Each new student who comes to the study of the Arabic text of the Koran is more or less perplexed by the problem of the mystic letters which stand at the head of many of the Suras. Thus we find:¹

ALR    at the head of Suras 10, 11, 12, 14, 15.
ALM    "       "    "    "  2, 3, 29, 30, 31, 32.
ALMR   "       "    "    "  13.
ALMS   "       "    "    "  7.
HM     "       "    "    "  40, 41, 43, 44, 45, 46.
HM'SQ  "       "    "    "  42.
S      "       "    "    "  38.
TS     "       "    "    "  27.
TSM    "       "    "    "  26, 28. 
TH     "       "    "    "  20.
Q      "       "    "    "  50.
KHY'S  "       "    "    "  19.
N      "       "    "    "  68.
YS     "       "    "    "  36.

And after listening to the variety of interpretations one's Arabic Sheikh can provide for them, one is inclined to say, as Mohammed himself said of other matters in the Koran, "none knoweth the interpretation save God."²

Both Moslem and non-Moslem scholars, however, have been diligent in seeking an interpretation, and the inquiry has exercised such different types of mind as Avicenna³ and Siyuti⁴ among Moslem scholars, and Sprenger⁵ and Franz Buhl⁶ among Christians. It would be profitless to set out in detail the speculations of Moslem commentators. It is sufficient to notice that they fall roughly into two types, (1) those who treat the letters as mystic signs, and (2) those who attempt some rational interpretation of them. The Turkish translator of Ibn Khaldun, e. g., represents the first type. "God," he says, "has placed these letters at the head of several Suras as a sort of defiance; it is as if He had said to them, ‘These are the elements of which the Koran is composed, take them and make of them a book equal to it in style,’"⁷ or according to Zamakhshari as Ibn Khaldun quotes him — "They indicate that the style of the Koran is carried to such a degree of excellence, that it defies every attempt to imitate it; for this book which has been sent down to us from heaven is composed of letters. All men may know them equally well, but this equality disappears when, in order to express their ideas they want to use these same letters combined."⁸

A variety of this type is the attempted symbolic interpretation of them. Thus Siyuti⁹ quotes some who held that Q stands for Qaf, the mountain that encircles the earth, or for the sea on which the throne of the All-merciful rests. Others again, he tells us, find a numerical symbolism in them, e.g., ALMR equals 271, to which various mystical significations can be attached,¹⁰ while others thought they were mystic words of heavenly language by which Gabriel used to call the Prophet's attention, or words which the Prophet used to call the attention of his hearers. Others still found in them specimens of the heavenly original of the Koran, or exhibitions of the phonology of the Arabic language.¹¹

An example of the second type is Siyuti, who in his Itqan¹² gives various suggestions as to the possible meanings, basing his remarks on Ibn 'Abbas and other early authorities whose works are now lost to us. Thus he quotes Ibn 'Abbas to the effect that the letters KHY'S at the head of Sura xix stand for the five attributes of God,¹³ Karim (gracious), Hadi (the Guide), hakYm (the wise), 'alim (the Learned), Sadiq (the Righteous). So ALMS stands for Ana 'Llahu 'rrahManu 's-Samad, (I am God the Merciful the Eternal). Baidawi also follows this line of interpretation, e.g., in his commentary on Sura xiii he says that ALMR means Ana 'Llahu a'liMu w'aRa (I, God, know and see).¹⁴ Under the same group would fall the suggestion that the letters signify the rhyme or rhymes on which the Suras were originally built e.g. the N Suratu, 'l-Qalam (lxviii), would indicate the -un, -in rhyme that runs through it.¹⁵

Practically all Moslem interpretations insist that the mystic letters are part of the original Koran as it was revealed to Mohammed, generally basing their claims on the words that so follow the signs, "These are the signs of the clear book." (Sura xii, etc.), the obvious reply to which is that they always stand at the head of the Suras and never in the midst of them, though many are confessedly composite Suras.

