
= Lfts^S ^ i_a« J (recto, 1. 3) is better Arabic than Hebrew. 
The peculiar idiom itMO by toy (79, 5, below) = iulc i-ASj 
(verso, 31) is somewhat nearer to the Arabic. Construc- 
tions like iwi Krm»iai> ru»» taw nson bz) (78, 6, below) = 
J-aS bj^i Jt I4H y>\ aJU x^*. 3 (recto, 6), or njDD bxaw Diana bit 
tea Nin (78, 2, paenultima) = ^J'b Uio j\sa t &)/&} f***?j (recto, 
15), sound rather Arabic than Hebrew. The Biblical 
expression lm^lD pN (79, 10) applied to a town is probably 

1 See about d»d: instead of <tc, p. 760, n. 6. 

8 The Diccm of the Edition is most probably a correction. On the MS. 
containing the Hebrew text (now in the possession of Mr. Epstein, Vienna), 
cf. Marx in ZeitschriftfUr hebraische Bibliographie, V, 58, 


suggested by the Arabic s\S, meaning both a country and 
a city, a word which our text actually contains (verso, 15). 

In view of the first-rate importance of R. Nathan 
Hababli's text as an historical source I give in the fol- 
lowing also an exact English translation of the Arabic 
text, accompanied by some explanatory remarks, for the 
benefit of non-Arabists. 

It will easily be seen that the Hebrew is not a slavish 
translation of the Arabic, but rendered a little freely, 
taking more regard to the spirit of the Hebrew language, 
and here and there even assuming the character of a para- 
phrase. The deviations from the Arabic will be noted 
more fully hereafter. Here I should like to bring into 
prominence only the most important of them as far as they 
are of historical value. 

Rector L. 1, nxjn rVin confirms the additional words 
of the MS. (Neubauer, 78, 31). L. 4. The addition Vil 
1N33N^N gives us a most valuable indication of the geo- 
graphical position of Pumbadita (see p. 756, n. 3). Ibid. 
The great chronological difficulty of the Hebrew text, 
ascribing forty years to the presidency of Kohen Zedeq 
(cf. Gratz 3 , V, p. 391), is removed by the Arabic, which 
merely states that the quarrel took place in the fourth year 
of his presidency. L. 6. According to the Arabic the 
Exilarch claimed also the right of sending dayyanim to 
Khorasan on his own authority, whereas this had been 
done before by the academy of Pumbadita 1 . L. 10. The 
Arabic text has fortunately preserved the names of the two 
sons of Natira — Sahl and Ishaq. They are of course 
identical with Sahl and Ishaq of the Natira family men- 
tioned in the Genizah fragment published by Harkavy (in 
Berliner's Jubelsclirift, hebr. Abteilung, p. 34 ff.) 2 . The 

1 Cf. Neubauer, 85, 17 : among the privileges of the Exilarch is also 
counted the fact that he was inwro pn DrrtN h'sto. 

a I venture the hypothesis that the author of the Natira fragment is 
the R. Nathan of our fragment. Both show a most intimate knowledge 
of the things reported about and a great love for details, and both display 


designation of Natira, the son-in-law of Pinehas ben Yosef, 
as " the father of Sahl and IsMq," is of great interest. For 
this expression is intelligible only if at the time of the 
report Natira senior was dead, and was therefore better 
described as the father of the well-known brothers Sahl 
and Ishaq then flourishing. This is in perfect agreement 
both with the fact that in his report on Saadia (Neubauer, 
80,6, below) E. Nathan speaks of the sons of Natira, who con- 
tinued the anti-Exilarchic policy of their father, and with 
the statements in Harkavy's fragment, according to which 
Natira senior must have died 916 (see p. 749, n. 2). As it is 
stated in the beginning of the report that R. Nathan had 
witnessed only in part the events related by him, it is 
most natural to assume that he was only a contemporary 
of Natira's sons, and an eye-witness of the struggle between 
Saadia and the Exilarch, while in his account of Kohen 
Zedeq he relied on what " he had been partly told." 
L. 13. The additional word "lint?, " several months," is not 
without importance in fixing precisely the chronology of 
the events reported there (cf. Gratz 3 , V, p. 393). It implies 
that 'Uqba had lived in Karmlsin some time before he 
happened to meet the Sultan. L. 17. Instead of the 
indefinite *]bn the Arabic has " Kisra," and mentions also 
the name of his concubine, both in perfect agreement with 
the Arabic sources (see p. 757, n. 10, and p. 758, n. 1). 
Verso : L. 6. According to the Arabic 'TJqba was prohibited 
from entering only the city of Bagdad. It seems to be 

