Skip to main content

Full text of "The ritual of Eldad ha-Dani reconstructed and edited from manuscripts and a genizah fragment"

See other formats



^ tip Qtyalagitui $ 





BM 720 .S6 E6 1908 
Eldad, 9th cent. 
The ritual of Eldad ha- 
reconstructed and ed. 



^m or m 

(* NOV 15 1 











In January, 1904, Professor S. Schechtee, President of the 
Jewish Theological Seminary of America, kindly drew my attention 
to some Genizah fragments which bore on Eldad ha-Dani, 
knowing my interest in that peculiar and much maligned char- 
acter of Jewish history. Whilst preparing the fragments for 
publication, I found that the material grew unproportionately 
large and warrented a new edition of the entire Eldad Halakah. 
The execution of this more comprehensive idea involved fresh 
difficulties and confronted me with new problems. As a result 
of my investigation I publish this pamphlet which, I hope, will 
contribute to the solution of the difficult Eldad problem. 

I take pleasure in expressing my obligation to those who 
have farthered me in my work. I am indebted to Professor 
S. Schechtee for his kindness in placing the Genizah fragments 
at my disposal; to Dr. Eduaed Baneth, my former teacher, for 
undertaking the weary task of correcting proofs, as well as, for 
numerous valuable corrections; and to my friends Dr. Louis 
Ginzbeeg, Dr. Henry Maltee and Dr. Alexandee Maex for 
useful information and stimulating suggestions. Dr. A. Cowley 
kindly provided me with photographic copies of the Oxford texts 
used. Dr. A. Cheistman and Dr. M. C. Schaae gave me valuable 
explanations of some anatomic and pathologic discussions in the 
Ritual of Eldad. To them also my sincere thanks are due. 

Berlin, January 29 th 1908. 


Frontispiece— Genizah Fragment T-S Loan 110, University Library 
of Cambridge, folio 2, verso. 

Preface vii 

List of Abbreviations vni 

Chapter I, Introduction * 

Chapter II, Eecensions of the Text 

Recension A 9 

Recension B I 7 

Chapter III, Plan and Contents 25 

Chapter IV, The Language of the Ritual 29 

Glossary 32 

Chapter V, Place in the Earlier Halakic Literature 46 

Text of Recension A 5o 

Text of Recension B y9 

Appendix on the Eldad Legends 1° 5 

List of Corrigenda !29 


A Eecension A. 

B Eecension B. 

G Risdlat al Burhdn etc. by Samuel ibn Gama', MS. Oxford, Catalogue, 

by Neubauer no. 793 1 . 

P MS. Parma, de Rossi cod. 327, 22. 

T-S Grenizah Fragment T-S Loan 110., University Library of Cambridge. 

Glosses on the Mordecai, MS. Oxford, Catalogue, by Neubauer no. 678. 

Je. Jellinek's copy of O. 

M Mordecai, JSullin beg., quotation taken from Babbenu Baruch. 

MS MS. of M, in the Library of the Jewish Theological Seminary of America. 

R Responsum, editio princeps [Mantua 1480]. 

Oxf Cod. Oxford MS. Hebr. d. ll., Catalogue, by Neubauer and Cowley, II 

no. 2797, l,n. 
T-S Genizah Fragment T-S Loan 94. University Library of Cambridge. 
ed. pr. Responsum, editio princeps [Mantua 1480]. 

Sh. Shalshelet ha-Kabbalah of Gedaljah ibn Yahya, ed. Venice 1587 p. 37b. 
Pet. Cod. Petersburg II, Collection Firkowitsch 1261. 

E. A. Epstein, Eldad ha-Dani, Pressburg 1891. 

Fi. H. Filipowski, Liber Juchassin, London 1857, p. 207 — 208. 

Go. D. B. Goldberg, Epistola R. Jehouda Ben Koreisch, Paris 1857, 

p. XIX— XXI. 
Mii. D. H. Miiller, Die Recensionen und Versionen des Eldad Had-Darii, 

Wien 1892, in Denkschriften der Kaiserlichen Akademie der Wissen- 

schaften in Wien, Philologisch-Historische Classe. Band XLL 
Reif. Jacob Reifmann, '•inn Vibx "1SD ^ D^riK Q^in, in Ha-Karmel, 1870 vol. 

VIII, pp. 254 et seq. 
St. M. Steinschneider, Schlachtregeln in arabischer Sprache in Geiger's 

Judische Zeitschrift II, pp. 76 et seq., 297 et seq. 

Gl. Glossary, 
no. sections in B. 
$. sections in A. 

end of a line in T-S Loan 110. 

beginning or end of a text or MS. 
1. line. 

Capital letter in left margin indicates MSS. and texts used. 

The first letter indicates MS. on Avhich the text is based. 

The number of the section given in parentbesis in B, refers to the 

corresponding paragraph in A. 



More than a thousand years have elapsed since Eldad ha-Dani 
appeared in Kairwan, and stirred the community by a remark- 
able account of four lost tribes of Israel, and by an equally 
remarkable Ritual on Slaughtering which he represented as "the 
Talmud of the Four Tribes". An interesting phenomenon in 
the Jewry of the ninth century, Eldad has not ceased to interest 
historical study of the present. In spite of this long lapse of 
time, however, there is still the utmost diversity of opinion 
among writers concerning the personality of Eldad and the 
authenticity and significance of the writings that bear his name. 

Such extreme diversity of opinion among writers, as exists 
with regard to Eldad, is a rarity in Jewish history. Thus Stein- 
schneider, writing in 1850, * thought that the work ascribed to 
Eldad had an apologetic purpose. In telling of the existence of 
a Jewish state his object was to refute the Christian theory, that 
the disappearance of Jewish temporal power was a proof of 
God's rejection of Israel. 2 I, B. Levinsohn and Jellinek regarded 
the Eldad writings as a polemic against the Karaites. 3 Pinsker 
adopted the opposite view.* He argued that Eldad was a Karaitic 
emissary from Eastern Babylonia, who invented a fabulous tale 
as to his origin in order to conceal his true mission, which was to 

1 V. Ersch und Gruber, Allgemeine Encyklopadie der Wissenschaften 
und Ktinste, Leipzig 1850, Zweite Section XXVII, pp. 390, 393. 

2 L. c, v. also idem, Die Geschichtsliteratur der Juden (1905) pp. 14 et seq. 

3 V. Bet ha-Midrasch II (1853), p. XXIX. This view was also held by 
Harkavy, V. note 75 in Graetz' Geschichte (Hebrew edition) III p. 267. Cf., 
however, Harkavy's "Skasanija jewreiskych pisatelej o Chasarach i Chosarskom 
zarstwe", Petersburg 1874, p. 6. For Levinsohn's view see his Bet ha-Ozar, 
Wilna 1841, p. 245. 

4 V. LiMute Kadmoniyot (1860) p. 109. 


propagate Karaitic doctrines among the Rabbanite Jews of the 
Maghrib. These writers agree at least to the extent of attributing 
serious motives to Eldad. Graetz, on the other hand, considered 
him "a braggart and a secret Karaite", "an adventurer and 
charlatan". 5 This opinion was also shared by Chwolson, although 
he disagreed with Graetz in regard to Eldad's Karaism. 6 P. F. 
Frankl went even further and described Eldad as "a rogue and 
a swindler", "devoid of any higher purpose", and as a man "whose 
deception was an indisputable psychological fact", and as "a com- 
patriot and counterpart of Ulysses". 7 Neubauer took a similar view 
and characterized him as "a cunning emissary of the Karaites"? 
and as "a daring impostor, crowned with an unexpected success". 8 

In Reifmann, however, we find an entirely different attitude. 
He denyed the authenticity of most of the current Eldad literature, 
and held that the real personality of Eldad was unknown to us. 9 
Metz suggested that Eldad's Ritual on Slaughtering represented 
the religious practices of the Falashas. 10 Epstein, following this 
suggestion, regarded the writings of Eldad as a description of the 
conditions of the Falashas in Abyssinia, or of a Jewish commu- 
nity somewhere in Southern Arabia. 11 

This diversity of opinion among writers is to be attributed 
to two causes. The anomalous character of the literature 
ascribed to Eldad naturally aroused the ill-will of writers who 
found themselves baffled in the attempt to fit it into known 
literary categories. The character of Eldad himself added no little 
difficulty to the task of the historian. His personality offered 
itself as a fertile field for folk lore, and legends soon clustered 
thick around him, until the original became scarcely recognizable. 

s V. Geschichte IV (1860) p. 243. 

s According to Gottlober in his Bikkoret Letolodot Hakkaraim (1865), 
p. 65 Note 2. 

7 V. Monatsschrift fur Geschichte und Wissenschaft des Judentums 
(MGWJ.) 1873 pp. 489, 484, 491, 1874 p. 555. 

8 V. Where Are The Ten Tribes in Jeivish ? in Quarterly Revieiv (JQB.) 
I (1889) pp. 109, 110. 

9 V. Debarim Ahadim c al Sefer Eldad ha-Dani in Ha-Karmel VIII 
(1870), pp. 254 et seq., reprinted in Ha-Boker Or VI (1881). 

io Judisches Literaturblatt 1877 no's. 40—43; MGWJ. 1878 p. 398, 
1879 pp. 135 et seq., 184 et seq. 

11 V. Eldad ha-Dani, seine Berichte iiber die X Stamme und deren 
Ritus, Pressburg 1891, Introduction. 

A more important cause for this lack of unanimity is to be 
found in the greater readiness of writers to pronounce judgments 
than to investigate critically the grounds for their judgments. 
Without entering upon a careful study of existing sources they 
accept uncritically all the literature attributed to Eldad, in the 
form in which we possess it, 12 and hasten to make categoric 
decisions. In the case of Reifmann we have a scarcely less 
summary rejection of most of the Eldad literature, and the 
adoption of a rather agnostic attitude towards Eldad's person. 

In view of this confusion of opinion it is clear that judgment 
on Eldad's personality and place in history must be based on a 
less superficial and a more discriminating examination of the 
literature ascribed to him. This literature is both Halakic and 
Haggadic. The more important is unquestionably the Halakic. 
Eldad appears in Kairwan bringing with him a Halakic work on 
Slaughtering, which he termed "the Talmud of the Four Tribes". 13 
This was made the subject of an interpellation addressed to the 
Gaon Zemah [ben Hayyim (898 — 905)], l4 and is preserved in epi- 

12 Cf. e.g. Frankl's words in MGWJ. 1874 p. 555 that Eldad „ini 
Wesentlichen dasselbe in derselben Weise erzahlt hat, was und wie es unter 
seinem Namen, sei es als Halachah der zehn Stamme, sei es als Erzdhlung 
noch heute "kursiert". 

1 3 V. Responsum ed. Epstein p. 5 no. 6 rumiStt Bnpn \\wbl onto imbnn 

f tti nnbvf Ti&bnn nsp»i uans» irrnx ntya lab tn&V usrom. Too much im- 
portance is not to be attributed to the word n^pttl in this context, which is 
otherwise also obscure. It seems that the Ritual on Slaughtering constituted 
the entire "Talmud of the Four Tribes". The citations in the Kairwan Epistle 
are taken from this Ritual exclusively. If there had been anything more to 
this "Talmud", it is inconceivable that nothing in the part outside of the 
Rules on Slaughtering should have been thought of sufficient consequence to 
communicate to the GaoD. The original text of the Responsum probably 
was '131 r6m rpfcn 13 v*w *bb nnbv niDbnn nrotf n», and not, as the reading 
is in the ed. pr. '131 nittbnn nspttn ttarot? no. ri^pon may be the interpolation 
of a scribe who concluded from '\2\ niJTfc 'in ynv llbx n "IDIKI (ed. Epstein p. 7 
no. 13) that this "Talmud" dealt also with other matters than slaughtering. 
If a word like nspttn was in the original text at all, this was probably nxpOB. 
This is the reading which is given by Jellinek, BH. II p. 108; cf. also Weiss, 
Dor IV* p. 112 line 2. 

14 According to Rapoport, Toledot R. Natan, note 11; cf. also Stein- 
schneider, Geschichtsliteratur p. 15, Epstein pp. 9 et seq. The period of his 
office is variously given as 889 — 895 by Graetz, Geschichte V 3 pp. 243, 456; 
as 882 — 887 by Epstein in The Jewish Encyclopedia VI p. 571 s. v. Gaon; as 
884—891 by Halevy in his Dorot ha-Rishonim III p. 282. 

tome in his Responsum. 15 The Ritual itself, no longer extant in 
its complete form, has recently come to light in a number of 
manuscript fragments, so that it is now possible to reconstruct 
it almost in its entirety. 16 

The other part of the Eldad literature is concerned with 
the lost tribes of Israel, (ten tribes in one version and four in 
another), 17 and with the adventures of Eldad. This Eldad Hag- 
gadah, or legend, is found in the same Gaonic Responsum. A 
much more elaborate form is preserved in the various Recensions 
of the Sefer Eldad. 18 It is the Haggadic material, mainly 
the Sefer Eldad, which until recently formed the basis on which 
opinions were founded. The Halakic material has been rather 
neglected, probably because the Ritual in its more complete form 
was not available until recently. Yet, because of its unquestion- 
able authenticity, !8 a the Halakic work of Eldad is of greater 

15 V. infra, chapter II. Recension B. p. 17. 

16 V. infra Recension A. pp. 9 et seq. 

17 The Responsum speaks of only four tribes, the Sefer Eldad gives an 
account of ten tribes. This discrepancy should have sufficed to arouse suspicion 
in regard to the authenticity of the Sefer Eldad. If Eldad had indeed brought 
such astounding news as that of the rediscovery of the ten tribes, we may be 
sure that the Jews of Kairwan would undoubtedly have been as much inter- 
ested in this information, as in the divergences and peculiarities of Eldad's 
Ritual and that they would not have failed to report to the Gaon in their 
Epistle that ten tribes had been rediscovered, if Eldad's account had not 
restricted itself to four tribes. 

18 The latest study of this part of the Eldad literature is D. H. 
Muller's, Die Rezensionen und Versionen des Eldad Had-Dani, "Wien 1892, 
published in Denhschriften der Kaiserlichen Akademie der Wissenschaften in 
Wien, Phil.-Hist. Classe vol. XLI. 

18a Eldad himself wrote only Halakah, that is he wrote the Ritual on 
Slaughtering. The older rabbinic authorities never mention any work of 
Eldad, other than this Ritual. It is this Ritual ()^in ntTm?) to which Abraham 
Maimuni (d. 1237) refers in a reply to an inquiry as to the home of the ten 
tribes. The passage is contained in MS. Halberstam 56 and was first published 
in Geiger's JZ. X p. 127. It has since been published also by Neubauer in 
Kobez c al Yad IV p. 62 and by Hirschfeld in Berliner's Festschrift (Hebrew 
part) p. 54. A Hebrew translation of the Arabic original is to be found in 

Kobez c al Yad I. c. and in Epstein's Eldad p. 71. It reads n^KDn^W ^ib» n$pl 

; Hirschfeld (?) taND^Ki) bkb^k! nbpii n:y «nissn y6n ybm ntrnw r6 *r\bx 

Hirsch-) b^tibibx rtex MB (Hirschfeld rob) na» nTOK&ta kjpk (J. Z. B«sn«^Ki 

{•jVl •'B (feld ^«*?lb«. The r6«Dl here mentioned, refers to the Gaonic 

importance than the other, namely the Haggadic, for determining 
his character and significance. Moreover it is of interest for 
the history of the development of the Halakah in general. These 
considerations have led the present author to devote his work 
primarily to the study of the Eldad Halakah. 

By way of introduction to the examination of the Eldad 
Halakah, I trace here briefly the progress of critical study and 
opinion of the Eldad Ritual. As early as 1873 P. F. Frankl 
held out the promise of "an analysis of the confused Halakot 
of Eldad", t9 but this promise was never fulfilled. Jellinek's 
study led him to the conclusion that Eldad was not the author 
of the Halakot attributed to him. 2 ^ Pinsker found evidence in 
style and subject-matter for a theory that the bulk of the Halakot 
were of Karaitic origin. The rest, he thought, could be traced 
to the Samaritans. He ventured the suggestion, however, that 
Eldad might perhaps be the representative of an independent 
sect. 21 Without naming Pinsker, Graetz adopted unreservedly 
his views and arguments in favour of the Karaite origin of the 
Ritual. 22 The Rabbinic elements he explained as additions 
"shrewdly adopted to avoid giving offence to the Rabbanites". 23 
The weakness of the pro-Karaitic theory of Pinsker was exposed 
by Schorr, 24 and the anti-Karaitic arguments presented by Gott- 
lober in criticism of the view of Graetz were sound. 25 I. H. Weiss 
similarly laid stress on the Rabbinic-Talmudic character of these 

Responsum, and not, as Epstein thinks (v. BEJ. XXV p. 31 note 4), to "the 
epistle of Eldad to the Jews of Kairwan and Spain". 

From a passage in a manuscript of the Miclrash Bereshit Rabbati which 

opens with the formula used in Eldad's Ritual bl pi p JWirr nii« 

ffttjp vh\ 'T bx Kip" 1 "»B» *6m D^ny xbl p»» ]JV n»« Epstein concluded that 
Eldad had written a treatise on "loans" (v. Magazin 1888 p. 82). It seems, 
however, that Epstein himself later abandoned this view, as there is no 
mention of this in his Eldad. 

The Responsum of Abraham Maimuni cited above shows that at least 
as early as the thirteenth century Eldad was credited with an account of ten 
tribes in place of his original four tribes; v. note 17. 

19 V. MGWJ. 1873 p. 493 note 1. 

20 V. Bet ha-Midrasch II p. XXVIII note 3. 
2t V. Likkute Kadmoniyot pp. 108, 180. 

22 V. Geschichie V note 19. 23 L. c. p. 452. 

24 V. He-Saluz VI (1861) pp. 62—63. 

25 V. Bikkoret Letolodot Hakkaraim pp. 64 et seq. 105 et seq.-, Chwolson 
and Fiinn also rejected the Karaitic theory, v. ib. pp. 112 et seq. 

Halakot, although he agreed with Graetz in his general verdict 
on Eldad. In Weiss' opinion the elements in these Halakot, 
which are in harmony with Karaism, had been borrowed from 
Eldad. 26 Neubauer adopted Graetz 7 views and again gave promin- 
ence to the Karaitic theory. 27 

The controversy centering mainly about the origin of this 
Halakic compendium thus lasted for forty years without making 
any progress. The appearance of Epstein's thorough work, in 
1891, marked a great step in advance. 28 Here for the first time 
the Halakic elements of the Eldad literature received as careful 
study as the Haggadic elements. Epstein published all the 
Halakic texts, then known, with an introduction and copious 
notes, and made a comparative study of the Halakot of Eldad 
with the Rabbinic, Karaitic, and Samaritan Halakot. His main 
theory was that Eldad's Ritual constituted a collection of the 
Halakot of the Falashas. This view has not, however, been 
accepted by the critics. 2 » It is to his credit none the less that 
he made the first thorough and scientific examination of the 
Halakic part of the Eldad literature. 

In the mean time, a most important addition was made to 
our extant texts for the study of Eldad. Among the valuable 
Genizah finds which now form part of the University Library 
of Cambridge, Doctor Schechter discovered two important ancient 
fragments, one of which he identified as part of Eldad's Ritual, 
and the other as part of the Responsum, mentioned above. These 
fragments have thrown a new light on the character of the original 
text of the Eldad literature. For the Halakah the discovery of 
the Genizah fragment meant the discovery of an older and more 
perfect text of the Ritual than any we have hitherto possessed. 
It has also made possible the reconstruction of the Ritual in a 

26 V. Dor IV (1881) p. Ill note 25, p. 112 note 27; cf. also his review 
of Epstein's Eldad in Ha-Hoker I (1891) pp. 158 et seq. There Weiss seems 
to adopt the view that the Ritual is an independent eclectic system of Halakot. 

" V. JQR. I pp. 105, 110. 28 v. note 11. 

29 J. Halevy in Revue Critique d'Histoire et de Litterature N. S. XXXI 
(1891) p. 463 opposes the Falasha theory. He points out that the ritual of 
the Falashas is very simple. He argues further that it could not have been 
written in Hebrew. The Falashas were brought to Abyssinia as captives taken 
in war by the Ethiopians. They did not understand Hebrew, and in their 
liturgy they used either the Septuagint or an Aramaic Targum. At a later 
time they adopted an Ethiopic-Christian translation of the Bible. 

much completer form, the preparation of a more exact text, and 
a more accurate determination of the interrelation of the other 
manuscript fragments of the Ritual. 

Both, on the basis of the Genizah fragments, which are 
here published for the first time, and a careful examination 
of all existing texts, the present work seeks to present, a 
text-critical study of the Eldad Halakah. I have attempted 
to reconstruct the Ritual of Eldad and to restore it, as nearly 
as possible to its original form (Chap. III). Furthermore, it 
has been my endeavor to show the relation of the text of the 
Ritual to the epitome of it preserved in the Responsum (Chap. 
II). A special chapter has been devoted to the peculiarities of 
Eldad's Hebrew, and a Glossary presented explaining the unusual 
words and idioms employed in the Ritual (Chap. IV). And 
finally the attempt has been made to indicate the place of the 
Ritual in the earlier Halakic literature (Chap. V). 

The utmost reverence has been shown for the readings of 
the manuscripts. The endeavor was to retain them in their 
original form, wherever it was possible. In editing a text so old 
and so peculiar from a linguistic point of view as that of the 
Ritual, we cannot expect accuracy and elegance of style, but 
must content ourselves with a merely intelligible text. Numerous 
emendations of Epstein have been gratefully adopted. These are 
indicated in the notes by the letter E. Many emendations of 
his, however, had to be disregarded, because they were deemed 
either inadequate or superfluous. Variant readings have been 
preserved in the notes on the text. 

In Recension A the more difficult Halakot are given in 
translation, because this was found to be the simplest and briefest 
way of indicating my conception of the meaning of the text. In 
the notes in Recension B, special attention has been given to 
the interpolations in the text, and detailed study has been made 
of the development of the original text into its present form. 

Incidental to my study of the Halakah, I publish in the 
Appendix (I) the recently discovered Genizah fragment of two 
non-Halakic sections of the Responsum. This fragment is of 
importance for the history of the Eldad legend. The comparative 
study here made of the Genizah fragment and the parallel texts 
throws light on the process of accretion and interpolation through 
which the present text of the Responsum was developed. At 


the same time it corroborates the results obtained from the con- 
sideration of the Halakic texts, that our text in the editio princ&ps 
is full of interpolations and alterations, and is by no means 
authentic throughout. Incidentally, the comparison of the various 
texts with one another brings out the place and importance of 
the recension of the Responsum preserved in the historical work 
Shalshelet ha- Kabbalah, a recension which has been hitherto 
neglected. A comparison of its text with that of the editio 
princeps also reveals a number of significant interpolations in 
the latter. 

The text of another and less important G-enizah fragment 
which originally formed the introduction to Eldad's Ritual in a 
Halakic code belonging probably to the Gaonic period, is 
printed here (Appendix II). The part preserved in this fragment, 
however, deals with the Eldad legend only. The texts of both 
Genizah fragments have been annotated, not only from the point 
of view of text criticism, but also from the point of view of the 
relation of the parallel texts. 

Attention is here called to the points in which the present 
edition of the Eldad Halakah differs from that of Epstein. 
To the texts published by Epstein in his work on Eldad the 
Genizah fragment has been added. Instead of parallel texts one 
continuous text is presented, the basis being the Genizah frag- 
ment where it is available, and the oldest extant manuscript 
where the Genizah fragment fails us. The author is glad to 
acknowledge his deep appreciation of Epstein's studies and to 
express his indebtedness to them. 

In closing this chapter, I wish to sum up here briefly the results 
of my study of the Eldad literature. Investigation reveals that 
Eldad himself wrote Halakah only. The study of this material 
points to the conclusion that the Halakot contained therein are, in 
substance, the dietary ritual observed by the Jews of a particular 
country. Though this country cannot at present be definitely 
located, it can be shown, nevertheless, that it was within the sphere 
of influence of the Palestinian Talmud and of the Arabic language. 
The form, however, in which we have the Ritual can, properly, be 
attributed to Eldad. The task of the future student of the 
Eldad Halakah will be to determine from the evidences of 
language, content, and relation to other Halakic codes of the 
Gaonic period, the country in which this dietary ritual had its origin. 

With regard to the personality of Eldad, the conclusion is 
that the only authentic source of information is the account of 
him and of his reports preserved in the Responsum. Here again 
the recently recovered text of the Genizah fragment proves 
clearly that the narrative in the editio princeps and in current 
texts has suffered much from later interpolations and alterations. 
The narratives found in the Sefer Eldad, which exists in various 
recensions, and is ascribed to him, cannot be trusted. Eldad is 
the hero, not the author, of the Sefer Eldad. The legends con- 
tained in this work should be traced to their sources and studied 
in their relation to the older Midrashim, before the Sefer Eldad 
can be used as a source of information on Eldad. At present, 
we must rely exclusively on the Responsum for our knowledge 
of this anomalous figure in Jewish history. Moreover, in view 
of the character of the text preserved, we must take no little 
caution in the use of even this Responsum. 



The Ritual on Slaughtering of Eldad ha-Dani exists in two 
recensions. One of these, which we shall designate by A, is 
either the original, or a version very close to the original. The 
other, which we shall designate by B, is merely an epitome of 
the Ritual, prepared in Kairwan and sent to the Gaon of Sura, 
Zeinah [ben Hayyim], for his opinion of it. 


That recension A was known to the older Rabbinic author- 
ities, is proved by citations from Rabbinic works which Epstein 
and others before him traced out (v. Epstein's Eldad, p. 137). 
It was known to the older Rabbinic authors as ytin.T 1DN mbn, 
a title derived from the opening words p p yt2nJT 1IDK. An 
erroneous interpretation of v '«, the abbreviation of this title, 
gave rise to the designation ^fcW p« nlD^n, by which name it is 


referred to in the works of later glossators and commentators. 3 o 
We do not possess it in any single complete manuscript. We 
are obliged to have recourse to a number of manuscript frag- 
ments, in order to obtain a clear idea of its structure and 
content. The manuscript material on the basis of which we re- 
construct it, is as follows: 

I. G. Citations from Eldad's Ritual in ^ JNJTD^K rfttfDI 
)Wt6n n^Sn ("The clear treatise on the ritual slaughtering 
of animals"), MS. Bodl. Neub. no. 793 », the author of which, 
according to a conjecture of Halberstam's, 3 * subsequently con- 
firmed by another manuscript, 32 was Samuel ben Jacob ibn 
Gama 1 (JJD}) of MNp in Northern Africa (twelfth century), author 
of a supplement to the 'Arab.™ This ri^NDI, which is very 
valuable in the history of literature, has been described in detail 
by Steinschneider. 34 One of Ibn Grama c 's chief sources was 
Hananeel ben Hushiel of The citations from Eldad, 
Ibn Gama c probably took either directly from a copy of the Eldad 
Ritual made by Hananeel 36 or from quotations in Hananeel's 
works, perhaps from the latter's commentary on Hallin.* 1 Ibn 
Gama c assures us in one place in his ri^MDI (MS. p. 100 b) that 
he is quoting the* manuscript of Hananeel literally. 38 Hananeel 
lived in Kairwan, the scene of Eldad's activity, one hundred years 
earlier. It is, therefore, probable that this scholar had utilized 

30 Azulai was the first, as far as I know, to maintain that the ms^n 
b»*\w pK was the work of Eldad, v. Sliem ha-Gedolim s. v. pN nra^n 
hx-wr. Graetz was the first to suggest that the title bUTM* pK rvobn was due 
to the misreading of the abreviation •>"« rwsbfl in yenrr n&N m^n, (v. Ge- 
schichte V (1860) p. 452). This was proved conclusivly by Reifmann, who does 
not seem to have known of Graetz's suggestion, (v. Ha-Karmel I. c. p. 280). 

31 V. Ha-Karmel III (1863) p. 215. 

32 V. Neubauer, Catalogue, Additions to no. 793 1 on p. 1154, and idem, 
The Literature of the Jews in Yemen in JQR. Ill p. 619. 

33 For this author v. Steinschneider, Die Arabische Literatur § 105 and 
Buber, in Graetz' Jubelschrift (Hebrew part) p. 6, where the existing portions 
of Ibn Gama £ 's Supplement are published; for the name cf. Steinschneider in 
JQR. X 514. 

34 V. his article Schlachtregeln in arabischer Sprache in Geiger's JZ. I— IV ; 
cf. also JQR. XIII pp. 457 et seq. 

3 * Cf. Steinschneider in JZ. I p. 314. 

36 This is Steinschneider's view I. c. pp. 240, 311. 

37 V. on this commentary Berliner, Migdal Chananel p. XVII. 

3 8 V. A. note 30. 


Eldad's .Ritual in its original form. This assumption is in a 
measure confirmed by the fact, that the few quotations from Eldad 
occurring in Ibn Gama c which may be examined by a comparison 
with parallel texts, can be proved to be considerably older than 
P (see below), and cannot be much later, if at all later, than 
the Grenizah fragment T-S (see below p. 12). 39 

The quotations from Eldad in Ibn Gama c 's work were 
marked in the manuscript by Steinschneider in 1851. 40 Goldberg 
in 1857 published these quotations, as indicated by Steinschneider, 
in an inaccurate and incomplete form, however, in the '1 ri?«D1 
Bmp p miiT, edited by Goldberg and Barges, p. xix et seq. 
In the same year Filipowski also printed them in no more 
accurate form in his Liber Juchassin, p. 207 et seq. i{ A better 
edition is to be found in Epstein's Eldad, pp. 99—104. This 
also is not entirely free from errors, as it is based on an inaccu- 
rate copy made by Neubauer. 42 In preparing this edition I have 
used a photographic copy of the respective parts of the manuscript 
itself. Ibn Gama 's quotations embrace 

1. Part I of the Ritual (HtoW niD^H, §§ 1—6), in its entirety. 

2. Part II (mints ITD^n), several regulations concerning the 
lungs (§ 9d, e, f, h) and the kidneys (§ 20). 

3. Part IV (.Tn iTD^n), a regulation governing the slaughter- 
ing of iTH and *)iy (§ 36). 

II. P. The longest continuous fragment of the Ritual known; 
MS. Parma, de Rossi, cod. 327, 22. Steinschneider 43 was the 
first to point out the relation of this fragment to the Ritual of 
Eldad. This codex was completed in 1290 C. E., by Samuel ben 
Joseph ben Samuel ben Joseph in Uncastillo (bwp ]1«), a place 
in Aragon. 44 This fragment contains 

1. Part II (JYIDnB IIDta), complete (§§ 7—25). 

39 A § 20 which is intentionally reproduced in abridged form by Ibn 
Gama c , cannot be regarded as disproving the view in the text; v. ib. note 203. 

40 V. JZ. I. c. p. 310. 

4i V. Steinschneider's Corrections in JZ. II p. 297—301. 

« V. Epstein's Eldad p. L. 43 L. c. I p. 311. 

44 V. Catal. de Rossi I p. 181, Epstein p. Ill note. For a fuller des- 
cription of this MS. v. Horowitz, Bibliotheca Haggadica, I (1881) pp. 16 et 
seq. The sufferings to which the scribe alludes (li«» nW7\ p \frv V&rro b*7\l 

D^vw) may be those due to the oppressive legislation of Alfonso X 

of Castile (1252-1284). or, more specifically, to the imprisonment of many 


2. Part III (D^UB), almost complete (§§ 26— 33e).« 
The whole of this text has been published by Epstein 
(pp. Ill — 121). I reprint it with emendations and correction. 
Unfortunately I did not have access to the manuscript. When- 
ever the manuscript is referred to in the notes, the reference is 
to the reading of the MS., as given by Epstein in contradistinction 
to the emendations which he proposes. 

III. T-S. A Genizah fragment T-S. Loan 110, University 
Library of Cambridge, 2 leaves of vellum, 20 x 18 cm., 29 lines 
on each page, in old Oriental square script, belonging at the 
latest to the eleventh century. 46 The handwriting is quite legible, 
*7 and 1, however, cannot be distinguished. The leaves are much 
damaged, and for this reason the text contains many lacunae. 
The fragment contains 

1. Part II (nisntD JTD^T), about nine paragraphs (middle of 
§ 17— § 24; middle of § 33b— § 33e) and 

2. Part IV (,TH m^H), almost complete (§ 34— § 38). The two 
leaves that belong in between (== § 24 end— § 33 b middle) are 
unfortunately missing. 

I have used this Genizah fragment as the text of my edition, 
wherever it was available, for the reason that this text is as yet 
unpublished, and also, for the more important reason, that it is 
older than P and 0, and probably older than even G. The variant 
readings of the parallel texts have been given a place in the 
notes. I have indicated the end of the line in the Genizah 
manuscript by a vertical line | . Wherever the text of the Ge- 
nizah fragment is defective, and readings from parallel texts are 
adopted, it is indicated by brackets. Where the inserted word 
or phrase rests on mere conjecture, it is indicated by a star. 

IV. 0. An extended quotation from Eldad in Glosses on 

Jews in Castile and Leon, and the extortion of an enormous sum for ransom, 
v. Graetz, Geschichte VII pp. 40 et seq. 

For the identification of bvwp \\H with Uncastillo, for which I am in- 
debted to Dr. A. Freimann, v. Zunz, Zeitschrift p. 134, Jacobs, An Inquiry 
into the Sources of the History of the Jeivs in Spain, Index locorum s. v. I can- 
not share Horowitz's view (I. c. p. 1) that ^nffp pan is Castel Buono in Sicily. 

« The heading 'ttl p p yvwb np"12\ n&W nw^n indicates that part I 
(ntrrw JTD^n) was at one time part of P. 

4 <> The handwriting is of the same character as that in the autograph 
letter of Hushiel ben Elhanan (end of the tenth century), published by 
Schechter in JQB. XI p. 643. 


the Mordecai by a German Rabbi, MS. Oxford, Neub. no. 678— 
"a veritable mine for tbe Halakic literature of the middle ages".*? 
Attention was first called to this quotation by Dukes, 48 and later by 
Steinschneider.49 A part of this quotation (§§ 32 and 33) was first 
printed by P. F. Frankly and all of it by Epstein (pp. 132—134). 
Both used a copy made by Jellinek. I have used a copy of the 
quotation made by Edelmann 51 which is in the Library of the 
Jewish Theological Seminary of America. From this copy it is 
not possible to make any important corrections in the text given 
by Epstein. This quotation from Eldad contains: 

1. Part III (D^US), beginning with § 32c and 

2. Part 1Y (iTH JYDto), § 34 and the beginning of § 35. It 
is evident that the original, from which was copied, did not extend 
beyond this point. The scribe therefore supplied the missing con- 
clusion of the Ritual (Epstein p. 134 no. 5—7) from the Responsum 
(Recension B). Moreover, there is even some indication in §§ 34 
and 35 that the text which he followed was incomplete, and he 
was therefore obliged to take one passage in § 34 (v. A, note 
528) verbatim from B. 

V. M. A brief quotation in Mordecai, in the beginning of 
Hullin, cited from TH1 im At first sight, this quotation would 
seem to have been taken from the Sefer ha-Terumah of Baruch 
ben Isaac of Regensburg, but it does not occur there. Presum- 
ably it comes from the no longer extant Sefer ha-Hokmah of 
Baruch ben Samuel of Mayence (d. 1221)." ffhe quotation, in 
its present form, contains only the first half of § 1 of part I 
(HttW mr^n), and proceeds at once to part III, where, after 
the introductory sentence of § 26, it reproduces the D^Ufi enume- 
rated in § 31, but in a rather condensed form. In its original 
form in Mordecai the quotation from Baruch must have been 
very much more extensive, for it appears from Shilte ha-Gibborim 
at the beginning of Hullin that Mordecai had quoted, in the 

47 V. Neubauer in MGWJ. 1871 p. 174 note 2. 

43 T 7 . Literaturblatt des Orients 1850 p. 768. His reference is rather in- 

49 V. J. Z. I p. 310 note 13. so J. MGWJ. 1874 pp. 548 et seq. 

5i V. Edelmann's edition of Caphtor iva-Pherach p. XXIX. 

52 For this author v. S. Kohn, Mardochai ben Hittel in MGWJ. 1877 
p. 565, 1878 p. 42; reprint pp. 102, 108. 


name of "]nn '1, "the entire Ritual of Eldad". $ 3 This statement 
must, however, in all probability be taken to refer to parts I and 
II only, which precede part III, dealing with D^UD (v. A, note 2). 

The quotation from M is too brief to be given a place in 
the text. I have, however, inserted it in the notes (v. A, notes 2, 
289, 303, 332) correcting the text on the basis of eel. Alfasi, 
Constantinople 1509, and a parchment MS. of the fourteenth century 
in the Library of the Jewish Theological Seminary of America. 

The four texts, G, P, T-S, and 0, supplement one another 
in a way that makes the reconstruction of the entire Bitual of 
Eldad, with the exception of the conclusion, possible. Moreover, 
for §§ 9d,e,f; § 17 middle— § 24 middle; § 33b middle— § 35 and 
for § 36 we have two parallel texts, and for § 20, § 33 b, middle — e, 
even three texts. In the present edition the text which serves 
as the basis is designated by the initial letter in the left margin; 
where parallel texts exist these are indicated by the other letters. 
The beginning and the end of a text are indicated by two parallel 
lines || . 

Of these texts the nearest the original is, in all probability, the 
Genizah fragment. 54 That the Genizah fragment has a more original 
text than G, can be established with some degree of certainty, 
although only one section (§ 36) is to be found in both texts. In G, 
Eldad's terms are replaced by the more usual later terms. Thus in 
place of NfiD and ")EO — the latter is in itself a later substitute for 
Tints (v. notes 551 and 552) — we have *V)DK and VT1D, instead of *6n 
D^m ttT miPp the simpler phrase nnniD (v. note 562), instead 
of the curious phrase Kin ^ly "O, the simpler reading *)iy bl) (v. 
note 564). Likewise we have in G in § 9e the Rabbinic-Tal- 
mudic form fcOTD, where Eldad must have used *pD (v. note 74). It 
seems, however, that both T-S and G go back to one earlier text, 
for § 36 in both texts contains the characteristic mistake tib DN 
"Wpn, where the sense requires the reading Tlfc?pn DN (v. note 555). 

That T-S is more original than P can be readily and con- 
clusively proved, for here we have parallel texts for considerable 

53 nrorc bmttn p« nvbn nm hi t/ftnb n^nnn -p*'" 1 -jnn mi awn w»m 
mn v6n. 

s< Among the peculiarities of T-S is the affixing of pronominal plural 
endings to singular forms of nouns and verbs. So § 17 (note 125) Ntfn msnim 
n^sji nx, Prww n« ; § 23 (note 257) .Toma^, Pianist; § 24 (note 263) mtsn n«; 
§35 (note 558) rrtftn, G nv»\ Note also the frequent use of the matres lectionis. 


portions of the Ritual. In place of the constantly recurring 
formula in the fragment "?D*T fc6l Kin nsifi or tofcOl «ffi IBtt, P 
reads in most cases simply nSID or "IBO. The prolix enumeration 
of domestic animals in § 20 of the Genizah fragment Kin "W DN 

nmoD me ^k ^ ma dki Minn Dvn &ntw itr« miaa Tit? ^>k ^ 

iniDD ^K ^« ^K D«1 initM D^?y HJ b» D^J> *t* DM TOM ^J> b« ^ DK1 

nmea firm to riWn dki nmw nry ^« nrp dki imaa w*n b& ts^ri dki 

is abridged, by substituting the phrase D^ID JYIDn^n to pi after few 
nni03 ms (v. note 183). A similar abridgment is adopted in an 
analogous case in § 23 (v. note 238 a). Biblical forms of the 
Genizah fragment frequently give way in P to Rabbinic-Talmudic 
forms; thus *Al is used instead of ^ (v. T-S § 23 note 249), 
wVd instead of n^ta (T-£ § 20 note 176), the Mishnaic "?3K instead 
of Vk (T-£ id. note 180), K,Tt? instead of nTW (T-£ § 24 note 
265), the participial ending \* instead of D\ The characteristic 
vulgar Arabic construction blW vby TIDK (v. Glossary s. v. D^inn) 
is abandoned for the simpler infinitive construction b)2tih TlDK 
(T-S § 33 c note 465). Similarly 1BO replaces the term WB, 
which Eldad used (c/. T-£ § 17 note 127, § 18 note 137, § 19 
note 173, § 20 note 202, § 21 note 210, § 22 note 222). Instead 
of mm (T-S § 35) the more usual nniDIK occurs in P (§ 17). 
In addition, the numerous corruptions of the text, omissions, 55 
and misplacing of passages which a comparison with T-S reveals, 
show clearly that the text of P is considerably later than the 
more perfect text of the Genizah fragment. No direct relation 
between the two can, however, be established. Here again certain 
errors 56 and peculiarities 57 , common to both, would seem to indicate 
that the two texts are to be traced back to an earlier text. 
O shows in turn a much later form of the text than P. In it the 
original style is still further obliterated. Thus "Dll (P § 33 a 

55 § 9h and § 10 beg. are entirely wanting in P, (v. notes 104, 104a). 

56 Thus K 1 ? instead of \b in § 33d (note 480); the negative particle vb 
before inn in § 22 (note 225 a). 

5 7 Thus Kin in § 18 (note 143), which is found in both texts; likewise 
the spelling of the second person perfect singular with n in nnrtns and nnnni 
in § 22 (note 215; for this spelling in the Bible see Gesenius, Hebr. Grammatik^, 
§ 44 g, in the Yerushalmi, Schlesinger, Das Aramaische Verbum im jerusalemi- 
schen Talmud, p. 13, in the Tosefta, Zuckermandel, Die Erfurter Handschrift 
der Tosefta, p. 9, in the Midrashim, Theodor in MB. W.J. 1893, p. 208 note 5); 
the omission of the & of "p&0* in § 23 (note 272) ; the omission of the article 
of D"J>& at the end of § 22, where both texts read in the beginning D"y»n. 


note 414) becomes IttfcO; TO1 (P § 33b note 436) becomes n»K; 

nawn n«r nn»^ (P#>. note 432) itsnw iro6; ot by to Vnj «»n 

(P &. note 439) 1WJ> Kton; ]«2Ja Kin TOK nynn (T-tf § 33 e note 503) 
]«sn ^« Win TOK njmn etc. inserts the copula, wherever the 
general usage of the neo-Hebrew requires it (v. § 32 c note 364 
DmntDl; § 33a note 401 p«1; ib. note 417 XV) \ § 34 note 521 I^SKI; 
#. note 523 n&1 etc.). Moreover tends to improve upon the text 
in several instances, thus substituting XW1 for HD in § 32 c (note 
358) and omitting the superfluous word nnpi in JTDiyjD nisro 
^ rttWl napj (§ 33 a note 415). also takes other liberties 
with the original text, abridging it in places for the sake 
of simplicity, thus e. g. instead of *lt5D Tfipm ]tt am Oitsn^ 

mnio«i D^n lian^i ttrD&y nn&ty nn» nnn& it reads o^n utsnt? 

nHDtyn (P § 33 b note 440); instead of n» 1*6 ,)xb niSfc ty "D^ 
in« D^n nanism *6 «m it reads simply xbl )xb ms» ty in^Bf 
l&rwn (P § 32 d note 476). Similarly in § 33 fewer examples 
are given than in the original text (v. note 431) and again the 
collective term niaiyni is used instead of the rather extended list 
of fowl mentioned in the original text (P § 32 c note 345). In 
the same spirit the unimportant sentence ny nn&TJt? n^ nJTJTI 

)bxn D^yn l^nii n*n d,tjp hot om« mW njypa ^m (P § 32 c 

note 339) is omitted, and a Halakic discussion occupying four 
lines that is not essential to the thought is left out (§ 33 b note 
444). Furthermore, the substitution of the Talmudic phrase fiTTl 
D^ xb) m (§ 32 c note 343) for Eldad's plainer words is evidence 
that this text is to be assigned to a later date. 