exact knowledge of the whereabouts of the Natira family. It would be 
quite strange that at the same time there should have existed two 
different men writing on Bagdad, on the same period and the same circle, 
almost in the same style and manner. I think it is rather probable 
that both fragments belong to one book bearing on the " history of 
Bagdad " (see p. 753). As the number of words on a page of both frag- 
ments is almost the same (the number of lines on a page is not indicated 
by Harkavy), I should not wonder if it were not the same MS. The only 
difficulty is that the date suggested by the Natira fragment (about 930, 
see Harkavy, 1. c, p. 35) differs from that implied by ours (about 945). 
But perhaps there is some mistake in the numbers. 


a misunderstanding on the part of an editor, who mistook 
i>33 for Babylonia (see p. 752 and p. 759, n. 4). L. 8. The 
Arabic has three or four years, instead of the four or five 
years of the Hebrew (79, 14). L. 9. The Arabic confirms 
the reading of the edition VTH p (instead of 1*in in the 
MS.), and shows also more exactly the degree of their 
relationship : their fathers were brothers. L. 10. According 
to the original one of the chief motives of Kohen Zedeq in 
refusing to recognize the new Exilarch was his ambition 
and his dread of David ben Zakkai's interference. 

In spite of the fact that some of the events reported by 
R Nathan Hababli go back to oral information, his trust- 
worthiness and exactitude can hardly be overestimated. 
He is thoroughly familiar with the things he reports about. 
He knows them down to their minute details. Even where 
he has to rely on oral reports his exactitude is admirable. 
This impression made upon us by the Hebrew version is 
strengthened still more by the Arabic original. As said 
above, he knows the sons of Natira by name, and mentions 
them in the order of age. He states exactly the geo- 
graphical position of the places he is speaking of. He 
knows for whom the statues near Karmisin were erected, 
and also gives an exact description of them. It has not yet 
been noticed that his report about the figures is confirmed 
by the Arabic sources in almost every detail 1 . These cir- 
cumstances are of great importance, for they are apt to 
confirm still more our confidence in him, and give him the 
advantage over the younger authority Sherira where he 
is in contradiction with him. 

It has already been mentioned that the beginning of the 
fragment is most unfortunately torn off. On the right- 
hand corner, however, we can still discern the words 
p pnDN 13N T«b6n Kn. But the handwriting is different 
from that of the MS., the ink likewise, and I do not 
think that this mutilated superscription can claim great 

1 See the notes to the translation. 


authority. At any rate, it is obvious that the fragment 
began here, the lines on this page beginning a little lower 
down than on the next one. It has been convincingly 
maintained by Isaac Halevy (Doroth Harishonim, III, 3, 
p. 75 f.) that the paragraph on the " prerogatives of Sura, 
as compared with Pumbadita" (Neubauer, 77 f.), preceding 
the account on 'Uqba, does not belong to the report of 
R. Nathan Hababli, which is introduced by the words 
\nn \n: '"1 "1DK "IBM. This view is strongly supported by 
our manuscript, for otherwise it would be a strange co- 
incidence that R. Nathan's report in our MS. begins on 
a new page, and in a way distinctly marking a new para- 
graph which introduces the report of R. Nathan. On the 
other hand, it may easily be seen that the language, which 
is an excellent Hebrew and highly characteristic, is the 
same in both. Besides the general congruence in style, cf. 
expressions like p. 77 penultima Dnjn ni^3Dne> *T]/ and 83, 6 
bnpn run moan, 78, 18 bto nmp nben and 78, 33 rhs>w iy 
Brpra ntaop, 78, 19 orwa mt?a tew and 82, 7 rrwa nnwxni 
DfW3. We have, therefore, to assume that the two pieces 
were worked out and pub together by a compiler. The 
hand of a compiler can also be seen in the introductory 
remarks [run |ru '1 nox "msw, 78, 30, 83, 5, 86, 13. As to 
the question whether this compiler was at the same time the 
translator of our text, it can scarcely be inferred with any 
certainty. The fact that the Arabic "ttOJO bim K$K (verso, 6) 
is rendered in the Hebrew by ba nia^D ba DW vbw rather 
points to an editor who mistook ^aa for Babylonia and 
added the words mai>» ^33. Perhaps the conjecture may 
be allowed that somebody who was writing on the history 
of the academies compiled different sources l , one of them 
being R. Nathan of Bagdad, who had written on the history 
of Bagdad (see p. 749, n. 3). 