There are indications that may be traced to P or a group 
of manuscripts going back to P. These indications are to be 
found in certain errors, such as |HD *C QK1 (T-S § 33 d note 
467) and omissions such as «i (ib. note 469) and blX* xb) (ib. 
note 472) common to both texts, also in the fact that both alike 
use certain forms which differ from those employed in T-S as 
rpte (T-S D"pte ib. note 473 a); UX) (T-S DK ib. note 479); monn 
nam TOa (T-£ ni»nn TD IYl»rD t&. note 489). This does not 
preclude the possibility that was influenced by a manuscript 
closely related to T-S. Evidence of this is to be found in certain 
forms common to both and T-S in the use of which both 
differ alike from P. Compare e. g. mm DKH of T-S and 
(§ 33 d note 468) with MiM DKH of P, and D'W'Oyo or )^y» of 
T-#, (t&. note 483) with ]my& of P. 


During the stay of Eld ad in Kairwan an epitome was made 
of his Ritual, in order that it might be submitted to the Gaon 
for his opinion on points in which Eldad's Ritual differed from 
their own. The oldest source for this epitome thus far known 
is the epistle of the community of Kairwan to the Gaon Zemah 
[ben Hayyim]. This epistle in its present form is, however, not 
older than the second half of the fifteenth century. It was first 
published together with the decision of the Gaon by Abraham 
Conat [Mantua, 1480]. 58 Our text is based on a photographic 
reproduction of the copy of this editio princeps in the Bodleian 
Library in Oxford. 

As this Responsum was the only continuous text known 
before the discovery of manuscript P (see above p. 11), it 
naturally formed the basis of all opinions of Eldad's ritual. 
Frankl, never, suspecting the genuineness of the text in its pre- 
sent form, was troubled by some of its discrepancies. In his 
prejudice against Eldad, he attributed these inconsistencies 
to the character of the author. To cite an instance: in no. 10 
several sections have been lost from the text (v. B, note 28). As 
a result a question is followed by an absolutely incongruous 
answer. Frankl's comment is that Eldad "has here given a 
decision quite unequalled in folly". 59 

A much more exhaustive examination of this Responsum was 
undertaken by Jacob Reifmann. He was the first to subject 
the Halakot of the Responsum to a critical and perhaps even 
hypercritical analysis. He found no less than nineteen passages 
taken, as he thought, from Talmud Babli, Hidlin, some of them 

5 8 Small octavo, 14 pages, n. d., n. p.; v. Steinschneider, Geschichtslite- 
ratur pp. 16 et seq. Attention was first called to the existence and importance 
of this text by Luzzatto in Literaturblatt des Orients 1846 p. 481. This editio 
princeps is to be found in only three libraries (v. Jewish Encyclopedia, s. v. 
Incunabula VI p. 597 no. 7). It was reprinted several times, v. Steinschneider 
I. c. Of the more recent editions the one by Jellinek in his BH. II p. 108 is 
reprinted from a poor text appearing in Zolkiev 1772, octavo. Epstein pp. 83 
et seq. gives the editio princeps, using a copy made by Moses Gerundi. An- 
other reprint, absolutely uncritical, is to be found in Isidor Zinger's Eldod 
ha-Doni, Podgorze 1900 pp. 5 et seq., v. Steinschneider I. c. p. 174. 

59 V. MGWJ. 1873 p. 489. For a similar example of Frankl's criticism, 
due to his uncritical use of the text, v. infra p. 100. 



verbatim. 60 In the same way he thought he had found passages 
paralleled in the older Rabbinic literature. His general con- 
clusion, therefore, was that the author, or compiler (TTDfc), of 
this ritual was an ancient impostor (pip "OB^tls 1H8) who had 
gathered his material from our Talmudic and Gaonic writings 
and other flalakic works no longer extant. 61 These he used 
either verbatim or in a slightly modified form. For some of the 
Halakot he claimed the authority of Moses or Joshua. He 
ascribed the whole collection to Eldad ha-Dani, and styled it 
"Halakah of the Ten Tribes". The citations from the V 'K HID^H 
which we find in the older rabbinic authors, Reifmann thought, 
were not quotations from Eldad's ritual, but rather sources from 
which this impostor drew. The latter, Reifmann implies, must 
have lived at least before Jacob ben Asher, the author of the 
Turim, who died in 1340. He bases this view on the fact that 
in four instances there is complete agreement —verbatim et lite- 
ratim — between the Responsum and the Tur and he therefore 
concludes that Jacob ben Asher made use of the Responsum. 62 
This interpretation of Reifmann, as we shall soon see, is erron- 
eous. It is, however, to his credit that he overthrew the view 
which had enjoyed universal acceptance before him, namely 
"whatever is ascribed to Eldad is authentic, i. e. was actually 
written by him". 63 Reifmann pronounced the entire Responsum 
spurious. In this he went too far, for, as we shall presently 
establish, the passages which led him to regard the Responsum 
as a forgery, are interpolations. 

Reifmann's investigation was followed by Epstein's study of 
the Responsum. Epstein scarcely took any notice of Reifmann's 
results. Confronted by the patent contradictions in the Responsum 
itself, such as the existence of two totally different decisions on 
the same Halakah, and the glaring discrepancies between the 
Responsum and Recension A, he sought to bring about harmony 
by means of slight textual emendations or forced explanations. 
He attempted to maintain the genuineness of even the most 
palpable Talmudic interpolations in the Responsum — some of 
which Steinschneider had already pointed out — and resorted to 
the hypothesis that Eldad had become acquainted with the 

so V. Ea-Karmel, vol. "VIII pp. 264, 269 et seq. 61 L. c. p. 280. 

62 V. Recension B., nos. 9», 9 2, 10, 16 || Tur, Yoreh Deah §§ 35, 53, 55; 

cf. notes 26, 27, 41 ; Reifmann I. c, p. 286. es y Frankl in MG WJ. 1874 p. 552. 


Talmud in Kairwan and, therefore, incorporated several Talmudic 
Halakot in his Ritual. 

Epstein thus left the critical study of the Responsum where it 
had stood before Reifmann's essay appeared. Neither did I. H. Weiss 
make any advance. He did, indeed, point out an additional 
discrepancy between the Responsum and Recension A, but this was. 
to him, only further evidence that Eldad was a mere impostor. 64 

There is, therefore, no need of an apology for undertaking 
anew a critical examination of the text. To sum up the results 
of our investigation: 

1. The form of the Responsum has undergone extensive 
alteration under Talmudic and Rabbinic influence. Frequently 
the less familiar words, phrases, and constructions of the original 
have been replaced by the more usual terminology of Talmudic- 
Rabbinic literature. Thus there is substituted in the Responsum 
DID^p for By (v. B, note 15); TD for 11JPBO IV (v. B, note 5); 
mru for nip (A, § 10 note 105; v. B, note 21); Mb IDKfcy tol 

iniM Mb ynn rrapn (v. B, notes 52, 53) for iniK onnn tok toi 
•ina Dipan nnb wik wi b$w*b hnpri (A, § 32 c) etc. 

2. The original Halakot of Eldad have been changed com- 
pletely in a number of instances, so as to make them agree with 
the Talmudic or Rabbinic Halakah. Thus, e. g., where the original 

text read Kin TintD Viyn i?inan 2pM (A, § 21), the Responsum 

now reads nsifc rajD ^IHtDn 2pM, the very opposite, the change 
having been made to bring about conformity with Hullin, 55 b, 
and the current Halakah (v. B, note 33). Similarly the D«1 

npnn vty pa tibhjd ik *?yto mtsp nnsi n^ni nn« n*ta Ksnr 

^DfeW Kin TintD of A (§ 20) is changed in the Reponsum to the 

nsits ,noin nap nyi np*n idd ny nrapnty k^d pi of Hullin, 

I. c, and the Halakic codes (v. B, note 36). For ISD&n 1K&T "3 
K1H 1BO nwn bxWZ niDD pD^D )^ ^ in A (§ 16), which deals 
with the decision concerning fractured ribs, we find substituted DK 

nsn& pro mbran n^m^s rontsw n»nn n*re», in accordance 

with Hullin, 52 a, and our Halakah (y. B, note 38); for the Ha- 
lakah of Eldad concerning an ox that has fallen into a pit (A, 
§ 24), the corresponding Halakah, taken from the Tur, Yoreh 
De'ah, § 58 has been substituted; compare also B, no. 21b, with 
A § 32, and B, note 51. 

e* V. Ha-Hoker I p. 163. 


3. The Besponsum has been enlarged through numerous later 
additions, some of which are altogether inconsistent with the preceding 
Halakot of Eldad, and even with the corresponding HalaJcot in A. 
Of these interpolations some were suggested by the Ritual itself 
(comp. rPDloVp in no. 1, v. note 3; HDHn IK in no. 24, v. note 80). 
Others were taken from the Talmud (cf. )bbft mb no. 8, v. note 22; 
IB? Dl? "W D^D IS no. 22, v. note 58) or from some early Ha- 
lakic code (cf mtSO 1TD fi^on OKI no. 2, v, note 6). Thus, at 
least in six instances passages are taken verbatim from Tur, Torek 
De'ah. It is, of course, evident that these interpolations were 
taken from the Tur, and not, as Reifmann thought, that the Tur 
took these passages from the Responsum. 

Interpolations from the Tur in the Responsum are most 
frequent in two places. At the close of no. 9 there occur in 
succession three interpolations taken literally from Tur, Tor eh De'ah, 
§§ 53, 55 and 56 (v. note 26). Again, at the close of no. 17 
there are likewise three successive interpolations from the Tur, 
§§ 59, 58 and 60 (v. notes 45 — 47). The interpolations following no. 17 
may be very readily explained. A reader or a copyist who 
missed the Halakot on flaying (rniVjn, no. 18* — Tur § 59) and 
on apoplexy etc. (nttis&ni jwiy»m mn nnn«, no. 19* = Tur 
§ 60) supplied the deficiency from the Tur (§§ 59, 60). The 
third passage (no. 19) represents a substitution of the Halakah 
of the Tur (§ 58) for the Halakah of Eldad on the case of an 
ox that has fallen into a pit (== A, § 24). The adoption of 
§§ 59 and 60 from the Tur suggested also the adoption of the 
Halakah of the preceding paragraph of the Tur (§ 58) in place 
of Eldad's Halakah. 

The position of the interpolations after no. 9 (9*, l, 2, 3) 
is inexplicable to me. They very plainly interrupt the continuity, 
for the discussion on the trachea and the oesophagus (no. 9) is 
properly followed by the discussion on the lungs (no. 10). This 
sequence is also indicated in no. 8 ropi tDfcSHn Blip 3pi ♦ ♦ . 
'pi n^n ropi nmn ^:) p m:min. It is remotely possible that 
in the original text of the Responsum there occurred a passage 
corresponding to the DSJfn "DBtt or the D*y P*nn inn^i of A, 
§ 8b, d. The word "DntJO suggested another Halakah beginning 
with rnntW (no. 9*, l) and led to its introduction into the text 
at this point. 

There can be no doubt, however, that these passages are 


later interpolations, taken verbatim from the Tur. The first and 
third (9*, 1, 3) absolutely contradict nos. 16 and 17 in the Res- 
ponsum, passages the authenticity of which is vouched for by 
Recension A. The second passage (no. 9*, 2) does not occur in 
Recension A, and it is incomplete. The conclusion of the para- 
graph in the Tur (§ 55) from which this interpolation is taken, 
occurs as an interpolation between nos. 16 and 17 (v. B, note 41). 

Other disturbing interpolations from the Tur are found in 
no. 10 = Tur § 35 (v. note 27), in no. 12 = Tur § 36 (v. note 
30) and in no. 14 = Tur § 44 (v. note 36). These also may be 
explained as the insertions of a reader or copyist. The absence 
of a Halakah on the NTH in the description of the lung in no. 10, 
suggested the interpolation of the appropriate passage from the 
Tur (§ 35). The meagerness of the Responsum at this point, 
even as compared with A, may have afforded special motive 
for this insertion (v. B, note 27). The interpolations in nos. 12 
and 14 are similarly due to the absence of the usual Rabbinic 
Halakot on these subjects in the Responsum. The passages are 
likewise taken from the Tur §§ 36 and 44 respectively (v. notes 
30 and 36). 

It is thus obvious that the text of the Responsum has been 
worked over repeatedly before it assumed its present form. The 
steps in the process can not well be traced now. Stripped 
of its most patent Rabbinic and Talmudic interpolations, and 
of alterations in style and substance, what remains is an approxi- 
mation to the original epitome of Eldad's Ritual. As such, 
aside from its intrinsic value, it is highly important as a means 
of controlling Recension A. 

In our text of the Responsum passages or single words which 
have undergone alterations either in style or in substance are 
printed in smaller type. Interpolations, taken from the Talmud, 
the Tur or some other code, are printed in Rashi script. 

Comparing the Responsum with Recension A, it will be seen 
that in the Responsum sections are missing to correspond with 
§§ 7, 8a, 9a,c,d,e,f,g,h, 12, 14, 18, 22, 25, 26 to 31, 33 in A. 
(On § 13 and § 19 v. B, notes 20 and 29, and on § 27 v. B, 
note 80.) 

An examination of the sections omitted shows that in 
general these passages are the non-Halakic parts, transitional and 
introductory remarks, and other more or less unessential state- 


ments that would naturally be omitted in an epitome. The Res- 
ponsum is practical. It seeks primarily to give a succinct state- 
ment of the Halakot that differ from the generally accepted 
Ritual. This accounts for the omission of passages like § 7, 
which is merely introductory to part II, and of § 25 which forms 
a transition to part III and does not add anything to the infor- 
mation contained in § 24 = B, no. 19. In no. 21c, which cor- 
responds to A, § 32 c, the stories of miracles and the purely 
academic discussions are likewise discarded and only the Halakah 
is given. The frequency of purely academic discussions in part 
Hl64a maY> perhaps, account for the omission of a great deal of 
this section of the Ritual. It is not clear, however, why only 
§ 32 has been retained of the entire part III, and the long, 
purely Halakic § 33 is missing. Especially strange is the absence 
of regulations so characteristic of this Ritual and so strongly 
emphasized by Rabbinic authorities as those contained in §§ 26 
and 31.65 

Other omissions can be explained as due to the fact that 
the text has been mutilated and parts of it have been lost. In 
no. 10 the incongruity of the answer to the question, referred 
to above is a clear instance of this (v. B, note 28). In the same 
way the absence of Halakot, such as would correspond to §§ 8 a, 
12, 14, 18 in A, may be accounted for. It is not unreasonable 
to assume that what now remains of the Responsum, after the 
interpolations are removed, is less comprehensive than it was in 
its original form. 

One other point to be noted, in comparing the Responsum 
and Recension A, is the condensation of the material in the 
Responsum. In one instance, three sections of Recension A (§§ 
35, 36 and 38) are condensed into one section of the Responsum 
(no. 23), the topics being merely indicated. In another instance 

6 < a Cf. e. g. the argument that the ntrrw is a nfrwi (§ 26), and the dis- 
tinction drawn between nnt&n nfcrm? and nb^m nrrrw. 

65 This section treats of cases in which the act of slaughtering is per- 
formed by a person, who is naked, unclean, blind or leprous, or by a 
mourner, a woman, a youth (i. e. one who is not yet 18 years of age), or 
an intoxicated person. Further there are here cases dealt with in which 
the act is performed by one who does not face in the proper direction or 
in which the act is performed at night *. e. in the dark. 


two rather distinct sections of A (§ 34, «TH WD and [§ 40, WD 
*py]) are combined into one section of the Responsum (no. 25, 

*)ijn rw wd). 

While the comparison of B with A reveals many gaps in 
the Responsum B, B, on the other hand, supplies two Halakot which 
are missing in A. One of these is found in no. 13 of the Res- 
ponsum treating of a case in which the spleen and the ribs, or 
the spleen and the reticulum, have grown together (v. B, note 33). 
The other occurs in no. 17 of the Responsum, and treats of a 
case in which the tPTOn y)2p is missing (v. B, note 43). Another 
Halakah missing in A is perhaps contained in no. 9. This deals 
with the case of a tumor that has formed between the oesophagus 
and the trachea (v. ib. note 25). 

Obviously it is not always easy to determine, with accuracy, 
which sections in B correspond to given sections in A. In B 
I have given in parenthesis the number of the equivalent section 
in A. 

We must next consider the relation of the concluding section 
of (v. supra p. 12) containing the closing chapters of the 
Responsum with the corresponding part of the text in the editio 
princeps. The parallel arrangement of the topics dealt with in 
A, the editio princeps of the Responsum, and in in the following 
table makes evident that the editio princeps has preserved the 
original order much better than 0. 

iv. ppM] m mhn. 


T-8. 0. § 34 TVT\ WD 

„ „ §35 ray mat* 
„ - „ mm 

mo to 
w _ § 36 n»n:in 


„ — § 37 DTI ■WD 
„ — § 38 mfilto IMP 

[-njroi] rrnn 

Editio Princeps oj 

the Besponsum 
no. 22 nm ^D^D 

no. 23 msits n»in 

= A§§ 35 + 36 + 38. 


rrrei msita 

= A §§ 35 + 38. 


A B 

[T-S.—§ 39 rb ffWn] ■ no. 24 ML? fA Bhfr ntl *pj> v^D 

DHPK1 "OP ! D^«1 

[ „ — § 40 *py WD] no. 25 ^jiyi ,TH WD mn WD 

So far as the sources admit of comparison, it is evident 
that the editio princeps follows the arrangement of the material 
in Recension A. Both A and the editio princess preserve the 
same order of arrangement, §§ 37, 38 corresponding respectively 
to nos. 22 and 23. Moreover no. 23 of the editio princeps 
enumerates the topics in the same order in which they are dis- 
cussed in Recension A, DSjm rrraBOl = A § 35, noiJQl = A § 36, 

upon in^o rrnn Kin p norma «in "it?*o mnon bm = A § 38. 

This is further proof that the editio princeps, at this point, 
follows the arrangement of Recention A. 

From the fact that the arrangement of the material in A 
and the order of topics given in the editio princeps coincide for 
the sections for which we can control the one by the other (§§ 
35—38 corresponding to nos. 22 and 23), we may infer that the 
order of treatment was the same also for the subsequent sections. 
On the basis of the editio princeps we can thus indicate the 
topics, in the order in which they were treated, in the part of 
Recension A no longer extant. 

Apart from the fact that the Responsum retains an arrange- 
ment of the subject-matter closer to the original than the 
arrangement in 0, there is also internal evidence to indicate 
that is of later origin. Thus, in the section of 0, which would 
correspond to no. 25 in the Responsum (*)1jn n^n WD), the WD 
?W1 are eliminated, because they had already been treated in 
the earlier section (§ 34) of Recension A. Similarly, the omission 
of noun in the section of 0, which corresponds to no. 23 of the 
Responsum, may be accounted for by supposing that the topic 
noun had already been treated at length in that part of which 
would have corresponded with § 36 of Recension A. This part 
of has, however, not been preserved. Why deviates 
so widely from both Recension A and the Responsum in the 
arrangement of the subject-matter, I have not been able to de- 

Linguistic evidence would also tend to confirm the view 
that is of later origin than the text of the editio princeps. 


Thus in a number of instances uses the neo-Hebraic in place 
of the Biblical idiom. So e. g. instead of DM *0 (B no. 23), tib* (v. 
note 75); instead of H^ (ib.), lib* (v. note 76); instead of b» fcO (B 
no. 25), b*X Kl (v. note 90); instead of 1111 and lb in (jK no. 25), 
n»81 and I 1 ? n»N (v. notes 91 and 99) etc. is valuable, however, 
in as much as it checks the text of the editio princeps (v. especially 
notes 57 and 58), and preserves for us, in some instances, (v. 
notes 80, 82 and 84) a better, and even older version of the 
Responsum than the one represented in the text of the editio 
princeps. It would seem probable, therefore, that is based on 
a text that represents an independent development of the original 
text of the Responsum, and although, as a whole, it is to be dated 
later than the text preserved in the editio princeps, it does, ne- 
vertheless, retain, in some instances, more original readings. 



The Ritual is divided into four or five distinct parts. Each 
part opens with the introductory phrase n»0 '•SO Jtt p JH8W 1»« 
mmn "S». This phrase is found only at the beginning of a part 
or division, and serves to mark off one part from another. 
Throughout this work there is a definite schematic arrangement 
of the subject-matter which is unique in Halakic literature. 

The first part of the Ritual (I) is here designated by T\lbb7\ 
ntaw, a title justified by the words Dnow onto ntm* nwton 
by which this part of the Eldad Halakot is introduced in the Res- 
ponsum. The second part (II) I have styled msifc ITD^n in accordance 
with the introductory words Dr6ff msnts JYDto "jm pffi occurring 
in the Responsum, and the caption, ninfi nobro. asm n»yon m 
Tin t6k "1 b>&, used in 0. In the absence of any descriptive 
terms in the references to part III, I have used the word D^UB 
which describes the subject-matter. Over the fourth part (IV) 
I have put the heading TPn nilbn, warranted by the words, bxp) 
7V*rb& DK3HK '•fi «iw with which Ibn Gama c introduces his citation 
of § 36 from this part. Owing to the incomplete state of the 


text of the sections on rpn and *)iy in A, it is impossible to 
determine whether the *)1JJ JYD^n originally formed a distinct 
division in the Eitual (V) or only a part of the ATI niD^M. 

The arrangement of the subject-matter in the Eitual can 
best be shown by the following summary of its contents: 


§ 1. Introductory statement regarding the five rules of 
slaughtering and an enumeration of the rules. 

§§ 2 — 6. The five rules of slaughtering discussed. 


§ 7. Introduction to the subject of Terefot. 
§ 8 a. Perforation of meninges. 

b. Fracture of the skull, involving no injury to the meninges. 

c. Fracture of the skull involving the injury of the meninges. 

d. Paralysis of central origin — gid (v. note 57). 
§ 9 a. Color of the lungs, in disease. 

b. Anatomy of the lungs, normal and anomalous formations. 

c. Atrophy of a lobe of the lungs. 

d. Partial union of the minor lobes. 

e. Complete union of the minor lobes. 

f. Union of a minor lobe with a major lobe. 

g. Union of a minor lobe with the trachea or oesophagus, 
h. Union of a lobe of the lung with the heart. 

§ 10. The [other] organs, the perforation of which renders 
the animal unfit for food. 

§ 11. Perforation of the oesophagus. 

§ 12. Perforation of the trachea. 

§ 13. Perforation of the heart; swelling of the heart (due 
to traumatic pericarditis). 

§ 14. Adhesion of the gall-bladder to the peritoneum; cho- 
lecystitis; healed perforation of the gall-bladder. 

§ 15. Union of the lobes of the lungs with the dorsal wall 
deformities of the thorax. 

§ 16. The ribs. 

§ 17. Fractures of the extremities. 

§ 18. Injuries of the spinal cord. 

§ 19. Formation of a blood clot on the omentum of the liver 
("two livers"); absence or atrophy of the liver. 


§ 20. Absence or atrophy of a kidney. 

§ 21. Absence or perforation of the spleen. 

§ 22. Perforation of the intestines. 

§ 23. Domestic animals that have been attacked by beasts 
of prey. 

§ 24. Domestic animals that have fallen into a pit. 

§ 25. Domestic animals that have fallen into a stream; 
remarks introductory to Part Three. 


§ 26. Requirement to face northward during the act of 
slaughtering and the reason therefore; disqualifications for the act 
of slaughtering; animals that may be slaughtered for food, but 
not for sacrifice (illustrated by §§ 27 — 30). 

§ 27. Two-headed animals. 

§ 28. Animals having deformed extremities. 

§ 29. Animals afflicted with congenital arthritis. 

§ 30. An animal with one of its extremities shorter than 
the others (as the result of a fracture). 

§ 31. Persons disqualified for slaughtering (continued from 
§ 26) and conditions which vitiate the act of slaughtering. 

§ 32. Cases in which the act of slaughtering is not required 
in order to make the animal fit for food (a, b and c below) 

a. The young born while the parent animal is being 

b. Foetus, which has only half emerged from the vagina 
at the time the parent animal is slaughtered. 

c. Miraculous story illustrative of a. 

§ 33 a. Ben Pekuah, and evasions of the prohibition of 
£3 m\ m* (Lev. XXII, 28). 

b. Prohibition of tt3 n«l im«. 

c. Prohibition of atofc Htfr (Ex. XXIII, 19). 

d. Disposition of the carcases in a case of 132 JIM IfllK; 
intentional violation of this prohibition. 

e. A method of determining the relation of the young to 
the parent animal. 


§ 34. Distinguishing marks of clean and unclean game. 
§ 35. Applicability to game animals of the regulations con- 


cerning domestic animals that have suffered fractures of bones 
or injuries of the brain. 

§ 36. Applicability of the Hagramah and the other regu- 
lations for slaughtering except c Tkkur to game animals and fowl; 
permissibility of c IMcur in the case of game animals and fowl. 

§ 37. Covering of the blood (ffin '•'IM) {Lev. XVII, 13). 

§ 38. Applicability of the regulations concerning inspection 
(Hp^) to game animals [except as regards Nikkur]. 

[§ 39. Slaughtering of double-headed game animals]. 

[§ 40. Distinguishing characteristics of clean and unclean fowl]. 

Owing to the incomplete state of the text of A, the contents 
of §§ 39 and 40 and of the the concluding part of § 38 must 
be inferred from the Responsum (nos. 23 — 25, v. supra pp. 23 — 24). 
The Responsum is likewise our source for restoring the Halakah 
on the union of the spleen with the ribs or with the reticulum 
(B, no. 13) in A, § 21; and likewise for restoring the Halakah 
on the absence of the D^Tll ywp (B, no. 17) at the end of § 17 
or between § 17 and § 18 (v. supra p. 23). The reference of Rabbenu 
Nissimesa (Alfasi, on Hullin III s. v. Hfllta *)ljn *p ntDlfctP, cited by 
Epstein p. 138) to a Halakah of Eldad concerning fractures of the 
legs or wings of fowl, indicates that § 40 was followed by at least 
one more section dealing in a manner analogous with § 35, with 
fractures of the legs and wings of fowl, and, most likely, also with 
other Terefot in fowl. Similarly, the reference in Sefer ha~Te- 
rumah (no. 8 cited by Reifmann p. 103, Epstein, ib.) and in 
Haggaliot Maimuniyot (HfcW m^n II, letter n) to a passage in 
the "'iin Tl^K "01 ins, which does not occur in any of our texts, 
affords further evidence that even the T\WTM JTD^n of Eldad, in 
the form in which we possess them, are incomplete. We are 
therefore not yet in a position to reconstruct the Ritual of Eldad 
in its entirety. 

In conclusion, a few words should be said of the arrange- 
ment of the subject-matter in the Responsum. The Responsum 
contains the same general divisions as A, and on the whole adheres 
to the arrangement of A in detail. In the Responsum, parts I 
and II are, moreover, distinctly indicated by introductory phrases 
descriptive of the contents. In part I we find the same sequence 

65 a An older source than Rabbenu Nissim, in which this Halakah of 
Eldad is referred to, is the Sefer ha- Ittur (Lemberg 1860), II p. 13a: vijj&tf 
*^jn ^ide> xinw -atMn rp to "wto pnsn T&nn s :nn v6k "w. 


in the presentation of rules of slaughtering as in A. Part II, 
however, does not always retain the order of A. Passing over 
part III which preserves only a single section of A, we note 
that part IV again follows the arrangement of A rather closely. 
The following are the sections of A preserved in the Bes- 
ponsum. The numbers of the sections in A and B indicate the 
extent to which the two texts coincide in arrangement, and also 
the extent to which B reproduces A. It will be observed that 
part I is complete, and that part II is relatively full. Part III 
is remarkably short, part IV on the other hand is relatively 
complete, except for the conclusion, where the Halakah on *)iy 
cited by Babbenu Nissim (v. supra p. 28) is missing in the 
Responsum as well as in A. 

I. nos. 1—6 — §§ 1—6. 

II. no. 7a = § 8b, no. 7b — § 8c, no. 8 = § 10, no. 9 = 
§ 11(?) } no. 10 = § 9b, no. 11 — § 19(?), no. 12 = § 15, no. 13 = 
§ 21, no. 14 = § 20, no. 15 == § 16, no. 16 = § 17, no. 17, no. 19 
— § 24, no. 20 = § 23. 

III. no. 21a — § 32b, no. 21b = § 32a, no. 21c = § 32c. 

IV. no. 22 = § 37, no. 23 = §§ 35 + 36 + 38; no. 24 = § 27 
+ [§ 39], no. 25 — § 34+ [§ 40].«s b 



The language used by Eldad presents one of the most difficult 
problems. If we could determine, accurately, the character of 
his Hebrew, we would possess a very important clue to the 
country from which Eldad came. The material for solving the 
problem is, however, most meagre. It is limited to the Kitual 
and to the few fragments preserved by the community of Kair- 
wan in their Epistle to the Gaon, and by the older lexicographers. 
In the remarks that follow, the aim is not to set up a new hy- 
pothesis as to the home of Eldad's Hebrew, but only to give a 

65 b Tjy" e nee ^ scarcely point out that such phrases as Mb "i»N Ti$n, "njn 
WIK li'jKty are part of the form of the Responsum, serving merely to introduce 
the material taken from the Ritual. 


brief characterization of the language of the Ritual. A glossary 
is added in which the derivation and meaning of the pe- 
culiar words and idioms, used by Eldad, are explained. On the 
whole, I am in agreement with Epstein (p. IX et seq.) in his 
views on the language of Eldad. 

The general characteristics of the language of Eldad may 
be summarized as follows: 

1. The Hebrew, viewed from the standpoint of the GJ-aonic 
idiom, is of an archaic type. It approaches more nearly the 
Hebrew of the Bible than that of the Mishnaic or of the Tal- 
mudic-Gaonic literature. Nevertheless it manifests certain dia- 
lectic peculiarities of its own. It bears the impress of a language 
spoken or written by Jews at some time and in some country 
and of an independent development there, similar to that of the 
Halakah written in it. The language in itself does not give 
evidence of having been created ad hoc. The old lexicographers 
never suspected its genuineness, or they would not have cited it 
as authority for their explanation of Hebrew words. Thus Jehuda 
ibn Koreish seems to refer to this idiom as authority for the 
explanation of the word ]Vffl in Psalm VII, 1, claiming to have 
learned the particular usage from "a Danite" (cf. Glossary infra 
s. v. and Steinschneider, Die Arabische Liter atur § 35). Dunash 
ibn Tamim, who lived in the middle of the tenth century (v. 
Steinschneider I. c. § 36), presumably the author of an ancient 
commentary on the Sefer Yedrah, had a theory that "Hebrew is 
a pure Arabic", and he asserts that he derived it from "the 
Danites". (v. Steinschneider, Introduction to the Arabic Literature 
of the Jews in J. Q. B. XIII pp. 306 and 315 et scq.) Ibn 
Gama c made a detailed study of Eldad's vocabulary, and arrived 
at new conclusions, which he asks the reader to examine 
thoroughly. 6 6 Even as late as the second half of the thirteenth 
century Tanhum Yerushalmi, who lived in Asia, referred his 
explanation of Lam. IV, 8 to the usage of "a Hebrew, living in 
the desert, known as the Danite, on account of his relation to 
the tribe of Dan", {cf. Glossary infra s. v. 122 and Pinsker, 
Liklmte Kadmoniyot p. 180.) 67 

66 In the Risdle itself, as far as I can see, Ibn (jama explains only the 
etymology of ons> (v.Gl.s.v.)and the use of npm instead of rnnn (v. p. 66 note 104). 

67 Cf. also Hasdai Ibn Shaprut's remarks on Eldad's Hebrew: "mo n\T, 
i:»b ntya uh nn bi) p"nbi -m b& mot? «-iipi mnsn (v. Epstein p. 70). 


The following may be regarded as dialectic peculiarities in 
the language of Eldad: 

a) His preference for the Hiphil form: W ,tf&yn ,p3 
tt^ttOn fWtbn ,nypyn ,1D3n3 ,BT»p^ (v. Glossary, s. vv.)\ ttw 
(§ 33 e), instead of ISST etc. 

b) Unusual formations derived from Hebrew roots: t^Dyn 
from tDV; Pp"^ from trp 1 ?; bv? from «teiy = «tel«; itilDfi from 
iDD; nans from &TS; D«ns and DnD from the adverb DKHS; P&n 
from BWI; nvrtin for nnn; (v. Glossary 5. m). 

c) The use of Hebrew words in a sense in which they do 
not occur elsewhere: D*tt, "sinew"; iTBi, "span"; pSD, "skin"; Tip7, 
"interference with the free movement of an animal's legs during 
the act of slaughtering"; nip, "reticulum"; (v. Glossary s. vv.). 

d) The use of words of Syriac origin (sometimes in Hebrew 
formations), which are not found elsewhere in Hebrew. Examples of 
this are mnn = .^^; BOin = -^^; nWIDB = I^j»qj», (v. Glos- 
sary s. m). 

Epstein (p. x) has collected some of the archaisms which 
Eldad uses. His list may be supplemented by the words given 
on p. 15 supra. Attention should also be called to Eldad's use 
of poetic expressions in the Bible, such as ym to denote "to 
kick, to trample", and lp*\p "the pate" for "head" (v. Glossary s. vv.). 

2. Eldad's Hebrew abounds in Arabisms. The influence of 
Arabic is undeniable. Epstein concludes, therefore, the testimony 
of the people of Kairwan to the contrary notwithstanding 6 ^ that 
Eldad knew Arabic and could speak it. This conclusion does 
not follow inevitably, for there are other instances of Jewish 
writers who did not know Arabic, but who, nevertheless, used a 
Hebrew style which showed Arabic influence. The language 
of Eldad contains moreover Syriac as well as Arabic elements, 
although the influence of Syriac is not very marked. It would, 
therefore, be logical to assume that Eldad knew also Syriac. 
It is, however, possible that the language of the country in which 
Eldad lived, was influenced by the Arabic. 

The influence of the Arabic is shown in the adoption of 
a) Arabic words, some of which are given a Hebrew form: 

68 Responsum (ed. Epstein p. 5 no. 5) W* in« W1 l^SX »ttn v6k m 


innDnD" 1 from l^c^; V^M = g^; p&D from ^U-o; B^fiVn from 
JaiJ; and probably t^DII from Ja£;> ( v « Glossary s. w.). 

b) Arabic idioms which, at a subsequent time, became rather 
common in the later Arabicized Hebrew: niTin ty 113V 1 P = 
^. 1 J* )js& &oS\ b*ntrt nipri im« Dnnn IBM corresponding 
to the use of f Jk. cum J* personae; hence irty 11DX in place 
of Hebrew )& TlDK; compare, however, Mishnah, Erubin VIII. 4 
(v. Glossary 5. m). 

c) Arabic constructions in grammar and syntax. The 
placing of the demonstrative pronoun before its noun is one 
instance: ^Dlpan n«t 18 pKH nw (§ 22, T-8,P)\ T\WTW7\ n«T in*6 
(§ 33b, P); ovn nt (§ 23 T-£); ™nWn nan (B, no. 21). 

Other instances are the use of the Sifa (indeterminate relative 
clause): nfi1B ,Al3 IBOl DW ^T TIJ& ^3 P 5> m & ^ D ^ in § 29 
(P), the use of the Tamyiz (accusative of specification): 1VT DM1 
n«"^n m»D DniDI TOWD my^TI in § 15 (P), the omission of the con- 
junction V = JS as in vulgar Arabic: in« \nb& ty taw TIDK1 in 
§ 33b (T-S), nn« nyn om« tew "pty 11DK1 in § 33 c (T-S), the 
elliptic use of *iy in the same way as ^^- is used in neo-Arabic, 
particularly in the African dialects, (cf. hatta 'qui, je vais parler, 
J. Oestrup, Contes de Damas, 1897. Glossaire p. 157 s. v. rlatta). 

The evidences for the genuineness of the Hebrew of Eldad 
noted under no. 1 make it impossible to regard our text of 
Eldad's Ritual as a translation from an Arabic original. As 
Epstein has successfully shown in his criticism of Rapoport 
(Mafiberet he-Aruk. ed. Stern p. xi column a), such an as- 
sumption is entirely out of the question in the case of the Res- 
ponsum, the text of which is free from Arabisms, with the single 
exception of rb^bn AND. 

The following Glossary gives the words and phrases of the Ritual 
which are either obscure in meaning or are used in a peculiar sense. 


l. did! — inis mn ♦ . ♦ jap tnoi ik pp tyw d*o (§ 23). 

Assuming DID! to be the correct reading (v. A, note 237), we 
may perhaps account for it as a mutilation of D^DT = Greek 
SaauTtoix;. This word is used in the Septuagint to translate rOJIN 

69 mptt occurs as feminine in Job XX, 9; II Sam. XVII, 12. 

70 nb y b is here erroneously treated as feminine. 


(Lev. XI, 6, Dent. XIV, 7; cf. de Lagarde, Librorum Veteris 
Testamenti Canonicorum Pars Trior Graece, G-ottingen 1883 I. c). 
Of the same origin is the Ethiopic Dasippda, species leporis vel 
cnniculi, v. Dillmann, Lexicon s. v. col. 1098. If the reading D1D*i 
is the correct one, this name of the animal may be derived from 
l/"bn, "to break." 

Epsteins's emendation DID (p. 115) is impossible in the context. 

2. f3T1 a) "to notice, to perceive". (In Biblical Hebrew the 
Qal is preferred) in this sense :y!»n DIpB pn *b D«1 (no. 12); 

nnoi rrV« pn *6i DDirf? rrtn "ppym (no. 6). 

b) "to understand" (as in Biblical Hebrew): pn at TH 
nttKl pi p ytW (no. 22), or "to explain, to interpret", as in 
'Abodah Zarah 45 ytob plK "OK »^pV ^1 "l»« (c/. Weiss in 
Ha-Hoker I p. 160 and Bacher, Die exegetische Terminologie, 
Leipzig 1905, I p. 8—9). 

c) "to investigate, to examine": nn« nytt ^ ^ B** P* (no. 13); 

nana, new Dip»n «to mrun ropa dm nte«on Dipa p^ (no. 4 
cf. B. note 11). 

3. onnn: Dip&n orfc otik ^nn tonart riipri im« onnn ipk toi 

in« (§ 32 c). Dnnn = Arabic ^ c?wra ^Js. personae. The joining 
of a subordinate clause with the main clause without the con- 
junctive particle ti = Jf in 1HK jn 1 ?^ ty tafcO 11D«1 (§ 33 b) and 
in nn« nyi nniK ^3*0 vty ™« (§ 33 c) follows the usage of the 
vulgar Arabic. Cf Spitta-Bey, Grammatik des Arabischen Vulgar- 
dialektes von Agypten, Leipzig 1880, pp. 424 et seq. 

4. lsnaru 'W d^t ym v6m m isnon vb\ mty» dmi (§ 6). 

Vfnft is used here as synonymous with t£OTl or D^Tl D^l BOtH 
and Bjnnn or D'H" 1 '! D'to'Q tsyn. In the Bible ffi» is a poetical 
expression, "to smite, to shatter". If the text is not corrupt, 
the reading is ^nBH; v., however, note 39. 

Epstein (p. 108 note 17) traces it to the Arabic ^s? Jac- 
tavit pedem suam animal jugulatum (v. Freytag 5. v. IV. p. 154). 
But the Arabic equivalent of Hebrew pn& is J^, both phoneti- 
cally and according to the meaning (cf. also Assyrian mahasu, 
Delitzsch, Handworterbuch p. 398), not ^sr°. In view of the 
biblical usage of f!"!» it is not necessary to seek an Arabic 
derivation for this word. 

4. fcrtsyn: rbmn nyn&n nnn» jwayn dki (§ 5); ntrtayn •oi 
V?wi dhw ibbjnv (ib., v. note 28 a); nyn»n nnn piro aw d«i 


Kin Tina tftojm »bl ffcrttl (ib., v. note 23), a verb (Hiphil), denominat- 
ed from the biblical tt& which Eldad uses in place of the Greek 
derivatve D)12b)p, commonly used in Talmud, Targumim and 
Midrashim (v. Blau, „Stndien gum althebraischen Buchwesen", 
p. 170, note 5). Neubauer, Journal Asiatique 1862 p. 206 note 1, 
wrongly traces it to the Arabic l£ "to slit'". 

6. mt: piro t ty ipsirn dtoi (§ 30); woD Dim yiroi 
(§ 18); niDrona owrn hcmmi and ,T&ip»:i nnaiy moirn (§ 17). 

Dlt "sinew", is evidently a Hebrew translation of Aramaic fcCO 
(from /nu = Dlt). Dr. Ginzberg informs me that nit in the 
writings of the middle ages denotes "the circulation of the 

7. ^mm man did rrton rwmm (§ 6, Gy, [pmnni =] 5mm 

rtftfYI iTT (§ 36, #), where T-S has »nnni. The context indi- 
cates that the meaning is: "the animal kicks with its feet". Gf. 
Neo-Syriac Jti5^ agitavit (Smith, Thesaurus s. v. p. 1366). 
This word is rare. Perhaps the reading in both places should be 
wy\T\T\\ a reading which is actually found in T-S § 35; v. next 

According to Steinschneider, J. Z. II p. 300 n^nim is a 
mutilated form of Talmudic riBOTD or nBOBO. Neubauer, I. c, 
compares Arabic k>^ "to bend a bow". Epstein is misled by 
the corrupt reading tfSWl of P in § 24 which he emends to 
Vtim (= Hiphil of fffi*i = D31 "to shake, to move"). In order 
to obtain the meaning required here by the context he would 
have to emend to TOSTTtt in § 6 and PSim in § 35. 