1 The piece on the prerogatives of Sura to be found in the dvan ponr, as 
quoted from the nchnn wian of Samuel Hanagid, differs too greatly from 
Neubfiuer, 78, 3 ft, to be a quotation from it, or vice versa. Both probably 
go back to a common source. 


The question whether the report of R. Nathan was put 
down by him in writing, or, as Halberstamm (Yeshurun, V, 
p. rbp note) maintains, was transmitted by him orally, can 
hardly be decided. Judging from the expression 1J2rOB> 1»3 
(86, 5) or rbwh mrot? 1IM (87, i) we would rather be inclined 
to assume the former. I have the impression — it cannot 
claim to be more than an impression — that the words 
Itfm 1N3DN (recto, i) "the history of Bagdad" represent 
the title of the book of R. Nathan, to which, if my conjec- 
ture (p. 749, n. a) be correct, belonged also the fragment 
on the Natira family 1 . 

The Arabic original suggests many a new question in 
connexion with the extremely complicated problem of 
R. Nathan Hababli. But these I have to leave to historians 

Text 2 . 


nijn rnn nd» ix-in isaas p p s 

K^TBK b>H bin jttt ^K ni1N"li>N PDpty JH^K DNl] .... 

rw vpv p pre |ro •bit Nnivoa b>y w tb fiT'na pD riDK[n^K] * . . . 
Knrwi ypi jn ba po n bjno n»N*N '•a "»n3jkJ>k vn Nnnaoaa 

onp^N *a jndn-13 rwn }« -jhi jndn*i3 -oi »by m-ran s^nais 5 

>!>« twiiwoK rbtap yoii pn xn^K ina* ksvhjj pi xnnaiDab 

Nruta^ rb>5? p pn nh^n iii' 1 tapy ntnttti bsn NiriaoB 

pa ibi >q yaxia KriHaoa jn nosh xr6xiDx:i -lnisriDn 

*)DV n^jj pnv fro -ixwk p jtoi snnaoa nya* B>xn pis 

fan jjb psnosi bno ia« trvtw nimn jit nanai diwb p 10 

1 Harkavy in a note to the Hebrew translation of Gratz (III, 291, 
n. 82) mentions an Arabic fragment bearing on the struggle of Saadia 
with David ben Zakkai. It would be of great interest to know the 
relation of this fragment to the report of K. Nathan Hababli. 

5 The MS. has no vowels at all, and very rarely diacritical points. 

5 The letter before the p is perhaps n. A remnant of pnD« ? 

4 There is space for one short word, perhaps 'Vil. 5 Sic. 



ttoja v$$ pcwip n5d> ircio *btt tapy •<£>: jn >btt ibzbtt mil 

nob® J3 D*10J> JtOI JNDN13 pno ^ NOD NiT-p-lt? »B DN-N PI3 

i>N i-p jn ^n mntj' jwo-ipa Naps dnpnb tnw na^ k>n[-i] 
ba *bt< nNhS" ama jx fni jwo-ip ^n Nfflbno "itnaa jo jnb!>[d] 