8. PDinM^m rr n« (P pbw) pwn (§ 24, T-S); Bomm 

^ri^iTT (0 iinm) (§ 36, T-S), meaning same as Vim. Cf.SyrisiC 
^oyl, usually followed by <n-aJo5, "to wag the tail" (v. Brockel- 
mann, Lexicon Syr. p. 124, Smith s. v. col. 1373). According to 
Sachs (Beitrage I p. 175) vim is = Talmudic two, derived from 
the Greek xspxo? (which is impossible; see, also Kohut, 'Aruk TV 
p. 328 a and Jastrow, Dictionary I p. 670). Here, too, Epstein 
would emend to BtoWl and Painm. 

9. vi^: run im« rwr\ w p< ^« (§ 34); rw D^ys * 
T^ (*.); » wi )^n twin nyT mn dk tea (§ 27) etc. (in 

Biblical Hebrew the Qal or Niphal is used) = TT "to chew the 

I find the Hiphil of V\* in the sense of "to smart" in Joseph 
[Benveniste] Vidal's Hebrew translation of Joshua Lorki's medical 

work mty»n DU (v. Steinschneider, Hebr. Ubersetmngen p. 762): 

nr\w s*bi Tibp mm rrwi \wbn piw oyts Kin nirrtan (quoted 

in Kaufmann's Die Sinne p. 166 note 9). 

10. BhpV 1 ! mo to lDYlp t^tf Winn (§ 8d): "The me- 
ninges begin to seeth or bubble i. e. to form gas" (v. A note 57). 
If t^p^t? is not a mere scribal error for p"6tW (Epstein p. 112 
no. 3: pl'W), the root may perhaps be related to the Arabic 
SsJ "the camel has her mouth covered with postules" (v. Lane 
s. v. p. 2671), the postules looking like bubbling water. The more 
probable derivation of t^p^ from the Hebrew noun Plpte, "bubbles 
of the rain", is suggested by Dr. Ginzberg. 

11. irrnDHDM Uin p irrriDnD^ (§ 37): "Make^the blood flow 
out of the garment", = Arabic 'jLs?* and !£&"**, "to flow from 
above, to pour fourth", (v. Lane s. v. p. 1313), Ethiopic zahzha 
(v. Dillmann, col. 1039). It is, however, possible that ln^OnD" 1 is 
merely a different spelling for IHTOnr, from /nsns, "to wash, 
to clean." Cf c Aruk s. v. 

12. nnys wwi by niay p "ipnni )«^n npu uro* ibno (§ 35), 

— Arabic ^ * ( _ 5 -uxJ\ J^f?, "a thing passes or takes effect or is 
applicable", (v. Lane s. v. p. 484). Cf. Cuzari ed. Hirschfeld p. 24 

(I. 50) nrttmrwB "by nt^a ny *b dnj; pn di^k j«a jd nnii and 

Hebrew translation, ed Cassel p. 43: JTJTO ■•0 [rather HSDl] *ttTI 

[read vnftii] witon ty non nnr *6 nap ni«o twwi. Pseudo- 

Saadia's Commentary on the Sefer Yezirah: (Arab. PfcO) "Diy r6lt nt 
D^S ^tPB (v. Steinschneider, Hebr. tlbers. p. 448 line 8). Jeshu c a 
ben Jehuda's Commentary on the Decalogue "I12JT (arab. ny) 
ISpf? (v. Steinschneider $. p. 144 note 282). Al-Ghazzali ed, 
H. Malter, (Frankfurt 1896) I p. XXIX: tf ^ JjUoY^ ^V 

A nny jvtyn :hdd «Tn i*? Vat^n ^ = ^^-i \ a^j\^ *ui AyJi 
OTUiw D^m Ti«n "pton Tinjm ids ivtyn Dipa b« yjmrm Qf. 

also i&. p. XXXVI line 11, Part II p. LXVIII line 8 etc. In 
later philosophical literature the Hebrew "Qy is frequently used 
in the sense of the Arabic jU-. 

13. nnnrp; narrow noron nnnrvs? (§ 6), "he tries to throw 

the animal to the ground". The verb nnnn seems to be de- 
nominative from nnn, just as the Syriac v-aI»z submissit, subjecit 
(v. Smith, Thesaurus s. v. p. 4424), from w-aII; cf. also Ethiopic 



athata, humiliare; athatta, dejicere, submittere (v. Dillmann, p. 553) 
and KEn 1 ? rpnnn»1 of Hal Geo 1 ,, ed. Venice 1548 p. 129 b, ed, 
Hildesheimer p. 512, [which Sefer ive-Hizhir, II p. 27 (n« nn»1 
TWtiTi) evidently misread for rrnNDl]. 

According to Epstein (p. 107 note 13): llp^D WX r\)T\nb p 

rtwrwn nnDD nmxn \ia ty twin n&nnnt? ^)b nam (T" 1 ,{ ? iTjw) 
iwa ik pan rrtan pyam. 

14. mm ma: maims nimrrD v nwsn ta>3in» tswnn n*nn dki 

(§ 8d), is unintelligible. I emend to HUTTD or nttWrrD, "leek- 
colored, green", (v. Jastrow, I p. 675, and A, note 57). 

According to Epstein (p. 122 note 5) DlfPO is an Arabic 
word, meaning "highland, hill". I cannot find such an Arabic 

15. DTnns: w*on xb) dvivd D^irrc d^d pntwio K2P dk (§ 13), 

"If there is exuded from the boil a light-colored (= straw 
colored), thin liquid, which is odorless", niro = JVrD in Biblical 
Hebrew, used of thin, fine oil, is here applied to another liquid. 
Epstein (p. 123 note 12) would emend to D'nVQ. 

16. wpbt wpta ^ntan np^i dki (§ 21) T-S; (P wp^n), used 
in the same sense in which KB^Ip is used in Hidlin 55 b (antonym 
of fcOfclD). Epstein (p. 115 no. 20) suggests )U?b)p2 as a possible 

17. rinsing: rinsing rrtai ik ,tt imntsw nnni n«i (§ 30), 

"If an animal has [at one time] fractured [one of] its hind legs 
or fore legs" (owing to the excessive weight of its body, v. A. 
note 329): nnsintt^ perhaps the Arabic J^dyJc from V^y^y, 
rebondi, arrondi par un bon point, (v. Dozy, Supplement p. 145). 
According to Epstein (p. 119 no. 29) nnsinia^ is equivalent 
to nnsinDI from V^KSI. To be consistent, he should emend to 


18. p^iidj pTDo innnn ma bw i»np nffjni (§ 8c), "If the 

meninges are ruptured", from V^TD, (Epstein p. 122 note 3). 
Cf. also Assyrian bataku, durchreissen, zerreissen (v. Delitzsch, 
p. 191). 

19. ybSfti to "itson dw r\yv y\yb ityn p y»:i» ^ a«i (§ 29); 
"oi nsyn nnno y»a»n dki («&.); d^to ^ *)ron jo nrn y&J&n dki 

(ib.): "If there is a collection [of pus] (i. e. an abscess forms) 
between the flesh and the skin causing the pus to spread 
in the flesh" etc. — Arabic ^J^° or j^sr 5 , "place of collecting 


and the like" (v. Lane s. v. p. 459; according to Dozy, p. 217 "a 
medical term"; v. also Epstein, p. 127 note 40). 

20. nwtoiDt niton kvi rrmfci»i nnton vxm rrw ^ (§ 8d. 

P), according to Epstein (p. 122 note 4) tolD = iniD or to», 
"fluid" (v. Levy iV. IT. TF6. Ill p. 39 and Nachtrage ib. p. 305). 
The MS. vocalises mvfol&l. The author may nevertheless have 
intended rprnVrttDl (Cf. also Barth, Beitrage zur ErJdarung des 
Jesaia, Berlin 1884, p. 4). 

21. n&tiinfc: ton n&&in» &n»_ rfaa ia (§ 11), literally u a 

piercing pin of iron", Polel of /o&n or Vain (v. Jastrow, s. v. 
p. 448), a quaint tautology. Perhaps ntOBinD is a gloss to^fctflfc, 
explaining the Hebrew tsntt by the more familiar Arabic s£*^, 
the "iron instrument [awl], used by sewers of boots, with which 
they figure or decorate the leather, or an instrument with a 
pointed extremity, used by binders of books and by others" 
(v. Lane s. v. k^ p. 593; cf. also Fraenkel, Die Aramaischen 
Fremdworter im Arabischen, p. 257). 

22. nunom ntonnyn&n nnna snw ni:mD»n "iinym (§ 5) 

ito Din K2n \Vlft D^n inns*' "WK, "and on account of the blood 
vessels" etc., c/. Syriac \^ojo and V^Soa — oopi-nec (v. Merx, 
Profcew des Syrischen Textes von Galenus "De Simplicibus" in 
^. D. if. G. XXXIX pp. 242, 249; Brockelmann s. v. p. 238; 
Smith s. v. p. 2727). 

23. nsmpa: \vby rompo ma to inm n\w (§ 26) 

vty mnipD «\T fc& inVTl, "his mind is not concentrated" etc. 
pp is used in a figurative sense in Psalm XLL, 7 "6 )1« pp 1 * in 1 ?, 
c/. Hashi ad loc. and also Weiss in Ha-IIoker I p. 160. pp is 
used in the sense of "attaining to consciousness" in the Midrash 
Bereshit Eabbati of Moses Ha-Darshan, quoted in Pugio Fidei 
(Leipzig 1587) p. 563, xb» rrenpjw nvn pi p yBn.T 1M1 no« 
'ini mt?n ^nate 1 ? nnpn *m jwann dik to wjn (read vty). Qf« 

also Epstein p. 68. 

Epstein (p. 127 note 34) derives the figurative meaning of ]>np 
from the Arabic 3^Z\y^ £ -^. or **AS e-^.. ^ we are to seek 
an Arabic equivalent for this use of pp, the more probable 
Arabic idiom would be dJla* £ J*<~^, concentre, retire (v. 
Dozy s. v. p. 300; cf. also JfZs, "state of concentration", v. Lane 
s. v. p. 2482), which corresponds phonetically to nsnipB. 


Y^p is also used in a literal sense in § 15 (P) of this Ritual: 

24. Pfiin&i nww tysnnD bwipi n«in dki (§ 8d), "But if 

you see the brain swarming with worms (Parasites)" etc. 125*1 
= Syriac Lsi , se movit, redundavit, corrosus est (vermibus), (v. 
Brockelmann 5, v. p. 359) ; nti^BI = Syriac i£s? reptilia, vermes 
{v. Brockelmann ib.). It is, however, also possible and perhaps 
even more correct to interpret JWS1 ^Sinfi as "soiled with 
dirt", from the Hebrew noun t?B*i which is a synonym for t^tt 
(c/. Fs. LVII, 20 and ttfin ^»1 B*B ]ni« nPljn Midrash Tehillim 
to Psdm XVIII, 8 ed. Buber p. 72 a). A plural IWB1, according 
to Kohut's emendation, is cited in the c Aruk (v. s. v. fc^ST). The 
verb £?3in& could without any difficulty be taken to be denomin- 
ated from the noun. 

According to Epstein (p, 122 note 5) iWBI means "move- 

25. IDana: mat* 1MTO UK -pK (§ 6). 1D3m is evidently 
Imperfect Piel of a secondary stem DBH. This form is the 
Hiphil of DBi, but it is here treated as an original stem. 

Epstein (p. 108 note 16) would emend to ID^Hl 

26. moa; WTO in*6 nte«DV nU3 «m (§ 3), literally "while 
a span of the knife at the end has not been used", i. e. not all 
of the knife has been used in slaughtering, a span of the knife 
being left unused. iTtDi usually "the act of spanning, extending" 
(bni«n rPM3, Mishnah Oholot VII, 2, Tosefta ib. VIII, 2). 
Here it is obviously used as a measure, "a span". 

Goldberg's incorrect reading (p. XIX) iTBJIKm obscured the 
meaning of the word. Of Filipowski's more correct reading 
rptttsm (as one word p. 208 col. a) Steins chneider says (J. Z. II 
p. 298: vielleicht zivei Worter, vielleiclit eine vox hybrida von 
HEX Frankl (M. G. W. J. 1873 p. 492) says sicherlich von HC3i 
gebildet. Epstein (p. 105 note 3) also takes rrfitt Km to be one 
word and emends it to iTBJrp, which he takes to be the Hiphil 
Perfect of T\M. 

If rp&atfni should be read as one word, an emendation 
would not be necessary; .TWH might be a form like JTOm, derived 
from the Hiphil of Mbl 

27. t^SiJ Kin Btoa *)iym rrnnt? *b (§ 36), "Since wild ani- 
mals and birds are full of life" i. e. they are very active. t?Bi 
is here perhaps an elliptical expression for t?Bi *»tyn or it is used 


like the adjective ]~*.±*2 with which the Peshitta renders PSi S}D 
of Prov. XXIII, 2. 

28. pso: Btra puD^ pso pa (§ li); dvtk mm p^ ina pBD 

(ib.) = Arabic jU-o, plural ,5^, "the inferior skin, the skin 
beneath the upper skin and above the flesh" (v. Lane s. v. p. 
1701, Epstein p. 123 note 9). 

29. hay* ;rmn toy -wrem msp nto«on n«i (§ 3) 

IIKdV iy toy mjn Min to:i PTI DK1 (i&.), evidently meaning in 
both places "thick, stout", cf. *tfoiy = *6DlK (v. Levy A 7 . T. Tf6. 1 
p. 74, III p. 646; Assyrian akalutu, Delitzsch p. 56), a small, 
dry measure (v. Zuckermann, Das Jildische Maassystem p. 41). 

Frankl (I. c.) explains toy as meaning 50 viel ivie "HID, 
viel messend — vgl. mo B^K J Chron. XX, 6, cf. also Harkavy 
in Graetz, Geschichte (Hebrew edition) III p. 460 note 153. So 
far as I know, however, TPTD does not occur, TDK being the 
only derivative from T7D which is to be found, v. Kohut s. v. I 
p. 114. Steinschneider, I. c. would emend to toy angeschwollen, 
aufgeblasen. Neubauer, I. c. says probablement le mot J^U, 
court et gras. I can find no such Arabic word. Epstein 
(p. 105 note 4) would emend toy from Arabic J**, "to be 
thick or dense", 

29. ^)p^ describes: 

I. The condition of the animal when its feet are bound, and 
is synonymous with Ipy, *V\py and JTTpy (which is a technical 
term used to denote the binding of the animal for sacrifice, cf. 
Mishnah Tamid IV, 1) K\m DHan ntffl br\ (G IpV) *lp*V DM 

wd niTim ked ipnm jwsn mp^ i»in nn« n»^ Kin tod mpiyo 

nvwp om m^m rrT aoinni rvbum tfv wpn dk 

♦"Di D^m d^t nn^p *6n nnw nityn dm ^ (§ 36 T-#, 0). 

II. More generally, the condition of an animal which is 
prevented from using its legs freely through any cause. Thus 
the term is used in the text to refer to an animal which is in 
such a position that it is prevented from moving its legs by a 
beam, or a wall, or because it has dug its hoofs into the ground, 
or because it is lying in a narrow hole in the ground. In these 
senses the denominative verb, derived from Tip^y is found in the 

a) Qal: TIDK rrtn MK Tlpyn&J> nam (§ 24 P), where 
the reference is to the slaughtering of an ox that has fallen 
into a narrow pit. 


b) Pual: nnpiya iT^ni (§ 6 G, no. 6 R). 

c) Hiphil: DDirQ rrtn ,TPpym (no. 6 i?). 

d) Hitpael: HTpn i« wn i« pto rnfcn nip^nii (§ 6 G) 
♦ppyn» jn n»n n^n mp^y n*n &6i pan rdwyo 

In Biblical Hebrew there occurs the verb BfHtfn, "to take 
root", which is denominated from the noun Khfc^, "root". In the 
same way Eldad seems to use the verb 1pJ> as a denominative 
from IpJJ, "root". The basic meaning of 10 in our text would 
accordingly be: "to take root, to be attached to the ground" 
(cf. the German festgewurzelt), and the more general meaning 
would be: "to be prevented from the free movement of the feet 
in any way". The different stems of IgJJ are used to denote 
various ways or conditions in which the movement of the feet 
may be impeded. (Cf. also the Arabic jAa I. "to detain" and 
II. "to cleave, to cling, to hold fast to something", v. Lane s. v. 
p. 2108). 

Babli Shabbat 128 b (quoted also in Hal. Pes. p. 142, Hal 
Ged. p. 513) expresses a caution against allowing a hen to touch 
the ground during the act of slaughtering, lest it may fix its claws 
in the ground and thus cause "llp^. The passage reads ^28 TOK 

*?td rrb bTi ^s vt Kjnan ,Tp:6 im^n^ r6unn tawi p wi 

At the end of 'Ikkur (§ 6) Ibn Gama c adds the observation 
that this chapter should be called 'Ikkad and not 'Ikkur, for 
Eldad uses Tip'y in the sense of binding (k>y^ jJUJI). He seems 
to imply that Eldad uses the verb Ipy in place of Ipy. In proof 
of this he cites TVT\ ITD^n § 36, where Ipy is used precisely in 
the sense of ^TpJJ (v. above sub I). This view of Ibn Gama' is, 
however, mistaken, for Ipy does not in every case mean "to 
bind", as may be seen from the instances noted under II, c, d. 

Epstein (p. 108 note 18) seeks to justify the remark of Ibn 
Gama c by referring it exclusively to the passage cited from the 
JTn TVobtl. Such an interpretation of Ibn Gama c 's words: Kin 

d^ bi2 Dpy "w s»i« my n«n^« «in )« "»ty naaya 9di rfttp p 

B2"Ak1 IpJ^K ^jn nTpJ^K ^D iTS 'p&ynDK m*6 Bma is impossible. 
The closing words here indicate beyond a doubt that Ibn Gama' 
refers to the entire chapter on *np^> i. e. § 6 as well as § 36. 
Steinschneider (J. Z. II, p. 300) understood Ibn Gama* to mean 
that the original text actually read IpJJ throughout and that the 
form "lpy is due to the error of a scribe who regularly wrote 


the more familiar word Ipy, until he met with Ibn G-ama"s note. 
Epstein (I. c.) has clearly shown that Steinschneider was mistaken. 
A mere reference to his argument must suffice here. 

31. miDBJ rttlDS pan 1« (§ 6). "On ground full of holes", 
from /iDS (v. Levy iV. H. Wb. IV p. 68 and Epstein, p. 107 
note 14). Epstein's other explanation (ib.) HilDD = MJDB, "a high 
place" seems less acceptable. 

32. ntonsj ynsa nnro ntons ),w*nn m (§ 9d), "If there 

is a cleft on the top, of the width of a finger", a genuine Hebrew 
formation from l/ttTS ; cf. Syriac jL^£s, fissura, rima (v. Brockel- 
mann p. 286). There is no need for the emendation Hp^S (Ep- 
stein p. 103). 

33. Dns, a«nD: onm dtk inona i« (§ 2); rr&Kns *a (§ 3); 

niDH Dni« DDnS' 1 DK (§ 32 a). According to Ibn Grama c 's ex- 
planation D«ns p nnia ™na tdb™ (v. note l) it means "to 
take by surprise" (cf. Pinsker I. c. p. 108). 

34. ^nSJ 1) In the sense of "cord, thread": nyntDfc TrW DK 
MS ty ^ns *6fc nVlTl (no. 5, i2) (The parallel text A § 5 has 
1Jtt?n toirD instead of ^TlS *6»). 

2) In the sense of a "garment 5 ', most likely a "cloak" or 
"mantle" in IDini intsnfc^l *m IK vty "WK ^Tisn r\p* (no. 22 i?, 0) ; 

mn n« in hdddi ^nsn n« tew (#>. i2); isyn bx ^nsn it?K Din 761D1 

(ib. R, 0). Cf. Targum Onkelos to Gen. XXXVIII, 18, 25; 
Tar gum Yerushalmi to ib. 25, where ^>TIS is rendered by NS'Wty, 
"coarse cloak, mantle" (v. Jastrow p. 1543 and Rashi ad. loc, 

m riMna nnm -jntety). 

If, with Reifmann, we read tolD for *6liD in the parallel 
passage in the Or Zarua and Mordecai (v. p. 87 note 572), we 
would have a further confirmation of the correctness of the 
meaning here given for Wis. 

Following EaMBaN to Gen. XXXVIII, 18, Epstein (p. 96 
note 40) takes Ws to mean BW1H YT1D, "a scarf wound around 
the head". In view of the fact, however, that Recension A in 
the parallel passage (§ 37) speaks of two D^:o, "cloak" appears 
to be the more probable meaning of ^ns. Epsteins other 
suggestion (ib.) that ^ns means TitK, which he bases on RaSHBaM 
to Gen. XXXVIII 25, is even less satisfactory. 

35. ipipt ipip by ipip inn 1 ' DHpnpn pap*i (§ 27); jnn 
nn« n&TOn dmp ttirwi ♦ ♦ ipip by ip^p mpipn (no. 24 

R). Here Iplp cannot mean "pate, crown of the head". It is 


used for "head" in general as a synonym of fcPtfl. This usage 
occurs in poetical passages of the Bible (e. g. Gen. XLIX, 26 = 
Dent. XXXIII, 16; Psalm VII, 17, all parallel with fcTKI); cf. 
also Assyrian kakkadu, Kopf, Haupt (v. Delitzsch p. 592). 

36. msp: mmspn [ntsntw] nwb&n p) (§ 3); &nt? xvjrain dm 
nawn nisrp iy ($.); ntswn map 1 ? -jnmi (&.), according to 
Epstein (p. 105 note 5) = noA and 1»:i 1J>. Qf. also Arabic 
*UJL>1 "end, accomplishment". 

37. nip : t^aon n^pn i« D^y&n 1« mpn npM n«i 

nn^pm mpm D^ya to npu ^ mpn i« m-pn i« D^y&n (§ 22). 

Here nip is used not in the general meaning of "intestines", but 
most probably in the specific meaning of [JWiS] DID, Falten- 
magen, reticulum {cf. Lewysohn I. c. p. 38). From an anatomical 
point of view, too, this is the meaning of nip required also in 

no. 13 (R): b)mn p npb mpn dk mpb b)mn pma. 

38. pan: (Pyrtttm) nsntyn p&n hied ik Din m»n in n«XDi 

(§ 22 T-£, P), "If you find therein a fluid that resembles blood 
or the expectoration (saliva ejecta) in the case of consumption". 
Btoll from Y¥vr\, "ejection". 

Possibly in both MSS. fcHDll is to be read ; v. next article. 

39. pan: runts t^ann *6 d«i t^ann nob «r wsn dm (§ 13, 
P), "But if there is exuded matter, whether it be odorless or not, 
the animal can not be used for food". fc^Dll = Arabic l^, "filth 
or foul matter — that collects or concretes in the inner corner 
of the eye" {v. Lane s. v. p. 1156). There is, however, also this 
possibility to be considered, namely that t^JDII = D»V) is derived 
from VDJDI, "to tread", and that it has the same meaning as 
DOID which is almost a synonym for t^D {cf. Micah VII 10). 

Epstein (p. 113 no. 13) suggests that the proper reading 
may perhaps be tPBll. 

40. HBMBI see PSino. 

41. t?&fcJ>! mi m WW DK (§ 9a); (or WIST) WW [vb] D«1 

np^inn min nn (§ 9g, v. note 103); nuwn unmb mm hit *6i 

(§ 12); mm nn tSWl (§ 15); "to serve i.e. to supply the lungs 
with air, to blow air into the lungs, or rather into an obstructed 
lobe i. e. to inflate it". Cf Hal. Pes. p. 143: mm nn SJWD ]W n«n 

uwn mm nn (read ff»p») i^d^dd jw nn Dins mm p«i 

nfc^ KBB>; .flak Ged. p. 525 Kpn fcttflOPD *6l KBD'; $. p. 526: 

ktj nt?n^ xw jwin Kpn mn mmmwo «p *6i (c/. Epstein in Ha- 


Hoker I p. 328). Conip. also Hagigali 12 b VW& 1W \tr% and 
Leviticus Rabba IX 2, rwm W n nWBMDff. V. also S. Fleischer, 
Nachtragliches zu Levy's CTiafcZ. Wb. II p. 578 a. 

42. nrnnn: nnw noom te rayn nvnnn dm (§ 18); iy 
mjhi (0 rnnn) nvnnn nyypn -ipk (§ 33 e, T-£, 0). nvnnn is^used 

here as a preposition (= nnn) "under, below". C/l Syriac Ua-J~^ 

43. tra^n: t^np] D^n we na prrv (§ 33c), "He shall 

rinse his mouth with water and spit it out". tya 5 ? = k£J "he 
ejected or cast forth his spittle from his mouth" (v. Lane s. v. 
p. 2666); cf. also Ethiopic aflasa (II, 1), expellere, ejicere (v. 
Dillmann 5. v.). In view of the fact, however, that Arabic £, as 
a rule, corresponds to Hebrew tt or 2, the emendation tffi^n 
suggests itself. Dr. Grinzberg calls my attention to Talmudic 
t^a, "to discharge, to vomit", which is frequently used, c/. e. g. 
Jer. Berakot VI. 10* 

44. tutsan: d^dth ntym w*o tawrn (§ 36), "She tosses 
her head about" etc., from Vwii (Epstein p. 109 note 19); cf. 
also Ethiopic meta, vertere, reverter e (v. Dillmann s. v. p. 214). 

We add here the three words, which are cited in the 
Responsum, as specimens of the peculiarities of Eldad's vocabu- 
lary (Epstein p. 5 no. 5). The text of the ed. pr. at this point 
is particularly unreliable, two entirely different readings beeing 
found for two of the words in codex British Museum Ad. 27 129 
(Miiller p. 18). 

1. tsnam = b\tb% according to Erankl (M. O.W.J. 1873 
p. 491) durchdringend, scharf, bei spdter en audi vom Geschmack 
beissend, pikant, v. Ducange s. v. See, however, Neubauer in 
J. Q. B. I p. 110 note 1. Dr. Ginzberg, in a letter to the author, 
suggests that the word is formed from t$nvi by the insertion of 
a D and that it means: "dialectics, explanation" (^la^a). Codex 
British Museum has an entirely different reading at this point. 
Instead of the words tTOVT b)tb^ of the ed. pr. the reading is 
D^DIO pip KaaPIK^, "a saddler (or shoemaker) is called by them 
D^pltt", from VoDI, "to cut, to crush, to break into small pieces". 

2. mpn = Tia^, according to Frankl (1. c.) = np*l, the 
name of a place in Naphtali (Joshua XIX, 35). The use of 
mp'1 for mas is due to Eldad's misunderstanding of Megillah 


6 a ''TIBS It npl and consequent identification of npl with TUBS. 
This explanation is ingenious rather than plausible. Reifmann 
would read here XWp21 instead of mpn, for according to Tanhum 
Yerushalmi (v. infra p. 44) pn is used in Eldad's vocabulary as a 
synonym of IBS. To be consistent Reifmann should also emend 
DEX to m^S or IBS. Metz in if. #. TF. J". 1879 p. 187 notes 
that according to Elad "Kurze Schilderung der Abessynischen 
Juden", 1869, p. 23, rekus among the Ealashas corresponds to 
our JlBlto. This meaning, however, would not throw any light 
on the word here used. Herr I. I. Kahan, in a letter to the 
author, suggests the Arabic yJ^ or jylo which means "to be 
empty", as a possible basis for Eldad's use of 11B*S in the sense 
of JYlp'H. Dr. Ginzberg, on the other hand, considers mp'H an 
artificial name for TlS^S or t\)y which, according to Hullin 27b. 
is created out of p\>^ (pp~\Tl p tiTiHW *]iy.) Here again Codex 
British Museum has the totally different reading )b pip "llBD*? 

3. fcnwn = !W, according to Frankl (7. c.) from xovOopuC«> 
= xovftpoCo) or Tov&pi£o>, murren, murmeln, von der unartikidierten 
Stimme der Tiere gebraucht Der Nebengedanke an in mochte 
Eldad vidleicht bestimmen, bei dem Worte wie Kiwn an die 
girrende Taube zu denken. See, however, Neubauer's criticism 
of this explanation (ib.). Reifmann (I. c.) simply emends to 7U1. 
Halevy in Revue Critique d'Histoire et de Litterature vol. XXXI 
p. 462 connects tflflm with KllJvn, giving the following explana- 
tion: fcOlJVn is used in connection with ]^sn {Men. 35 a. KTirvn 
^DD r»P»V Tttbn ptem) and l^Bn, in another place, are spoken 
of in connection with the wings of a dove [viz. Shab. 130 a J^sn 
D^BID by* ywbtito *pJ *)1i p^S], hence Eldad's identification of 

*nnyn with mr! 

In conclusion we give here two words which do not occur 
in our present texts of Eldad, but which are traced to him 
by the older Hebrew lexicographers. 

1. IBS = p2*l: In the commentary of Tanhum Yerushalmi 
on Lament IV, 8 the meaning of IBS as a synonym of pDl is 
given on the authority of the usage of Eldad. D»sy by WDy IBS 

cy 6 J-^ u (3 5-*- ^ ij* */*> JA ^^ W n P^ J^° JH^ 
a..~~J (read ^iljJb) ^yjJb ^i^jo ^Jl ^l^> ^ cr^r?* J^-) 


p Mfc? ^J\ (read <^<*J). (Tanchumi Hierosolymitani Comment. 
Arab, in Lament ationes, ed. Gul. Cureton, London 1843 p. 37; cf. 
also Pinsker I. c. p. 180, Frankl I. c. p. 483, Epstein p. 73). 

David ben Abraham Alfasi, a Karaite of the tenth century 
(v. Steinschneider, Die Arab. Literatur § 47) in his dictionary 
called Agron, s. v. 132 likewise gives pm as the meaning of IBS 
in Lament IV, 8. He claims the authority of the Mishnah for 
this usage. We may surmise, however, that he takes this ex- 
planation from Eldad, in all probability, indirectly, through Ibn 

2. iV at? or rp UP = pDJfl TB, according to Herr Kalian's 
suggestion going back to the Arabic ^y^Uo, "traiter une affaire". 
In his Dictionary, Abulwalid explains ]W of Psalm VII. 1 to 
mean "occupation, pursuit", quoting as his authority Ibn Koreish 
who claimed that he had heard Eldad use the word in this 
sense: Jyo .yljJi J^JI £*— *>\ ^ Bnp p JTTW ^ l***j 

nuiutr ty • *rvb \vm c^ x ^f*- s-^ <3 /b D3Pr kS{? nn naw 

(v. JQftft aZ-ZMZ p. 702,' Sefer ha-Shorashim ed. Bacher, p. 497; 
cf. also Pinsker I. c., Frankl I. c. and »np p n^KD\ Introduction 
p. XXIX, where, according to Neubauer (Z. c), the correct 
reading of the MS. is ^bJl J^-/ — plural, not singular. Cf. 
also S. Eppenstein in M.G.W.J. XLIV. (1900) p. 487 note 3. 

David Alfasi in his Agron gives the same meaning for TOtP, 
probably taking this explanation likewise from Ibn ICoreish. 
Frankl, I c. pp. 494 et seq. points out that in the extant frag- 
ment of his Bisale, Ibn Koreish (p. 99) discusses the meaning 
of the root ,"W without any reference to Eldad. Perhaps the 
reason is that Ibn Koreish had mentioned Eldad already s. v. 
"732, in the part of his work which we do not possess. 71 

It is to be noted that the plural nnP is used in the same 
meaning of "occupation, pursuit" by Maimonides, Yad, Tephillin, 
VI, 13 )Dtn ^nnn WViBn VttWD -rtjn (v. Bacher, Aus dem Wbrter- 
buck des Tanchum Jeruschalmi p. 138). Cf. also Rashi on Prov. 
V, 20, and Epstein pp. 71 et seq. 

7i The fact that Ibn Koreish quotes Eldad as an authority for the usage 
of Hebrew words, has led Pinsker (I. c.) in to the errorneous view that 
Eldad himself was a Hebrew lexicographer; cf. also Schorr in He~Haluz VI 
(1861) p. 63. 




The Ritual of Eldad has been declared by many to be a 
fabrication, a system of Halakot which Eldad invented from 
motives unknown to us, constructing it in eclectic fashion, borrow- 
ing his material arbitrarily from Rabbinic and Karaitic sources 
alike, and embellishing the medley with a few phantastic Halakot 
of his own invention, in order to give the whole the semblance 
of originality. This view arises from the difficulty of explaining 
the sudden appearance of a system of Halakah, the origin and 
source of which cannot be discovered. It is, however, to my mind 
an even greater difficulty to comprehend the psychology which 
would account for the "invention" of a system of Halakot — a 
phenomenon, so far as I am aware, without a parallel in 
Jewish literature. The invention theory is thus a Incus a non 
lucendo. The description of gid in A § 8d — a very frequent 
disease among sheep — and the explanation of it as caused by 
parasites in the brain (Taenia coenurus); the account of the 
swelling of the heart in § 13; the fine distinction drawn 
between simple and purulent pericarditis, and likewise the de- 
scription of cholecystitis to be found in § 14; the suggestion 
in § 19, that it is the formation of a blood clot on the 
omentum of the liver, which makes it seem as if the animal 
had two livers; and likewise the suggestion in § 20 that in 
the case where only one kidney is found, it should be weighed, 
in order to determine whether it is sufficiently large to perform 
the function of excretion normally carried on by two kidneys; 
the method suggested by him in § 32 e for identifying the 
young of an animal; the observation that all species of doves are 
distinguished from all other species of birds by the possession 
of a double crop (B. no 25) — all this, of which no trace can be 
found in our Halakah, cannot well be the product of an individual's 
fancy. It registers the careful observation of facts in the course 
of generations, and rests on the long experience of a people in 
its practical life. 

Internal evidence of this character would go far to prove 


that this Ritual represents a body of laws and practices actually 
observed at some time or other and in some country or other, 
though we cannot at present determine when and where. At a 
distance from the Babylonian and Palestinian centres of the 
Jewish Diaspora, an independent system of Halakot could readily 
have developed. Even in the absence of all schismatic tenden- 
cies, 72 such a system, though basing itself on the Mishnah, would 
nevertheless develop features of its own, having their origin in 
an independent observation of animal life and in the distinctive 
customs of a given country. 72 * In the course of time the need 
would be felt for the codification of the body of Halakot thus 

It is important to note here that in the period of the 
Geonim, or at least as early as the time of Yehudai Gaon (760—764 
C. E.), the work of systematizing religious observances and setting 
them forth in short manuals, for the purpose of giving defin- 
iteness to religious practice, and facilitating instruction began. 73 
The Ritual of Eldad, who lived towards the end of the ninth 
century, may, therefore, be regarded as a Halakic manual com- 
piled under the influence of the tendency toward codification. 

The divergence between Eldad's Halakot and ours seems 
less peculiar if we bear in mind the period to which Eldad 
belongs. The Halakah, particularly in regard to HBW and 
mSHD, had not yet become entirely fixed and rigid in the period 
of the Geonim. It was still in the process of development. In 
Gaonic literature there are Halakic regulations met with nowhere 
else, having found no place in the subsequent codes. 74 The 
sudden appearance of Halakot at such a time and their subse- 

72 To the evidence hitherto adduced in refutation of the theory that 
Eldad's Ritual is of Karaite origin we must add the polemic of the Karaite 
author of the ninth century, Daniel al-Kumsi (v. Poznanski in ZfHB. VII 
(1903) p. 108) against one of Eldad's Halakot (v. B. note 102). 

72» j t j s wor th mentioning in this connection, that among other reasons 
in the Or Zarua, II. § 432 p. 177 the pecularities and divergences of the 
Halakot of Yehudai Gaon from the current Halakah are accounted for by the 
assumption that his halakot originated in a strange land and were brought 
to Babylonia by captives of war, not earlier than a century after his death. 

73 V. Miiller, Handschriftliche, Jehudai Gaon zugewiesene Lehrsatze, p. 4. 

74 Cf. e. g. the two passages occurring in some Responsa of Sharira, 
published by Ginzberg in JQR. XVH, p. 270 1. 21 )bvib nwK *nm tirm *m 
w-\$2; p. 271 l. 1 »ann Ki-aoo *nn vb nam kbw. 


quent disappearance, without any abiding influence on the general 
Halakic movement, is therefore not altogether incomprehensible. 
It is still more intelligible, if we assume, as there is ample 
reason for it, that this system of Halakot represents the 
divergent practices of a country or district somewhat removed 
from the highroad of Jewish life and Jewish religious practice, 
and therefore not lending itself to the ready acceptance of the 
Halakot based on the practices of the Palestinian or Babylonian 
Jewish community. 

Though we cannot say what country was the home of Eld ad's 
Halakot, there is evidence that the place of origin was within 
the sphere -of influence of the Palestinian, rather than the Baby- 
lonian, Talmud. I. H. Weiss 75 pointed out that the Halakah, 
which provides that every animal slaughtered must be subjected 
to inspection for each of the eighteen Terefot, is in accord with 
the Halakah of the Palestinian Talmud and not of the Babylonian 
Talmud. 76 The recently discovered Seder Kodashim of the Pa- 
lestinian Talmud 77 affords additional material enabling us to 
show the extent of its influence. We note here the following 
parallels between the Palestinian Talmud and Eldad's Ritual: 

Eldad Yerushalmi Hullin 


•an mp *nniy «n« p. 3 a. 

bs vh» n mbn im vb\ *b p 

W p rrnn nsan kw mbn 

75 V. Ha-Hoker I p. 162, Dor III*, p. 286. 

"6 7. § 8a. This injunction is old as is evident from the fact that it 
occurs in Targum Pseudo- Jonathan on Num. XIX, 3 ^bib "HDMona rwpTOl. 

77 Published by Salomon Friedlander, 1906. It is immaterial for us here 
to know whether this work presents the original text of the Seder Kodashim 
of the Palestinian Talmud or whether it is a collection of quotations, gathered 
from the existing orders of the Yerushalmi, for the purpose of reconstructing 
the Seder Kodashim, which was lost at an early period. Though scholars of 
authority have produced weighty arguments against the genuineness of this 
work, nevertheless, I ventured to point out the parallels, given in the text 
and if the presentation of these parallels should in any way contribute to 
the furtherance or advancement of this controversy, the author will feel 


Eldad Yerushalmi Hullin 

yxn nan ks nniD n» yTp nn« 
td nawn wwi dm § 31. «^i *w pan P w P*™) a ™ 11Mn 

[km r&iDfi] n»« typtanD k^>i tnia^ pan pro ponw 


win Kin nov 'i ibk p. 6b. 

«DD^n i^sk pnna pte«»i «n^s 
nvn I 11 ! ptamt? tb)yb p ^ni «y 

nam wrwni i&. «^n in pa *n idki nW» pa 

♦7&ynn n^ioe ♦KBtyn "tn»i torn 


inn^s ^m toWa «»id ib. 
rtiDB nam ntswm i&. tnn rrt Din <ton paaiai «»»< 

,mdtc 2 )^ Wa wn 

wm *6a ew dm § 26: 
Aws pirn 1 ? natsn ynt naatya 

tya np&iD ns*B 'i p. 15 a. 
D^man tow ny &w *6 np 
3).naian ^a» ruso 

^>a« [Kim] maa atsm Km dki #>. 
■»a i^a« w iKtet? iy taints 1 ? i 1 ? pK 
^ injn mrra typnn K\m ntown 
•131 n»ts> in 5 ? mm vty maipa dik 

3) Comp. Mishnah Berakotlll 4: 
.'131 too win in 1 ?! unsifi np too 
*ip ton Hint? nam r6am now rr«n ib. 5 

.nsp 11 k"tk p^oa s i6 

4 ) Comp. Tosafot 'Erubin 65 a 
s. v. -ran: -ranan ^ nitonn fern p» 
mntsn n w » run «i ^oty avon nnv ^ 

.tn pa «to 

■■ail p^Dpl K1K W p. 53 a. 

hdk iron nry*? on Dt?a ppn? 
ayto n» nioa 1 ? iidk iron ,&ints6 
p"? pnp pin p Din Kiano^ k 1 ? 
map n«t to (read *ytW) WW 
(»"a *n mjw) p^ *6i miatsi 
*p« t^a mi myn *pifc» p no 
4)tya iai mjn *pi&& lyx 

i) Comp. B. Ratner's arguments 
against the genuineness of this passage 
in Ea- 1 Olam (ed. by N. Sockolow) I. 
25, July 2, 1907 p. 324. 

2) Comp. Or Zarua I (1862) 
§ 368 p. 103: nniKO nm ate kdid ^aa 
]^p^ipt3 nt^ntsnt? wpu ai*\» 'ta^a r»^ 

While the subject-matter of the Eitual represents the reli- 
gious practices of the Jews of a particular country, the form, the 


remarkable systematic and schematic arrangement, is doubtless 
the work of an individual (v. supra p. 25). There is no reason to 
doubt that this individual is Eldad. To what an extent Eldad 
followed earlier models in the work of compilation is the next 
topic for consideration. 

The relation of the Halakot of Eldad to the Halakot Pesukot 
and the Halakot Gedolot has been treated exhaustively by Ep- 
stein. 78 He did not, however, reach any definite conclusion. 
There are striking similarities, both in form and in substance, 
between the Halakot Pesukot and Eldad's Halakot. The deter- 
mination of the exact relation between these two Halakic com- 
pilations is a very difficult problem. In the first place, the 
manner in which the decisions of the blind Yehudai Gaon were 
transmitted, makes it impossible to determine accurately which 
of the Halakot originated directly with the Gaon and his im- 
mediate disciples, and which of these are of a later date. In the 
second place it is still uncertain whether the present Hebrew 
text of the Halakot Pesukot is original or represents a translation 
from the Aramaic. 79 It is, therefore, scarcely possible to do more 
than to point out in detail the remarkable identity that exists 
between the two in subject-matter and arrangement. 

The comparison of Eldad's Ritual with the Halakot Gedolot 
is equally difficult. Here the parallel passages are usually 
widely separated by quotations from the Talmud and by Halakic 
discussions. Moreover, the original character of the compilation 
is obscured and the language is Aramaic. In general, the simi- 
larity between the Eldad Halakot and the Halakot Pesukot is far 
more striking than between them and the Halakot Gedolot. This 
may be readily seen in the parallel arrangement of the passages 
from the three works given below. 

The similarity of Eldad's Ritual and the Halakot Pesukot 
is particularly striking in HtDTO txobft (v. infra p. 51), both in 
subject-matter and in form. It is not, however, possible to prove 
the dependence of one work on the other. Therefore, we must 

78 V. Ha-Hoker I pp. 326 et seq. This article of Epstein has been re- 
published in his rrnps nvUK, Cracow 1893 pp. 20 et seq. 

7 9 This is the view of Halberstam in Halakot Pesukot ed. Schloss- 
berg, p. II. Cf. also Briill in his Jahrbiicher fiir jiidische Geschichte, IX. 
(1889) pp. 129 et seq. (Briill I. c. p. 130 note 3 gives examples of mistrans- 
lations) and Epstein in Ha-Goren III pp. 67 et seq. 


either conclude with Epstein, 80 that both are to be traced back 
to a common origin, or, as it seems to me, we must assume that 
the arrangement, which is common to both works, had already 
become stereotyped. 

Eldad Hal. Pesakot ed. 

Schlossberg p. 135 
et seq. 
nBW mate 

ntsw niato 

nai 'ok § l. 

two s a» ]u p jwiiT 
irnian ■•©» (Di&y p) 
^np"6 naitn to 

*ibk mirr an ids 
iraty nats te 

Hal. Gedolot ed. Hil- 

desheimer p. 508 

et seq. 

yb)n naw mate 

an 1D« p. 510. 

as in H P., '131 H3& te 

.in&vufla bv*b tdk m^nt^o tei^> iidk 
jnewnttin|nfr«i nia^n *on it '■ki 

,rrni»i:in ,rpniTte 

na rwrow § 2. 


nip^i n»in ,mte 

jua nrD rrw 

•>3\n rrw i&. 

nttTwn ^nnnty ^nn*n pa ^an 

«Dii« «n«i rwraa 

nionan ioa«i diik aai 

itio nteaan ntea i« nana "ikis» it ntym p nr '■tei ,tmki 

te vriDn tea i« n^sii irryaw 'i« na s « n&rp *i«is 

njm ntea i« ray '^aatyia ivopaon teai aa^ jvnya 

m« mans ik it te i« ius ^ "hid ^ '« ^t jawae 

ny iTa nnw anaia man nyvs naaty rvh aw «Tan «n« 

nana naw ny^a pao nteaty is wa te tea "M ^ rp*p by 

tea ann -jtei mna majw i« nteai iTa rbpw\ iTV p «rao 

-iijn?a n^aa ^an iTan «ai yaw i« pa ant^i «vao «ap '•aa w 

onn *pit? *6ty ]ity«nn antsn «nty n« antsn «n«i aanriK ""oa ••« 

♦ ♦ ♦ ♦ l^ia nanana^ntsniytio *m ** awi man 

noai npib np*i 2)nna »wn antra iiywa 

naa p nn«i noA noai npT^ np win^ 

teA niDK intt^nty 'iai noa^ 

♦ , ♦ ♦ ♦ «nDna nm«» 

i)Hullin9a. 2) read 


1) omits n^B3» i« 

7. Z. c. p. 328. 



no nwom § 3. 
rftnaon *nm ♦« ,, n 
ran nonnni n«o mn 
&rran nn p onwrn 
nn« oysn po^on ^ni 
Vn«n *6i «\n nom 

Hal Pes. 

pn nana no-n 
b»wi noiy nno mm* 
nnoi nnnnnpnnrpn 
ty it nrm "tomi rnn 
T^vt? DTipi pon 
pDs: iKisn n« kw 
nn« oj?snnonn"i«is 
mtrpn ns irnnt? ^on 
nm nonn nnw i« 
DTp mais pDSii rmn 
ik pD*? iwn 1)1^^ 
nom&6 nonn ya-inp 

noraiKraty pontes 

mm i^k tan arrow 

nni»o toi^ td*o hdtt 


Hal Ged. 

p. 510. The same 
in Aramaic 

Knvn 1 ? 


no ,Tnn6n § 4. 
nonnn omty *? wrty *wi 
tran^ napn p ia r\^h 

■wato Tyn pt* dh^i 
otsn 1 ? -1K126 oxyn p i« 

dki «\n nsno ntyo 1 ? 
*on nsno noo^ ow 
. ♦ ♦ ♦ inntyy non d^im 

pn n^3 min p.511. The same 

DiDi12)nnoom^n\nty in Aramaic 

Tiy ^n ^ao twn 
j-oik po^ nonn 

nyw IK JO^D*? JO^D 

nonn -iktc by nos ik 
tani^i onnno po DinJi 
nan nn tw rtoo i« 
Dmro 1« nnn^ i« 
ty r6oo nni« npmi 
po Dinii nonn ira 
tsntsn n^oo nroa nnn 
bv pD T^ n^P n 
oyo oyo nwsn nnnn 
ikixo it mar *6l 

J) ed. I^ty. *) Sefer ive-Hizhir 

2) Read &m» mo WW ed. Freimann Zev. p. 12 b 

and paragraph, .iBm? 1 ? FflWian n« ymnisf IK 



[na] mraron § 5. 
nbru nyaaa nvt 
nynaa nsn& nbyabi 
♦mints naabi rftru 
ymK jjw vby anwm 
nyna nnna yniyasK 
nnna aintsn nbri:i 
bto «b n«i royasan 
by ibSK «b« yn-i« ty 
D^ity by s« ik tfbfcy 
tai ♦♦♦♦♦♦ m»a 


fioZ. Pes. 

na^n p. 136. 
taiw rriW pa wa 
twin wi nnn nyaan 

nnai * ♦ ♦ 

nna^nana ^tanit^ty 
nnn niynxa yni« 
Hainan n y a & 


■■a nawi p. 141. 
anaoi nana &mt5W 
Diabipa napb nb 

ay nb nnaa 3 )dki 

nb •'styai nana atsnin 

nbyab naa p aiabipa 


Hal. Ged. 
p. 512. The same 
in Aramaic. 

antsn )«a na^n ♦ . ♦ 
w*yh nb anaai «nvn 
♦nana K&t&ni *ma nb 
na nb w *ai 
annn p i)KDabipa 
♦nana «b^b 

na i-mip^y § 6. 
nanan nnnnw kyi 
,Tb:n nipynji nantpb 
rrvpa ik wn ik pan 
pan is pan nabana 
«bi amtyn tonty miaa 
nan «bi ninab ni&a 
in nan rferi iip^y 
lbia mn laty ppyna 
nnatr: nn nnn pma 
nana nnpiya mbm 

naa naiy nw pa 
mn naai nana amtsn 
4)tm«iTnty ayi) nia^ 
npyn: nana bt? pa^a 
]na nn« ib^aKi n^a 
nnai aintw amp 
ntsraai 5 )Oin^ rrn 

i&. The same in 

nynab \-un\npKi 

maaaipa^ananab wbiaa i« «yn«a 
nnpw Tia: aipab tyiai KtsmbiTnnnai 
aipaaiK pan n ^ n 
api nt?*n worn niaj 
nana ^a^D miKt? 

i) Ed. Brow. 
2) This entire section of 
nimn is quoted in Ara- 
maic, in Samuel ibn Gra- 
ma's Bisale publ. by 
St. JZ. II p. 77 e. 
s) ed. DM. 4) e d. ttrm. 
s) This passage, occurring 
later in the right con- 
text, is quite out of place 

i) ed. KD&fcpa. 

2 ) Sefer we-Hizhir II 



Hal Pes. 

DTip ma^D npym 

)n» mK i^bki taints 

ten bpi ik nip ik 

*iid«i npT> ••in i^k 
♦nann nniKD te^ 
In addition to the parallel passages noted in the Glossary 
s. v. t2W attention is called to the following parallel passages 
that occur outside of nBW mn^n. This list could, no doubt, be 
greatly added to, if we had the original text of the Halakot Pe- 
sukot or of the Halakot Gedolot. 


pv niai tyKsonsun §9a. 

♦iro nsia ni^n 

nM t p*iib it m* dki § 15. 
♦ ♦ ♦ ♦ ♦ n«nn ))»b pirn jsnn 

vnyn tonan np^ dm § 21. 
T,ia mK isn dk ^n ynnn k^i 


i»ik nn» ni^> ♦ ♦ ♦ ♦ § 36. 
tbo nrnm K&a ipnm )Ksn np^ 
iwpn dk Kin t^si *]iyni n*nnp ^ 
.toki trtD&m tsntrni n^ten .TT 
rrteii ,tt Bomni twwi ntym 
nrn« mryn dk ^ nnwp am 
taintym D-ten d^t mwp «te 
♦'Di n^tei fy TDyn 

nnmy&n rnnpan pant*] § 32. 
rail D^n DiTyofi d^m pKsr vni 
mK mpfi *w ty oniK o^e 
n&w lniai men dhik DDns' 1 dk 
nmnaa i»k 

#aZ. Pes., #aZ. GW. 

n»m* n«n Hal Pes. p. 143. 

i)Kjnn^> ik nwn 5 ? 

♦nsite "6k te 

nan bw KmK a. p. 140. 

♦nsi& jsnn nnnKt? 

TWK teK JlaZ. Ged. p. 536. 
ytn *6i in ^nyn noaion 


BmtWTi iM. Pes. p. 141. 

ftypp *pn te tb^ te ik ^iin 
nnK' 1 ,)n^tei iniatrn jn^tei 
yw k 1 ? pK^i [vmynsK] p j.Ttei 
nun *pn ns^n np^ niit? mbd 
.n s iip nb ]w ?jn fcoprw ty^nm 

in Knrai Hal Ged. p. 561. 
innte wi yr ■>« m^ nV aw 

1^1 Kinn pBi naw^ moil 

n&wn ntsn Kin khd'o: in wi 


1) Read wnn^; comp. jHte#m 47 b. 

ansra p»i Npnios Knitsos wen nan wi 
and Yer. Kilaim II 8, p. 28 a, where 
Kinn is explained as Npmo. 



g. Mrman s s» (onay p) wd '•bb ju p ywvr uai b« §1. 
2»in&w» ta*6 now nt^w rtefm jw wki ^*nt^ rawi ^ 
^mipyi rpmB^n rrnn^n rrniDm rpnvw xnwm riD^n jn i^ki 

* The original form of the introductory clause which recurs at the 
beginning of every chief part of this ritual, probably read originally nta« 
mYMn '•&& ntWD '•BD JH0W\ as is clearly stated in the Responsum, ed E. p. 4 no. 3: 
nmnan '•bis n»ta , Bia(sic<?'jnn'' niax «*?« nan di» dp p« onto -n&Vn toi, 
further i&.p. 5 no. 6 : (sic "Vjn& *6i na^fa ^jnta j6 ddh m» a T3ta law 
n»& s b» ytsnn^ *bb wb p mbn ^n bik -p «*?« mta^n -tyae k^i 
n 113 in ""SB; e/". also If: yenn* "iBN"i nj>; Responsum (B. no. 22 note 59): nip 
JWIiT 1BX and the Letter of Hasdai Ibn Shaprut, quoted by Epstein p. 70: 
*bd nt?B s bb yann s ■»&» fcp tap p ^>*oany ibik .Tn p ntotow tyivtb nBjni 
mi 3 in. It is, however, possible that the original formula contained also the 
patronymicon )1J p after jwi.T, thus reading: miMfl -fits ntfB ^BB pa p yenn" 1 1BK. 
Such is, in fact, the reading in Recension H., ed. Mii. p. 65, 8 e and in Midrash 
Bereshit Babbati, quoted by Epstein p. 68; cf. also Recension D. ed. Mii. 
p. 64, 8 e , if pa p is not itself to be considered a later insertion. 

It is from this introductory clause, that the whole ritual was cited as 
*"K nir^n. As an old interpolation in this formula, we must regard the title 
lam, found in our present texts, before yenn*, at the beginning of each part 
A as well as in B (nos. 1 and 15); further in M, beg. (v. note 2) ; in Recension J. 
(ed. Mii. p. 65, 8 e ), and also in two quotations from the Responsum, con- 
tained in the manuscript fragments of St. Petersburg and of the British 
Museum which Miiller published on pp. 17 and 18 of his work. The later 
stages of this tendency, to enlarge the formula may be seen in the still more 
expanded forms: mvun ^bb on ray p nana s bb pa p jwi»t U3"i na« (A I G.)- } 

ian*i nana s bb (A. II. P); nt?B n^K s bb (Recension B ed. Mii. 

p. 64 no. 17). 

2 M: ^n& p TibK n wirm ntrm? noto 3iro wtn -jnn iwi nnD 
(i¥S. omits this name in the text, supplying OTI v6k '1 in the margine) 
•n^ nnun ^3 srrcun s b» nana to jwrw "i"«: roa* 'va Kan (This 'n 1 ? is clearly 
a misreading for '^=ton»^; <?/". also MS. Petersburg, Firk. 1261, [Mii. p. 48]: 
(vwnva bKlvb] 'idk now ma^n jnv ir[«i] ^«n»^ naitn b .-niaan to 


§2. ik mik inosNi n^nrin nna 1 ? nnan ^nnn ,iwi <na iTmw 

in m te njrwi ntea in wy te -mon te* in itb nteNan ntea 

mn -jten ninN nam naw nyw ^ its nrran nnm din »\nanB 

•o minn rtSi te Din ^bw Nte jnwnn niirwo ntea -mri ten 

Vwan Dy ^san teNn Ni?i ri5i 6,Nin ^Bia ^ ^ n ^ ^ Si 

§3.&rwm mi nanam inb mn nteN&n «nn» ♦nvi no mnranri 
non rra&i *teNn n^i nvj nam nnN D^sn pa^n inm anan ns p 
teNn nVi N\n nam nnN ays:> pa^an inm ^ansn srp&Nns *d niao 
tsd nsiat? nanai> nan nnN aysn anan ,TaNnB ^ maa non na*?i 
inisb narw nteNan dni s.[n^DB] nvt ^b^i mem am te ^ni awns 
meo naw inN^ nteNa^ waa Nm nnN ayBn antsn a^ays nansn 
a^ayB nniN anan nnm io^y nNism map nteNan dni .team N\n 
Nin ina aw dni .te&n Nin una 'wrapa uprantw] n^ten pi 
*6i ijwn te nteNan nteai ^nj -ijw vtei m*b ^V 10 ^5> myi 
dni antra n^ n^tei n-otjn nnN ays arron pma mn nteNan ram 
Nin ^n dni .teNn *6i Nin np\>n nawn ^mp iy ant? nymn 
teinn te lawn nvp iasm nasn te nteNan nteai m las n«« tet 
n*?i iwi n^DB nannn majn nteNan nteni naw ■o Nin -qd-j ny 
awi my te nteNan ispi iasn i*pw dni ,«7in man teNn 
".tewi una nawn i2m^ -]nmi a^ays ntrten ^nm 

. . . . na» te y^n" na«T nj; ia> lna^ntwa ^d«^ "iid« no s n» ma^n ynr urn (The 
text of Jf is continued in note 289.) 

3 MS. rrnnp'jtt, evidently a false analogy with the feminine plurals of 
the other nouns, but correctly mip^y in § 6. In MS. p. 77b. (E. p. 99) the 

quotation is introduced as follows movi noni Nay p bn^ttty "^no p t6k *?pji 
pmioVn "fi D^n^«. In the MS. the letters of the first nr>m are dotted, to 
indicate deletion; the correct reading follows. 

* MS. Arabic «o. 5 y. Gl. s. v. 

5a St. (p. 297) emends ^stf; a Passive is, however, desirable; *]BW may 
be read as such. 

6 Lev. XVII, 14. MS. omits te and reads iwnn instead of 1BO, E. 

7 Dcwf. XII, 23. In MS. p. 78b-79a (E. p. 99—100) the question is 

introduced by: y&* UTO omiotea ^>«"i^ s Da^ )1DnT «B ]y ^?pi ^i«?« m^ '^ 
and is followed by the remark: ntyto rriji ^nn na«l D«n«) p nnia intona YDSm 
pn ^ayo mm DnYxa m nyten ^a» niwio^«. 

s Supplied by E.; Go. p. XIX and Fi. p. 208a 1. 10 supply njyio. 
9 V. Gl. 8. «. io y. Gl. s. v. 

11 Supplied by E.; St. p. 298 proposed &ntP; Fi. supplied noniM after n^nwpa 

12 7. Gl. s. v. 

i 2a St. p. 298 Kmwi, MS. clearly v^an. 

13 Probably corrupted from nan or ]na^. E. suggests that ]m be taken as 


own pntewon noiDi nonnn isenw ^mw »\n no nwrin §4. 

nan& ntDtt 5 ? tow» dm iwi nsnta nbytb tantyi "d^tAi nip 1 ? Dan 5 ? 
rena Tpa*i nno'oa DipDD ntaaon np t^vn» nan dVim *i*4Wi 

a verb, in accordance with Dan. 1, 11. This is unsatisfactory, unless the text 
be emended to nam. 

u In MS. p. 82a— 83a the quotation is introducend as follows: t6k '"11 

*ajMri« inn s s tyi *■«. 

15 E's emendation Bint? (p. 101) is unnecessary, ^nnn, followed by a 
participial construction, is not uncommon; cf. e. g. Abot de B. Nathan, ed. 
Schechter, Text A. p. 2a: D^BIK nbiy *ma *?a V»ay l^nnn (v.Briill, Jahrbiicher 
IX [1889], p. 135) ; Lev. Bab. XI. 75 ait» ?iBp& ^nnn. 

16 Probably to be emended to V 2 ™, by analogy with Bttnf? nipn p W 
and -ik^ 1 ? osyn p w. 

17 The meaning of this passage is not quite clear, nsy seems to cor- 
respond to np-iBtt (Eullin 113 a), while 1K«, that part of the neck, which is 
in front of the npn&B, seems to be designated by trTJ^ nip 1 ? Btt^. I would 
translate the passage : "If, as he began to slaughter the animal, the knife slid 
in between the oesophagus and the trachea, or between the trachea and oeso- 
phagus on the one side and the jugular veins on the other, or between the 
skin and the flesh (v. note 16) or between the collarbone and the front part 
of the neck, be it the oesophagus, trachea, or veins etc." 

i7 a kti nana, hbb^ Bintsr dki is omitted in E. p. 101, Go. p. XX 1. 8. (V. 
St. p. 298, d). 

i7'> In the MS. the words nip 1 ? mtib n«ref? osyn p w ixnbi myn p» 
are inserted between D^IKI and TVXX. A repetition of n»a in the margine as 
a catchword under nip 1 ? clearly indicates that the insertion is to stand be- 
tween D^M and n&a, not after nan n^tlO (Fi. p. 208 col. a. 1. 30).— This inter- 
polation must, however, be rejected, both because it destroys the unity 
of the phrase imtSW n»a dViKI which has its parallel in B no. 4. n»a tb\* 
pin and also because the remedy given after 'VT n»a D^K1, is most probably 
meant to apply to all cases of mbn, and not only to those enumerated 
in the interpolation. The latter is plainly a repetition of the words (1. 4) 
nip^J Bun 1 ? i«reb osyn pa ik TOato *nyn pat?, omitting, however, the last word 
1ftn*9\. The purpose of this repetition which, to be complete, should begin 
with "iai p IK nip 1 ? Bttnn fa, seems to be to restate the condition, described 
in the beginning, in order to recall it to the reader, before giving the remedy 
»131 ^nnt?^ ntsa nbw. In the MS. a long line is drawn over the words (1. 4) 
D^Ti^ nip*? &wb na^ nsyn p ik "\»n^ Tiyn per, where they occur for the 
first time in the text. This is probably a sign of deletion made by one who 
considered the interpolation an improvement of the text. 

Go. p. XX is very inaccurate omitting the interpolation (and much 
more) without any comment. E. p. 101 gives both passages in parenthesis 
and (having been misinformed by Neub.) adds an incorrect note (p. 105) 
on them; St. p. 298, d suggests a radical emendation: 1K1S 1 ? D2J?n pa IK "WaVl 


r\wnwn DipDn no^a n^K»n isdk naT Bintw tiki ,ttmt^i n\ity 
Kin nsnts ntswn mp»» nbynb n^K&n D^n dki ^wi Kin una 

♦n*nta D^^n ntawn Dip»» nts^ dki 
oip»a nrKi iNftb nivn p nte«»n w^n «o nm& rnVnn noi 
i9 a nsitDn ^nba bs i&te»n xb roi nfeia kvti ktj nmn n&w 

2o,nntswD n^mni h^kd 
§5.n^ni n^ntDD nana nty&^i n^m njnto» km 21 [n&] rrniD'Wi 
n^nj njna nnna rmjnsK yniK ]m? vty anit^m M«rrYin& nt^i 
ik wbw by i^sk k^k ymK ty ant? xb dki niynsKn nnn» Binm 
vftn niKtri k^>i n^n^n nyntA nteK» nyon dki ♦,tto dtip by sk 
nnn» ipk nomnn inn dki .Kin una 23a n «n^ napn j» nnK nanin 

♦nana ^yurn ama rtnwi nyn&n 
y:nK nnn& asn&^yn dki ♦nana 22a^p n-wyn nt^nsy toi 
nnna zsnts^n dki «*Kin mnta nVrran nynton ^k a*mpiann manm 

26,K\n nsits niDmn vmbwz n^nn nymn 
jna *rav d*to pvinp jn no^ kti nmts namn ty i&Kn *si 
unit? 27a nnDnn tsintrn ntoan nynan ty dki nVrun nptsn m ^nn 
nynton nnn» tsmt^ 28n^nD»n unym Din k^ k^i Din wnbw 
t&nwyn ^1 .1^3 dtt Km i^n nnun inns 1 ' -wk nimi 

i« iim , ? , i yijm p» nea noob Biner* dki km nano nty&b ontfi d^t^i napb tsisrm 
151 wwr noa obiKi km nana nap^ Btsnai n«is^ osyn pa. 

18 E. unnecessarily emends OK. 

isa MS. and E ^3«n. is Denf. XIV, 21. 

J9a nnon instead of nanon is probably more correct, cf. Hullin 128 b. 

20 MS. fol. 84 b — 85* introduced by the words : Kin -a *n *aite TT?K il 
miVK. 21 MS. and editions omit no. 

22 MS. had originally m»3, as in § 36 beg. nTBD ne&te nVvtt nynooi, 
but on top of it, as a correction, nnno, so also Fi. col. a line 3 from bottom, 

Go. p. XX i. 18 [snnw K"i]. 

22a So MS., the < being mater lectionis. " V. Gl. s. v. 23a g MS. 

24 St. JZ. II, 298 suggests the reading mpiann; followed by E. p. 107 
note 10, who would also read by for *?K. 

25 E. 1. c. would emend km nETito. 

26 E. Z. c. would emend Kin Tints, in both cases his reasons for the change 
are not convincing. A manuscript of the Montefiore Library (Cat. Hirschfeld 
cod. 108 §483 fol. 48b) contains the following reference, nt&^tsai D^pn Bmtrn 
\yv) D"3i ti^db 'vaa s i»a baK «nn« nyaaa piD nin^Kipa 'W\ nn»a ini3M» nitoipa 
V'2J? "inn m^K. This Halakah does not agree, however, with that of Eldad. 

27 MS. on, not DK, as E. reads; his emendation on is satisfactory. 
2 ' a So MS. 28 MS. rwrnotwij v. Gl. s. v. 

28a MS. n&-BjM E. 


njntsn nnn **]m2 ant? d«i ♦kvi nsia 29 napa i^«n d*Mn istayrv 

p«n rrtn 32mpj;nsi rttonwb n^n^n sinnrovty K*»n n» mip7 §6. 
*6i ttnwn ant? s^ids pan i« pan ro^io ssrrvpn i« wn i« 
sspinD to Din n;st? ppyno jn nan ?rt;n np^ n*n *6i y\mb naaa 

29 MS. nipa E. 

2 9 a So correctly MS., not as E. has p\nn. 

so MS. fol. 100a— 101a, introduced by the words: "S "11 *rb* T&& ni 
rm KD a83^>N (read Kin) mn. The quotation is followed by the words: 

iak&Mk rrim *pnn xsnn nn^pa nnn ^>«::n wan ttsa nnnai (not n&, E.) k& p «nn 

amy papnts^K (MS. Kfiitya) Dibyis [vi, E.] itniy nay v^k (MS. pasted) 

rra wnanaK nd n-empn wnia« nun nbssa n^« «jt»3 npi «w« pa na^a ony 1 ? 

pn «n^B «:^> nm!?« p nvna pkbk ^ "ja taps jkdb^x «nn *s> d s ^ n«yai (MS. rpa) 

button niana (sic MS.) p" 1 «nan«i ty^ anpn p ^>y v«nna»^« D&jn yxnnDs'rK 

nea^K r6aKp& y& k$k nsitn'jHa na& yanp s «?k nafc6a ^y ^pi pV aa«n «:«i «nyea^ 

,-Mk tw }K naa&aanDK k£x n^y^B. "This is the text which I have found in the 
manuscript of Rabbenu Hananeel. I copied it literatim. The expressions which 
Eldad uses, are [not] known to us, etymologically, on account of our imperfect 
knowledge of the (Hebrew) language. God in his goodness and on account of 
the merit of our sacred fathers, however, has revealed (for us) that over which 
we have meditated and we have explained it. And not in this place only, but 
in many places of Eldad's ritual, we have found pleasant original ways and new 
important interpretations, (which are better) than those of our predecessors. 

And perhaps He who gave them (to us), will grant us (MS. vocalises p" 1 ; 

I read p 1 ) a postponement of death because (the problems) are so many. 
And I request the reader who wishes to understand me, to satisfy himself 
with such an examination only which is the result of his thorough study of 
the subject. Then only will he understand that which we have brought to 
light, if it so please God." 

si V. Gl. s. v. 32 v. Gls.v. 

33 So clearly MS.; perhaps a fern, sing., based on the regular plural mvp 
of Vp, "wall", or nomen unitatis of Tp, "a piece of a wall". Go. p. XXI 1. 7, 
Fi., col. b. 1. 18 read mip3; E. reads HTpa and emends mipa. 

34 V. Gl. s. v. 

35 MS. pro or possibly pro; the former being the reading of Go. 1. 9, 
the latter of Fi., 1. 21. St. p. 300 emends pins as in § 5, end, (s. note 29 a); 
E. has the far fetched emendation pya = a^ya, v. p. 107 note 14*. 

36 So Go. 1. 10, E. p. 102 no. 6; MS. mrii6 mm6; evidently the writer 
forgot to cross out the first y\n»b. He intended to write nnx^>, as is shown 


Dial roan ai& irtn "ramm wetn aipan p nrna -jtwai rrtn 
uw *wm ♦sin lints n^ .T^ni stto nayanm te Din ■jisty 
■tippn p ™^ sa^na ii« -p« rrtonn ita^ tok iy itant? 1 ? ^ 
n^nnn vnn vi»tw «sn did inn mm i»i mntsn inttnaw ins'? 
«ortm w 39«nan a to w dni »rtm ra toyr^ tok ny ifcna 

41 !|nsnai Kin -np7 

by the fact that in the second y\n»b he dots the superfluous 1, to indicate 

37 So MS., Fi. 1. 23, Go. 1. 12, E. 1. c. mwnmj St. 1. c. n»Tm; v. Gl. s.v. 

ss V. Gl. a. v. 

39 MS. ixnnan; the dot over the * seems to indicate deletion; possibly 
Hiph. isrn»n was intended; if umtan is read, it must be taken as Hophal. v. 
Gl. s. ©. Go. 1. 17 raoen ; Fi. 1. 29, E. 1. c. (misinformed by Neubauer v. 
Eldad p. 108 note 17) erroneously give ttrnon as reading of MS. The n of 

WJV»n in MS. is indeed far from clear. It might be taken for n ; so St. I. c. 
reads rarrBn, the meaning of which is doubtful. 40 MS. lfcni. 4i MS. 

p. 103b— 104a, introduced by the words: Ntt n«2^X Kin "*> *n ^K v6k ii 

•m After the quotation from Eldad, Ibn Gama c remarks: 5di n*?1p pa Kin 
naKb ernn o"f? ^nn mp^ -bd< k»:k may axn^K Kin ]« ^j> n«jyo (Epstein ty) 
pato ■»& di^>k iKxnaxa iya n^p Wjn aai^Ki ips^K ^aya nTpj^K s ay» ,tb V»j?nDK 

nwta DKanK *S KX*k ^Kpl .d^j?k n^>Ki i. e. in Hebrew, more accurately than 

E. p. 102 no. 6: i^« «"ip" na&x ijwn ni» way ^ai iaits6 (not *pw, E.) non nt 
pnnm iwpn [pioa] rena m^yn puaa ia penem ^a jjp t^"ia tb n"£ia np^y 
«Tn m^na p oa i&ki jnr ^m '*|iaa Din riTsjna nt nh« mai 1 ? ny& R-ani. Then 
follows a quotation from the n^X DK3nK (§ 36) and another remark of Ibn 
Gama c , (MS. p. 104b) which reads: n^KlpK -j^ib (not n^lpR "^IKS E. p. 103; 

nor -j^ia, Go. p. XXI 1. 22) ,TPpy aKa map axa'jK xin ]K. Correct E's Hebrew 
translation: nTpy njw i^x nytsm nt *a iaix nm, as follows jit nan viai nbxi 
nTpp nvt^ iS'k Kin ijran (c/. St. p. 300). 

The reason of *pan Din nTSjj is given in the cases of iTMt? (§ 2), non 
(§ 3) and n&lin (§ 5), but not in the case of 1lp s y (§ 6). For the meaning of 
Ikkur v. Gl. s. v. For the parallels in Hal. Pes. and Ba7. Ged. v. pp. 53 e£ seq. 
Eldad and our halakic Codes agree that "Hp^ may be caused by an impediment, 
which hinders the animal in the normal use of its feet. The essential difference 
between Eldad and the Codes, however, lies in their divergent conceptions of 
the nature of *T)p7. According to Eldad mp7 does not depend upon the 
manner of slaughtering the animal, but upon the fact that the animal's feet 
are not free to move about during or after the act of slaughtering. 



41a maiB xvchn 

p. tsfnan 43^ 42(um) rwn ^» p p ytjnrr 42^on ibk §7. 
inrva BUa pan ^ ^s ty im« rrapn 44 p^ nt^s dti -ibo ^ 
mat? urn rwnsb rrzpn n»b'^ ♦m»na iw ^^n^ y6k B'oa m&n -jato 
♦dtmd ten tow «to nzpn ^zb o^pi ^ant^ vrw •ban mania ityy 

♦msntsn na pp-nn urna nan §8. 
^k»p too i« j^ tob hid to mip npi dk ♦awin man Vwa a. 

"•[wn nana] di to D^Ti i»j; yiyno «nn hid to ienp 'o 
bwvi ty ^>pDn i« *]n^«n i« pan n&nan n« jks nyn nan d« b. 
ntpD 1 ? mp& nantw *6 inanm i&ipDn tdid mo to lonpi osyn -new 

♦«in mnD 47 m» to i»np to Dnvn 
innnn ma to ionp ntyjni 48^ntjo npjw ^^P n PP n innty: D ^ 1 c * 

.*rn nana sonn« nip»o nw& 49^^ 

4i a P has the following heading ^p» ]U p ytnT 1 ? npnai «no s n» nia'jn 
: ^dd nwb rabn main "&» «ii nwo <sd 

42 7. note 1. 

43 MS. S SD1, but correctly ''fiia in introductory formula to III § 26. 

44 MS. Br, E. "W\ D^ is preferable, as it is nearer to the reading of MS. 

45 P in several places erroneously "jl^T, but Piel *]^T is undoubtedly 
intended by Eldad, as appears from § 23 (T-S). 

46 Added for clearness. 47 Sc. Dlptri DIpD lintfi «*?. 

48 So MS. The text is very obscure. The possible emendations are: 
1. Delete ppn, treating it as an error of a copyist, who made the correction 
DVipn, but failed to erase ppn. The difficulty is that Dllpn *ianB>i is an anomal- 
ous expression. 2. Transfer the copula 1 from nwyil to onpn, reading DN1 
Vntw npjtt Dnpm ppn "nn»i and take ppn in the sense of osyn. This, of 
course, is still more difficult. 3. Read "DB> instead of ianB>i on the supposition 
that the latter is due to the analogy with nnntfi of § 8d. Dinpn ppn natf DK1 may 
mean: "If the horn [sc. of a bull] pierced the membrane [sc. of the brain of 
another bull"]. § 8c seems to speak of the goring of a bull, hence the resume 
at the beg. of § 8d: nyn nan «*?l D^y bwq -on»i *6 D«l referring back to 
§ 8 b, mn« nbna nnii «to referring back to § 8 c. A fourth emendation is 
suggested by Dr. Ginzberg. He would read: Dllpn p»n lDn»i DN1, with re- 
ference to § 8b. beg.: pan n»nan n« ]KX nyn nan, although good Hebrew 
style — which Eldad does not write — would require pNa Dnpn "iant?i DK1. 

49 E. separates innnn from pmaD and explains innnn nitt bv ittinp ntfjttl 
as nvfch f?B1i DK. § 8a— c seems to speak of three distinct cases: 

a. perforation of the meninges, 


a.mna nam "nroa «^i ny)*\ mn xb) osy ts\snn *nr»BO *6 a«i 
■a ♦bpkvi pnanty t^V nni« ntantyi nn« ms ty ^nn nm« man 

n^n to d^d hid to l&np ty «s»ai 55 nftnrro mffirro it 5 *;w£n 
♦naiB 5 7,niD to i»np s*wpbw b^nnn ,nton vby jno *a 

§9. a. 

nop «ipi a\nn n*nn -o * pin nn« n»n n*nn mpa *6 n«i 

b. fracture of the skull, in consequence of a blow, where, however, the 
membrane is not pierced, 

c. fracture of the skull, where the animal has been gored and the membrane 
is pierced. 

E. sees in this section five different cases which he states (p. 122 note 3), not 
very clearly, as follows : m» bv l&np DK (J .ynBO npjtt DN (n .ppn -nntw DM (« 
Kin dx (n .npn i« pna in en dn (t .hdo^ tew ia«n ,innnn n»jw iDipDa nmy irx 
.nn« Dip»a nawft 

so Pregnant for nnx DipD 1 ? iDipDD. " MS. nim 52 y m Gl. s. v. 

53 Possibly "itto is to be supplied, viz. tffl mxvfcl&l Dl^a twnn mm »3 
[^P3], D"6t?a. Dlbtya in the sense of "wa does not occur elsewhere in this Ritual. 

54 Y. Gl. 8. V. 55 7. Gl. S. V. 

56 V. Gl. s. v. and read t^p^.tf or perhaps tf^pn*. 

57 The text of § 8d is badly preserved. The following is a mere attempt 
to make it intelligible, nwrma or nirma rm^sn »B"«id twnn .-win dm 
,*wa[a] or [°V] to[ia]& d^ds ma to iDnp ^y KStosi (instead of MS. maima m&ima u) 
to ronp (MS. trp^ff, E. pi^p) p^tr» Winn, n^an vty (MS. jno, E. jm) p*o '•a 
hs-id ,mo. 

If the reading, suggested, is correct, Eldad speaks here of paralysis of 
central origin or gid, a disease very common among sheep and due to Taenia 
coenurus (v. Miiller's Veterinary Surgery, translated by John A. W. Dollar, 
London 1903 p. 111). Translate: "But if the cranium is intact (not broken), 
the animal not having been injured either by the shepherd or by another 
animal, but you nevertheless see the animal turning to one side, you must 
examine its head after it has been slaughtered. If the head {Arachnoid) 
and its fluids are normal, the animal may be eaten. If, however, you find, that 
the brain is swarming with leek colored worms (parasites, or, according to 
another interpretation, given in G. no. 24: that the brain is soiled with 
dirt), and there is to be seen on the meninges something that has the appear- 
ance of meat, when cooked in water (fibrinous exudate, literally like water 
cooked with meat) and when you put salt upon it, gas begins to generate on 
the meninges (lit.: the meninges begin to to seeth, bubble), then the animal 
is not to be eaten". 

So far as I know, this case of paralysis which is accurately observed 
and described here, is mentioned nowhere else in rabbinical literature. 


men rpttoiK r»m M*an dm ♦msnan ^ jwan n«r na o mm 
Msan dmi #im nana nn na ss^t^ dm i^sm n-iwn man im Din 
dm i^s« atea pabn 59HDM ms&h dmi ,iwi nsna rrwn pT mai ty 
i^sm rnnia& an?n man ty M*an 6o D «i ^n nana rrn na 5^^ 

♦nana nn na ss^^ dm 
tfrm* dm .n^ap ty dwi nw ty b6» mmfr eipiM a>an b. 

npto M^D^ DMI ♦KVl HDltD p"6 ^Mttt^D DWH IM ^MD^ p^a tf^n 

nts6tsn jwa nyana Msa^ DMi Aafcn mn nine !?«d»d ntrttsn t^b 

♦mvi nana ^MfctfB 
pTa map nnM nKsaai pans dvw ^Mat^a d^ik ^» msb^ dmi c . 

.to*m Kin nints annn pTa im mn 
G,p. 65 ppt?Mia tsn 6*jn«^ ]tiM pp^a 63 mpm 62niitDpn D^tMn dmi d. 

♦Kin nina ynsM 67^nna "nans 
nn jna ^man 70 l ni * ^ ) ns 69 nia*iDD maapn eso^tMn dmi e . 
jtim^ )nm p jrwpa 7 *KaTD dmi Kin i»a "vnanii ^m^Ka ijnaai 
nt?p so^tm! d^d '^n 7 §^d 7?dmi mn ^am p'ns "ipsa *6i 
84j,Tra dmi Min Tina ainn laa ssjrpva mn m^i 82tikb nnM sijns^ 
s'nnjnsi ijnaa m 1 ?! nna sensing main main s*w2iy rwiM laa 
9i»nsia 9o^itMa sgnipm mainn ss^^ rftaKaa 

68 V. Gl. s. v. 59 E. nniK. eo MS. tyi, E. 

6i P uses the more usual form p«, plural piK, D^IK, nillK, which is 
the Aramaic WW (HuUin 47 a), while G (v. § 9d) and B (v. no. 12) use the 
Hebrew put, pi. WW; ef. also JTaZ. Pes. (p. 143) nmn ^n« Tit* and n<an» -pn-n. 

62 p ^aepn rmisn. 63 p mpn. 64 p p^ p«. 

65 p nrpfffcca. 66 P omits nans, v. Gl. s. v. 

6T p ami. es p nui«n. 69 p pi»D. 7 <> P \\» by pa. 

7 1 MS. in the text m&n, on the margine correctly rrsn ; P rr&n. 

72 £ MS. vrrt-KD. 73 P omits mawi. 

7 4 P omits flTa K3TD; K3TD of (r is a substitute for Hebrew ^no, which 
alone is genuine with Eldad. 

75 P pa*? yi« fa iy"isi x 1 ? dm. 76 P v=d npn. 

77 P D«. 78 P omits "pD. 79 P inserts mn after fi. 

so p Kin «H 81 p ljnsm^. 82 P in«. 

83 P DiWn. 84 P Dn^a. 85 P »nay. 

36 P nnnsniv 87 P nips ijn&n: »^. 88 P n«n»n. 

89 p trpan. 90 P D^^sn. 

9i P adds ^am l6t; 6r MS. p. 125a-125b. The quotation from Eldad 
is preceded in G by the following citation from R. Hananeel (published, in 

part, by St. I. c. p. 307 e) p p^n'jN ]«3 «i« MH nblpl KiVp3 ^«p R bxiin 11311 

pp Vd^ pnD3 n 1 ?^ wmai Vip ins npinsta «|«"iB«^m won^w '•a i^ ^misin^H jwm^ 
Ta D"iiK:n lrnmi n&nei ]mD3 «to «nn v s di nnVn^K vd mnmf?« is mpm ni«D «i« 



f. wp 9^id n«nn d« rfcran 93^ rtpian roopn sapta 9i a n« 

g.njnsai m no 102^^ Qm ^ ^ -^ i+tnpvn mtsp p« dm p. 

•wn nana npnnn nnn m 103^^ [*6] d«i .tawi nnta rp^ao 
b. i04||a^ ^n nsitD n^ npm jnan dki g. 

^Kl^K ^pj »fil ittania tm ^prn^K «in ■•& r6 pBKlfc. The c ^Lrwfc (ed. Venice 

1553 p. 10b) s. v. ''ilK also quotes this view of Hananeel as follows: xbw n""l trvs 

na-iB mne ^pieo pp^yi yu&Ka *ai« [read s 3nD] "are i^bk an ^d an ntep pnoa 
(according to Shittah Mekubbezet on Hullin 46 b) [ik] it *aj ty it nipm DK p» ^31 
ns-iD Win n-B^twi Dy mWKin. So correctly, not as Tosafot Hullin I. c. s. v. 
pnD3 ^>3K and Mdadanne Yom Tob ad loc. are quoting : IK m»n Dy n:tPK"in 
jwton d$? rrwn. 

The quotation from Eldad closes, with the following remark of Ibn 

Gama c (MS. p. 125b, E. p. 103 no. 7) (not yw\) ]Ki nao^K na ijn d 1 ? *?lpbK aim 

$>t ^«iin wai blp"? naiWB }K3. Ibn Gama c means to say that there is no 
traditional basis for the view, set forth by Eldad, although it does agree 
with that of R. Hananeel, in as much as both agree that if two lobes of the 
lung have grown together in any way, the animal may not be eaten. Ibn 
G'ama's last remark is perfectly clear, any emendation would by superfluous 
(against E. p. 110 note 24). 

9i a P DKl. 92 p pan. 93 p repeats pan before nbmn. 

94 P inserts DK before "pD. 95 p inserts Kin after "p. 

96 P "pD DK. 97 p D^Dim. 

98 p pinsy, so also in § 8e (see note 85), either in both places a mistake 

for away, or perhaps Arabic influence, cf. O^yJL*. " P nsito. 

ioo p omits D'pij? 1 ?; G MS. p. 127a, St. p. 301h, E. p. 103, 8. The passage is 

introduced by: ^Kp HSU win Ktt bpi s iK^K mb» 11 and is followed by this remark: 

^rmibK p ]»«3^« pKtn"?K^ (MS. npoin npta) npoin p *i3K ^j>» na -ik-ik na« p&M 
jdsk Yipa k"?k Kfin^DKi p pn" 1 d"?i. ioi MS. and E. nwn. 102 MS. n»e»\ E. 

103 MS. and E. km nana npnaa nnn na wo» dki. E. supplies [njnsa k^i] 
after np^na. The reading of the text seems simpler. 