«nj-iN3 tin* nhjd into jNnayr yoii -iN»ni tfoio ^n 1 . . . is 
aipjo prion? -np pi>yo oi>io r6 jsnat? n6d* jwid ho spj ^[n] 
noi '•a p6bdk joi nm^i pT'D rinw kids rhre res nas[^] a 
n«a jo S> t no ^3j^»k jo h3M '•ixa im nvw 3ip:o Dia n-i[ix] 

13NB DV 5>3 'a ffSrtoiw ^K 33T' JND^K JN3B Tori |0 313*1 

r6 tyn mjn has rb njt6ni hdndn eppiiwa xapp nDB3 20 
njh *M^N3 jio isa noes "ijy anSta jno^k anto snaionDN 

Nin '•b DNpNB 3nN3^N "]hi [D]TB S>1nS>N TQW xb NJ?[na] tfiPN 


row jnb$>d& ni33 anto^x "D"} fl3D 1JJ3B Nn[>0]i D? 


[fU^NBB wd^n ho xijn "I'to n[V] 

noised jo bBUm jno^n mo33 

'303. "nno now ^y r&N3"iN3 

nt"im ^» ihi doj?b wii ^ia r6 {jjdb 

'. . . n^v mpjn frwri *ai jn ^n iaoh 8 ii>i» n^a mas *a tay 5 

na pnx dn^dn^nb nh^di jw xi3N 1*033 bHr< t&& "ittfnwja 
3-iio^N ^k i*Di noaa >i>y iaxa piB^« \nnba jo [*i]b nW D^a 

1 n is doubtful. Two letters may have preceded. 
1 The whole word is quite doubtful. 
3 Space in the MS. 

* Slight traces of letters justifying our reading. 

* Sic The place on the line is torn off. 

6 We would expect rrom, but the n is impossible. The i might also be 
a s. 

7 One letter missing at the beginning of the word. Might be n or <d. 

* Sic. 

' Slight traces of t and p are still discernible. Not impossibly Tip?to« 4 


i?i otby jk »i>8i in pD a bho fb&yo notvibts noNpNB 1 
JN31 niDip^ napy By ja vn "at p w laoxaB nS&N ^y 
ropy i>ya jd spai noo-ifo ^» rr nn»*i[p]n 5n' n^> pix jna 10 
i>Ni a-ra ija "6k ntid ^n anai ixiaaa ntid fiaTio niD*ipB 
2 ii>T3B nnDNn npri ma-isa* jn omoto ^-ia tm^n 
yiiD ^n N11D jd dn^nfo yoj yo jwn^Ni a-pa s ua 
fai Nnya itnaa fibaij? »b nane> rfrno [n]spi»N ni? Jjnp'* 
[n«]"iN3B nh?x ktodi "at ja -m i$>a \m Swdk nrio x-iid 15 
[n^]i nnDNn pnv jna jjbnt ih yDj >bi ' nnDKn npyr 
1 [«iny»fo] ni>a B*n *d*j jsai j^d i ira "fri 4 dnpn nhjd »p ^n na^ 
6 [jfa«] nnn ho pata ntd r6&<5> snn s^na 1 ' *m «b«nra^ 
8 . . d Dt6aa ^ap b nna^ jxai nW .t!?k 7 ^4>» r*o jk ^k jmd 
9 ["6y] xnxyai am >^y rftspo nana i>ap i< nW>ta ^n nnsa 20 
10 Wta tfti d-it nat«VK n^y ejpi jk ^n pis jna asiax 
pi^n nhan aaDi n[*]io n^y Dtbyi pnv jna naxiriDNa 
T$>« nhn jk ^k bap 't» t&i* nnna nd naviDfo dk-iso nb 7 pa 
Dtn inton jk ybtt ana ni> 7 p tnna mix e»ki ni> iwpa 
ro"6» t6) n«ii ppm iin ^n naxJNB n^» ipyni nib^N 25 

Translation 11 . 

I ... of the history of Bagdad of what he had been partly told Recto, i 
I . . . the Exilarch 'Uqba of Davidic origin came to Northern Africa, a 

1 Space in the MS. s The i extended below the line. 

3 Possibly '33. * Sic. Read M DNpH. 

5 Supplied from the Hebrew snu. Possibly also ironto. 

6 bn still visible. Supplied from the context. 7 Sic. 

8 D not quite certain. There is still place for two letters. Perhaps tod. 
See p. 760, n. 9. 

* A very slight trace of a letter which might indicate an ». Probably *>3 
which fits in exactly. 

10 A stroke at the end of the line in the MS., and a space at the beginning 
of the next line. 

11 Prof. Theodor Noldeke, with the kindness so characteristic of him, read 
the proof of this article and added a few explanatory notes. These are 
included in brackets and marked at the end by N. 



3 I [He had occupied ?] l the supremacy for many years, the number 
of which he (R. Nathan) could not make out. Kohen Zedeq ben 

4 Yosef a had been in charge of the academy | in Pumbadita (i. e. 

5 Anbar) s in his CUqba's) days for about four * years, when [ a differ- 
ence of opinion and quarrels 5 broke out between them, on account 
of the jurisdiction of Khorasan. For the jurisdiction of Khorasan 

6 had in olden times belonged | to Pumbadita, whence *the dayyanim 
used to be sent thither 6 , and all the tax 7 on her revenues used to go 

7 to I Pumbadita. c Uqba, however, wished * the dayyanim to be sent 

8 to her (Khorasan) by himself, in order to take possession of her | and 8 
get hold of her revenues for himself alone * to the exclusion of 

9 Pumbadita 8 . But Kohen | Zedeq, the head of the academy of 
Pumbadita, protested against it. Among those who assisted Kohen 

1° Zedeq * against him ('Uqba) 8 were Yosef | ben Pinehas and his son- 
in-law * the husband of his daughter 8 Natira, * the father of Sahl 

n and Ishaq 9 , together with some of | the most prominent men of the 
land. At last 'Uqba was banished to a place called Karmisin beyond 

12 Bagdad | about five days' journey to the East, * on the way leading 

13 to Khorasan 10 , while Amram ben Shelomoh | was the head of the 
academy of Sura. 'Uqba had been in Karmisin * for several months u 

14 when the | Sultan left Bagdad 13 to take a pleasure trip to Karmisin, 

1 See p. 753, n. 4. 

2 The Hebrew has erroneously pidv mi or rpv n fa. 

' nun^N td missing in the Hebrew. The identity of Pumbadita and 
Anbar is confirmed by the itinerary of Benjamin of Tudela where -naVs 
(ed. Asher, II, 69) and "arte (p. 53, J. Q. B., XVII, 292, Tnrtw) are to be read 
■iun 1 **. [I notice that Gratz 3 , V, 389, n. 1, has already this correction.] 
In accordance with this the conjecture of de G-oeje, Z.D.M.G., 89, 10, that 
Pumbadita (on account of the reading ■ma'w) is identical with iiil 
is out of the question. On the position of Pumbadita cf. Berliner, 
Beitrage zur Geographic una" Etknographie Babyloniens, p. 57, and de Goeje, 
loc. cit,, p. 10 ff. 

• The Hebrew ruffi 'o. Perhaps the translator misread jo,1 for ^k><\ . 
See Introduction, p. 749. 

5 "jA, pi. .j-i. querelle (Dozy). 

• Missing in the Hebrew. Literally : from it (Pumbadita) a dayyan 
would go out to her (Khorasan). 

7 aJLJ tax. Cf. the explanation in Dozy. 

• Missing in the Hebrew. 

9 Missing in the Hebrew. See Introduction, p. 749 f. 

10 Missing in the Hebrew. 

11 Missing in the Hebrew. See Introduction, p. 750. 

12 The Hebrew only ita. The reading of the edition tan -po shows that 
the editor no longer knew that tan was Bagdad. See p. 752. 


because there were old monuments 1 . . . | . . . 2 and springs 8 and 15 
trees, and all the saffron of Bagdad used to come from there. 
Outside of it (Karmisin) 4 , | (at a distance of) nearly half a mile 6 , jg 
there was a place called Shafaran 6 , which had a hanging 7 hall of two 
men's height in excavated 8 form 9 , | containing the statue of Kisra 10 17 

1 Arabic ,IS|. In the same way Ibn Rustah (tenth century, ed. de Goeje, 
p. 270) in mentioning this place says i^iS y^sJl (S)j)l A?T *->jU \'M>J' 

The rest of the line has hx ^N "to the." But the following word is 

1 See p. 754, n. 1. 

3 " Springs," literally "waters." For a similar description of Karmisin 
cf. Ibn Hauqal (wrote 977, ed. de Goeje, p. 256), l$*S &Ja) iliA* -J*. 