104 a MS. p. 129b (E. p. 104, 9), introduced by the words "iKlVK bpi "SI 

and followed by the remark: (MS. K311D) nsno bp* Dbl npm n*?lp Ktt ]K ]o: ]mi 

KfcmnK S D ymn^Ki yKtw^K D3n '•B Krmnn s ip"» ko nniKys. "We mean that 
when he uses npm instead of n31"iD, he merely follows his habit which cor- 
roborates our assertion concerning the doctrine of belching and suppuration 
in either of them (i. e. in the heart or the lung)". I do not have, at hand, 

the passages of the MS., to which Ibn Gama c refers. *LL^ (v. Lane s. v.) 

is probably the condition, described in Hullin 46 b as KtfllKT nKn; Ja>iy, or 

more clearly Jai^l , said of the flowing of thick, purulent matter (v. Lane s. v.) 


p.nypn aip^a tasmn naipa on n» ^« D"apa 104a § 10 . 

wfr^bn aip^a D'-pnn fcpa man nap* 

nn«i p^> tik pao io?mpsD ^p i°6tmTi wm BBmn naipa §11. 
one ata« 1« yip rriawo p^> Kin new losvjinnnn apa dm .dvtk 
nta« nopia D'wyn p w pn oy xvaia? nta« in *?na io9n»tDino 
pso pa rvbzM rv ll3 "marpa pSn n2psDn apai &Bna mytani py 
pfcn psoa apa wo .towi *6 «vt nsnta jw i»a ntyyai win pso^ 

♦i?ain sin ibo ^aiD-ma apa *6i 
niaian wgnotrt win rrr? *6i n«nn niais pa npn apa dki §12. 

astf dm h-imb a*?n n«T tok ny a^n ty new atai apa dm §13. 

may refer to the Talmudic *6ai&; </. JBmWmi 55b ^dbi rur-a npn *6ana nni 
M'toa, and ib. 48a "iff 3 D"Ot D^fi ,"1B Kbilfc K^tt. § 9h is missing in P. 

104a The beginning of § 10 is missing. It must have read in substance, 
that, if certain organs are perforated, the animal is n&IB. 

105 Qf. Mishnah Hullin 42 a zbn np^a ,mo to nnp np^a B»in nn^pa 

Disn ppm np-a man nap'a nnpn nnp^a rap-at* rwnn Mn rva^ 

yin"? np^tf niDiDn ivm DD»n .... nnp'W n^asn. It will be observed that all 
the pierced organs, mentioned in the Mishnah (on no ito D^npa), are enumer- 
ated by Eldad in § 10 with the exception of mta to n)y np^a and nnp^atP rw»n. 
These are omitted because they had already been treated at length in §§ 8 
and 9. In B, however, only mo to Dinp nip^a is omitted from the enumeration 
(v. B no. 8), this case being the only one, treated earlier (viz. in no. 7). D13H 
nnp^atP jwaan and y\rb mp*a» nioran n*ai DD&n, of the Mishnah, designating 
parts of the stomach (v. Lewysohn, Zoologie § 62), are here referred to 
collectively in nn^pn nsp^a (assuming this reading to be correct, v. infra). E., 
supposing that Eldad wishes to enumerate at this point all the niB'iB, con- 
cludes that the enumeration in § 10 is very incomplete (p. 92 note 12). In 
this he is mistaken ; Eldad treats here only cases of pierced organs (D^pa). 

In spite of the striking coincidence of the phrases nnfin nnp^a nnpn nap^a 
ppin inp^a of the Mishnah and n^pnn mpa mbn nnpa n:ppn nip^a of this Ritual, 

'n2pn mpM and not ni^pn is, in my opinion, the original reading. In support 
of this view I call attention to the following: 1. B no. 8 has after &tsnn Blip np s a 
the parallel expression m»*in *M3 p rnmn 7\1p% substituting, as usual, maia 
for nip (v. B note 21) ; 2. B likewise omits nrpn nnpa, although it actually 
has pp"in rapa, mtan nnpa. 3. In A itself the discussion of Btsnn nnipa (§ 11) is 
immediately followed by that of napn nmpa (§ 12). 

106 E. unnecessarily emends ttttnn^. 107 y m Gl. s. v. 

108 Probably an analogous form to , awiK (end of §11). Or is it possible 

that the writer should be influenced by the Arabic ^ilXa^ ? 

109 V. Gl. s. v. no MS. nmn; E. suggests XPSVt. 
in E. emends jwnm, since yy is masculine. 

112 MS. Btsnn, E. "3 MS. -man, E. 



D^inx d s d ]wn» «r dk nni« nnsn jw m»a ibn n"?n ty 
d«i Braan u»jo 116 «r 115 t?&n d«i «in ibo w«an *6i mowna 

§14. ne^nsno t^«nn *6 

H7Q« man i^mrrp ty ^i m»a nty« D^n p tei m»n napa 
itson m^ra irn»i runtw dm ♦nBTB *nro6 udd *)^w *jy in np: 
toy lau *a ttoiw Kin ipa in*6 Ksnnm men ap: nasaa dki .naifi 

§15. ♦Ksim K\n nton nsm« bw isua nana 

^ n\n *6i n«nn p*6 pm jam rrn p-n 2 ? it hsd^ dki 
nraaipa niy^n vn ik ^o^nto-rt nmn nn\n is it o^aaw DipD 

H4 V.Gls.v. 115 F. GHL s. v. 

lie MS. *»"% E. 

ii6a Eldad speaks here of a swelling of the heart, due to traumatic peri- 
carditis. He distinguishes between a simple or serous pericarditis on the one 
hand (its^KSn xVi Dvnns n^ins n^o pntwitt NX 11 on, "If there come forth from the 
swelling [tumor] lightcolored [strawcolored], thin liquid, having no odor'", 
v. Gl. 5. rv. nos. 15 and 39) and purulent pericarditis on the other (iwan DX 
'121 BTNan K*? DK1 BPKan U»» KX\ "If there come forth matter [pus], whether 
it be odorless or not"). 

neb MS. naan mrrn, but correctly in § 19 nasn rnnr. 

U7 Neither the context nor the pathological condition, described in § 14 
beg., enables us to determine, whether iaan nw means "the caudate lobe of 
the liver", as Moore (in Noeldeke's Orientalische Studien vol. II pp. 761) 
would interpret this term in the Bible and in Rabbinic literature, or whether 
it means "the lesser omentum of the peritoneum" (reticulum jecoris), as the 
Vulgate and some commentators would have it (v. Hoffmann, Das Buck 
Leviticus, I. p. 165). Either interpretation is anatomically possible. The 
context, in which man rnnr recurs later, in § 19, does not throw any light 
on the question, since there, too, either interpretation is admissible. The real 
difficulty in our § 14 beg., however, lies in the obscurity of the relation of 
the passage Tins'? Uttfi fpjJTP "ty 12 ap3 D« in its context. It seems that in 
these words the condition already described, is again repeated by way of 
resume. E. rightly supplies ll"n before DK1. In this sense I venture to translate 
the passage as follows: "If the gall-bladder is pierced so that it drops the liquid 
contained therein upon the caudate lobe of the liver (or upon the lesser 
omentum)— [now, I repeat], if there is a hole in the gall-bladder so that the 
liquid flows back through it — then the animal may not be eaten." The con- 
dition, thus described, would be biliary fistula. 

The second case, mentioned in § 14 (nt?an nilTD initoi nantM D«), is a 
case of cholecystitis. This likewise makes the animal unfit for food. The third 
case (nnx 1 ? KJnnm nn&a npn n»S»a DN) is the case of an animal in which the 
gall-bladder had been pierced, but the perforation had been healed. 

us MS. nans bv iaur m« bv wu "a. 119 MS. d»\ 

120 MS. bmn. E's emendation 7\Wn is given in the text for want of a 
better one. 


121-pD dp nm *b dk n«Ti jsna ]an nni« nm* jsnn !?« jana 
nrntn nw» myVsn wi dki ^bo myfcjn *?k rotm npm nnm *Ai 
m 58^otyi nniDi mn^i *6 nann dki ♦am nana nmn hidd 

Aain am una nnn 
'id *p^ d« ^k»p» T&y d^pi v^d mtry b^p ^«n niy&m §i< 
dm ♦ilso ^kdpd 'ti j^d 'n 12^ dki .*ibo ^kdpd imtii pana 
jto3 j^d pc ^ 1SDD2 nas^ o ♦nsiD ^«oty» "^ )^d ,-wy raty 

♦am *ibo iwn b«Dtyn 
masyn .d^d ni»syn rrpaty mas Kin *6 i^«n niy^n nTnty § n 
inmi Dsyn nnBu dm ♦nntso n&D^i nniDiKD nana ntyD^i nmriKD 
t-s, i24^3«i j*^i «in* nsio di n| "sod^i ^a^ro |j ^^ pv^n 
niDijm -ntn *6i mm Dipan | {osyn nnnjnnonn *?« nroa nans 
i26TDo^rn «*rraM n« | j«tyn} nanim nsnnm *6i jroipan nnajiy 
omn | {i«}soii 12 9^nm maiy wn dk i^sm iss^sn «m wntma 

i3o.te«n k^i am nana nranina 
*6i i32rr»inio | jiDnnmi isirrjania by nnnib nam nma dm %u 
n|oiy "wran tarn iksdi i^ip-toi issniBntsn rrnvm am inm 
nynn nni« ^nyn bpm dm ^ 7 ^n^^ mh nm» i^nm | ffc6i* 
i*raai isemtsntyi "SjTirwrn tarn inm | {fc6}i D^nan "^nmi p«n i« 
rarw 139 ^ dm | i38^3«^ gin -line laipan naiy 137 nij-ityn} em 
ja i43«in i«ntn rnrc{*»n ipTn*} "^nm i 35 <TTOn Dm no^^ 

i2» -jod an "adhesion", <?/". § 9e (6r and P): ]t1K ni3l»D nutopn DWttn DK1 
'121 }NK ^. E's emendation *pD is unnecessary. 

122 MS. ^Kttfftt 'ai pona 'n, E. 

122a Here begins T-S. Loan 110, p. 1 recto. 

123 p DO iv 124 p omits tew «^ K1H. 

125 p ntPBJ; the meaning of the phrase is: as long as the animal is able to 
raise itself without any assistance, cf. Shabbat 94 a, 1W n« «»W s n; WBl=lt3Sy. 

126 p omits maym. 127 P n»3. 12s P omits temt 
129 p itfam. 1 30 P omits te«n «bi wn. 

131 y.^f n^mto in two places (see following note), but D^non in the 
third place. 

132 t-S rraniDb. 133 p wiian»i. I34 P omits ipm 

135 P ?ni»n in three places, but twice n*r\w in the same § 18. 

136 p omits ^nm «^i, but inserts ifiipoa after no^. 

137 P sin wd. 138 Pnsn. ■ i39Pi3nnn. 

140 P l«»OM. 

141 p^n: «bl n»iy which destroys the sense of the whole passage. 

1* 2 P omits nwi nmt^n ipnn\ 

1 43 «in is not expected, in fact it is somewhat out of place here. Possibly 
Kin is a scribal error for Ninn, a substitute for ntn, either being correct. 
Note that P also has Kin (vfvnn p Kin inn: k^i). 


»in*j una ^ron rpty ^« awinp yim dm | \i*»*miffl b& mnn 

iwi nana inro Dip& usrwc | nsyn b» *)finn p dki i44^f«n 

i*}n« Dsy | nrwr niaoin b» 'fltfn "*nvnnn dm i«.taKn *6i 

l4 8.te«n *6i rcn nana uao i^Ditn j*-jnnii 

19. isiniTin p i« maiyn p i« | )tren p in ip£tfi p isoant? 149^ 
as I Kin ii»ni3 d« i5«rnm^ »wptta' 1 am nan dhm ^ty nn i^mm 
qijo ;m issntrim •'ten "ntoift ^DHinn ^t? jn s issfain nd&*} 
i6i«mn leoonn o^n *joti ij*i*k»; iy | jtswn ^Dn^ i^ntrim tdi 
iiKty {dni | nsntD D«Tjty ^^jy. dk issn^im ^faa d^dh ■ot? ty 

nn« dki ta*n Kin njnia jina} iwi in« i63 D ^ D g<| ^v^g nm p 

«7«B}p'» DIH ,T,T *jm i«1^K D^M *«W «*V{iT HjO "OBB {TB1K 

i63D»y ft ■p^ ipnn ni}D:> {wy}i | {iron mnr i« pm ^m 
♦ ♦ ♦ ♦ D^ty | {*DHn3n naty dki bsw)* «in lints} in« law m« 
i69*DiTi»*} ii«^ ft nsitD nnpjtt 7^2* d\jb6 nnnn , , . i63(?)Dnn 
"low "ote^n «"?i «m nsna ms ksdj *6i ipy ta | ipon ntow dki 
"Oafcw Kin 173-nina | iron n«it» rrenya i« 172 n^D 

20. i-en^D dk {ft} K{s}»a *6i ipnn p ik j«sn p i^dik ^nt? dki r-fif,p,| 
nt? i8o^« ^ i79^ n iw dm i7sni«i^ pun «i{n} nan | i"nn« 
nniDD nis tso^a*} ^ | n ia dki «inn Dift tant^i ibw isi^rnoa 
te | ^« dki im»D my na i« troy ^ dki miDD toy ^>« ^ dki 
d«i nniDD nt^ ^ ^nry d«i ini»3 tr^n ^n ^«n d«i ihidd ^« 

i43a p DS j; n# e. emends nsy.T »" P omits ^3X^1 «in. 

145 p nt ^«2. 146 p omits ^3«n nVi. 147 v. Gl. s. v. 

148 P omits the entire concluding phrase from nvnnn DK1. 

!49 P n«i. 150 p inserts onx after on». 

i5i P rnsiyn p i« nvnn p is. 152 p «sd:v 

153 p p*n\ 154 p om it s mtnb. 155 p ««? ^ d^bo «^n d«. 

156 p qvos. 157 p \V«n. 158 p n^na. 

159 p nt^mn. ieo p nnn. iei p omits Kinn. 

162 p ^ > # 163 p DJ3-., 164 P «n\ 

165 p Ql Ti». 166 P ft«. 167 P «Bps W# 

i 68 One letter after n s 3^ and two or three letters after Dnn are illegible; 
the meaning of the passage is obscure. 

169 In P the whole passage, beginning with D*HM ViKB^ DM1 ^3M*1 etc. is 

170 p omits ^>3Mn M^ K"n. »« P tM. "3 p nnD. 
173 p n^ 3> 174 p omits ^1 Min. 175 p ani^ DIM. 

176 p joViS (Responsum M^D). 

177 O substitutes the short phrase nriM M^M ^M DM1 for the whole passage 
from the beginning. 

178 P omits mM-i*?. no p omits Min. »»o p ^«. 
i8i P omits W1D3. 132 MS. my. 


D^m» «pa '"nvbn | "W 189 ]n s i warn issmaD isT^n m 
i»5Qiri^n f^fiD i94n}n«n moiy 193 «^ D « 192 D^f2T«tt *pn rvp*)taro 
mDv^n 19 Hiitt n{n«n npnn d«i| "ete^i Kin Tina D^:t«»n 
nnsij iitm *"nnK «wta »stf d«i 19 » | ta{*»n *6v*} tfn naie 
203,^^ «in 202 -^nta npnn vby 291 f« 2oonanyD w ^njra m&p 

r-^p.207^3^ «in nine (?){2oe j 205^ ^n&n 204^: dk §21. 

^inan} np s i dki ***&&* *6i Km nsrws | ^lt^n ^ntan spa dm 
♦ [ Kin 2io-nnt3 *in« fran dk ^ sran *6i void 
npnm to«o woo awn} wpn * n*y»n ik w*anpn np^ 210a DKi §22. 
ik 2t4*n*|Bn»n 2l; wn ro»D ik dvj mas 2 i 2a n 2i2nn«{soi} | |im« 

183 p shortens to dVis nianan *?3 pi after nnitt3 PftB b^«. There can be no 
doubt, however, that T-S contains the original reading, for, in the first place, 
the diffuseness of style is quite in keeping with that of Eldad and in the second 
place the Responsum has still preserved a part of the enumeration (cf. B no. 14: 
IVIJ^ Tiny, "h:6 "Hi, btyb bty). G omits the whole passage between nn« n"*?3 
and np*l. 

is* g np\ 185 G nvte. 

186 q ntsnao instead of rt&naa nnsnan. 1S1 G «\n» for K\n -i»«. 

188 # in&niD. 189 G d^i. 190 G omits wten w. 

is 2 P omits the second D^tKfc *pa nvbni. This, although it is intelligible, 
as it stands, may be a mere dittography; G, to make the text clearer has the 
paraphrastic words M» *pn nnKn D^tsn after D'OtKD *p. 

193 p omits k\-i. i94 P nnmv nn«n. 

195 T-S MS. in both places D"J»n against P, G (D'JWn). 

196 p, # omit ban Kin. 1 97 6? imms. 

198 P, ^ omit *?3Kn K^l IW1. 199 G omits H^ KSft" OKI. 

200 a substitutes l^&K for MR fclTO. 

201 p <k ; a omits apro rty y». 202 P W3. 

203 p, G omit ^K"! HVT; G, MS. p. 149 a-b (E. p. 104, 10), intro- 
duced by the phrase ^KpD W\b» t6k n KttKl and followed by the remark Kini 
«:naj? TJD ^ip. Ibn (jama does not quote literally here; he gives merely a 
condensed paraphrase of the original. He even departs from the arrangement 
of his original, quoting the second case of abnormities in the kidneys before 
the first: "iDi nvbi tip np- nn« »b» y* d«i ninta n»nys i!?s« naep nnsi rtna nn«. 

204 p ^ W . 205 «|^ D . 

206 There is room for one or two words in the lacuna. I am, however, 
unable to restore them. 

207 p omits bin Kin. 208 P wpto. T. Gl. s. v. 209 P omits ^Kn Kbi K^n. 
210 Reading in T-S uncertain; P im. 

210a T-S Loan 110 p. 1 verso. *" F. Gl. s. v. 

212 P riKSBi. 212a p omits 13. 213 7. Gl. s. v. 

214 p pnwn; there follows in P a mere dittography of »»n 1»3 13 KJHSi IK 
^nwn, which E. failed to recognize as such. 


iniK 2i5nnnns {ins} | ronTin 214a m»n i« a^na 2i4a^ D2 ln ^| Di 
*6i *rn nsitD siempn ^ nn^pn i« D"y»n t^nn itow 2i5a Dfc< 
as&n dki 2is f ta«n a6i km nsnia dtt man «*&•< jd«i} | 2i7toan 
K2P *6i 22o«sm\n nwiron ai^n^ i{»j« | mpan n«t i« pan n«t 
>n 223^^1 W in 222-nnto | trsnn *6i n^ins d^d dk ^ 2210*] u&b 
♦m&to> 225b^p^ 225a ^n nnpm « 8 mpm d^d ^ 224^^ 
§23.1^ wn ««swpn 22en}»n rron ntoun nmt^ rwni | «n 
{nj;i npn rrty «i2n ity« n»nn to | Dn»iK rn ♦p p ywm ^ ty 
p p jww 227^1^ -inn l an 5 ? trrs nap *6»n ton ny wn ,tyidk 
2so^« 229^^ rrnn p »mh¥i 227b n ^ n ^ 227anonnn dm pnb i&jki 
p toy ty rtjap 23in«n wni nam ain^ \d ♦wn itrn ■o wm | np& 
^ ndd aw 2S2a^ 232^ p toy irw | mod ^ni ainji ijpn 
wnn i»n n« 2 32b^ rp ^ | ^b Sinn» jiysij n«tn p to:i toyrw 
ny ton 234a finty 1 nibp 23i,-pn» tato^ 234*in« 233^^ vn& p^ 

*1K *J1tD}P 237D1DT IK ]1tDp tyW D«1 236^^1^ 235^p pp^i -^ 

«^ ijtik 24onan 239 tow tm ik ^m {238a j | 2 3sn^f ^ 

214a p m 

215 P nnnna. nnnns is an archaism for nnns, <?/". § 13: nbn ^y xsiy dx 
nmx nnsn pnw m&3 abn. V. also p. 15 note 57. 
215a p w . 216 p na-pa ix 3ip3 ix. 

217 P omits ^Dxn x^i ion. 

218 P omits the entire clause, beginning with mn m&3 Xl^ DXl. 
219 P yni; E. suggests $mi 22 ° P xsnnn. 

22 * T-S omits m. 222 P ntw. 

22 3 p omits !?3«^ xin. 224 P wpa. 22 5 P npn. 

225a p n\il; T-S, P have X 1 ? before Kin, which makes the passage un- 
intelligible. Eldad clearly refers to ma (§ 14) as a parallel case, which reads: 
,dik bw 1&U3 nrsna ^ isw ^ .^d« s i «in -1^3 ,in*6 xannni nioa 3pi nxuoi n»\ 
KBnn^i n\t n'jin; c/l note 118. 

225b p U p iD# 226 p n &3. 

22 e a So P. E. emends nii-an; if so, emend further naa (T-5) to nb3 (P). 

227 p » Bfi< 227a p nfina< 227b i n y.^ wr itten above «M. 

228 P omits «\-t. 22 9 P -u^>. 230 p mw< 

"t T-S twice n«n, usually rrn. 232 T-S npan. 2 3 2a P a 1 ?. 

232b p ^ nv 233 p BW W^. 2 34 P in«b. 

2 34a p nawa. 2 35 p mp n«. 2 36 p \&nvb. 

237 p Dion. As the manuscript does not distinguish between 1 and 1, 
either reading, D1DT or D1D"1 , is possible in T-S. V., however, Gl. s. v. Dim. 

238 T-S 7\yh. ny^*, may be a scribal error for Thy, "mountain goat" 
(Prov. Y, 19). 

238a rphg i acuna must have contained the name of one or two animals. 
P omits this series of animals and proceeds from }Ep Dim directly to PTi IX. 

239 P omits bm. 2 <o p mi. 


ftayn tan: vb\* mod inwj | ifctai 24ijj^n yn i«n ubd tona 
vty*} | taafi rrn*} ^3 *w d« ♦ nj; tan m\wb mi& «mn 

244^j*^n*} ta« 243DHO TMH Dil D^IVI in{*1total242 *y. M 

son nts^n ma 345 dk ♦jny tan itonits6 nniD ainn Dvn m:i | Tn}im 
*6i w»o {nitA}ai {D^nn wm ^np | ifsp rrtjy 246*^*5 na&p 
kvi dki ♦nity atanty ij> | n&niB6 niDK mrun *6i dvh *«m nta« 

IK D^ | 250^Tin 249a-|n«^ 249^^ JVl^l 249^ 248,^01 ^1 ITDI^D 

| tidk D«m d^ nai&t? nn« it3mts6 inia wsnfpn 252^1 251 ^ntyj; 
ina 1 ? m^i n«t 255^1*6 mnyna dki ♦mt? 254«^»nty *iy ntamt^ 
pi Diyn ty Kin pi *ny tan 257n&niB6 iniD 256^1 | npnjyrMy 

258,ni^nnn ty Kin 
26ojp« ymm* 259, t # , , , ,-hvtjs ninn nn^ni nnn ta: | w {dki} 
n&{nn ijeik 262nn«n nat? | i&nity{ty t^k m nn iitnb 26i^y> 
*?k itt i&ik I nn« Knni 263mt?n n« innK 1 ? {tidk rrtn n« -iijpynty 
I 265n\Tty 264j^i^ vn t>ki ijroa *i*t3nt^ ta« .ini}« lBnan inn 
tiki intaw Djna im« Dnanpi} 266&itan DTnia m itaKj^ ngnta 

270Qijtyi« 1^18 269Q^ty^1 1H1K 268Di tt nity | 267^11^^ 

Ditoi id*t n« 272-jis^ | d-ib inji^} n^Jo} wi dk 27inmn m>n&n 
275itaj"ii; | it hk 274^3-in^i pKn 273^ ^ m{*i*^i in»tw K*n} 

241 p jas. 

242 There is room for one or two more words in the lacuna. 

243 p omits the entire passage from nnitt KWn {*^JW1 ^nsa »b)*\ to ^OK 
■VWI1 j*liw*}. For this reason the text of P has heretofore been unintelligible. 

244 p omits -ntwr. 245 P dki. 

246 Supplied on the basis of E's emendation of &6l (P) to 831. 

2 47 p omits m. 2 *3 p n^on. 249 p -61. 
249a p -^k. 250 p omits amn. 251 So P. 

252 P omitS D"fi\ 253 P -|^ n . 254 P «^«.^. 

255 p in«. 

25 6 E. would omit m^l; but the words mnynatP ~\nnb are equally super- 
fluous. The whole phrase may be retained as a somewhat prolific repetition 
of the condition stated. 257 T-S ,T&mB6, P lBne6. 

258 p D^nin. 259 There is space for one word more. 

260 P nWn *?13^ «H 261 P te\ 262 p nDK. 

263 j 7 .^ rr-itao, P omits n-\t»a n« ^3t6. 264 p ^^y, 

265 P K,TP. 266 P D «.^ n# 267 p WW?*. 

268 P J<onW. 269 P ^3»1B1. 

27 Space for one word more, possibly cnna or D^ptn. 

271 P inserts ^na after nisn. 

272 T-S "y\V\ P *]t»\ In spite of this coincidence there can be no doubt 
that *pfit^ is to be read; cf. the parallel passage § 6 1^3 mn "p£>t^ onoi. 

273 P p^ # 274 P ^£5-1^1; V. Gl. S. V. V^H. 2 ?5 P l^T H81. 


8b\ 27snni nnirv *6i Dsy A m^ *6 dk I 277j',npB^*rty unsbp. 
to ks»:> d« tas ♦itso fnp rant?* dm 279-,^^ rm^n tain c inm 

.nsito nnu& nni i« Tint? nsy 
§25. *jron b$ n^n ia own ^ rtarto zsoynwi ^nsn ^ w dk 
ten nsn 281 ,«pty on btnw p»» "nan ni»^ *6p vty asoawnrn firm 
n^n tatm^ man on 2 ? jrpff VTD^n pno rm vty I'iwi pi p ytsnrr 
i«n /m np» n« ^kpkp 282a nj; ytnT nr6 -m 282,^nn ara6i 
^ ty "7™** ^ke6 i:«n nans iV roi taai wb ^k vT»!?n 
unn w» nm« ^w Aran ta^n yntos m«i td^> ^ im anw 

♦vnnn nisnsn i^m m ntao^jo pavii nty»te Va&tyn m l^na *6 
ik m?a^ rm o^yn j» ts 28*nt ^ nai w*n ram ntya onia V«t? 
?yna ujtik ^w nn« m ty ^ mitiqi ?nmi 1« pasi ik my»^ 
pn ♦lntsw rfnos 2S5Ksm ^ nt ^ ant^ ton to *a dp6 in 
i»i« nn« am ♦ps^ D^nn^D rn aseonn own wan ^to 11 into 
nnt&n tt ty ^i« &nt&n to rAana nts^n^n n»« r6ann ntown 


276 p ^p"iin. 277 p in&tt&'P. Here ends T-S Loan 110 p. 1 verso. 

278 MS. nn: nm. 

279 The text has bin, which destroys the continuity of the sentence. 
Evidently the copyist was misled by the bin, which occurs four words later. 

280 MS. jnan, E. 2 soa Probably with omission of the N = IKTTini. 

28i Subj. of vty on is nnpn. 

282 To harmonize the Halakah of § 25 with the Halakah of § 24 we 
must here suppose that the "iin is full of water, in order to make this cor- 
respond to the 7-ni into which the ox sank. Another solution is offered by E. 
(p. 125 note 27). He takes the words nm toint^l in the sense of D^mtw IttD 
"inn, and he adds by way of explanation "linn mnvb "WWl Dipt? p*o nn ^b. 

282a For the elliptic use of iy see p. 32. 

283 MS. nt b», E PJPK b*. The emendation in the text is based upon the 
"is nt *nb, which occurs three lines below. 

284 MS. nm, E. 

285 E. reads r6lD5 ksmm IS nt ^^> and explains (p. 125 note 29) IS ^on 

286 MS. "in-ian nnn D^o^n ^-i^ vnar ptn. 

287 J,er. I, 11. Our paragraph is incomplete, for the decision in the 
Halakic question submitted to Joshua or Moses, is not given. The infor- 
mation at the end, however, forms a very appropriate transition to the third 
Part of this Ritual. In the absence of a parallel text, this fact alone, that no 
Halakah is given, does not seem sufficient warrant for the conclusion that 
the text in its present form is corrupt or that it has been shortened. 



mron <bb rwo ^b» pa p ytsnrr ass^m n&« §26. 
p w rmoitan p nttm row n&w owi tent^ to nnta to 
pan nn« dk ^snt^ rpn wmb) ^» tow 289 4 «<m Vub itmn 
p«n on« d«i ibwi nyo pa to wrraa usn tnpn p 291 miD 
jiBsn EHpm arm 293 nm dki 292 prm^ rant?] mt&V tanpro my» 
ntonn "o ♦Dm 2 ? idw mm anpm pssn ana dki 294^^ ^nty 
Kin rotea ovn teTO* to into now nn« dki ♦wn 295in«n mawm 
296nDnv» t ? pty»i rty jnanw vn nnton *a *on id pip vty n^nnpo iw 
paring dd^ Try ^D pton by 297^^1 tjnpa b«n^ to unton 
vn^poi nio» trw oiki maa nifi« t?to wo[yji] 299nnran 'i» «in 
^nnn »w^ ^ nt^s pton nt ^k -din py 301 vr\Tpi dim A 
•mn mta aip&n mn ntyn on« to unto tA p^n d«dw rami 
•p:r *6i tDnt^ dm ,Bnp^ sosj-oBny ny tantw ^Kitrt iid« -p^ 
dm iiflD ttid «to orw dm ♦'wb any mm &w dm ♦win tors 
yyrrb itoi jnt roa^is nym «to tsnty dm ^wb maty Mm ana? 

288 y. note 1. 

289 M (continued from note 2): MS. Va) new &in»<» ma to jranrr -iB*n 
«in bus (lf& *pjD pa rrna pa) *py i« ff»n i« nBnaa pa nawi nr*w (hbtw onwn 
tair «"?. The words ^KTO* to which ikT omits, may be part of the original 
text. Similarly rpy IK and fetf K 1 ? of ikf may be original, although they are 
missing in P. 

290 MS. Dn&to, E. 291 MS. nnso, E. 292 E. 
293 MS. on«a. 294 MS. pftta. 

295 Perhaps to be emended iroc. 296 MS. nanyB^, E. 

297 E. unnecessarily emends *piyBl. 

29s The words natB na^s"? "p"iy ^31 which sound like a quotation (?), are 
unintelligible in their present place and form and disturb the context. E. sug- 
gests the reading of "pTH*, with reference to Berakot 55a, which seems to form 
the basis of our passage: mtBn 3TOn jrt a\Tl « s ny TiX Kto^n un^W ^V T'lKfin 
'1 >\nbw2 d^di nattoa nns 'n ^s'? nw« inbtyn nt ^» naTi a^na^ maji n^B« wbiy yy 
wn^» v»ajn ^«nw Vv nsDD natc o«p p"fifra» pt hi xrmv\ nan nt^x 'ii 'aw 
v^y nB3B m» b». 2 " MS. nattani, E. 

300 E. 301 MS. intsipi, E. 302 Ez. XLI, 22. 

303 MS. l&nff^ and UB^, but in all the following verbs the singular 
forms are used; likewise in M, which reads: b& ni&V IJ? BW» S (M5. &b) *b\ 
(MS. omits «w) mt\ buz n^na «^>a Bin^ djo (MS. nbwm) n^sn b^b'? empn 
nawi ynt nas^B pnmn &b b«i ^wb -itid v^v [rrn] 8^1 anty nxi te ony oint» s dki 
ibax ^a Bin^b iidk "?a« ^ws Bntri. 


305i«te^ iy tsint^^? )b )■«« 304^« [«im] n^nn new *on dm ♦tos 
vto 306nsnip» d-ik to wjn n\w ppnn *onn nawn o .itoa ^ 
at? nto«i o^joto nnnn ity itynn km nnwn "d nw in^ nrn 
1*6 d*o n&wn by yin* y» yiv n\T ">bb 307^nto vs ^ nnwi 
nn^> so9v«»[t>] -^to 'pw nn« dm sosd^ Dvito^ nmt 'dik mnnn •rto 

•inn v '^ to-rcr^ ntoj*6 "»e*o — *on imp 1 ? 
ntow .n^na n^w pm nn«on mbto pn j^nna y»n n»m 
j^mp rn *6i nitons did nisnnn n« 3 10 ppnD rn o^nnn Tn nntan 
nn^n dv nnto ny ton Din^b miD n^n*6 to« .]mpn nyn dk s n 

3 "j»w ni« ton 'nnn 
§27.n»nn nai .miiwa nann it to *onn 3i2n^n*6 toi^ tsiwn to« 
n&nnn dk ♦Binan jn» nt \s 3^^^^ v>^ n ^> tw n»nn mwo 
D^t«i wy Drw D^«im D^ty D^*nn -rum d*w D^ann njnn 
pan \w .dwi "oty n^ tw nannn 313^^^ ntontw typm m^to 
]n"»i 3i4^p^p ty ^p vnn^ DHpnpn pp^i ntotri r^™ n&a^ town 
dm ♦jrton j^n t?*n to n^i Dint^i 7^1 nn* orw to rr 
n^w «n dm ,toan *6 [Mn] nana sien^Byn ^ p^yn sisq^ 
3i6ts^n ik m$ 3i5a n -^ ^innnn town dm .ton tod pinn irton 
jwi twin awn in inn 8«}wi awn njrr Mn dk to« .tod 
town nyY *nn dm to« sis/p^ t to Tto *"» 71W ^ w rni 
3i9nnn mw v,,t ^^ , » '^ itysi 1 ? toetyn n ,r npn "ot «to ^s^ ^ids 
sao^i^D nnmi *)to ^d ^s^ 'ity 319a ^w mt «to i: 1 ? j^d d^h^i 
to»tyn npm toDiy ty«m ]w wm d^ni ^ty 321^ tw n»nn '^s^ 
32ia t to«n «^ Dn&yn «b n^os to»t^n n^n') 

304 MS. ^ONn n^nn. sos MS. wxtrtp, E. aoe 7. Ql. s. v. 

so? D ew ^. XXVII, 7. C/ 1 . Pesahim 109a. 308 ^ J. XXU, 19. 309 E. 
310 MS. )npno. sit D ew f. XII, 15. 312 MS. n^3«B. 

312a (y Tar gum Pseudo- Jonathan on Dei^. XVII, 7, Menahot 37 a. 

313 MS. I3TO\ 314 y. Gl. S. V. 

315 Probably to be read DTi, Piel. 

3i6 MS. inBnayn and o-eyn. 317 MS. nn», E. 

318 P S . CXXI, 5. 319 ^^z. XV, 6. 320 p s . XCI, 7. 

321 MS. "ft E. 

32ia The ritual distinction, drawn between the left and the right side 
and the preference given to the former over the latter, which is unexplicable 
to Epstein (p. 125 note 28), is closely connected with the mooncult or the 
mooncycle. {Cf. A. Jeremias, Das Alte Testament im Lichte des Alten Orients*, 
Leipzig 1906, pp. 29 et seq.). In the mooncult the West is considered as 
Kiblah. In accordance with this Kiblah the Temple in Jerusalem was con- 
structed. Even as late as R. Joshua ben Levi, who lived in the first half of 
the third century (C. E.) it was believed that the Shekinah was in the "West 
(B. Batra 25 b) and the "Oiyft bnft was generally considered in the Midrash 


jfrwb n»ai jnip 1 ? ^ds wbn wi d^t b6p rft tw noro tea § 28. 
idd W2ftt6 wn w^n torn n« d^t ^t*n D^n tsAt? 76 w n«i 
p*6 ran *6 TTttp «\n dk tea ♦ban *6 iwi rAras arrow oAnn 
^do w^irtr nonn dki ,nm 3 23DAm mrro *piy d nwnwb nmo 

D'fciBa Dnym*o rrnmriK ty ro^m no« ^dd nnin *«n dmi § 29 - 
trasn mtan&w "in*6i ♦d^ rwi&ff "in*6 n^w*? inio niton nnn b» 
326 a^n *pyT y\v^ ib^ r n 325 ^» ^ D ^ *^ nn i™** P mn nn ^ nfc; 
ynmn dki ♦nana n^nn te «nsyn nnn& y&j&n dki ♦nana Ad -itynn 

♦itso d^th b« iron p mn 

as the permanent dwelling place of the Shekinah (cf. Bischoff, Babylonisch- 
Astrales im Weltbilde des Talmud und Midrasch, Leipzig 1907, p. 106 note 3). 
Now then since the West is Kiblah, the right or upper side, i. e. the North 
is favorable, while the left or lower side, i. e. the South as also the East 
where the Shekinah does not dwell (B. Batra I. c), are unfavorable. Eldad 
bases his theory concerning the favorableness of the right side or the North 
upon the fact that in the Bible always the right side is mentioned in con- 
nection with God and the righteous. (Similarly Plato [Leges IV. 8] connects 
the right side with the Gods and the left side with the demons.). The same 
view is found expressed in the Talmud as e. g. where it refers to the placing 
of the pillow-end of the bed facing toward the North as a remedy against 
miscarriage or as a means of bringing about male birth. (Berakot 5 a.) 

The right i. e. the favorable side should during the act of slaughtering 
be upward. This represents not only the view of Eldad (cf. also § 25), but 
also that of the Shi'itic ritual (v. Van den Berg, Minhadj at-Talibin, III. 
p. 298). Now since the right side of the animal should be upward during the 
act of slaughtering and since a Kiblah is required by both (v. Van den Berg 
I. c. p. 299 and S. Keijzer, Precis de Jurisprudence Musulmane selon le rite 
Chdfeite par Abou Chodja, Leyde 1859 p. 55), it follows that only the "West 
can be considered here as the Kiblah. 

In an older form of the mooncult the North had the same significance 
as the West had later (v. Jeremias, I. c). Just as e. g. the Temple of Nippur 
was facing the North and just as the North was the direction of worship for 
the Mandeans, so was the North according to Eldad (§ 25 end) the Kiblah 
for the Israelites in the Desert (n"inn»B V71 mn DW imaa ^mt^ WW pn 

322 MS. pto, later correctly p*6. 

323 MS. D^Tl; the change to D^ra is the only possible way of retaining 
the clause '131 *piy »3 in its present place. If D^TS is retained, the whole 
clause "131 *piy <3 must be inserted after too& '131 D S T V&V T\b WV nam ^>3K 
n^D*6 ntwi p-ipb. 

324 Text mV\ E. mtos. 325 V. Gl. s. v. 

326 MS. D^3, v. Gl. s. v. yoifc. The disease of which Eldad speaks here, is 
congenital arthritis, caused by an encysted abscess which, in turn, may bring 
about a generalized infection. 


§3o.ty nDin rurrn 327nna-in&i? rrtoi w »tt mnsw n&m dki 
ton dni H*n ntoos y& t ktt dk ^i«n ty ^nn *6 •i o by y^n 
n« rpiia ntfwn m«i ♦n&w t ? ibo town ton dk to« toe p»tj 
■vina ]S\rm r by ipsini 328 DW tn[V| rrrym xb) i»ipon KBpvr nsyn 

^"idb iton D"»»ion to pip^ to« m»»*wi 

§ 3i. n^iDs jnison rtoofi dt:q n^os «oioa rAiDS nam nowm 

nrwi dki • BinB6 inio isn mtymi mntoi 330 norm to« • ^m 

mifi^ *6»w iy &irw *6 ijsn t» nion ono t» rrcw t» mrwn 

liDD ^n ipd <a vb t\w d^ibp vty myp iro6 jpt tbi nay ipy 

*% nyiyipto ma nto«oni uyjnni vti v»ty 33in^mpD K\n *6 injni 

332,te»in *]inn s *6 ^ naian nt^ntyn «nn ^ it» ntosen jmn 

§32.vm D^n dh^dd Dnton ptrer wn 334nn:ny&n rvapan ^\prw] 

a ormn man nm« 3360^5"' dm 335/v™ nap» nt? ty Dm« d^tid 

♦nnintaa iom naw 
b.npn Dnin m^n ikbw in^n *«p tornt^ nto«on j^ ^ to« 

327 7. Gl. S. V. 328 MS. QnMKi; V. Gl. S. V. 

329 The Halakah of § 30 is very difficult. It treats of an animal, one 
of whose legs has become shorter than the others, as a result of a 
fracture. I take it to mean the following: If an animal has [at one time] 
fractured [one of] its hind or fore legs, the fracture having been caused by 
its fatness [i. e. the excessive weight of the body had occasioned fracture, in 
consequence of a violent contraction of the muscles after some sudden or too 
vigorous action], so that the animal walks on three feet, one foot not touching 
the ground, then if it [the injured limb] is the right fore-leg or the right hind 
leg, the animal is pasul. But if it is a left leg, the animal may be slaughtered. 
After it has been slaughtered, the leg must be examined in order to deter- 
mine whether a firm, bony union has taken place (lOlpon KSp^n DSjm n« ^pTn) 
and the fragments of the bone have not moved from their proper place, 
(nTjmn according to E. = Targumic inyfiK, v. Levy, TWb. II, 203) and 
whether the sinews occupy their proper position on the leg. [If this is so], 
then the animal is "itso, [for the case is one in which the fracture has healed 
properly]. 330 I n e. non is marked as doubtful. 33 1 y m Gl. s. v. 

332 m (continued) : nnx 1 ? \p\ to ono to (MS. ntwtn) nttfx to ntrrwn tdki 
n*?« hi [ed. Const, nut] Kfttya tnoim m» n'"> vbww ny ny:m mv 'a v*?y nay» 
riTilS ]l^ni xb) nmnn. Cf. also the anonymous quotation from our Ritual in 
Jacob (ben Samuel) Zausmer's, Bet Ydakob (1696) p. 25b, Resp. no 25: bn« 
^a«n b"w nnsty p'tb yp na» ]\tr ^pa nins npri'rp T'nno ••j.iion ^n^n 
'"idi nm n'" 1 DTp nv:m nity nwa& p )ptm. AVith reference to ono no^nt^ it is 
worth mentioning that the heathen Bulgarians considered it a crime to^ eat 
the meat of an animal that had been killed by a Eunuch, cf. Haberland, TJber 
Gebrauche und Aberglauben beim Essen, in Zeitschrift fur Volkerpsychologie 
XVII (1887) p. 356. 

333 Supplied by E. 334 MS nnmyom, E. 
335 MS. -irW. 336 v. Gl. s. v. 


mrm oma ■warn imsn ybwr\) am» «r ita>« iran *jnnm ntea&n 

po.ruvm a^tey a na ksoi maiyo »*nwn ms arw m« »»p»| < 

D^teyn itea™ D«n nmap w ama mte rwpn tok nv nn&w ny 
«im mw. 34 °inD a^an n^ omty mm mm npa nos wjni 339 ^«n 
343«in Dir newa mm »JBW»T&m 342 in:ifco» om« aty s^Dmty »w 
'nti 346,Dn^n:a itefcM i«a s^m^n riai owsm antjon 34 4n« nan 
naT -w«a nnniD a^teyn wi 347 tntya» ^a*6 irv nt?« ny tea 
p*isn mm 349 *u&k now umntaiaa urna an™ .yijn h» ***&» 
i«a omit? 353 D rn« «^in ssantairny 351ms ^o» *a 36onat? man 
mm nte i«a arnwi rvaa i&y 35 5DW n vn tdk mn^n M«ri* 
nn« ^»tyi 359ono te*6 sssna ar6 357^ s56,dmd Dntwm D^teyn 
Da'? v« 36iDni i»w wat? nisiyn ^« wd seonvw nyw Q^nsn p 
365nn 364D^nta an ntsmp iten 363D-.teyn o 352 ono ^a&6 nrcn 
Kb nwy D^Di ^Di noa i*n 3«mn y&t? tpk p*un 366 ia h 7 Ak*w^ 

337 Beg. of 0. The heading reads: "1 bv WSHB ni^m «Sb3 nttfyBn nt 
vs\ nte. 338 omits mrm. 

339 o omits from ib»n wbiyn I'njmi to ny nrvni. 

340 o inBi. Of. Misnah, Hullin 74b : n<:ty won p i^bk now mw bt-i 
'Ol m»a tnim. See Frankl, JftfTFJ, 1874 p. 414. 

34i O omits DiTfy »nn* MUR1. 342 omits VtftnB. 

343 D^l «^ Q*3 ?PJW, a quotation from 6. Megillah 18b (= Tdanit 12 a, 
Ye&amoi 54 a, JVidda/i 63 a) '121 -vn «"?i vn n^i *bn nv -»« m nBK -on w DiB:nB. 

344 o omits n«. 

345 O omits rujrn roi n^nsni and has instead briefly nwym. Our passage 
makes a carnivorous animal of the ostrich, the food of which is principally 
herbs, seeds and fruits, although it eats also birds, reptiles and insects and even 
small mammals (v. A. H. Evans, Birds, London 1900 p. 29). This information 
is, however, no more startling than the report in the Midrash Shemuel ed. 
Buber p. 51a (= Yalkut Shim'oni ib. § 123, 'Aruk s. v. noj») that ostriches 
(mi^jn) were fed on the body of king Agag (I Sam. XV, 33). 

346 o Dri^li b*& D*K3 (Jellinek's copy vntw) mt>m. 

347 o omits from rrm to nnBttB. 

348 o tfj^i b* (Jell.'s copy ^«n) bt>« s nn "OT "i»«3. Frankl ?. c. i»*o 
injn ^« t»"« t^Kn *an\ The reading which Frankl gives as the text, is pro- 
bably the proper emendation. 349 UOK -pots ttmn&(?) irby. 

350 p omits rav. 35i O Kinn; Jell's copy omits mnn. 

352 o omits n&inff. 353 omits onw. 354 ^ ^ « 3 - 

355 ^tynn vnw. 3 56 nraion niaiym. 357 n»n «*?i. 

358 nwn. 359 n^nn p t ?i3« 1 ?. 360 nana iwnt». 

36i O omits on 1 ? now. 362 pB. 363 omits D-^j^n. 

364 D^Tin&i. 365 omits nn and inserts ^D^ before b&"\VF. 

366 p nn. omitting it altogether, inserts instead 

367 b"» after -Qnn. 


torn?} •o irra no« rrnn 371*3 rinn "onrt 369jm^ tnvi sesmpn 
i^» vn» a^m 374qi^ 373 i»« 372-in^b m^i ^n ib) na&e K2P1 nana 
jo 377^^ 376Dntyin *6*6 rrapn jna *6t? «*n»n wn Dai iba 1 ? 
nn« dki ♦hstd 379^^ nitrn ibd i»d D^ayn ibo mri 373^ i^nn 
382Qi-inn nt?K ssi^i ssono Kin *6i Kin no •o imn no« am no 'din 

3361^ 1D« 385a in « DipBn D^? 385^18 "ITIH 3S4^1ttH? JY'^pn 383^11* 

a^n la 5 ? t«i nvfan n^n 3 38^^» -idk 887*^nbn 387 ia^> Tnm Din 
idd -6in— j-ibd it?« Kin nr la 5 ? Tnm nan 390 ^ ids a^n 389^ 
392n»D Hint? nt ^md* aniD inasn t?k n»D \nn Ka V« iDKaty 
397nj;n 396^ m ay i^n 395^ Tnyn 394^^ im« 7^0 393nn«ty 

§33.mpo "6 40l r« ^ 400^1 399niofi k^i nnpa sgs^nty t^« dki 

ses p omits n"spn. 369 o has '« = nn8n Din before jnia». 

370 O omits *vrb. 371 o mint?, omitting 'O. 

372 P nn8, but 374 D ^. 

373 p D ^-i qib^ v» s nn8. 

375 Probably to be read Kin; omits it altogether. 

376 has mtn after outran. 377 ^3^. 
878 nnyn before *bv. 

379 O 8M mtsa "itynnw; Frankl (IfCrWJ 1874 p. 550) sees a reminiscence 
of the \T&n ^n of Abaya (Eullin 75 b) in this story with its two miracles. 

380 p no 8bi imn n&8. O Kin n&rs 8^>i Kin no -o inn n&8. 
38i o ntwo. 382 y. Gl. s. v. 

383 O omits W1K. 384 p mits ^K*i»^. 

385 O omits W1K. 

385a cf. Midrash Deut. Bab. IV nmi n"npn *,niK nDK d*w nmn nntaiK lan 
nn8 Dipan pvim. 

386 O on 5 ?. The original probably read la^y, in accordance with the 
following forms; cf. also § 33c, end. 

ss? nnb. 

387a O/; ^inan n8 "j 1 ? vnnn, Midrash Tanhuma, section Shemini, ed.Buber, 
p. 15b; YaZfari hn-Machiri on Ps. CXLVI, 7, erf. Buber p. 142a; Yalkut 
Shimoni on' Ps. /. c; Lev. Pa&. chapter XXII, 10; Frankl, MGWJ, 1874 
p. 414 would erroneously substitute here tn of B. 

388 p )yfy nrh. 389 omits fy». 

390 P again why, O Mb. 39 1 Num. XII, 12. 

392 P omits n»3 8inB>; owing to a homoioteleuton O omits from Kin HI 
nbm nt33 ni^K to iok nnito. 

393 o nn« i»k. 394 omits n^s 1 ? mi8. 

395 p po^i, O pia, Frankl ?. c. -paa. 

396 P omits 1138, E. 397 P nyx 

398 O taints. 399 O omits na&o. 

400 n^in. 401 vki. 


xb) 405UHK \n» *b\ 404 u&& toa*6 typm ^unty [im«] 402^^ 
n«i ini« w 1« 409Titjn 408^ id« 407fcnty isy« nnjn i&nB6 406^ 

^n ^1 niD» 417 «^ 4i6nm« Titans ^ nmi 4 15 nnpi mmyD nam 
421&CP wm 42o«i^d to*6 typn« lis^ttfl im« ^iA t$>pn» 418^ j^i 
423^a^ty A rron 422a ^ n « mm i&mwi *6 rroaiK rriwn mmrwb 
ron 428^ -irn» 427^1,-1 426 *6 425^^^ 424nninan •£»} rw» •'so 
♦lmrtBD id« maw *a into«ni ^p»n npip ty 429 im« 
nn»p nn» iroty ^oDrrnjWDm Drmn»BD Drrop&» n^nwn toa b. 
rat tunown n«t iro6 43110*0 bod i« io«n *u i« ioki toy i&rwi 
437 on^ 436nm 435, nm« bmb '*rvbn ^«i yaw 434^ 433 wm 

402 b*W\; Jell's copy VlM, which E. emends to HJM, Frankl, (/. c. p. 551, 
note 1) to bllb. 

403 o nn^y mi8. 

404 int?3 ^ish 1 ?, P ono nba«^ Wp3i, which E. emends to nno. The scribe 
intended to write UOO 'jlSX 1 ?, as is evident from the same phrase: t?p38 0^181 
1300 bl38^>, where O again has n«Q ^38^ (v. note 420). He was misled to 
write DHO, through the immediately preceding nnoy. 1^38^ is a mistake for bvsb. 

405 o omits mi8 *w 8^1. 406 o bw rvn. 407 O oin^tp. 
408 O omits 'MP. 409 P "IW. 

410 i e y. XXII, 28; O omits m n«i ini8. 
4ii O omits 8inn. 412 O *vtbr\b. 

413 O omits \)i p. 414 O -1081. 

415 o omits mpi 416 O rrnomyi. 

417 O 8^1. 418 O "JW. 419 o d"?18. 

420 O VW3, v. note 404. This seems to be the reading of O's source and 
differs from P. 

421 O ini8 vnm = vwvnm, not, as E. suggests, =on8 mmon, or as 
Frankl, (1. c, note 2), rm8 vnon. 

4 22 o lorn? 1 ?. 

422a q om its from moi8 rronni to in8. 

423 p uyaew ib WiS\ which is to be emended to lyottftf, E. ; O *b Win 
nvm "SO onyo»»: Frankl (£ c. note 3) emends: — \b 1081 ytPirr 8 ? jw.T "TO^no 
nyattw — mn. 

424 O omits ni13Jn ^soi, but inserts the words \b 1D8 after ntPO; subject 
either "in8 n^bn or Moses himself. In either case we would have to read 
Unb "108 instead of I 1 ? "108. 

425 o lam* 1 ?. 

426 p a^l, E.; O TnO -^8. 427 O omits 81.1. 328 O ^>38. 

429 o iron. 430 O nn^apoa Dn^ioni DnnotPi. 

4 3i O omits 1081 »33 18. 432 O 10n»W instead of WWII n8t. 

4 ^3 O 183. 434 O "?S8. 

435 O omits nni8 *?8^ r-i^n bw. 436 O no8i. 437 O i 1 ?. 

■wn ropon p 44onr«n ia&w •owi 439^1 ty «n Vru wen "swmai 
't3i« nn« TD^n *42 # ninD«i D^n ubwi M'unov nnaff nno nana 
nntjoi n^an d^idd i»ik in« Total taK* *6i ipan «in ^ids 
|imn Ayn ♦man Dntwi nino^n ni^os 'oik ^td^i mn»«n 
444niinn ^sd too •bo p jn jwijrr ivnp iok wap p y^im 442at,a T _£ )j 
446mrm* tontri dik 445^0 dk irai | \****nm *w*n\x iwb\ 
bm 44sm« Dvn l {loTOn} *6 now nn« 447a wn nn« nnn 447^^ 
tdki inn'' i^:A tid« ta« in« ovn ita«n »b im n 1 ? 449,-6^ 
innty 4 52VJ5B virr 451^^ orriK fo*r xb) ina jn 1 ?^ ty 450^ 
453p wn bwm »b n^nn m ***.)&* ^ n n na iwan | *6 riSpri 
4S3,n»Di n»D nn« ty ran | own pn tos ta« iok atao 

C. I YltV to*0 TO31 454b n ^^ 454aV*irP bSW 454^ m ^^ 453b ^fl 

457 n«t "mKi top or6n inrrspm 7s n« 456prnm "watt ^nnn 
458t^aVn$ n^n i,tb n« pnr te« j*ra jot* i« n^n | dni ibo tow 
46i,Tim dk 46onr-n^i TOn town to« Ab | mio 459^1 tow n«t in«i 

4 38 O tt'Ol in consonance with the omission of CTtoVn. 

439 O DW Wtfy xon. 

440 omits from 130W to I3ra», but 

441 adds nn3BD after ninbKI. 

442a Here begins T-S Loan 110 leaf 2 verso. 443 p omits ]13 p. 

444 O omits the whole discourse from *mx Vtbfi to mDtfl , BO and inserts 
p (Jell.'s copy *p) before 1b^. 
444a umK. 

444b omits n»o. 445 p nST. 

446 T-S nw«, but in all other places ninim 

447 y.fif mim; in all other places 0^331; P 0*331 nmtrs; ninoxi 0*33. 
4 47a omits »n. 

448 L €Vt XXII, 28. In the quotation begins with rw IX nitfl (the rest 
is missing), in P with 135 nxi mix, in T-S with i&ntwi $b. 

449 T-S r6i3«; P *mx ov3 ife«n tab n^3« «*? nmn -in *6 !?3«. 

450 So T-S, cf. p. 32; P VlDt6. 451 p omits DW3K. 
462 O omits from biw "11D«1 to Hff D^tMK. 

452a jja.^. XXni, 19. 453 p rn . 

453a omits from 3 { ?nn nt to the end of the paragraph. The meaning of 
the clause is: It is not permitted to boil the meat of a suckling (iok 3Vno )3) 
in the milk of its mother; how much less is it permitted to boil such meat 
together with the meat of the mother. 

453b p t, n a. 

454 p ibn ix. 454a omits vnn- bm\ 454b p nrn:i w# 

455 p omits from ^ n»31 to n3"333 ^nnn. 

456 J 7 -^ y-<nini, but in all other places, the Qal form is used; P y mini. 

457 p-p. 458 7. Gl. s. v. 

4*>9 P omits from 3*?n 0X1 to 1t2>3 b^H\ 460 p n3*33 IX. 

46i P omits n^-l ox. 


464ynT sb d«i irrs pm 46 3d«i mrr 4 62nm« ^d*6 v\b\y nno« 

466,nn« nyn orris «5^i vty iid« | irr»a 
dp n}r itnn myn:n nn« Dm 46srpjm | D«n «7j tt ntyi «n d«i 
4711^ -itson «in | ^s jrtwn 470^ w Ton tome i« 459*0 n{t 
vty p^ny&i iron «3a a ^ 473pjn|tn inw &nwm «"tew hVi 

477^ 476Nin HD 475^6 ,^ 474f|«fi ty "D^ fc^*6 H^ HDD | ^1p 

dm "o vty y& «in weri 480^ nipjsn | 479 D « 478in« Dm itan^n 
nw 484ini« 483j^i^jo i ^snpyai 4 § 2 Dn^ 48im« ty^ 1 ? d«i nipte 
| 488monn 4 8 7 nittnnD w\r\)x 485^-.^ ni»m *6 ^ dki mam to 
•rconi 492^ji B n TO^n tyn}V wnn 491^^ 490^^1 489m»nn to 
-io{k dki ft iniK} toi^ na k^ 494Q« tsnwm | BniEft jnr 493^^ 
djjs ^opft &n|Bw jnn mmo-wro n^nn | 495^^ nty« torwn i&w 

496j|,niT^ | i^s«i ntrntsft «in tooa n{^y 
r-#, 0. 1 DTOp& 49sn{pn ^« iki:p nnmi) njyi&n pomp vn 4971^1 

■pi 5oo* D n*^T fy «i^ ato} mm pi rontmn pa 499Q^naw 
mn «^>* ann*} o^a | sos^^ ^n -raw nynn *<»rpri | soidki ^a 

462 p ^3^ omitting DniK. 

463 p Q K> 464 pm\ 465 p ^3^. 

466 p -iriK; omits the entire § 33c. 

467 p itjnp pita d*o; torn? pn dki. 468 p rum. 
469 P, omit K3. 470 p typ. 471 omits to. 
472 p, omit ^ *6l. 473 o omits jnta WW. 

473a p ? r pt, D- 

474 nTO. E. questions the correctness of the reading ms& without 
sufficient reason. 

475 p omits the second u6. 

476 O omits Kin n& 1K^ altogether. 477 ^n. 

478 Xev. XXII, 28; P again quotes from ua ntO ini8 on: ef. note 448. 

479 P, O DK1. 

480 T 7 .^ an d Phave *b; paraphrases, as follows: tanun inapfifi n\l DN1. 
48i T-S nn«^> pw. 482 p ugnp; tantr. 

483 p jnM)>&; D^avfi. 484 p una mi«. 

485 p again jnMyo; O wy\ 486 y.^ omits WUt. 

487 omits morofi. 

488 T-S n^nnD, a dittography. 

489 p and nana na:3 nana. 490 omits wi. 
491 p vny m 492 omits b^Sn. 

493 O omits "\wani, but has Kinn before ]r\V, reading tomvb ]nv «Wn ^W*&n. 

494 p inserts i6 before «Sfi\ 495 p omits Janty i»«. 

496 omits from "'U tnw DK enwni to the end of d. 

497 o vnttn. 49s o s yn. 499 o )^ni»i. 

500 O '•"V; E. supplies nt, evidently reading nt ^l 1 ^V. 

501 O nx. 502 o omits ,Tn. so 3 jtrcn ^«. 



nyytyn | nt?K ny {507*m&ron *D , ony *D , ynn ?*n^nn ^m *nwan pa 
sio^ni nbmn ♦DfrrniD'w nn^D ww ni*}«i D*ryn sosjypnnn 
traT jirnfey tei mam} ipam 5120^3 trat 5 ud,t;wk | ma^ao 
np* 1 ityj« B«napon p aw 1 ? su&ppy *ipk | 5 i 4 (?)»**» rrni s«Ynro 
♦D^n rip** »b 5 i 7 D« np 1 * dki ^dk n p^ ^af? Q^n 


iTn r\d?r\ 

trroain {'•sa ™» •£»} y^irr wni 5* 7 dk 
34.nvm 52ii^s« "o-itysn | nnn& napri ]n: mas n*ro 519 «n 
523HD 522^^1 1 i«m )wti 52l Mp«i iiam •osn ^{« *)n* i^«n nvnnp 
mfi n*}^{*y»*} toik n{*n« «n* sa^iarf? *i^«n *nvna} a won 
\ik y» bm nra ntyai hdib d^ibd now nn« | no-is oneo! 
n*}ia ia tr ♦onnn pa ina vnx fowio ••a*} | ms-v-^ «,rn« nmfi 

so* I n y-# there are visible after the lacuna the letters arm which I 
take for the end of the word jrpnn^in, the « being a copyist's error for D or \. 

505 r»3, according to Frankl, I c. n&X 

506 VD3D, E. ^DIO. 

507 The lacuna has been filled in on the basis of the inferior text 
which reads: ]KM n^ny on o^nm nin&Kn p (In Jell's copy, according to 

508 O vnnn, probably to be read Tnnn = nvnnn of T-S. v. Gl. s. v. E. 
emends nnn, the reading of Frankl, I. c. 

509 A mere attempt to fill the lacuna. 5io O trtym D^mn. 
sit O ]rrni&N. 

sis O Dl^D Wiv ttb, which rendered the entire § unintelligible. 

513 o mv in\ 

514 Three letters KV(?), are still visible after n\"l1. 
514a ppn. 515 oipttn. 516 Hb). 

517 O niHISK. 518 O TOK. 519 to. 

520 O inserts ^»TO^ after nvinB and reads TOM instead of TOin. 

521 O ^"BMl. 52ia r.^ ^ 

522 D eu t XIV, 5; O abbreviates the whole list by a "m after mam. 

523 nbv 

524 According to O the lacuna should contain the words DV3r6 nrm 

KfeB DK line DS nnn, in addition to the words n« an, which are suggested 

by the following TO1K n . . . However, there is space in the lacuna for only 
four words. 

525 T-S leaf 2, verso. 


ma man noi rw Kin * t\* 5 26 | -p^ tnr D^ysfD 

527,1^ n*nn -i&j>k nwi to npnn i^bki | ipnm ]«sn «in jrao |^d 
ntDfitoi nty&to | 53 0D^i to« 529^n 628n«B& 527a n ty&to n*yw 
nnb) on | ma n4y» jstym mn«m tonn nan now nn« d«i ,"6 en 
K<n nawD 532n«nn nt ssinonaiD jy« ^nn^ M<*rfowb& ttw 
n^ty frxb v« ittmri ^«m •osn roeo dhidd ipnm I j«»n "'ton 

534 4 n l ?y»to D^rty 1 1^«^ jw nty&to 

537q^j; rrrap 536jnn -nt^ n« i^«n nvnn 535^^ i^bki §3? 

T-& -w dm l^aai 539anTtya ma^i raDnoi nana || 539^^1 1 sssnninfc 

54©to« tbd to»tna jwm pana rrwy ipnm |«sa | wtoee D^ton n 

•tow *6i «\n nm& l B4ipm ty D^ton | nrpnt* *pDv too^n dk 

526 I do not know, how to restore the text, as the whole passage *3 VW 
W Kin is very obscure. 

557 omits from DVDr6* 1^«n* nvm to 7\b n*nn 1»K and introduces the 
words ntyfc'? ES^AHV by the phrase )b vr* DN. 

527a nbynb. 

528 O KfiB. The whole phrase of 0, KfeD, ntyfcf? D"r» \b w DN seems to 
be taken literally from J2 (v. B, no. 25) and substituted here as an epitome of 
Eldad's text. 

529 O omits K\"T. O alone inserts the words U"!& n^fc*? D s rt? ^ p« D«1 
after KftD. The rest of this paragraph is missing in 0. 

530 T-S D^ni Eldad seems to be of the opinion that all game animals 
have tusks. In Rabbinic literature we find pa*a ascribed to the camel 
(Eullin 59a), the dog (Shab. 63b, B. Kamma 23b), the bear (Gen. Bab. 
chap. LXXXVI, 3, 4) and the lion (Targum on Psalm L VIII, 7, Joel I, 6. 
Hebr. mynto; cf. also on Job XXIX, 17, Prov. XXX, 14) 

530a cf. Eullin I. c. 531 Sic. 532 T-S K*-a 

533 The K is suspended in MS. 

534 It is possible to retain the second n'rjJDVtt, if we take the whole clause 
to mean that neither sheep nor cattle and the like nor gazelles, rams etc. have 
upper teeth. As simpler reading would be nDDbfi. 

535 omits JD». 536 DJ-Q. 537 asyn. 

5 38 T-S rnanto, in all other places correctly mwnD. The copyist may 
. ^> j, 
have thought of Arabic ^£> ; c/., however, Tosefta, Kelim B. M. IV, 15. 
tfinisn *Tin» "QlMtf mart bv rrnSD, where Levy (IV, p. 451) and Kohut (VII, 
p. 275) emend nam 

539 End of 0. We have rejected the words D2J>n p^Dl "QWB WXD DK 
%HpM IDiy. If we accept E's emendation of p"D1 to pwi, they seem to refer 
to § 17: 'iDi nana on n dd^i wwa *mk pwn inmi asyn intM dki. 

540 T-S 1^S«, a repetition of the preceding i^B«l, which destroys the 
sense of the phrase and contradicts § 16. 

541 T-S ^KatSto, clearly a dittography, cf. § 16. 


5 42ipnm proa inpu wa* -iimo ftwi | nvnn mo bw wnp aps dm 

5*3,nvnn ty my p 
1 36. n&»to r6na nyntoDi nana nty»^ n^na nyntoa | n^a i^b«i 

547-pbo rnpi^o «\m antyi nvrn bry | 5 «Tp^ ^dn j ^ S k 544^^3 r-s, #. 

5ola n vnni 55lKfcfc I 550-|pm ]«^ llp^ ^IfclK HflS 548,^ 547a^ n 
«TT 555 "Wpn DK 554«in 55Sa^^ 553^1^1 iYfVW ^ 552?^^ 

559^-inni d^dth ntyni sssn^^i B *' h B i, aDni 557 &ntyni | sserr^m 
w mwp «^n setnnis mryn dk seo^ | nnwp am rrbrti rpT 
x>« 564njDna tai *on *)iy *o | rrtn ty sesTiayn bimm ^d^ti 

37.^1 nj j «i »67p^ i^«n nvnn | p arrown r-s. 

in^itr ^p ^ ntyy n» | nay Daiy i^s« m ksit 

vwitsn nm* *m dm j } dot | utyrn Din no^i two 

ony *i*«w sin ^d-o roia J } anyi | Diny ses^^i 

wi laan | •wi ty rp&rnw rrnn { } owan no 1 1*6 

b*n I Din fy "tan «m jm Din ^ido { } man Tinn 

542 "T. § 8. 543 y t Gl. s. v. 544 Cf. § 5 a. 

545 Beginning of G. 

6 *6 6r Ipy. The reading of the verbs np^, mpljHD and nip^y in T-S can 
never be certain, as MS. does not distinguish 1 from 1. I follow here the 
somewhat clearer reading of MS. G., v. Gl. no. 30. 

547 <2 nittD. 5 47a £ om i ts m . 548 £ n »f?l. 

549 6? i«. 550 & omits ipnm. 551 q -udk. 

55 1a a rrim 552 # -, nM3e 553 y_£ ws1J?1i 

553a y; @i # s# ^ 554 ^ more correctly K\1. 

555 and (t, both iwpn S %t ? DK; the negation, however, contradicts the 
following rwmp om. 

sse q omits rrtoii. 557 (5. tain^m. 557a y Q s# v< 

553 & n»«n. 559 # innni, v. Gl. s. v. enm 

560 E. emends '•S^. More correctly ""B^ might be taken = ""B^ in the 
sense of ^yzb. The meaning of this paragraph seems to be as follows : Since 
game animals turn their heads around and rear on their haunches, when 
they are slaughtered, in spite of the fact that they are bound, it is certain 
that they would rise up, if they were left unbound, while they are being 

56i G nntyn, omitting nni«. 

562 Q nnnifi, instead of D^J-n DH' 1 rrntyp t6a. 

ses t-S T»ym. 564 q n &ra tei w fav 555 a moynt^. 

566 G MS. pp. 104a— b. For the introductory and closing remarks of 
Ibn Gama c to this passage v. note 41. 

567 In the lacuna some thing like a B is visible; probably the text read 

•■1B&, cf. Harkavy, Responsen der Geonim, Berlin 1887, § 32 p. 12: *ibw 

rain pnt* [nn» =] ^oai ]m. 568 MS. i«ori. 


pa b* in« «n Ton pp* { * • •jam m»j»o ntson 

fnT in i^> { } d^t inn jo "°ii-pnDnD , 'i nay m | anp 

blDB { |»M DM1 ISJp liTDD'' M1HH D^&n I^SM D^n I^H 

^Btyl { * ♦ ♦ ♦! nDDn nipri inn *6 "0 d^ms | mow Kin 

"l/iDyn I 1HDD1 ID! n« 

Awi nvnn J *np*}nni nrpm nprrja ta« §38. 

*nn*}Dn fnpnm* *np*}^m nvtan | npnn 

♦ ♦ ♦ 1 572nn« n^fco tan ^fm&n npnni* 

♦ ♦♦♦♦♦♦ 

sea MS. pana. 570 y. Gl. S. V. 

5vi Lev. XVII, 13; <?/". .HuWw 88 a noa^ w D*aa*a laDS* bw inD3i pan wn 
nsjn id£ mi^n *ten nx rtpj ef. also, JETaJ. (M. p. 569 (nnn mdo roata). 

The sense of this paragraph is clear, in spite of the fragmentary con- 
dition of the text: If a man wishes to slaughter a game animal on a stony 
piece of ground, where he cannot find a handful of dust, he may take his 
garment, burn it and cover the blood with its ashes. If he has two garments, 
viz. a coat (nil) and a mantle (^n&, v. Gl. s. v.) he may burn one. If he has 
only one, what is he to do, since he is not allowed to pronounce the 
blessing, if his body is naked? Let him tear his garment and let him slaughter 
the animal upon one part of the garment, while he covers his body with the 
other part and pronounces the blessing. The meat of the animal is then fit 
for food. Afterwards, however, he must take the garment to a sandy place 
and squeeze the blood out of the garment and cover the blood with earth. 
He must also wash the garment and cover with earth even that water, in 
which he washed the garment. Under no circumstances, however, is he per- 
mitted, to cover the blood with stones, for the Bible makes no provision for 
such an expedient (cf. Hullin I. c). 

It is more than probable that this Halakah on nnn '•wa, as it is found 
here in Eldad, is the one referred to as "gaonic" by Isaac of Vienna in his 
OrZarud I (Zitomir 1862), mn *ID*3 tYOhn § 396 p. 109: p^BKn KnD^IKJJIB^ 

ijv^b nny yam anaa K3<« <k noam am nan pnw 'an nna nam noam wVo *pw 
-rajP«> * 3 i n «nvn n\in ik kbijj mnn ant nan kw *6n a"« «nvn p ix kbw p 
Reifmann p. 280) K'pnaaa i"« N-n33 atan 's»m tonts^ pan ">b la-'pn 
emends j6anoa, corresponding to anaaa) Kfian^ 'BO 131 Din MD'3 ^ n;nattl 
a»n n^a ^on (Reifmann «fenD) «^naD K'nn i« «n^3 n*nn «on s tnsjn 
3"p .on p nai»3 p onn wa pj>^ pan ia^pn '•am nans *6a noam M'tsa 

.DUiaan pp 1 ? 

The passage is also quoted by Mordecai on Ilullin IV (onn "nD'a), 
beginning D^aiNJn wis^na 3inD mjH; c/l Epstein p. 96, note 41. Reifmann I. c, 
on the contrary, considers Mordecai as the source from which the "old 
forger" drew. 

572 § 38 is incomplete. It must have originally stated that nvn are 
treated like nifcna in regard to all msno, with the exception of nip^a. For 
the content of the rest of the ritual, which is missing, v. Chapter II, pp. 23 
et seq, Chapter III p. 28. 



iny\ar\ "»fi» nm ^so yaw i B wn i»« i(§i) 
!?id*6 iid« nt^m? niD^n jnv law ^ant^ mm nam tqii to 
jpnwwi rrnrtn ^rrniDTi irnrw in&TW rvobn jn 1^1 ♦inawo 

♦Dinns •'Dm rrnipy s^nswip 
ntoa&n tea i« mik io^«ni namn on«n rip" 1 *a lna rrmw 2 (§2) 
»wt> oft r6iDa nins hew »*« nr\m ms ty itid tea 1« it» 

6 *o->tD iw 

1 The Halakic part of 12 is introduced by the words : nrbv TiWTW nia^na 
D'HOW. For the meaning of the various types, used in this text, v. supra p. 21. 

i a V. A note 1. 

2 A § 1 bvnwh nam fe. TON of B probably to be emended to rtM». 

2a U here rPniDYl (c/1 also no 24, note 83), but in no. 3 regularly 

rrowYi, nomn, nram. 

s As the text stands, it would appear that n*D10^p should be considered 
a special ntDTltP nal?n, which would enlarge the traditional number to six. 
This number, however, would be inconsistent with Eldad's own statement 
(A § 1), for according to him there are only five ntrrw nia^n. Moreover in R 
Wtobp does not occupy a place as a special nBTltP Jia^n, but, as in A § 5 it is 
only a subtitle under nonan. Undoubtedly the word iTDI&'rp here has been 
inserted by a reader or copyist, probably the same one who consistently 
changed the reading of B no. 24 (0) wonan IK litT^rm IK inDlin IK inntwi DK 
impjHi IK inDliabprn to impjm IK lnona^pn ik, thus again adding Dita^p as a 
special 'w mbn to the traditional five. On Dlto^p in this text, v. note 15. 

4 raB by *niD bsi 1K; note the conformity with iM. Pes. p. 135: W 
ras bjmiD 'bzm, while A § 2 has ra^y by mion ^aa ik. 

s A § 2 mriK nona ntrnt? mjr»a np; in B there is substituted the less 
characteristic 'Ha of Mishnah, Hullin 32 a and later Codes. 

s mtSO 1U2 o^on OKI does not occur in A. It seems to be a later addition, 
inserted under the influence of some code. Hal. Pes. p. 135 (= Sefer we-Hizhir, 
Lev. p. 12a) likewise read mtw m ^jptw nn» K? OKI; Ha/. (M. p. 510 
(== Sie/%r we-Hizhir ib.) mtw p^B nWB; Maimonides, Yad, Hilkot Shehitah 
III 2 ntoos inBTW m H3ta riinB nn» DKl. The very same phrase occurs also in 


3 (§ 3) nn« Dysn nma mam tcb mn rfaKon mrr dm ina jwram 
♦nomn kti n«r ^nni ann *»m "jVnn 'osyn ty ntowon ntea 

.te«n k 1 ? *wk ntaan wn mrbzin nnom 

4(§4>nta«»n ^isn did rw*n nntyro noron mna dk in» rrnnin 
ntyate n&w dn aiimun pi tstrin pa mmon nto«on «nm rrty 
ni?3«on rip dk iopn n»D d^ik »r»nb nnn» n&w dki anaia 
2I p->jnjt5 wps cf> ♦ntewo^ mp& jw natron jsik fy Binm n&ip»» 

5(§5)D« rms n&aii rtiDfi ntyrin nbm nystsD ma UiTiwitfi 
tei "rite mnan .mao rras ty ls^ns «te nVn:i nyntsa twi 

♦ntoos 15 did^pd mtyyn ntrw 

the "Sefer ha-Oreh" (sic, not ha-Orah, v. ZfHB. X (1906), p. 169) Hilkot 
Shehitah, § 116 ed. Buber p. 155 (=Pardes § 212 erf. Warsaw 1870). It is note- 
worthy that the entire § 116 of the Sefer ha-Oreh, found only in Azulai's MS. 
(v. Buber, ib. note 1) is taken word for word from Maimonides, Yad I. c. 
Ill 1 — 19. This fact has been overlooked by Buber. The passage is poorly 
edited both in Buber's edition and in the Pardes and should be revised on the 
basis of an early MS. or of the ed. pr. of Maimonides ' Yad. 

7 The text would become clearer, if IN were supplied before nbaKOn n*?Si 
DXjjn by. The word WSy refers undoubtedly to a bone of the spine, the 

meaning of the whole phrase being: " or if the knife be applied to a 

bone of the spine." E's copy reads njwn after dsjj which would destroy the 
sense of the passage. 

8 Bead nn^aatt, subj. nnD'm i. e. "such a procedure would render the 
animal n^aa." 

9 Ed. pr. mtw, which E., according to A § 4, rightly emends to nana. 

10 So ed. pr., perhaps npyn has been lost before pan; E's emendation of 
n»a for nftD is not satisfactory. 

11 n&Tlty Dlpfta N^tP nnanan nspa DS has been pronounced an interpolation 
by St. (p. 300 note 2). In ed. pr. this clause is separated from the preceding 
paragraph on nn^n by a space, extending over an entire line. The Halakah, 
"laid down in these words nana nWTW Dlpttn xbvt finnan nnpa DN, is inconsistent 
with that of no. 8 (= A § 10) bmn »b) ton n^ioa nx^n *aa p nnanan nnpa. 
E. p. 90 note 5 seeks to harmonize the two passages by limiting the applica- 
bility of the first halakah to the time of slaughtering. In order to retain 
this questionable clause it would be better to adopt an emendation which 
introduces only a slight change in the obscure words. Emend n^K&V Dlptt p*» 
to nteK&n mpa pa s i. In this form the clause ntrn» oipea *bv nna-ian nnpa ds 
nsnta is the necessary complement of the opening of the sentence Dlpft pai 
n^3«ttn. For the meaning of pan which occurs frequently in this Ritual v. 
Gl. s. v. 

12 Ed. pr. rp&m 13 y. Gl. s. v. 

14 The passage is obscure. In ed. pr. there is a large space between the 
words nnt?3 and n&nan, indicating the beginning of a new thought. Read: 


rrton Tpym yp i« "irv i« p« ^« nam!? nant? dk ina mmp^ 6 (§ 6) 
»b t&x rbim ^n n«t nnpiya n^m nnoi n^« pan *6i amr6 
nta nawam nation kit did ik ito Din k:t Dna pn dki ta«n 

16 *t>£->v ©»:iw npM oft PTOSO pim Dlpa ^K Dlpan 


n *6i man Din Dan *6i ^nntya ^^ ^ m&n ^p in ^ nnD 7a(§8b) 

mo to ianpi laipaa naiy osym man ted twin «2P )w dm b.(§8c.) 

^♦n^ios Dty naw jtob i« m»» man Dana Din dki it?a laipaa naiy 

^in pa 2o n ^ an napa tmn niip npa \yir\ npa man nnpa 8(§io> 

noTiff bai [nbiDB] nb nonan or h^db mo^ipa mwyn now ^a [ja]i n|a noian 
,-6idb Dio'ripa iTwyn. 

!6 Talmudic-rabbinic Dlto^lpD (Hullin 30b and Codes) is a later substitute . 
for Eldad's to*j» (c/1 A § 5). Already St. (p. 298 note 2) bad suggested that 
in B no. 24 (note 84) inDlfc^pfi be a substitute for [VJnEPoyn. 

16 n&no DWODn npy: nxi. The Talmudic use of npjtt is entirely different 
from that of Eldad (v. Gl. s. v.). Steinschneider (p. 300 note 2), has 
properly pronounced the word an interpolation, wherever it is found in the 
Talmudic sense. In this Epstein has followed him, v. Ha-Hoker I p. 328; cf. 
however, his Eldad, p. 92. 

» This section of R is introduced by the words : Dn"?t5> mB"iB mabn "pi pm. 

17a Subject DBJjn. 18 Subject mo bv lonp. 

19 I do not think that 7b represents a new-case, not mentioned in A. 
While A § 8c explains the formation of what appears like a tumour on the 
meninges tracing it to its pathological basis (v. A note 48), B is satisfied 
with a mere description of its appearance (man nwa tfioa HS* 1 \>TW D«). Further- 
on B distinguishes two cases (a: ntw loipoa noiy mo to lonpi toipoa noiy nxym 
and 6: rftiDB DP n05T» ^nPO 1« WOO moa DJ3J Din D«1). These, it is true, are 
not treated as separate case in A, for nnx Dipoo nawo , . , mo to lonp jwjw 
of A § 8c is rendered here in B positively lOlpoa 101JJ mo to lonpl. 

20 The relation of the two Recensions to Mishnah Hullin 42 a is shown 
most clearly in the following, where the texts are arranged in parallel columns. 

B no. 8 
apn ppin npa men napa 
pa manan napa oann onp 
nbiDB nsnn napa n«nn *aaa 
ma^> a^n apa baKn t6i inn 
rman nap-a napn nap-a pa^ton aW b«n «b Mn 

onan .... npin lapw ,*oa^ mao w im« 

coon .... napw» mowsn 
.lapwty mwan n^ai 

Mishnah Hullin 42 a 

A §10 

mriin np^osi ot^in naipi 

na^pn aipw otsnn 


apw mo bv onp apw 

napi [or rather 


Mn n^a 1 ? a"?n 

aip^a o^pnn iapa 


. . . rnon» i« napww nx^n 



*b 22,ji P p , 3 i a in spa 2ia^ n *6i mvi r6iD£> n*nn mpa "rann 

"♦naa^ too tm 23<jni« p^n atoi te«n 
9(§n?)npi dm »ty^ "mru p yrw tcr dm pi win nnipa pi 

tdj; DMi n^iDfi dtop DMi mw wnvu^ *6i wb wi^ m 1 ?! «mwi 
"♦meo n&iya Tjnw dm nn m^i nn m 1 ? api *6i an tffwa 
9* i stir t)J PC if> ppw> nproc if> »wai (sic) ippp: if> tnnxn) *t napes 

.3*?t5 pf> pip -5C31 -?ir )>f> of> -?ipf> west) ©:rt>i .wcs O'T 
2 .t> piw|5i> pinpr f>if> ©ipp Ids f>ii tjrw r>*ia-? idpp: 

For the omission of mo to Dnp aipy, mo^Bn o*i3n 31p<i etc. in both Re- 
censions and the additional omission of rwin nip^J in A v. A note 105. 
Whether Mn ma 1 ? sfcl 3pJ should be regarded as one of the pierced organs, 
that are enumerated here, or whether it forms a paragraph by itself, as in 
ed. pr. and E. (v. p. 84 no. 11), corresponding to A § 13, the Halakah on the 
heart is open to doubt. I am inclined to accept the former view, for the 
following reasons: 1. A § 10 likewise closes with a?n 3lp^. 2. B follows the 
arrangement of A also in that it takes up the discussion of o»i immediately 
after completing the enumeration. 

21 B uses nn:ni, in consonance with Hullin42a, and 44a et seq., through- 
out, where A always has nip. 

2ia Comp. Hal. Pes. p. 140: nana )vrn mnn p nvbb nip ap^a. 

22 Recension A has only afcl 31p^; the words Mn rpa? are inserted from 
the Mishnah ib. 23 Exod. XXII, 30. 

2 * DewZ. XIV, 21. .HttWtra 114b. Possibly both quotations are a gloss 
of a reader to ^3«n *b or rather l^a«n N 1 ?; at any rate, A § 13 has nothing to 

25 The Halakah contained in this section is very remarkable. As the text 
stands, it treats of a swelling between the trachea and oesophagus which may 
result in one or both of these organs being perforated. The general propo- 
sition is made that a perforation in either the oesophagus alone or in the 
trachea alone does not suffice to render the animal unfit for food. This 
assertion is in utter contradiction with A §§ 11 and 12, where a perforation 
of only the trachea between the lobes of the lung or of the oesophagus only, 
so long as both of the membranes of which it is made up are pierced, suffices 
to render the animal unfit for food. I am, therefore, very much inclined to 
suspect the genuineness of this Halakah B no. 9. Perhaps the idea of a 
swelling rests upon a misinterpretation of the words A § 11: ynv 103 nt?j?:i 
and trachea and oesophagus were substituted for the white and red layers of 
the oesophagus. Even such a solution, however, is unsatisfactory. 

26 There appear already in the ed. pr. three Halakot between nos. 9 and 
10. Of these the first, 9* 1 '131 nonan *T nantM flatly contradicts no. 16 (= A 
§ 17). The second, 9* 2 >m n *no rrbin ttnm is without a parallel in A and 
the third, 9* 3 ^31 win flap tol contradicts no. 17. The three Halakot have 
been interpolated here. They are taken literatim from Tur, Yorek Be ah. 


p 7^3 7ia? x& pi nteoinD D^en wana rrato* n*rfr pi« rwon io(§9b) 

D^P DH DK1 27 '»35 7ii 0?>P pppiW f>if> ptft) 77p3 p7P1» '3'fa t>3Pp PPP 

Vtrin , « wm i»« iiw« Dnoi« iri i&k yi ♦nsnts 7KDtso rrc^Bn jnya 

rpyai Turaai man? nn« n^n viami nyifc to« anpfci w»n nn« dy» 

™.mpb npm 
oman n!?n mnnno iron nVn ty& -ma DnM ■ot* rA s nt«j& iiyi 11 (§ 19?) 
s ?:n nnnan ^t? jnai i«o n^on d^d ^d wirb rm ♦dsidi ppm 
in«n ntysrvn nn«n m^y^ dk id«i mnn dw Drpty Tisb^i ntymn 

isnm i't&k ik namn *p nnntw § 53. onro ik nanan *t mnwa no. 9* K 

rwhv n 1 ? »■•» ik ppya Bawaw ik mai? n»?» nb vrv ik ppya npya» ik nai 1 ? 

dm i-natwtw tibk iasy aspm nws b^t my y» bk tibk "Qtran bsjj.t, n-wa ct 
.inn n« pin i»ai my yK ,ain hk yam m#3i 

Bipa 73a Km nana n^j-i wnna § 55. ^33 Km nana rpVjn isnna no. 9* 2. 

jn maipa roaipa k^k ,w maipa niaipa k^k Bipa 

.nana ipBBaw ik \*vm nais iba: § 56. .nana paaaw ik "run yiap 701 no. 9* 3 . 

The similarity between 9* *, 2 and Tur § 53, 55 had been noticed by Reif., 
(p. 279). Why the copyist, compositor or reader should have inserted these 
three halakot at this most unsuitable place, is incomprehensible to me. 