* The hall, as well as the figures of Khosrau, Shirin, and the horse 
described in the following lines, are often mentioned by Arabic writers : 
Yaqut, III, 350 ff. ; IV, 69 f., 112. Mas'udt (died 956), Murug ad-Dahab, 
II, 215, Ibn Hauqal, 1. 1., Muqaddasi (wrote 985), ed. de Goeje, p. 339, 
Ibn Rustah, 1. 1., pp. 166 and 370. For particulars see the following 
notes. The Hebrew has here more the character of a paraphrase. 

5 Yaqut, III, 250, has one parasang, Ibn Rustah, p. 166, three, Ibn 
Hauqal, 1. 1., eight parasangs. There seems to be some contradiction. 
I am not in a position to state the exact relation between a mile and 
a parasang at that time. According to the Talmud a parasang was equal 
to four miles. Cf. Kohut, Arueh Completion, s. v. nciE. [This is also the 
usual way of counting in Arabic. N.] 

• The vocalization is not certain. I did not find any references to this 
place in the Arabic sources. Instead of that we very often find the name 
"qasr Shirin," "the castle of Shirin" (Yaqut has a special article on it, 
IV, 112), which the Persian king built for Shirin, and where she used to 

pend the summer, Ibn Eustah, 270. 

7 'liJl (Ji£ "he made the building . . pensile, i. e. supported above the 
ground by pillars . . . Hence they say «j»Ic. JojUI U*SJ, they dug 
beneath the wall . . . and . . rendered it (jliU, i.e. pensile" (Lane). 
See also Dozy. The hall seems to have been arch-like, since Ibn Eustah 
(p. 166) and Yaqut (III, 250 ; IV, 69) call it a taq, " an arch-like building." 
Cf. the following note. 

8 Arabic np:o, cf. the foregoing note. I am not clear about the exact 
meaning of this architectural terminus technicus. I think it means 
excavated (in the rock), and is identical with Ibn Eustah's statement 
(p. 166) that it was^s*^' ^J ,jfl-U ,jlL>, "an arch-like building hewn 
out of the rock." As aipm occurs a second time in line 18, I do not 
believe that the correction -npM would be justified. 

9 This word is doubtful. See p. 754, n. 2. 

M The king alluded to is the Sassanide Khosrau II Parwez, "the 


and the statue of Sinn, his concubine *, and under him (the king) 2 
jg on the ground 3 | the statue of a horse 4 , also excavated 5 and hollow, 
the water which came from the mountain entering through its 
jq mouth I and flowing out through its tail 6 . And the Sultan used to 
20 ride to this place every day on horseback. Then 'Uqba made up | his 
mind to place himself before him and to salute him. The first 
si salutation he addressed to him | was considered elegant by the 
secretary of the Sultan, and he wrote it down for himself. On the 
32 following day he delivered | another salutation, which did not re- 
semble the preceding. The secretary of the Sultan wrote down this 

23 (salutation) too. In this way he went on | [for a whole year], 
delivering every day a salutation which did not resemble the others, 

24 and the secretary | writing down a[ll of] them. After the lapse of 
Verso, i a year the secretary told his story to the Sultan, and that he | had 

not duplicated any salutation during the year, and he asked him . . . 

victorious" (in Arabic, "Abarwfiz," see the writers mentioned above), 
reigned 590-627 p. C. (Noldeke, Aufsatee zur persischen Geschichte, p. 124). 