" The words npnna K7« pawn maa maiy 'rai map nnK ya s 1*3 my nb «w 
"«ia IS*? ana refer to the little rose lobe of the lung, called Kvnn nn^tt^j this 
lobe, however, is not mentioned at all in A § 9b, although the structure of 
the lung is there fully described and all possible variations in the position of 
the lobes are enumerated. Nor is the rose lobe taken into account in the 
following sections by B itself. Furthermore the passage '131 11JJ Tib W\ inter- 
rupts the continuity of the first and last part of no. 10, the authenticity of 
which is vouched for by the parallel text A §9b nya* 1 by mbv r\\srh piK wan 
km nana ywb 'rKawa aviwn ik 7Kaw7 ^a mbwn ^nn s bk n7«aw by wnm. For 
these reasons '131 liy nb Wl must be eliminated as a gloss. It is, no doubt, 
also an interpolation, taken from Tur ib § 35: nnK K31K ]nr 1*3 11JJ T\b W1 
B s iB is? ana npnna k^k ni:iKn maa maiy ni^i trrwn Kn^ii^y nanpai naap. 

2 8 It is self-evident that the answer has absolutely nothing to do with 
the question. The latter states a condition which has its parallel in A § 9b: 
K\T "IB 7KBWB nwbwi ]^B nj;3nK KSa" 1 BK1. The answer, in our present text 
however, would correspond to a question involving a condition like that de- 
scribed in A § 9g '131 BW17 IK nap? np13T nJBp ]1K BK1; or in § 9c where the 
atrophy of a lobe is treated. Here we have a scribal confusion which can be 
readily accounted for; a question, pertaining to § 9b had been combined with 
an answer, suited to § 9c, the intervening passages having been lost. 

29 The text of this Halakah agrees with that of A § 19 treating of a 


12 (§ 15) ybn p -am not* it n» nn&n jrr» trytaA ramn )t« npM 

dk rora ns^i nmn yo ytarn td^i ntowo npn nnfewn nmn ny art 

♦rtoofi m»& rmn *w dki mtso nnn n&j> 

DipD pnn *6 n«i 30 ,(n^DB i*6 dki nnn tibjti nn^n n^y d«i) 
i*6 dki n^os nnn *«r dk mpo im« ty )bw pnn p jrv siy^n 

in which two livers [i. e. a clot on the omentum of the liver] are found 
in an animal. The method prescribed for ascertaining whether the second 
liver is only a blood clot, is the same as that given in § 19. The difficulty, 
however, in attempting to identify the two halakot, arises from the fact that 
according to our text (no. 11) one of the livers is "above the heart, and the 
other beneath the heart", an anatomical impossibility (cf. also Frankl, I. c. 
1873 p. 489). To make the two halakot agree, it would therefore be necessary 
to make the radical emendation 133n mm\ 

A possible way of retaining the present text presents itself, if we take 
Dn33 to mean "blood clots, looking like a liver'', an interpretation which is 
not improbable in view of A § 19. Two such blood clots might easily form 
on the pericardium in a case of sudden death. Against this, however, it must 
be urged that A has nothing to correspond to such a condition of the heart 
and that in B itself the entire discussion on the Halakah of the heart is con- 
cluded in no. 8. The former emendation is therefore preferable. E., aware of 
the difficulty, emends D*HM in two places to D'HII. This is inadmissible, for 
Kill does not occur in Eldad's ritual at all, the reference to it in no. 10 being 
an interpolation (v. note 27). 

30 The words oripn ibn dk pi mtw n\ xbi m npi canp "w r6 w s nmn 
jvtyn have no place here and no parallel in A. They are inserted as a quo- 
tation from Tur, Yoreh Be ah § 36 mw2 nt ma nt ap^ D"onp "MP fib v* nmn 
nTtW 1^13 p^yn D1"ip n^i I^SSl DrPitP 13prtP iy. Then the remaining words 
n^DS im dni nnn niftjn nn^saa n^y dki, seem to be a mere repetition of the 
preceding phrase nblDS na»o nnn air dni m»3 nnn nay DN. Their purpose is 
to reestablish the connection with the preceding part, which has been inter- 
rupted by the interpolation from the code. 

31 So E., ed. pr. j^on. 

32 The logical connection between '131 y!»n Dlptt pan m DK1 and that 
part of the paragraph which precedes the interpolation, has already been 
established by E. (p. 94 note 23), who says: DIpDn nmna 133 m DKff tfttttl 
ini« ^ pn }^ m is nnn n» m< dm nimto dp pna 1 ? jrr mi yforn f?« npma dpp 
Dipttn mim -p nn« atm pna 1 ? pyoi? i 1 ? n\Tl Dipan. Although A § 15 treats of a 
number of cases of coalescences, and of other conditions which are not taken 
into account in our recension, nevertheless no. 12, in its entirety seems to me 
to be merely a variant recension of A § 15. The present juxtaposition is 
sufficiently vouched for by parallels. Compare A § 15: pm pm pnn iVfl 

•una nnn na »Btsn pn» pn nnia nnr prab it d^ dki w\n 

biOl turn. The end of no. 12 '131 DlpD mix bv bv pnn p \W seems to be 

based on Hullin 46b. 


nan uaa «ri nn« nyta n an dn pi ina tonan )d nana i3(§2i) 
jnsrr *6i mpntnV n«n dk 0^26 Snnart pma .ibo 1*6 n«i ^db 
nip 1 ? ^inan pro J?yw nVi Kin *?idb d^sA npm uaa ikbw ny 
^inan d«i ibo ^inana np^> npn dn ntoaan mi« ■pnn n&rp «to 

ss/lBto to ^nan ^aJ nana vaya ^inan 3pi ♦VlDB mpn ]a np^ 

ipn p toy d« nn« «to dk ■o asm *6 ma niton p nisia 14 (§20) 
n« aton oy a^t«o *pa 34 ]ni« inni in« toya nito *w npn Kin 
ik 1 ? dki nt$o ss^nj;^ Tiny *tti *u toy^> toy *iton ;np tos Tiayn 

36 .nDi3 3ij?3 nj?i npnn tea ij> nrapnt? «^a pi xtpw 13 npf>p£3 iai:i 

33 The first part of no. 13, with its distinction between ruptures of the 
milt, accompanied by the emission of its fluids and ruptures without such an 
emission, is more specific than A § 21. The latter simply states: 3p s i OKI 
nn« 123 DK s 3 jwan vb\ vajD binan. The union of the milt with the (floating) 
ribs and the reticulum (concerning 3"ip v. Gl. s. v.) is genuinely Eldadic and 
has evidently been lost in A. Both recensions have in common Linton bwi 
1^13. But while A (§ 21), contrary to Eullin 55 a and the current Halakah 
clearly states: Kin lino V3JD Triton ap^i, B unmistakably says just the opposite, 
in agreement with Hullin I. c. naiD V3J73 ^inan ap"^. Unless we adopt some 
pilpulistic harmonization, we must assume that the text in B has been worked 
over under the influence of our Halakah. 

3 4 Ed. pr. in*. 55 V. A note 183. 

3 « The two halakot one nnxn nnpb snano oniay ona d^d nv^an dmi 
naiB 13 tin*n»3 baiil nt?an DDDDnitf are both lacking in A § 20, although we now 
possess this paragraph on n^bz in no less than three texts. It is, therefore, 
more than probable that the two halakot quoted, form another interpolation, 
taken from Tur, Joreh Deah § 44, where we read D^D [ni^sn] on DK1 
-m«3 Vrax nnp 1 ? .nana dwidi D"3t i^e* ik on^ji on dk ^a« m»a D"»3t d^d 
mspD3 DTK tinx^ dkp iy itP3n DDDDnarc bs s ip^> Kin m w nana jno 
ni'tpn* H"Via na» jaito nj? piayn d-di *ip^n y^t^ Kim ^am ddddhd 
nni^3 )K3D -in" 1 nana ^33 ^>iai Via3 ny npn n»n33 ^iDm v 13 ?' 1 rIto3,, n&na 

A^Wa n3iX?D ny Dili. The closing words '121 nrDpnff K^O ]21 flatly con- 
tradict the parallel halakah of A § 20, according to which the atrophy of a 
kidney to the size of a mustard seed ("?l"in) or a lentile (ntsnjj)> does not render 
the animal unfit for food, whereas in our text such a condition is not per- 
missible. Furthermore, the measures in A § 20 are *?"nn3 and ntsnjft, while 
here we find the Talmudic measures Viaa and 3ip. There can, therefore, be 
no doubt that our text is worked over either directly on the basis of Hullin 55b: 
n s iii s 3 3i?3 vy noi3 b)tt iy npn3 nrDpnt? rawn or, what is more likely, 
according to Tur I. c. In general no. 14 follows the Tur very closely in the 
arrangement of its subject matter (cf. Tur, I. c). 

If the above suggestion is correct, — that the words DnD nn«3 nnp^ 
nana 13 nnKntJO *?aii1 "lean DDDDniP do not go back to Eldad, but originate 
in an interpolation from the Tur, then the agreement of Eldad with 
Maimonides, (Yad, Hilkot Shehitah VIII, 26) in the explanation of Knip 1 ?, 



15 (§16) m^n w two *bb jmT tiw n»« um« 'hbA iA 37 io« mjn 

38. nana pm rvfarun rrmyto raiwav PlDrG nKED DN 

16 (§ n) 39^«i mtw nwato nntnao trtrp i« hto ton niosyn mo* 

w*i ww tfw .mBD i»ipon my osyn jarcn rAgmVi nnmnaB inty: 

071pC fatf 40 "?far 9W t>50t> 0>]5T> Df> PP'FC Ttt lf> t>PTC S71p of> WT 

41 »t>WDl t>tJ t>PTC 

i7 4 »ntoi to inm nto D*wn pnp tev nona u 1 ? now t$n 

I8*ypc: of) bf> •t>wp »lip ^pp 44 tnir ip:c if> o»»pp t>W)30 opcp: ofn 

45 *tv3C3 m7Ct> is 'dp ip w> twito 

which Epstein (p. 94 note 27) pointed out, is merely apparent. There is no 
reason for presupposing a common source for the two in some lost explanation 
of Hal. Ged. 

37 The subject here is Eldad. 

38 No. 15 again contradicts A § 16, where the strange, but thoroughly 
Eldadic proposition is set forth that broken ribs do not render the animal 
unfit for food, if the numbers remaining unbroken, are equal on both sides. 
The same principle is found with regard to the lungs, A § 9b. There, in the 
same way, an equality in the number of lobes, whether there be two or three 
on each side, is insisted upon. Contrast with this the principle set forth here 
in no. 15: nsna jam n^nan rrmyto nantwtp nana which agrees with our 
Halakah, Hullin 52 b and the Codes. This is, no doubt, due to revision under 
the influence of Hullin and the Codes. Here too E. makes a farfetched attempt 
at harmonization, v. p. 123 note 14. 

39 Ed. pr. *SM. 40 Sic ed. pr. 

* l The words '131 TQ» kxm, have no parallel in A. Their similarity 
with Tur I c. § 55 end— the beginning of that § we have already noted 
under no. 9* 2 (v. note 26)— nn-tt inK*? IK D"no nwjtf D« JKT pro mat? KStti 
D^no npyap yiTa Tint? naon Dipfc dk has been noted by Reif. I. c. Since this 
passage is an interpolation from Tur, it can neither be the source of the same 
Halakah in Sefer ha-Terumah, as E. p. 95 note 31 maintains, nor can it be 
the source of Tur I. c, as Reif. held. 

« Ed. pr. mto, E. 

43 This halakah has no parallel in A. There is, however, no evidence 
to militate against its genuiness. 

4 * Sic ed, pr. 

45 This Halakah, on an animal, whose skin has been abraded, which is no- 
where mentioned in A, is given on the basis of Hullin 55b as represented in one 
of the codes, probably Tur I.e. § 59: D1K «T3 p miy to BVM* Kim miDK mfon 
mtW ytoa na nTVitM DK1 ^in «T by pa. The only unusual phrase in our text 
is the measure nyima for the usual J?^D3. Nevertheless it seems questionable 
to me, whether the original Eldad text contained any provision whatsoever on 
this subject. At least we can find no evidence of it in A. 


nvp in by b&iw *py pi n^naa lponna kdp unn 1 ? i^ maa Dipnn nbaiw nana 19 (§ 24) 

46, nana i*6 dki nwa D^aa nate arn dki nisnan ^aa pna 1 ? *p* 

47 »mt;D p::iiwi p3Cirrt>i oTt) pnrf> 19* 
naa ejno w ntsDB inm nw ty n«t iw ra rrwn r6sun 20 (§23) 
ito*6 i»a mod *nwi *rin intfaKn ik iroai wn n&yi ww is 
a^nn i«nn ty by i« pm ty 4 stm ik an ^sr dki ♦in&wnp nyn 
wm« DnoA u 1 ? tbk Tijn «n» A atow ny ini» w« vpd imotei 
^wi u*ra lira ^ nrrai ty ^n« *6 'bki jwirr te u»y «d tsn« «a» 
1 5 ? w ony *« ^ty ^ rrtn nra o^nn "innAm np rrn ^rty 
♦m& 49 ^ k^p ny mnifi ny« n«m to*6 nni» nan 


«yi rrty nte«an n« jw ik mmy kvti ipnn ty dik ynn 21a (§ 32b) 
irennn 'in 5 ? ksw «wmn prf? wr vwri n^sn vsn -ww iniyn 

nw nai ♦inform mown *n mm 50 "i ty byi ti ^y imdb awn b(§32a) 
"♦^3fc6 nuoi imn&» io« ntDTiP nW>n nana 

46 Here again the original text which corresponds to A § 24 (animals 
that have suffered a fall) has been replaced by a reading, quite similar to that 

of Tur I. c. § 58: KB» ytwin , , . . ma 1 ? n^Mt* ik maa nana n^Bit? nana 

np^na *"j> nnn<ai nt*p nan ty aaroip *yiya ynn pi nna« ipanni 

nana 1*6 Dan ni»3 ni&na n ns ^oa note aroi dm ,%nn ^ naaa im« ipnans\ 

47 This halakah also, being without a parallel in A, is certainly an inter- 
polation taken rather from Tur, I. c. § 60 than from Misnah Hullin 58b. 
It is to be remembered that §§ 58 and 58 of the Tur have been used in the 
two preceding sections (nos. 18* and 19) of the Responsum. 

48 There follows in ed. pr. IB IK after TB3 IX, evidently a mere ditto- 
graphy of the latter. 

49 Sic ed. pr. 50 E. supplies rtn after "1. 

51 No. 21b is not free from difficulties. If lntontfn be taken in the sense 
of a command ("thou must slaughter it, if thou wilt eat it") and if '1 by ID? 
correspond to the usual technical term yp~\p i"y D^sn of Hullin 75b, then the 
text of no. 21b flatly contradicts that of A § 32 a, for in A § 32 a a njnpB p 
is unconditionally exempt from the requirement of nDTW, while here TttPTW 
is required for a nyipB p, as soon as he sets his foot on the ground. Again the 
clause nWn nana which contradicts A § 32 a, limits the dispensation in the 
case of a njnpfi p from nDTtttf to the one night in which he was born. The 
present text of no. 21b is unquestionably influenced by our halakah. 

Of course, it would be easy enough to harmonize the two texts by 
eliminating the ungrammatical nWil nana and by taking lnentwi in the sense 
of "thou mayest slaughter him." But such harmonization is not our task. 


c(§32c) n^m l^n 53 nnn 54 113 Din * 3 nnn 52 W103 ^ -rnn rrapn u 1 ? id«» tei 

♦no-oaty ib* niDnty nana 53 nnn &^ro 


[(*pin) rrn no^n] 
22 (§37) inDDi ion n» ^an ^ main praan m ty im« i^kp myi b. o. 

ipk 59a^nan rip' 59^«i p p yt&nrr pan rn w 1^ no«i awn 

52 Hebrew translation of Eullin 109 b n-nitt ]"? K"itf Kiann ^ loan ^D; A 

§ 32c has nnx mpan on'? vnn ,nnpn irv\& nnnn n»« fcn. Cf. also Lev. .Rafc., 
jDeut. Bab. and Tanhuma, as quoted on p. 76 notes 385 a and 387 a. 

53 Cf. Yalkut Schinioni on Ps. CXLVI, 7 (s. v. onwK Tn» 'n): nm» 'i 
D"in tid<k nnn ,-jb ^nnnn i 1 ? "motw n» nnn \r\i "i dpi pnv '-n «nx 'm '^ 'ii 
nina ^» nnna ntt>nn nana -no'K nnn pine *yiy w msny -ncK nnn pinoin jn^. 
A does not have this pregnant form of antithesis with nnn. 

54 Eullin 1. c.; Midr. Tanhuma, Shemini, ed. Buber, p. 15b (which has 
also bmvi besides "D3, v. A note 387a). 

55 Eullin I. c; Yalkut Shimoni I c.\ Lev. Rob. XXII, 10; Yalkut ha- 
Machiri. on Ps. CXLVI, 7 ed. Buber p. 142a. Frankl found in this passage 
which, as it reads now, is closely patterned after Eullin I. c, an additional 
proof that Eldad "in his attempt to inculcate the belief that Joshua had 
received from Moses a remark of Jalta (the wife of R. Nahman), was a mere 
impostor, playing on the credulity of his contemporaries". (M.G.W.J. 1873 
p. 489.) But Frankl failed to remember first, that the same thought appears 
in A § 32c in a form, independant of the Talmudic version and second, that 
this form of contrast, although used in Eullin I. c. by Jalta, is not original 
with her. According to Yalkut ha-Machiri and Yalkut Shimoni I. c, this 
antithesis was already used by as old an authority as R. Me'ir, in whose name 
R. Aha and R. Bisni transmit it. In Lev. Bab. 1. c. it is ascribed to Jonathan 
[ben Eleazar, a Palestinian Amora of the third century], and in Deut. Bab. 
IV. 6 to the pan. From what has been said, it is evident that this old form 
of contrast was not necessarily borrowed by Eldad from Eullin 109 b. 

56 Lev. XVII, 13. O omits the whole clause from beg. to 1BJD, which 
is due to the Responsum form and has instead simply JWi.T na« "ny. 

57 O nsy X2D vbv y*?D Dlp»3 rwi DK. The word *y»yD after J^D in B, which 
is wanting in O, might be the gloss of a reader or copyist who had J^D rpyo, 
Judges XV, 8. 11, in mind. *|*yD J^D as it reads now, makes no sense at all. 

53 O omits nsy dp 'W* d*B IK; A § 37 (T-S) likewise fails to take this 
case into consideration. 

59 O omits the entire clause, which again is due to the Responsum form: 
Thus O reads: 'di vty nt?x 'men np s isy ks» vbv j^d mp»3 rpn dx ytfi.T -i»k Tip. 

59a 7. Gl. s. v. 


Din "p^»i <*W2ro onn '•idd ty 65^ i 64tnn m n «3nM»i 

72«m p naraa kvtibno rtanen tarn »*jnyttn tayn rrvntym 23 (§35, 
":£m 7 «na^ .miTiDK «"0 mo ^uw *»Tip*n "-ate iTnn§ 36 -§ 38 ) 

snarwn nn& m ^ 8o*py i« rpD3 ^ a^an w rA tw rrn 24(cf. 
DiTi» &wi rmth ^awni nty»^ ptwi np*7p ty npnp n^pipn jnn §27 and 
smoTrn i« inntyn dn sim^ Kimtsw b»v&n ty*nn ♦nn« n&wn t§ 39 ^ 
tok jwi nnte itso impyn i« 84, nD ^p ni irvarm i« vn^nn ik 

60 oinan. 6i 11. 62 o omits 'rnsn n« ^n. 

63 noa*i. « 4 omits Din n«. 65 o -pm. 

66 O omits ttima and inserts 'paiNl instead. 

67 O omits "wk. 68 -isj^n Dip&b. 69 o noa»i. 

70 In Edelmann's excerpts from cod. Oxford 781 23 our Halakah is 
found in a recension which at times agrees with JR, at times literally 
with 0. -pa&i bin n« in noa&i ^nen rip* 1 nay p«iy mpaa Ktrra din 
nana «^a noaai nayn mprri ^nsa i»« Din -p^ei (=0) teiw nnn »iD*a ty 
.nmajn 'so u*n mws *&& v^n 11 n&iN pi pa: nox an 1 ? n^awn mwy into iann nt 

7i omits mixn. 72 o omits Kin. 73 o -ab. 

73 a St. (Z. c. p. 300 note 2) without any justification, emends to lip" 1 ?; v. 
the parallel text A § 38. 

74 pjra. 75 n^>n. 76 naVa. 

77 abrny. 78 omits mn. 

79 B condenses the three paragraphs 35, 36, 38 of A into one: nvatsa 
tsspn corresponds to A § 35, nonia to § 36 and nisriBn ^aai to § 38. The 
omission of niacin in0 (v. note 71) may be due to the fact that had § 35 
still in a comparatively complete form before him, but that for some reason 
he failed to copy it. For a fuller discussion of this point v. supra pp. 23 et seq. 

so R reads: D'Wtn "it* Dn^> vrvt *pjh rrn n&m. This is certainly not the 
original reading. Part IV of this ritual deals with rrn (and *)1J?), and not 
with n&na. From A § 27 it appears that the Halakah on tftwn *W nb WV nt>na 
(naiwo nana) is identical with that on &v*n *ivt nb w n s n. As the vrv nana 
D^xn ^B^ rib of A § 27 is not represented in B, it is quite possible that a 
reader or copyist wished to supply the omission by inserting here the words 
nena 1K. *py IN may be part of the original text. It affords a transition 
to the next paragraph which deals with apjf. si O omits toiwn Dflfc nt s «. 

82 It omits, owing to a homoioteleuton (ntato^— ntso^), the whole clause 
7\txh wn n»K bvmn v*r\r\ nnx nts^n^a a.T'W Btn»M. 

ss in\n»n in wdywi d». i2 awd O inomn, not vwnrm, cf. note 2 a. 

84 WD&'jpni, E WDlfi^pn IN, showing a later revision of the text, v. 
note 15. 

85 Sic R, ed. pr., O omits the whole passage from iwa to Wtron DN». 

86 DN1. 87 )bK bfi. 88 n^os. 


26 (§34 una 92 ^ 91 im yunrp 90 ^« «n ^uny ^n» in« nb id« iiy 
and [40]) 96bnnr6 p^o 95^jq ^«i 94Q^nn msijn nvn toki ^« T*s mwh 
hdib d^sd T3» ^« iwi ytsnrp i 1 ? wi "/nnon pm «»&n pa 
•o ^ ma ntyo P« tstoai Kin nois onso TtKi l 1 ? '»« rr& ntyoi 
ib w d« TTa 97a ^ lT nt p^ D y^, T iy:n ^ w .Qi«ris D^riDsn •ok 
99 im 98-ihd nisiyn mm m ••ana T*n 1 1 ? iai .k&b nty»^ d^p 
W d« iar» nnKBO tto io»«rr n» )»^d 100 yp^ wm A 
im.kbb nns kVk ^ p« d«i Tints nns 1 ? rosity nn« lo^rniKmo 

89 omits from the beginning to to. •• bxN. 9i "IDKI. 

92 O omits & 9 3 omits W* '•illK. 

94 o omits from nnm nisijn nvn msai. 95 jnv. 

96 omits bnsnb ptrn. 97 o -nnn *pyb kbb *|iy p. 

97 a J&J. ^r. K3* (E.), but later KiT (v. note 101). 

98 consistently omits the entire passage, dealing with mn "iO^D, from 
y»im lb TW to inn m&iyn, since the mn ''Jtrn have already been dealt with by 
0, v. A § 34. »9 nox. 

100 omits y«nm lion. 101 O has "iron after Km. 

101a Sic ed. pr., probably a misprint for rniKmn. 

io 2 inverts the order of succession, reading: nxjin nnK KbK lb yt& HK 
(so Jell.; Neubauer reads nnin, v. E. p. 136 note 7) roait? nnK lb w hki Knn 
"lints nrm by. The doubtful mun ought, most probably, to be read nm The 
text may be defective and may originally have read as follows: Kbx lb y*K nK 
mnn nnx by n2Bit? nn« lb » s nxi, Knn [nxnin is] nsia nrm. This emendation 
is based upon a passage in the Sefer ha-Mizivot of Daniel al-Kumsi, published 
by Harkavy, Studien und Mittheilungen VIII, Zur Geschichte des Karaismus 
und der Karaischen Literatur, Petersburg 1903 p. 187, which reads: niK nblKi 
*3 ynn nxta .ninixn «bi nrotso ms p by .nnn nmn xnn *jiy niK3 tints «\iy 
mm 'T by nm 131 mnB Kin p i«p nix ^ n»« »|iy ba imon ba 
«in p «bi bsx^b inia nsui n«n» ib ntpx ^iy bs iraix i» s i ,nb» mn 
nxno s d nwin on *a n^nna s ntr nn rwiii nxnnn s d not^in nm .nnto nyna «bi 

»in p «bi [AJ.Xs^a.5 a^oilS] nbswi nsa«p km nj«ii. 

ntna and mill are the two crops (the upper and the lower) in certain birds. 
Such is the view not only of the author whom Daniel al-Kumsi here opposes, 
but also of Saadia, who renders Lev. I, 16 with the words yn nnbsin jhjm 
nnniKp (v. his Oeuvres Completes, ed. Derenbourg, vol. I (1893) p. 144 and 
Harkavy /. c. note 6), £j.-o^&* being the upper, and <J^*ails the lower crop or 
the gizzard (cf. also Abba Jose ben Hanan's view in Sifra on Lev. I, 16 and 
Mishnah Zebahim 64b : noy ppnip bnm [ntn&n n«] nbnw. pp"iip here may 
refer to the second crop as well as to the gizzard, v. Jastrow 8. v. II, p. 1344; 
Dalman, Wb. p. 374). 

The authority whom Daniel al-Kumsi here controverts, is, most likely. 
Eldad whose view might have found acceptance by others, hence noPW nm 
'131. For Eldad clearly states that the sign of purity for birds consists in 


that they have two crops (not as E. p. 98 arbitrarily interprets the text DK 
s»a nrm »b» ^ v 8 nx ' 1 11nta nr ™ ^ na*w nn« vnano s nt? to mean one crop 
and a gizzard) and that the contents of the one, i. e. the upper crop "flow" 
into the other, i. e. the lower crop. 

Among frugiverous birds, only in pigeons is the second enlargement of 
the oesophagus sufficiently developed to be properly called a "second crop" 
(v. Gegenbauer, Vergleichende Anatomie II. Theil, Vom Vorderarm). There 
can, therefore, be no doubt that Eldad believes that pigeons are the only birds 
or fowl that may be eaten, probably because they were the only fowl offerings, 
brought in the Temple. The same view is held by the older Karaites, such 
as c Anan (v. Harkavy I. c. p. 67), to whom the Karaite Aaron ben Joseph traces 
the distinguishing characteristics of "clean" birds: D31 *?"Qn ^3 Wfi7\ *|KW 
vmiBK tatt), which are taken from pigeons (v. Aaron ben Elijah, Gan 'Eden, 
'Inyan Shehitah, chap. II, ed. Goslov p. 82b col. b; Adderet Elijahu, c Inyan 
Shehitah, chap. VII, ed. Odessa (1870) p. 111b, col. a), Benjamin Nahawendi 
(Harkavy I. c. p. 179), Daniel al-Kumsi {ib. p. 188) and others). The later 
Karaites are much more liberal and permit the use for food of all the wild 
and domestic fowls, not mentioned in the Bible as "unclean", so long as it is 
possible to identify them on the basis of tradition. Attacks on the validity 
of the characteristics, accepted by the Talmudists or Rabbinic Jews, for 
distinguishing clean from unclean fowl, are Kara'itic (cf. c Adderet Eliahu I. c. ; 
Gan Eden I. c; Neubauer, Beitrage zur Geschichte des Karaerthums p. 43 
(Hebrew Part) no. 6; Emunah Amen ed. Goslov 1846 p. 40a etc.). 


The most authentic or, more property, the only authentic 
account of Eldad and of the Four Tribes is the Kairwan Epistle 
to Gaon Zemah [ben Hayyim], preserved in the Responsum. The 
study of the Halakic part of the Responsum has proved con- 
clusively that the text of the editio princeps is considerably 
removed from the original. Many passages of the original have 
been changed in substance and in style, and a number of inter- 
polations have been introduced from various sources — notably 
from the Tur of Jacob ben Asher. Some of the interpolations 
are as late as the first half of the fourteenth century. In view 
of the fortunes of the Halakic part of the Responsum text, we 
may presume that the non-Halakic part 1 has fared no better. 
This general presumption is confirmed, wherever we are able to 
compare the non-Halakic parts of the editio princeps with parallel 
texts, strikingly so confirmed for the portion of the text con- 
tained in the recently discovered Genizah fragment to the exa- 
mination of which this Appendix (I) is devoted. 

The Genizah fragment is the T-S Loan 94, University Library 
of Cambridge. It consists of two leaves of paper, 16,5x12 cm. 
Leaf 1 contains 16 lines recto and 14 lines verso; leaf 2 contains 
12 lines recto, and 13 lines verso. It is written in an early 
Oriental cursive script. Its date cannot be later than the thir- 
teenth century. 

The fragment contains a) the end of the Epistle to the Gaon 
(leaf 1), and b) a passage of equal length from the middle of 

1 An English translation of the legendary part of the Responsum was 
published by Neubauer in JQR I pp. 104 et seq. 


the Gaon's reply (leaf 2). It thus preserves about one-eighth 
of the entire Responsum text. The inner quire is unfortunately 
missing. The text contained in the Genizah fragment corre- 
sponds to the following sections in the editio princeps (v. above 
p. 17 note 58): 

1. a) The Epistle to the Gaon, ed. Epstein p. 5 no. 7 1. 2 — 
p. 6 no. 9 1. 3 = ed. Miiller p. 18 nos. 7—9. b) The reply of 
the Gaon, ed. E. p. 7 no. 16 1. 2— p. 8 no. 17 L 1 — ed. Mii. 
p. 19 no. 16— p. 20 no. 18. 

The text of the Genizah fragment is further paralleled in 
two other texts, to be described below (see 2 and 3). 

2. Codex Oxford MS. Hebr. d. 11 fol. 62b— 63b, Catalogue, 
by Neubauer and Cowley (Oxford 1906) II No. 2797 1, n, publish- 
ed by Miiller (pp. 18— 19) 2 . This contains, however, only the end 
of the epistle to the Gaon, that is only about as much as is 

2 Professor Miiller has used the above mentioned Oxford MS. as a 
source for, at least, three excerpts. He has, however, failed to indicate the 
page in the MS., from which he quotes, an omission which is the more serious 
in view of the bulk of this Oxford codex (377 folios). These excerpts are : 

1. Parts of the Responsum. These are reprinted on pp. 16 et seq. of 
Miiller's work. It is surprising that there is no mention of the Responsum 
in the detailed description of the contents of this MS. in the Catalogue, by 
Neubauer and Cowley /. c. Can it be that Miiller is mistaken in regard to 
his source? 

2. The legend of the Bene Mosheh, in Miiller's work published on 
pp. 27 et seq. = MS. fol. 62b — 64a. Miiller's vague reference to "Cod. Ox- 
ford 2" is evidently to our MS. By the equally vague reference to "Cod. 
Oxford 1" (ib.) he evidently means codex Oxford 2399 which he mentions on 
the preceding page (26). Here, too, Miiller fails to give the page of the MS., 
from which his quotation is taken and even the number (2399) of his Manu- 
script seems incorrect, for neither Steinschneider in his detailed description 
of the contents of this manuscript (in Ha-Mazkir, 1871, p. 37) nor Neubauer 
in his Catalogue (I p. 842) makes any mention of the passages cited by 

It is this mysterious "Cod. Oxford 2" which Miiller uses as the basis of 
his text. Of our MS. Hebr. d. 11 (no. 2729, 1, n) only the variant readings are 
given in the notes. There is another codex, MS. Oxford 2287, 7 which Miiller 
should have used for the legend of the Bene Mosheh. This codex some- 
times follows codex Oxford 2399, sometimes our codex 2729, l,n. 

3. The so called Elhanan Recension, printed by Miiller on pp. 41 et seq. 
= MS. fol. 64b. Miiller (ib.) speaks of another Oxford MS., which he used, 
but he does not tell us anything about this MS. 


found on leaf 1 of the G-enizah fragment (see above). The other 
text which is available for comparison is that of the 

3. Shalshelet ha -Kabbalah of Gedaliah ibn Yahya 
(1523—1588), ed. Venice (1587) p. 37b. Here 11. 4—15 contain 
about as much of the Epistle to the Gaon as corresponds to 
leaf 1 of the Genizah fragment. LI. 25 — 27 contain only the 
first half of the part of the Gaon's reply which is preserved on 
leaf 2 of the fragment, the text breaking off in Shalshelet. 

There are, thus, three texts available for comparison with 
the text of the Epistle preserved in the fragment, and two texts 
for comparison with the first half of the passage from the Gaon's 
reply. Only the editio princeps preserves the text corresponding 
to the second half of this passage. 

Although comparatively insignificant in extent, these frag- 
ments are nevertheless sufficient to prove that the account of 
Eldad and of his country which the Gaon received from Kairwan 
was, as yet, unadorned with many of the most characteristic 
features of the story, which now form part of the Epistle. 
Thus, for instance, the Genizah fragment, which has preserved 
a simpler and much older text than either the Cod. Oxf. or 
the ed. pr. or the Sh., warrants the doubt that Eldad made 
any specific mention of the Bene Mosheh. Its version makes 
it probable that in his story, as in the earlier Midrashim, it had 
been the Levites and not the "Sons of Moses" who refuse to 
sing the songs of Zion in a strange land and are saved by 
the interposition of a cloud from the wrath of their captors (v. 
note on leaf 1 recto, 1. l).s 

In other points, too, the Genizah version differs from the 
accounts hitherto known. Here the Levites do not carry out the 

s Speaking of the Bene Mosheh I would like to call attention to a 
fragment, published by A. Harkavy in Ha-Gat (Petersburg 1897) p. 65, which 
relates the travels of Jacob ha-Nasi, a magician who lived in Susa between 
1240 and 1276 C. E. Jacob ha-Levi asserts to have visited the Bene Mosheh 
in the land of Kush and to have received from their prophet the power to 
do wonders by means of the devine name. The year 1276 C. E. is given by 
this prophet as the year of redemption for all Israel. Harkavy calls this 
fragment: wn mb» nbiQ i« "jpn nh b» rrvsn \w\vq khmh rnyoB "nso. 

Herr Kahan informs me that in the Siddur Ashkenazi ed. Venice 1645 a 
TirPn TP is printed with the following heading, rwo *J2 lipn "WK "iwn T» 
■ma nnio v 'y iibd px b» «nin q»w j^mbd in: 1 ? nn$?». 


act of self-mutilation upon which they have resolved. The 
divinely-sent cloud does not bear them through the air, together 
with their wives, their children, and their belongings, on their 
journey to ancient Hawilah; it merely conceals them from their 
persecutors in their flight thither (v. 1 recto 1. 7 et seq.). Eldad 
must, therefore, share the glory of inventing the story of mira- 
culous deliverance through the agency of a cloud (v. Epstein, 
JBereschit Eabbati in Magazin, 1888, p. 83) with some still more 
imaginative successor to whom we owe the development of the 
tale. That the country inhabited by the "Sons of Moses" is free 
from unclean beasts and reptiles is another feature of the current 
legend, which is absent from the Genizah version. Similarly, the 
most marvellous details of the Sambation story are missing. 
Instead of a river without a drop of water, constantly rolling 
down a flood of rocks and sand, a river which cannot be crossed 
so that the inhabitants living on its banks can hold communi- 
cation only by shouting to one another, the Genizah fragment 
tells of a stream the flow of which is swift enough to bring down 
sand and rocks. Whether in the original description given by 
Eldad this river stops flowing on the Sabbath, cannot be answered 
with certainty (v. note on 1 verso 1. 4), but the words "it rests 
on the Sabbath" (m mt^l) are strangely omitted from the Ge- 
nizah text. 

That interpolations crept into the text of the Responsum 
is certain. It is highly probable that the source of these addi- 
tions was the developed Eldad legend. The Recension of the 
Eldad legend from which these interpolations were drawn, can 
no longer be discovered. It is, however, significant that the 
Recension, which Miiller designates by H. (E. pp. 55—60, Mil. 
pp. 13, 30 et seq. 53 et seq.), is found published together with 
the Responsum in the editio princeps, the text of this Recension 
preceding that of the Responsum, (v. note on 1 verso 11. 11 — 12). 
The mere juxtaposition of these texts makes probable the in- 
fluence of the one on the other. 

The fragment serves to clear up the text of the Responsum 
at several points. In turn light is thrown on the fragment by 
the Cod. Oxf., the ed. pr. and the Sh. These three texts con- 
stitute one related group. The text of Cod. Oxf. is less trust- 
worthy than that of the ed. pr. 

The value of Sh. as a source has not been sufficiently 


appreciated. Epstein (p. 8) did not deem it worthy of a place 
outside of the notes and he pronounced the text "corrupt". The 
text of Sh. is, however, more accurate than that of the Cod. Oxf, 
and in several respects is more closely related to the Genizah text 
than either the Cod. Oxf. or the ed. pr. (v. note on 1 verso 11. 
10 — 11). Moreover, Sh. contains fewer late interpolations than 
either the ed. pr. or the Cod. Oxf. On comparison with the Ge- 
nizah fragment, it is found that the absence of these interpolations 
cannot be attributed to the intention, or to the carelessness of 
the copyist, (cf. preceding reference). The significant conclusion 
is that we may rely on Sh. for the correction of the ed. pr. even 
where we have no older text for corroboration. 

The comparison of the following passages in the Sh. and the 
ed. pr. will serve to illustrate their mutual relation. Where the 
agreement is verbatim, this is shown by spreading the type. The 
numbering of the sections is that used by Epstein, p. 6 nos. 11 — 16. 

Ed. Pr. 

Drmtosw t6k "i pojn .10 
nb nso u»» anyioty -wai wish 
p pnr sunn ja iy»w own 
m n ^« 'n ikw nnop Kami i» 

uto D^bsn nnD maii 
.p^ai& rn jnspfci 
^ywn ro im mm .11 
tnarh yvo oni bto* nunpo 
n^ni mimo ntyBttty *sb ftbw 
mat* nityn ^nsai p^im ^nt? 
ny tsnp&n w pi) ^»n m^ 
pi (roty td^> snp ?ro6 niDty 
^nnnty ^n mm iw p ^n ikip 
p«» i«2p ^fcwa &itos6 wk ite 
nnrni oty urn &)ib brntr* 
d^d-di nnty dotisi niiJi p« 

i Text not clear. Read: yitaia 

irtni V't UTiia-i nana E. would 

leave Wnana and read W*n instead of 
WKII ; Reif. (p. 287) emends WHITO to 


lvnifi ntapa iA w -o in 
a"*irn ifiD rwnty •pao pnr 
mmu inoni wi ii^k i«i 

/^sia onspai n^ty 

nnnio n^ntao ijni 
m*6 'n nit^n ^nsii p^im 
nni tyiD p*6 id!?h mim ^a 
d^t nnm p« nnrrA dp 
on 1 ?! linii mo bi n»bn 
nnm s d nnb ^jnni 'n niny 1 ? 
\nmam mini 

a nnatf waffl seems to be omitted 
by mistake, cf. the Plural forms: 
Itn and man. For the name and date, 
v. Rapoport in Mahberet he-Aruk 
ed. Stern p. XI col. a. 

b Read mato. 


Ed. Pr. 

nits te n«^a Dn s nnni p« 
rwra w n« nny 1 ? Dn^> nnii 
i*jnm nnn8& vnisD te nrc^i 
onnn wn nnna *a on^ 
i isd lawn mn^i minn 
.n? tt^k 
nr^ i»y uvnai mr\ 12. 
'ni nnroo *p by t) ^kibm ite 
■»t ty nn«i isaimn* s t ty 

**>b n'nten ten p tont^ mnn 
pnnn Diip tsna^ ^n isosy&t? 

DJD ^KBf ^ naia ♦D^fc? iT'^p 

iy p yoi 8^> mqv nt inn 
♦n^te mte 

pw VT^8 "I 1»181 13. 

,pini am ne^p rfrpo) wd 'in 
n^nm pin nrY'D minn pai 
nni»«n nrros te no«i mitm 
♦(pin *6k n^a nno minn 
imi ate8 nts>» "on ^1 14. 
pty'ma n»« Dn^pa p^no 
ntenp tni»n bi irnini Dn»i8 
dw ntyo **a d^ isaimna 
dh ten nnm^ ijnnt? jrai 81m 

♦11*78 '1 

iymn8 nra Diipi 15.) 
ni&nten ipD^rii on jyan p*6 
2^1^ '•Ba inp^ty mpan inatfi 

tern np» nnt> ins nip^so A 
nani? pi«n ten iw D^ntyn 
(♦ate ty iTvinn ipnnn yo^m nw 
Mityn by inarm ^>8i 16. 
,t6k fct> Dny»t? W8 ^nm 
pa ■•Dnni ten ^nn nnt? 
nn« n:ty» d'oip ^8W 
2 e. rwo y*. *ni jip^nn 


ite nrte "» n^nn no«i 
ten p »n» inn 8te ^>8itr 
ana^ T^n i»s^Dty ^ nvten 
^i«m d^p sr^p jmrrt nnp 
.lite mte ij; p VDi «^ 

^n ^in *n^« 'i io«» n»i 

nnai d^« ntyo ^n 01 
1018 n»8 Dm« ij-'pjo jrtano 
nten i^imnity tsni»n n«iin 
rw» ^n 

Mi^n by in»nn ^«i 
nn» nana Dnv»tyfcy *pVm 
^«ity^ \>ik '•onni ^nn ^onn 


Disregarding such amplifications of the Responsum as are 
mere embellishments and add nothing of consequence to the 
contents, we are safe in saying that \T]$b rti&& iy tSHpIDH YiD s )£1 
Hit? Ttib 2)"\p in no. 11 is a gloss. As Eeifmann (p. 287) has 
pointed out, these words have no logical connection with the 
preceding. In no. 12 the words WE'D} T by "11 mmD *p by !"1 
lrYJJK *T ty nn«1 'WDfiDK *? ty nn«1 may be a later inter- 
polation, taken over bodily from 'Eser Galiyyot (v. Jellinek, Bet 
ha-Midrash, IV, p. 133; Grunhut, Sefer ha-Likkutim III p. 2). 
The original text may have read merely m^Ji Mffp )yr\)^ tttin 
^Klty )bl. This reading is preserved in Sh. The source of the 
interpolator is probably the ancient Seder c Olam, of which the 
"63 r\vb* HDD was as integral part, {v. E. HilleFs Commentary 
on Sifre II, 43, ed. Friedmann, p. 82 a, note 46; cf. Eatner, 
Mebo leha-Seder 'Olam Babbah p. 123 and Marx in Z. f. H. B. 
IV pp. 98 — 100 and in Steinschneider, Geschichtsliteratur p. 173). 
The assumption of Epstein and others, that Gaon Zemah ben 
Hayyim quoted the Midrash c Eser Galiyyot, is therefore not beyond 
doubt (v. Epstein, p. 17, note 15, Eatner I. c. p. 49, Grunhut I. c. 
p. 13, p. 'n note 1). 