1 The pronunciation of the name is usually given as Shirtn with a (_p, 
but the codex of Ibn Rustah (p. 270) always has Sirin. This seems to 
have been the pronunciation in Bagdad. Shlrin being a Persian word 
(meaning " sweet "), there is the same difference in pronunciation as that 
between the usual Karmisin and Karmashin ( = Karmansah), Ibn Rustah, 
p. 166. Shirin was famous for her beauty (cf. also Mas'udi, II, 232), and, 
though a Christian (NOldeke, Aufsatze zur persischen Geschichte, p. 125), she 
was the favourite concubine of Khosrau. 

2 Since we know from Ibn Eustah that the hall was very high, a stair- 
case of 250 steps hewn in the rock leading from the bottom to the top 
lt^&\ (Jl JjlloJl Ji-il ^y*), it would be more natural to translate "and 
at the bottom of it (the hall)." But on the other hand, we know that 
Khosrau was represented riding on his horse (Yaqut, III, 250, 252, Ibn 
Hauqal, 1. 1., Mas'udi, II, 215), and R. Nathan has hardly left out this 

3 I am not certain about this translation. It might also mean on a 
pedestal or substructure. Besides, it would be natural to expect 'V? 
instead of <E. 

* The name of this famous horse was Shabdaz or (pronounced with the 
imala) Shabdez (see the above-mentioned sources. Yaqut has a long 
article on it, quoting several poems in its praise, III, 250 ft). 

6 See p. 757, n. 8. 

e In this detail R. Nathan differs from Yaqut (III, 250) and Ibn Rustah, 
who maintain that the water came from the figure of a man, standing on 
a side (Ibn Rustah). According to Mas'udi (p. 215), Ibn Rustah, and Yaqut,- 
there was a large number of different figures in the hall, of birds (Ibn 
Rustah), of men, women, footmen, and horsemen (Yaqut, III, 250). 


[ and the Sultan looked at him and considered his words . . . | and he a, 3 
granted him this 1 , and he returned. This, however, was a heavy 
(blow) for Natira . . . | effort to banish him, and did not stop intriguing" 4, 5 
until he was banished a second time and [decrees ?] were issued | 
publicly ' against him that he might never enter Bagdad *, and in case 6 
he entered it, Islam should be more appropriate for him 8 . | But 7 
no city of the cities 6 of the East could offer him a foothold, and he 
made up his mind to emigrate to Maghreb 7 . | And the supremacy 8 
remained unoccupied for about three or four 8 years, until it became 
hard | for the nation, and they spoke to David ben Zakkai, who was 9 
the cousin of 'Uqba on their father's side, in order to appoint him- 
But J Kohen Zedeq did not like his appointment, * being jealous of 10 
his supremacy, and being afraid on account of what had been done 
to 'Uqba 9 . | So the academy of Sura appointed him * in Bagdad' , n 

1 J^ - used of a king, " qui accorde quelque chose dans un dipldme." 

2 <__pVi l 1, " employer desmoyenssubtils, p. e. la ruse, la flatterie." Dozy. 

3 j^H, rv, " publier, promulguer, faire proclamer par un eerier public." 
Dozy. It seems that the reading of the edition psn D»n (instead of 
-iton n«a of the MS.) has something to do with the Arabic intoni. 

* Here there is a serious difference between the Arabic and the Hebrew, 
according to which he was banished from the whole kingdom. Since it 
is impossible that the translator should have mistaken into in this way, 
it is to be assumed that the misunderstanding goes back to the Editor. 
See pp. 750 f. and 752. The MS. adds also 79, is VKha toj which is not to be 
found in the Arabic, [toi perhaps misread for bin.] 

* The Hebrew has won to mn\ I am not quite certain as to the 
meaning of the Arabic phrase. Perhaps it means that he would be 
forced to accept Islam. [I know of nothing better though it looks very 
strange. The Hebrew translator has certainly taken offence at the 
expression. The excommunicated person is probably forthwith handed 
over to Satan. N.] 


6 The plural ,jljJb has usually the meaning of " cities " (see Freytag 
and Lane), and this translation is more in accordance with the statement 
that he was excluded only from Bagdad. The translator who has y-i.< 
jtoin too misunderstood jij . Cf. also p. 748 f. 

7 At that time in the hands of the Patimides who were entirely inde- 
pendent of the Eastern Califat. [The Arabs of Iraq, however, just as the 
Jews of Babylonia in earlier times, consider Syria and Palestine a part 
of Maghreb. N.] 

8 The Hebrew has four or five. 

9 The Hebrew has only rapw to isnp mrro 'dV 
10 Missing in the Hebrew. 


Thereupon he wrote l to Sura, to the ordained scholars and to the | 

12 students regarding this matter, and commanding them to bless him 

13 and to acknowledge his supremacy. And | the ordained scholars and 
the "tannaim" 2 , together with all the students, went down from 

14 Sura to a place | by the name of al-Kasr s , a beautiful town to the 
! 5 south of Bagdad, there being between itself and | Sura six* miles, this 

(al-Kasr) being the town of David ben Zakkai and (the place of) his 
j5 origin 5 . Then they blessefd him] | and acknowledged his supremacy. 

Despite all this Kohen Zedeq contested his supremacy, and [did not] | 
? grant him the least thing thereof, remaining in this attitude for 

about three years. And Nissi 6 , the head of the " Kalla " assembly, 
j8 [known] | as the Nahrawanite 7 , was blind 8 , and he used to mediate 

between them in order to settle the quarrel during those [three] | 

19 years, until he once came to him in the night. He used to open 

20 every lock by a word [of his ?] 9 | and he had opened that night 

1 Grammatically "he wrote" can refer only to David ben Zakkai. 
But the Hebrew is more logical in referring it to the head of the Academy 
of Sura, and reading niid rasr tarn. 

1 On the " Tannaim " at the time of the Geonim, cf. Lewy, fiber einige 
Fragmente aits der Mischna d«s Abba Said, p. 9, n. ia. 

3 There were quite a number of places bearing this name ( = castle), 
[Ours is undoubtedly qasr Abi Hubaira, which is often called al-qasr and 
lay in the near vicinity of Sura, see Yaqut, 4, 133, 17. N.] 

* The Hebrew edition has seven, the MS. ten miles. 

6 The Hebrew has wiVid yw, the Biblical expression probably suggested 
by the word jJb. See p. 748. 

* The Hebrew edition has D'DJ, the MS. ro. Since later on (79 
penultima) both the edition and the MS. (the latter according to the 
information of Dr. Marx) have 'D3, it is to be assumed that the editor 
took d ( D31 5JT13 to mean "known by miracles," regarding the fact related 
in the succeeding lines as a miracle. See note 9 on this page. On the 
name *D3 cmp. Harkavy, Zikhron larishonim, V, p. rap. 

7 From Nahravane, half a day's journey to the east of Bagdad (Neubauer, 


8 I found nowhere the euphemistic use of TS3. 

9 See p. 755, n. 8. The Hebrew has una "m (MS. Visas) %m Vs nniD rem 
which can mean only : he used to open all the locks in Babylon with the 
Shem (Hamephorash). This strange translation suggesting the wonder- 
working character of the man may have been caused by a misunder- 
standing, the translator probably having misread [d]dnV32 instead of 
DNtoa. The meaning is not quite clear to me, on account of the missing 
word. Perhaps he wants to say that being blind, and having special 
difficulty in opening doors, Nissi only had to call his name to have other 
people open them for him. At any rate, the original does not suggest 
anything miraculous. [To me it seems that the author refers to a miracle 


fourteen 1 locks closing gates, some of them [on] | the doors of 21 
Kohen Zedeq, until he stood hefore him and found him studying in the 
middle of the night. | And Kohen Zedeq was frightened hy him and 22 
strongly impressed by his coming and the reason of his visiting him. | 
And he said unto him, head of the Academy ! I unfastened no 23 
less than fourteen ] locks before I reached thee. | And he said unto 24 
him : What dost thou mean thereby ? He said unto him : I beseech 
thee to bless the Exilarch and acknowledge him. And he conceded 
this to him, and he confirmed his hope and did not disappoint him. 

I. Friedlaender. 

which at bottom has been caused by some misunderstanding. "To open 
all doors " prpbably meant " to overcome all obstacles." N.] 
1 The edition has fourteen, the MS. four.