In no. 13 '131 pjni nn nsnt? n^pD may again be a gloss 
amplifying the term '1JVB *t. It is conceivable, however, that 
these words are genuine and that it was the purpose of the 
Gaon to demonstrate the truth of tradition by the very fact that 
those tribes are practicing pin which is not mentioned in the Bible. 

I shall not attempt to decide whether the concluding sen- 
tence in no. 14 "Dl btt nnm^ IJWTty JVD1 is an interpolation or 
not. On the other hand, Eeifmann (I. c.) is fully justified in 
suspecting the genuineness of no. 15. This section is undoubtedly 
an interpolation, based on Temurah 16 a. Its purpose is to 
prepare the reader for what follows and to impute to the Gaon 
an explanation justifying the divergence of Eldad's Eitual from 
our own. The same motive appears in the other interpolation 

vby myp wins ana ybm &* m nbtw pirn *&2 jw wb em 

tlKH *p myon T™ mittl ed. E. no. 17 beginning; cf. also note 
on 2 verso, 1. 3. It likewise seeks to account for or to justify the 
discrepancies between Eldad's Halakot and the current Halakah. 
Evidently the reply of the Gaon was later found weak and un- 
satisfactory on this point and consequently the attempt was made 
to improve it. The words "76k TDD IfcfcOty im bz EHfi^ ]133 j"»H1 


m iriDH at the end of the Gaon's reply (ed. E. no. 17 end) 
may likewise have been added for this purpose (cf. also Reif- 
mann I. c.). 

In these instances the existence of parallel texts enables us 
to prove that the non-Halakic part of the Responsum has under- 
gone alterations and suffered from interpolations in the same way 
as the text of the Halakic part. There is therefore ample justi- 
fication for doubting the genuineness of the text of the editio 
princeps in other places, where this cannot be proved by citing 
parallel texts, and where the appeal must be to internal evidence 
alone. For the reasons assigned by Reifmann (p. 261), the 
following passage, which occurs at the beginning of the Epistle, 
is of doubtful genuineness: xbx llbl anttB&n Kb (ed. pr. )b)) »b) 

an wai d^sik mn« ante an "o pd nnab inyo i&x &towb 

DTllty (E. p. 4 no. 1). In view of Eldad's other utterance it does 
not seem reasonable to attribute the statement to him that the 
Danites recognized the supremacy of the Gaon. We must not 
forget Eldad's explicit statements, that the only source and 
authority for law recognized by them was God, Moses, and 
Joshua; that they did not know of our Talmud, and could not 
acknowledge its authority; that they had a code of their own, 
and contrary to the injunction of the Talmud (Sanh. 41a) and 
to tradition they continued to practice capital punishment in 
accordance with their own law. This would cast suspicion on 
the genuineness of the words at the end of the Gaon's reply: 

m^n bs by in«i nw&nn ten ^n by a^tena ant? nbx naiKtsn 

"Dl nmy an n& (E. no. 18). The Epistle does not say anywhere 
that the Danites pray for the scholars of Babylon or for the Jews 
of the Diaspora. The only place which may possibly be construed as 
an allusion to such a practice is the passage, cited above: $b) 

ante an ^ pd nrr^ "inya ipk wiwb »b» nnte annsan vb 
w*nw an va^ai a^tew irana (cf also E. p. 20 note 1). The 

authenticity of this passage is, however, doubtful. It is, therefore, 
not improbable that the entire passage, which at present forms 
the conclusion of the Gaon's answer (no. 18), is an interpolation. 
The answer of the Gaon, in the original text, probably ended 

with the words nna pyaai Tiaten pa mtyan p «\n nn« minm 
ynw ten (E. no. 17). 

Further examples can be cited, but these two instances suffice 
to show that the original story of Eldad is hidden by a thick 


overgrowth of later origin. Once this is pruned away, the Res- 
ponsum will be found to be not "a stupid forgery", as Reifmann 
thought, but an authentic historical document. 

In conclusion a word on Rappoport's view of the Responsum. 
In discussing the text of the Responsum preserved in Sh. he 
writes (Mahberet he-Aruk, p. XI, col. a) : "There were two Re- 
censions of the Responsum of R. Zemah. The beginning (sic) 
of one of these is to be found in the Sefer Salshelet ha-Kabbalah, 
the other recension is to be found complete in our Sefer Eldad. 
In my opinion both are translations from the Arabic. They are 
divergent at many points. It is possible, however, to correct 
many mistakes in one by referring to the text of the other. In 
several instances the Recension in the Sefer Eldad seems to be 
more correct etc." 

There are, however, serious objections to the view that the 
ed. pr. and Sh. represent different recensions. Comparison with 
the Genizah fragment reveals characteristic variations and cor- 
ruptions common to both, the ed. pr. and Sh. There might be 
more justification for considering the Genizah fragment an inde- 
pendent Recension, for the differences between the text of the 
fragment, on the one hand, and that of Cod. Oxf, ed. pr. and 
Sh. on the other are very considerable. Even these differences, 
however, are not sufficient to warrant such an assumption. The 
points of divergence admit of explanation along the lines indi- 
cated in our discussion of the history of the text. 

The other theory of Rappoport that Sh. and the ed. pr. are 
different translations of an Arabic original, seems to be equally 
untenable (cf. also Steinschneider, Die Arabische Literatur § 26). 
There is not a single feature in the language of the Responsum 
to betray an Arabic original. On the contrary, the language of 
the fragment is a simple, natural Hebrew, a comparison of which 
with contemporary style leaves no room for doubting that the 
Responsum was originally written in Hebrew (v. also supra p. 32). 

With these general considerations we turn to the text of 
the fragment. 


fol. 1. Recto. 

mb n»«i onnty nvbn by\ nwby l 
warn D^n n»y ]vx w» Mb iw 2 



foil. Recto. 

on *3 anwa DmnflDSK nra^ 4 

Dm d'od twf mynsa nnow 5 

pan *w Dm mi ro* tsnp&a wn 6 

w$n yiDi^ wpm ranpm n«oa 7 

(«) ijhmi n«26 D-i?jn w ?»pn cr6 8 

p« *?« an^poi Drrom on tt«n pv nnn^Di 9 

im«n ism n»npn r6*mn 10 

pyn -idj; ibw Dip&n n iroi Dipo n 

dst/yok d,-6 vi^di mnn nW>n 12 

toa typi op «mn nWa n*n "o 13 

am (sic) rwi ■o -dim bm ii« ipsn i*ni 14 

7, wi D^ani DTtom D^pns 15 

*im «inn nyn ds^di mijn 16 

/o£. i. Ferso. 

Dty ivn *6p DipD ^im p« bbm 1 

roin d^dk Wii vm ps ntoyo to 2 

yoty ^na ^ip A an -\pn p« ny Vim 3 

kVojv rot?:n nnn pima 1 ? ny 4 

^myi py wimd nnino^ \vy 5 

pnrv kVi rta wyb m« ^d 11 *6i 6 

rOt? ^SID IV IDIpD kVi inn 7 

to mis jwnpi nw idd -mm 8 

Dwtono im« jwip naiBtoi jvtsniD 9 

dw onrw rnDipa nrn TOn en 10 

xyo pnn ^pv wk D^ntyn r6w na« 11 

b« to pawn pinnty o» 12 

jw*n rva pim ibm -u ^nsa 13 

^n oy 1711 p ^a bx* mpnai to 14 

btmtp p*6ty o^ann 
pnms ]w nn« pts6i 

IID^n py^> bl* ]^D1D JW 

nr^ Bh <:> *pbTi ]twi wo) 
ayron n^ntt torn Dyta 




fol 2. Recto. 

)b n«ii nt topan pi in« 


in« oyta i*? run* m oya 


pjnnp ]my *)jh* [ni]*opDn 


D^y&n pi^n ifrtfo w nron 


irrwi (sic) nivno tan \in pi 


niDinoi nimnsi 


fol. 2. Verso. 

-prrm nniD&i n^ytoi 


mty&n pw ta cpiD^ 


pi»V pi»yi mno nm kito 


«sid nn« p*o uks^ to 


nmtan (read «*?«) *6i Dn»nn pa *pWi 


p»ta tan s in iniK pDiw 


^*w p« ^m jvaiK 


Dm&Kt? noi nnbw mnnn 


TiB^n orft Bf D^ntyn <n 


ny BfiD^ ita» D^annn "3 


yvbnn or6 rtro n^ntyn 


orw»n p«i np ppta 


mp& ■o ann nty inn 



Leaf 1 recto. 

L. 1. The conclusion of the preceding page is missing. Restor- 
ing it in accordance with the editio princeps it reads JVn ninBO] 
"Di D^ta tyi nrrty prntpnn ray tan^ ^ant^ ity unpnn. Here as 
well as in 1. 2 the text speaks merely of Levites, but the ed. pr. 
and Sh. both read TWO VD in place of D^. The version in the 
Genizah text is the same as that preserved in Pesikta Babbati 
on Isaiah XLIX, 14 ed. Friedmann p. 144 a. The Pesikta 
is the source of the quotation found in Midrash Tehillim on 
Psalm CXXXVII, ed. Buber p. 262 b or rather in Yalkut ad 
loc, v. ed, Buber note 1 on Psalm CXXII. In these Midrashim 


the act of self-mutilation is attributed to the Levites and not to 
the Bene Mosheh. There is evidence of direct connection between 
this passage in the Pesikta and the reading preserved in the 
text of Cod. Oxf., the words nnn T» (Mil. p. 18) and the con- 
cluding quotation from the Bible DK (read mri *IDK) mri n»N 
*W mwn D^BTP imPK being the same in both. It is, therefore, 
doubtful, to say the least, whether the original text of the Epistle 
made any mention of the Bene Mosheh. 

This doubt is not removed by citing the reference to the 
Bene Mosheh in the answer of the Gaon: nmi D^SS ntMD "03 *D1 

rbxw tm»n b"\ i^nm d^bik p» '»i« n&« an 1 ? ^p& jvano 

'131 Kim DW TWO ^n D^ TSanaOl In the first place the 
authenticity of the text at this point has not yet been definitely 
proved. The words 7\WC> 'Ol in addition to D^l^ would seem to 
be a gloss. Moreover the substitution of the W» *H for ti")b 
in the reply may perhaps have originated with the Gaon. On 
the other hand I do not attribute any importance to the omission 
of the Bene Mosheh episode in the St. Petersburg fragment (pt. 
II p. 10 1. 3 — p. 13 1. 3, Mii., p. 48) which contains a recension 
of our Responsum. The omission of this story may be explained 
by assuming that the compiler of the Eldad legends had no 
interest in the Halakic parts of the work and consequently, stopped 
at the very point where the Halakah begins, overlooking this 
legend, which comes after the Halakic part of the Responsum. 
It is therefore unnecessary to assume, as Miiller has done (v. 
p. 51) that this episode was missing in the copy from which the 
compilation was made. 

For the identification of the "Sons of Moses" with the 
Levites see my article in the Jeivish Encyclopedia s. v. "Moses, 
Sons of" (vol. IX pp. 59 et seq.). On p. 59 in the first para- 
graph of the second column the reference "Midrash Eser Galiyyot" 
ed. Grunhut in Sefer ha-Likhutim III pp. 13 ei seq. should be 

L. 3 et seq. Drum DnTlljnSK T\T0b D^pno; ed. pr., Midrash 
I. c, Sh. '131 DirniyasK innr>i ; Cod. Oxf. ♦ rann. The 

intention merely, is presupposed in the fragment text; all the 
other texts, however, assume that the act of self-mutilation is 
carried out. For an account of the biting off of a finger as a 
theme in legend see Cassel, Mischle Sindbad p. 18 note 1. For 
the phrase: TBO Dm D'OID tt*W niJDXK c f. Mishnah,Bikkurim III. 4. 


L. 6. n^K, ed. pr. IKTJ not y$n as in E. p. 5 no. 7 and in 
Mii. p. 18. 

L. 7 — 10. yiDJ^ Wpy\. These words are missing in ed. pr., 
Sh. and Cod. Oxf. (v. Mii. I. c). It seems, however, that they 
form part of the original text. In the ed. pr. and Sh. the epi- 
sode of the cloud which carries the Levites away, is introduced 
abruptly. In the Genizah text the Levites are in the act of 
fleeing and the cloud helps them only to the extent of concealing 
them from their persecutors (11. 8—9). This version is the same 
as the one found in a MS. of Midrash Bereshit Babbati quoted 
by Epstein p. 43: UTVm tyl Dfrt? HD31 pyn IT T&bt\ «W JW 

tin "ij> ^bn te w^im tsw Tioyi mpn nnb T«m nmm Drrrvum 

'121 "ipnn. The G-enizah version in the continuation of the legend 
is also the same as that in the above mentioned Midrash. In 
both the Levites wander on foot to the ancient Hawilah (1. 9 — 10). 
The, Sh. and Cod. Oxf. add the more miraculous feature 
that the cloud bears them thither through the air: DKBttl pyn Km 

rfrbi nw DTmm nb^rb M^im Dipm mksi nn^n« ay. 

The flight episode seems to be part of the original form of 
the legend. An echo of it is found in the epithet Diy tana? which 
is applied to the Bene Mosheh in the various recensions and tra- 
ditions (v. Mii. p. 35, E. p. 15). The origin is perhaps to be 
sought in an attempt at an etymology for Sambation: (Distant? = 
DW Bit?). This conjecture of J. Theodor, Bereschit Babba p. 93. 
Commentary 1. 13, seems to be confirmed by the text of recen- 
sion H: dp bv p anptt parity bim bw o^in mi rwo ua? myi 
ryfc IDifcP. Compare, however, Neubauer, Where are the Ten Tribes 
in JQR. I. p. 101 note 2. 

L. 11. '131 Dlp&n U mil. This detail that they settle in the 
spot where the cloud rested, is modeled after the Biblical story of 
Israel's wandering in the wilderness. In this form it is found 
only in the Genizah text. It may, however, be an interpolation 
even in T-S. 

L. 12. DfrrVOH nnb VlOTfl. This is missing in Sh. and Cod. 
Oxf. In the ed. pr. it is amplified into Kb 11BD ttb 10« "nj>1 

nrrninKa irnnai ttTroK. 

L. 13. D)n occurs only in this text. 

L. 14. inDI ^VU 11« "ipM 1KTI— probably an allusion to Isa. 
IX. 1. What is meant by a great and "heavy" light which they 
saw in the morning? Is it no more than the dawn? In Sh. "DD 


is missing; ed. pr. reads ^>TJ instead of *VIN(?); Cod. Oxf. reads 
*D3 PJH UW nW inifcO iy»P1. The variety of readings would 
indicate an early corruption of the text at this point. 

L. 14—15. 'n -kti D^oni n^TDm nyvt onn (sic) mn o. 

This passage also occurs only in the Genizah text. 

L. 16 et seq. It is interesting to compare the texts here. 


bbw *ini mrm njn dm^di 

Mm pit 

Cod. Ose/. 
^n) von Qipo n« nniot? 

.^mi p« Wiii (d^d ns'ta 

ed. #r. 

d^sn Ww via on 1 ? awm 

^lity ini d:tqd n nwni 

Mm d^3k 

nwi seems to be an early scribal error for 2PDffl or 3DQi This 
error was, evidently, in the copy from which Sh. was made, hence 
the addition of DMD, to make the text intelligible. D^D HS^D *ta 
in Cod. Oxf. is due to a scribal error. The words occur later 
in their proper place. In E. and Mu. the reading of the ed, pr. 
is incorrectly given as IHiH nK UTib 3WI1. 

Leaf 1 verso. 

L. 1—2. iniK ps D^iy» im DP ttn *6p Dlp». The river 
thus created for itself a channel, where none had been before. 
This thought is contained in the concluding words tibw DlpQl 
DblJJfc VJi DP !TO which are, at present, incomplete and not de- 
finitely related to their context in the ed. pr., Sh. and Cod. Oxf. 

L. 2—3. *ipn v« iy ^ni nnnn d\dk W[i]ii; ed. #n mini 
d^jd ^n ^ini D^n« %ia py ainn and later w nw ^ ^ia mini 

d^d ns^ta dip ^n *?ini D\na yntwi; #&. ^n ^ini d-oik Wii pyi 

D^»; Cod. 0#/. D^O ns^ta ^n ^>mi D^inK Wli mint Significant 
here is the later addition of D^fc (DD^D) ^n in ed. pr., Sh. and 
Cod. Oxf. Among the numerous recensions of the Edad legend 
H. is the only one in which it occurs again (ed. Constantinople 
1516 — 17 etc. v. Mii. p. 12 and p. 68). Our fragment seems to 
mean no more than, that the river was so swift that its waters 
brought down stones and sand. The Genizah fragment thus proves 
that Bacher's conjecture as to the original form of the legend 


was correct, v. Bacher, Agadah der Tannaiten I 2 . p. 291 note 1. 
Subsequently, however, legend made of this a river in which rocks 
and sand took the place of water. This is the idea of the inter- 
polation D^fi (D^tD) "ta. It is doubtful whether the words *?y 
Ipn pK (Job IX, 10), found only in our fragment, form part of 
the original text. 

L. 3—4. nnin pirns^ ly ynw bin* b)p )b tn (comp. Ezra 
III, 13). Ed. pr. bm b)p) brn »jnn; Cod. Oxf bm »jrQ; Sh, 

ni^pi tf pn. What follows in the ed. pr. inn JWB 7Vfi KtabNty 

lSBio n\n bnn ta>; Cod. Oa:/ tmswa rnn bnj inn pis KtabKty and 

#ft. 13S3» rWl ^tnn bv in l^SK Dfc? n\n I^W is manifestly a later 

L. 4—8. toy «to taijn py rnn^D nmno^ |»y xbnxv nntyni 
ian imp nnta> wia ny mipfi «Vi inn pinn 1 * «to v6k ewfe tn« 
mw; edjpr. rty Tin ntt«9»n pa «wn dv nnyw njwm na nnt?m 
nnt? ^«ta ip i>Vm fcswb te* di« jw py; £ft. nntrn bin v tan nn 

ni (7od. 0#/. preserves only the concluding words of our 
passage nnt? WIB iy. This fact was overlooked by Mu. (p. 18); 
as a result he reprints here an unintelligible text without 
any comment. The conclusion of this passage is found in the 
Cod. Oxf and in the Recensions of the Eldad legends (Mii. pp. 68 
et seq.) after the description of the river Sambation. The text 
there is very much like the text of our fragment mym ni nnt?m 
ato nbw uk «to vbx &xb toy dik jw jtyy aban&i py vby Tin nnt? 
utoat on. 

A very surprising feature in our fragment is the omission 
of the characteristic element in the description of the river, 
namely, that it ceases to flow on the Sabbath. This is particu- 
larly surprising inasmuch as this detail is old in the legend. Cf. 
Pliny, Hist. Nat. XXXI, 2; b. Smth. 65 b, Gen. Bab. XI, 5 ed. 
Theodor p. 93, and the literature there referred to, and Krauss, 
Lehnworter II p. 369. To the references given by Krauss there 
should be added ItsniOD — the reading is uncertain — which is 
found in a Hebrew inscription, cf Renan, Mission de Fhenice 
pp. 192, 856 and Chwolson, CIH. no. 64 p. 102. 

L. 8. Dwono iro« jwip Distal jvaniD im ini« p*mpi. The 

ed. pr. has Wtsno iniK pip liKI p^nD 1DW Kffil; Sh. Klpi mto 
)Vnnty Ttt; Cod. Oxf DU^taiD iV plpfl jvano VO n\T. The reading 
of our fragment is, therefore, not novel. In support of Bacher's 
view (I. c. p. 290, note 2) that p^nD represents the original form 


of the name and that p n Dl»D and )VD23D were derived from it, 
we cite the parallel development in Greek. "In certain districts 
where Greek is spoken there has been developed since the period 
of the xotvYj a homogeneous, nasal sound before accented, un- 
accented and aspirated consonants standing at the end of the 
syllable. As a result 2aj3(3aTts has become Sa^pa-cU, 2aji.{3aTsis, 
Sajxpaxt?, 2a^pado5?, Savpaxto?", etc. (v. Karl Dietrich, Unter- 
suchungen zur Geschichte der griechischen Sprache 1898 p. 92 et 
seq. ; comp. also Schurer: Geschichte III 3 p. 428). 

The origin of the word is still uncertain. The most probable 
etymology is the one which traces it to the Ethiopic " Saribat", 
corresponding to the Hebrew nafc?; comp. J. Halevy, Te'ezaza 
Sanbat, 1902, p. III. 

L. lo—ii. n»K dw nnmt? m»ip» mn m*D w\ The same 

reading is found also in Sh. 'D nm X)7W man imKD m»lpD Bhl 
rtDK. Originally this was meant for the dimension of the widest 
part of the stream. The insertion of the negative particle in 
the ed. pr.: (not nnb E. p. 6; Mii. )b ]*»& man im«D ni»1pD m 

p. 18) n»« d"wd nnn and Cod. Oxf. n»« d«w am law ♦ * . ♦ * 

makes this the measure of the narrowest part of the stream. 
The other Recensions of the Eldad legend (Mii. p. 66 below) tell 
that throughout its course the river is "a bow shot" {Gen. XXXI, 
16) in width H. e. 200 yards". H., however, makes the width of 
the river "220 yards". 

In the ed. pr., after the width of the river has been stated, 
we have the following: DnfiDB )*Ofi ttfctt DfcPD man TO O^Diy am 

Dmb ybw iaK *6i mby *pp» manp ^as» Dnioa am £« ay "6k 
dpd nmb ?b)y an *6i d^«. The words nniD« Dm allude 
to Fer. &w7i. X, 5, (Is. XLIX, 9 etc.) totf *6k IKS DmD*6 1»fc6 
)VtDniD maa D^>; comp. also £fta7i Bab. on II, 2; Pes. Bab. 
XXXI ed. Friedmann p. 146 and Recensions in Mii. pp. 66 — 67. 
The purpose of the words: "D1 DiTty *)"»p» imn» s aSD DniDK Dm is an 
attempt to explain why it is impossible for the inhabitants of the 
opposite banks of the river to hold intercourse with one another. 
These words, however, are not found either inSh. (man liTlfcO niDlpD tTl 

Dmb« ia« *6i la^K k^-d^d' 1 Daw p^i n»« 'D nni *ont?) or in 

our fragment. It is possible that this explanation has crept in 
here as a gloss from one of the Recensions of the Eldad legends 
(v. Mu. I. c. D, H, J, 8 n ; G. 3 d ). Furthermore the words am 

)b» dj; )bx d^bdidi ]«d» law d^d man -ran Dnisiy which occur in 


the ed. pr. and in the Cod. Oxf. (D0B D^SDDl 1H« 1X12 DHDiy DH1), 
but which are missing in Sh. and in our fragment, may likewise 
have their origin in one of these Recensions, (v. Mti. I, &, especially 
D. 8 n and ih B. 19 a ). 

If the description of the width of the river was followed in 
the original text of the Responsum by another sentence, — which 
is not very probable in view of the reading of the Genizah 
fragment — such a sentence must have been similar to that of Sh. 
DiT?K 13N fc6l Whs Xlb U^T D^« pb). This was then amplified 
in the spirit of the later more highly developed legend by the 
introduction of material taken from the Recensions, until it assumed 
the form in which it is found in the ed, pr. and Cod. Oxf. 

In the ed. pr. and in the Cod. Oxf. another extended passage 
follows this point: 

Ed. Pr. 

n»m fc6i np rm n.wn pai 
■o n« v dwyi D^pp *6i n«JDis 
nwim Dipm d^ks (sic) dn 
rem ite n« "6k ltoran dtmi 
dthv vn *6i nan pmnn Dn^> 
,n&nt& p ^n 

Cod 0s/. 
njn rvn xb (sic) on^n *6i 
d^ddii n^pty *6i naioD n»nn *6i 
)Wim i:6 mpni Di«x dk a 
on 1 ? nsoi 1^*6 i^k i^wi pjnro 
(?)nt?» •on nroo ^t? rra pun 

♦dp n» pyTi'' uk ]w 

In £7i. it is much briefer nnintD nvm nionn *6k DiTaa JW 
D,t6n ns tyn^ '« rmn n^ai^l. Our fragment contains only the 
concluding words of this passage v. 1. 11 — 12. 

L. 11—12. jwih pinnt? •w n^n pin py™ d^k d^sbtj n^i 

1^, which makes good sense in connection with the preceding 
and does not suggest any lacuna. There is reason to believe 
that the statement that there are no unclean beasts or reptiles 
in the land of the Bene Mosheh has crept in here from one of 
the Recensions of the Eldad legend, perhaps from H. (Mti. p. 63 ; 

cf ib. 8 c - d naoto n^n *6i k&b *)iy *6i (sic) n«ata w ddj; )w 

am ♦ ♦ ♦ . ♦ ♦ ipm ]tre t6n • t^iyns *6i nut *6i njn n^n k^i 

"iDi nnsipi D^jrm. 

It is probable that Recension H was, at an early date, 
prefixed to the Responsum; the fact that it is printed in this 
way in the editio princeps would point to an earlier practice of 
the kind. This circumstance may have suggested to the copyist 
or the compiler amplifications of the Responsum. The instance, 
however, in which the interpolations in the Responsum are more 


closely related to Recension H than to any other Recension are 
not sufficiently numerous, and the agreement, between the inter- 
polations in the Responsum and the text of H, is not striking 
enough, to justify the conclusion that H is the source of the 
interpolations in the Responsum. It is equally probable that 
the real source is a Recension which is no longer extant. 

In the Recensions the account of the agricultural life of the 
Bene Mosheh is followed by an account of their study of the Law. 
This explains the following passage which occurs in Sh. at this 

point: dpp!?k ns tynt? '« mm dh&^i. 

L. n—12. "Di w rvn pnn d^jw WK n^ntyn rfon. To whom 

Tlbti) refers, is not clear. The text in the fragment is nevertheless 
clearer and simpler than the corresponding text in the ed. pr. 
or Cod. Oxf., quoted in the preceding note (q. v.). Sh. has an 
incorrect reading: )W«1 JTOn pina D^Y 1 171 tib )1 *XXL 

L. 13. )W*n IVO pim. It is to be observed that ed. pr. 
and Cod. Oxf. both have the incorrect reading m pin *in*6 
•»iff. Epstein emends to ]W*n. 

ieci/ 2 ratfo. 

L. 1—3. nn« nip» bnnbty n^nm *?*ntj" p*6ty n^nn 

nn« pfc^l. The verb is evidently missing before nn« rtttflD; ec?. 

_pr. and £/*. read nn« mm a^it? iw p« wm Vnn ^ddh nm* 

n*l )1p s m. The phrase 11 ])p*r\5 in the ed. jpr. which probably 
means "with the utmost exactness" or "in the recension of Rabbi 
[Judah ha-Nasi]." is missing in T-S and Sh.\ it seems to be 
interpolated in the ed. pr. 

L. 4 et seq. W *3 ^71 DiTTO (read BT») W\ TiD^n \>tyb bm 
"D1 DJJtt nt^. What the Gaon meant, is now clear. The reasons 
which led Reifmann (p. 287) to doubt the authenticity of this part 
of the text, are now without force. The Gaon means to say that 
the Babylonian and Palestinian scholars use the one and the 
same Mishnah. In the exegesis of the Mishnah, that is in the 
Talmud they at times disagree, just as individuals differ in their 
interpretation of the Bible. This is the thought behind the 
obscure words of the ed. pr. im DJ>& Dn»1K i&n DH^fi D"»y^1 

nsii nt r&p&n is *np»n pr6 pntyvn n^n w ibd nn« nyv> )bbr\) 
•ma ayta i 1 ? nana nn ina nyv )b. Sh. reads here only w in 

D^Ttt^ni which is perhaps an old error for DHID^n. 


L. 9 et seq. "Dl STO! pynp JfcP *)yK pYI]*Op»3. The MS. 
reading is KlpDS. The plural ]fW would however point to the 
reading XYl^lpDl, which is the reading in the ed. pr. Moreover 
in the latter the connection with the preceding is indicated by 
"'SKI. In general the text of the ed. pr. is clearer at this point: 

btmr p*6 bn p ]m W w mron nynp pp m*npD3 "•ski 
n^piDsn Tinnn 'tiddii nrayts ^pra^m mavoai "imnsni nmtoi nnonn. 
#7*. reads: nrainoi mmns mm nnonn n«iiD msipon i^s«i 
a^Dsn Tinrn mioom. 

L. 11. nvrrPI nmiD. The reading of the ed. _pr. is nnonn 
HTPJYEH. filTHD is certainly a scribal error for rrnon and may 
possibly be explained as due to the copyist to whom it was 
suggested by the Talmudic phrase mHD rrairiD (v. Tosafot Me- 
nahot 32 a s. v. fcOTK Km, Miiller, Masseket Soferim p. 30), the 
word nraiHDI occurring in the next line (1. 12). 

Leaf 2 verso. 

L. 1. nniDDI, ed. pr. abbreviates ""ttDDSI, intending the plural 
and not the singular, the reading adopted by E. p. 7 and Mu. 
p. 19 and actually found in Sh. 

L. 2. D^plD^D, ed. pr. and Sh. read tfplDSn. The form D^piD^S 
is found in the Massoretic notes in numerous Yemen manuscripts 
of the Bible, cf. Ginsburg, Introduction to the Hebrew Bible, 1897, 
pp. 72 note 1, 85, 104 note 2, 106; idem The Massorah, compiled 
from manuscripts, 1880—1885, II p. 450 nos. 189 et seq.; M. H. 
Hyvernat, Le Language de la Massore in Revue Biblique Inter- 
nationale. Nouvelle Serie II, p. 519. 

L. 3. UNS^ ra pray prayi {Eccles. VII. 24); ed. pr. pray 
tafyT ra pIDy (comp. Joel II, 11); Sh. breaks off at this point. In 
the ed. pr. the connection is here interrupted by the following 

vty rajw rwre m& ^nm at* ro n^w pmn Wo uw no^ bn 

m«n *)W myian THH miBl That this passage is a gloss is 
clearly evident from its position in the text. The style, which is 
that of later philosophical literature, gives evidence of the lateness 
of the interpolation. The Gaon has already explained the 
divergence of the Eldad Halakot from the current Halakah on 
the ground that there was no absolute uniformity in the exegesis 
of the Mishnah. In view of this explanation it is superfluous and 


almost inconsistent to seek further excuse for this divergence by 
attributing it to the lapse of Eldad's memory. 

In the ed. pr. there follow at this point the words bn« 

yn:6 p« tt»»i "•oin^ "»« vby *on nn« min rtiwnn which are absent 

in the text of the fragment. The continuation is again the same 
in both texts. The words '131 «\n nn« mWl TOPOn to« are a 
statement made above that there exists only one Mishnah and 
that its text is authoritative. The genuineness of this passage is, 
therefore, questionable. 

L. 5. MS. tin, ed pr. bm *ana i^n *6) i^n hsid nnx yw 

"71»^rQ *6« (]1Bp 1^3 *y6n K^l. The words in parenthesis seem 
to be a later amplification of the text. 

L. 7. The text of the ed. pr. is here less clear: pK "Oil 

L. 8—9. Tia^n on 1 ? t^ n^n^n "o Dm»Kt? n»i. The style is 
here that of the Responsum. This passage is missing in the ed. 
pr. and the omission results in obscurity. 

L. 11. YTTD, technical term, "to edit" (redigieren, v. Bacher, 
Die Exegetische Terminologie, Vol. II p. 133). The text of the 
ed. pr. is corrupt, the reading being nns. The last three words 
in the passage UYIK Dn'Ofi DHt^ BTTIpn ]Wte TlD^Tin D^niS seem 
to be superfluous and suggest interpolation. 

L. 12. The ed. pr. has "131 *tb D^n DP U )W Htt. 


We publish here another Genizah fragment, belonging to the 
Eldad legend: 

T-S Loan 39, University Library of Cambridge. It consists of 
one leaf of paper, 19x14 cm., written in Spanish square characters, 
tending towards the cursive. It belongs to the fourteenth or fifteenth 
century. The vocalization of some words, some corrections and 
the completion of some of the abbreviations are, however, in a 
smaller and different hand. This fragment appears to be a stray 
leaf of an early ritual code which probably belonged to the 
Gaonic period. 

The subject of discussion on p. A. (recto) is nbni and HBlto. 
It concludes with the "catch word" at the bottom of the page 


y*?lto JWW which indicate the subject dealt with on the following 
page. I have been unable to identify this part of the fragment, 
but its pregnant style, suggestive of the decisors (D^pDIS), and its 
use of Aramaic would indicate that it formed a part of a Halakic 
code belonging probably to the Gaonic period. The superscription 
indicates that the KIHP st ?nB p T\b* '1 W2TO nWTW JTD^n 1SD 
pfi began on p. B (verso). The ritual of Eldad was preceded 
by a brief introduction, giving a summary of his life and of his 
stories. It is a part of this introduction which is preserved 
on p. B and which is published here, 

The same account, which serves here to introduce the Ritual 
of Eldad, is contained in Cod. Firkowitsch 1261 (St. Petersburg) 
and published by Miiller pp. 47 et seq. There it is in the form 
of an Epistle written by the authorities of Kairwan to the Jewish 
communities of the Diaspora. The account in this Genizah 
fragment is, however, intended as an introduction and not as a 
letter. For this reason there are omitted here such passages 
and phrases as would be in place only in a letter. Thus there 
is left out the epistolary heading D"0ptt brtpft WpnY| T3 JO 

poy "rtwb jnu ♦jKiTp nn&n omn nyn •uwp] bto^w 

"Dl ^na in YT[>K] or passages like nffil f»T\ Ht t6k "A l^Wl 
131 nt?y»n rrn pi p Wl nt 1^ TJm ]K3^ liwa For the same 
reason the first person of the epistle is changed to the third 
person of the narrative. Thus tsvty 1. 6 is used for li^V; 
M^m, ib.— should read Dw6tWtl— for ttlTS^Wn; tiTTCMK 1. 7 for 
U^IK etc. This change of person is maintained consistently only 
in the first part of the letter. That the epistle was the original 
form of our fragment becomes, however, plainly evident in the 
second part; comp. for instance 1. 9 M fiipn Htyy *ty) where 1BK 
is a later interpolation, moreover 1. 14 Tvtyl and 15 "WHIt, 1. 21 
li^lrt, 1. 22 Kb VSfih and the conclusion OTllTJ pp^ Kin 

Aside from this feature the fragment contains nothing of 
consequence. It is published here only for the sake of com- 
pleteness. Brief notes have been added. 


pa Minty ^na p tAk 'i tram* naw JitoWi tbd nt 

^nsii nty^i p Dm a^n&yn nnt? wai 1 

n^inn pan ^ nnA mya a^Ditsn 2 

rrra* Tyai av 'a ty av 'o a^iai nanpn 3 

ntrn ^nm dhjq mpi ^n:in a^ n^ tnri 4 

mnn^ im«in^ nfpri typ-oBo nam aysm in^ 5 

(sic) DD^anro nWn ^nn mya nn anty may 6 

di« us ptow rw Mnan ^ra anuria pan 7 

5in^D«i ntso byn "W ^nzs ^na py n\nty *w 8 

fra ty in« ins «ni aa riipri ntyy "'ayi ™ 9 

(read "til Jtt) T^ T»» D^ '1 WH1 anay ^« Di ^Pl 10 

••ittqsi ^yar» VoA ^y^n&y ny p« pan n 

(read ainn) ainn ninn am "dew ants6 Tianp ly "W 12 

a^asn n^n a^Tan antf "^«i aisi na pa 13 

niDa K\m p«s nn 1 ? atya ^gi aw ^ti 14 

]Anr *aa at? w*rn a^aya^ n^y lrrrw 15 

annrwi pi*n ant? annnai a^niK ^yn ie 

a^antyn i^« "o ™« Tiyi ana«i rnysa anty 17 

[?)na]wa tot? rny? naam nnin ana tr is 

anaann nKtwi roan pun amp 19 

i • m im ttyaawi isaa 20 

ponn t6]k "1 *ran 5 i nt in wjpwA uv6« 21 

[1]a« mno ttwra pp" 1 Kin a& toA nt 22 



L. 1.— The reading of Pet indicates that -ft has dropped 
out of the text after ^nBi. 

L. 3— Ty*tt is here substituted for the longer phrase found 
in Pet "Dl t6« '~\b l^Wi v. supra p. 116. 

L. 13. — Miiller's suggestion that there is something omitted 
after the words Dnsi "HD is not born out by our fragment. 

L. 14.-— niDtt, Pet has the same reading. "Mecca" is evid- 
ently meant. niDD occurs for "Mecca" also in Jephet ben Ali's 
Commentary on Daniel, ed. by Margoliouth, Oxford, 1889, p. 136 
1. 21 and 144 1. 17 and in the Oxford MS. of Midrash Babba de- 
Rabba, Catologue Neubauer, no. 2399 fol. 46 b, v. Neubauer in J. 
Q.E. I. 114 note. 

L. 15.— D^KJW rvty IPrrW, "on account of which— the 
sanctuary of Mecca— the Mohammedans shall once tremble". 
This is an allusion to Habakkuk III 7 po "tow niJTT pUT, pB 
being identical with btfyBtr. 

L. 16.— D^niK *>yi. This reading would indicate that D^ruo 
in Pet is due to a misreading of D^HK '1 ==» D^HN "fyD. 

L. 18.— [pB]WB(?); Pet reads "TCh pKB. 


At the close of this book there is printed a list of corri- 
genda, based on the editio princeps, to the texts of it as printed 
by Epstein and Muller. These corrigenda are to both (1) the 
non-Halakic part of the Responsum and to (2) Recension H of 
the Eldad legend, printed there with the Responsum. The correct- 
ions to the part of the text reprinted above which have already 
been given (v. pp. Ill et seq.) are not repeated here. 



Ed. Epstein 

Ed. Muller 





Line 1. 





Line. 1. 







5, 2. 













„ 6. 







57 55 












„ 8. 

















„ 14. 

„ 5. 







,, 6. 







„ 12. 
„ 14. 


















5, 22. 
„ 4. 


n 57 57 *• 

77 57 57 J7 

„ 7. „ 1. 

35 55 3? rx * 

37 37 77 8 « 

p. 6. „ „ „ 15. 

35 33 

73 35 

33 33 33 

33 53 n 37 

9 2 

p. 7. „16. „ 4. 


33 33 33 '■• 

p. 8. „18. „ 5. 

33 33 33 35 


p. 19. 

p. 20. 

„ 9. 

„ 16. 

33 37 

„ 18. 



Ed. Pr. 

•flaTJ, not *xwn 

w\nb „ x^n { ? 

13-qV „ i:na^ 

\vnv i"bb\ „ a^nbi 

inas a-anto del. 

omit, irrasrto after irrbsnto 

bHTWP not ^KIBT^ 

mrpa „ T'ab 

omit ana after aa^ 

yaniya „ o^aniya. 

D^vr „ a s jnv 

B"KSW not ^K*n 

jwat? „ pjm» 

minn „ rron 

nafr „ laV 

]"^>B „ ^IB^B 

omit li^> after itfb 

nai^> „ iat6 

mw» „ mw» 

nstpaa „ cnspa 

wipni jva „ p"anan 

*JK\T not "pKJl 

nna „ nnan n« 

pip „ pntp 

•\b „ mb 

DnBDO „ D^BDErt 

D s niD« „ pIDK 

D/T-ina „ nmjn 

pDv „ popa 

second D'naiN delendum 

^pio^Bai not "pDBa) 

nbyn „ nyn 

n^K3 newa a^y «-ia jvu6 n^nn ateai an. 


J5d. Epstein 
p. 55. no. 1. L. 

p. 56. 

5-6. p. 55. 

Ed. Muller 

.53. no. 1. L. 

33 !? 35 
57 2. „ 





Ed. Pr. 

ar6ai not n&r^yi 
thmv „ wbvw 


Ed. Epstein 

ny ity nnwrb to^oin t6i 
"pna lawn *b aviai Q^iy 

no. 4. L. 2. 

„ 5. „ 1. 

» » n 6 - 


„ 6. „ 1. 

» » »? "' 

» » » 3 ~ 4 - 

jrass toaa nxm D"pV| 

na nny [ena ma 1 ? nayo 

/nn:a p^av «ns 

no. 6. L. 7-8. 

p. 57. no. 7. L. 2. 

p. 58. 

p. 59. 














Ed. Muller 

ij? Tip nnw^ lE^oin t6 

n,nna awn t6 awa^ n^y 

,'wi ram "W ntn 

•jK\"n my ntn nnna uwn t6 

p. 57. 

no. 4. L. 



omits D^DUl 

after Detain 
not ijpjnt* 


n s a*w 




„ K'TIK 


„ trpoini 

„ 6» „ 5-7. 

naya nnss ntoai n"pb] 

^is na "nny [»ia nrtiS 

/nnitt |iVav» 

no. 6 C . L. 4-5. 




p. 73. 

8 a . 

8 C . 
8 d . 

nay& .mass train nana n«p^ 

"i^av ^is na nny w^ ••nm^ 


omtdi mai n^nn&i nvw 

not D'Dnnsrt maa onanai nine? 

^n not bn 

bn „ n^n 

an^> delendum 

omits ^n maa nm 

omit n"yo after nan 

omit ifcuw after *?n3n 

V '«D delendum 

aio" 1 not nw» 

yano „ yanta 

nsate „ x&ta 

cwppi nnap t» !?di 

p. 65. no. 8S. L. 
8 h 

notnva&p nna Vai wwp Tia *?ai 

9 e . 


10 b . 
10 c . 


10 e . 






itoi iKia not nam «ia 

omit nann before itaa 

1«T not 1KT1 

nyana „ nyanto 

D*YDt6 „ 0^10*6 

^nbtaa „ waoa 
omits l^p 

nna „ mna 

omit DW before D1K 

ma^aoi? not nwaoatf 

Ann „ Mn 

nt "jmK nttDl delendum 

nntaaa not naa 

nnain „ ontan 

^tatt „ na* 

omit ini« tiatri after pffl 

now not ntawn 


Ed. Epstein 
no. 10. L. 

» » » 
- 11. „ 

p. 60. 





Ed. Mull 




no. 10 e . 





not TQ 

not naiy 

„ 13. 




„ n&nte 

» 5J 



„ raw rat? 
„ nura 

„ 15, 



















Printed by W. Drugulin, Leipzig.