Modern European attempts at interpretation may be Golius' suggestion¹⁶ that they were scribe's marks; thus ALM would stand for Amara Li Muhammed (Mohammed commanded me), and KHY'S head of Sura xix would be the mark of a Jewish scribe and stand for the Hebrew. KoH Ya'aS (thus he directed)¹⁷. The most famous suggestion, however, up to recent days, was that of Nöldeke, who in the first edition of his Geschichte des Qorans (1860),¹⁸ suggested that when Zaid ibn Thabit was at work editing the Koran for official publication, and was piecing together the fragments received from different quarters, he preserved in some of the more important fragments thus received, the initials of the persons from whom he received them. Thus ALMR might stand for Al-Mughira, or TH for Talha and so on. Later, however, he went back on this, and in his article on the "Koran" in the ninth edition of the Encyclopaedia Britannica¹⁹ he followed O. Loth's suggestion that the letters go back to Mohammed himself and have some intentionally mystic signification. Loth's argument appeared in an article in ZDMG for 1881²⁰ on Tabari's Commentary on the Koran, where he criticises Nöldeke's earlier "monogram" theory, and states his own, opinion that they were due to Jewish influence (seing that they almost invariably appear in Medina Suras, where Jewish influence was strong, and not in Meccan Suras, where it was practically non existent), and of the same nature as the mystic figures and symbols of the Jewish Kabbala.²¹ Loth also thought that the key to the symbols would be found in the opening words of the Suras to which they are attached, a suggestion which, as we shall see in a moment, has borne most important fruit in more recent investigation.

Aloys Sprenger²² favoured the monogram theory of the letters, but not Nöldeke's variety. Taking the KHY'S at the head of Sura xix, he noted that this Sura dealt with the histories of Christ and of John the Baptist, and that it was the Sura which Mohammed's ambassadors recited to the King of Abyssinia. So Sprenger suggested that the letters there stand for a Christian symbol, much after the same style as INRI (Iesus Nazarenus Rex Iudaeorum) as used among Christians. Thus he would read 'Isa 'n Nasari maliku'l Yahidiyin, and taking the most prominent letters in each word²³ we find 'Isa gives 'Nasari gives S, malik gives K, and Yahudiyin gives HY; so KHY'S. All of which is very far fetched and improbable.

Hartwig Hirschfeld in the last chapter of his New Researches into the Composition and Exegesis of the Koran,²⁴ takes up again and makes a strong case for Nöldeke's original theory. Against the idea that the letters go back to Mohammed himself, he makes the conclusive point, that if that were so, then Mohammed must have had some share in the arrangement of the Suras, for it is obvious from a glance at the table of the signs that there is some connection between them and the numerical order of the Suras. And of course, all our evidence is directly opposed to the Prophet having had anything to do with the collection and arrangement of his revelations. Starting from the fact that there are only twenty-nine occurrences of these letters, and all of them at the head of composite Suras, Hirschfeld elaborates Nöldeke's original suggestion.²⁵ In every case he treats the AL of these letters as the Arabic article al and gives as his suggestions as to the names of the persons represented by the letters:—

M—Al Mughira
S—HafSa
R—Al ZubeiR
K—Abu BaKr
H—Abu Hureira
N—'OthmaN
T—Talha
S—Sa'd b. Abi Waqqas
H—Hudaifa'
'—'Omar (or 'Ali, or Ibn 'Abbas, or 'Aisha).
Q—Qasim b. Rabi'a.

Hans Bauer made a new beginning at the solution of the problem in 1921 in an article in vol. lxxv of ZDMG.²⁶ Otto Loth in the article we have already mentioned laid it down that the letters were certainly to be considered as abbreviations of well-known Qur'anic expressions, and himself made certain suggestions of a solution in this direction, e. g., that ALMS stood for Sirat AL-Mistaqim,²⁷ but his method herein is almost as arbitrary as that of Sprenger, and lies under the serious condemnation that it makes the symbols more unintelligible than the words for which they are abbreviations, whereas the essential point of an abbreviation is that it be as clear and intelligible as the words for which it stands. Is there any more certain way of interpreting them as abbreviations? Bauer thinks there is, starting from the fact that four (or perhaps five) of these mystic letters, viz YS, S, Q, TH, (and maybe N) are at present used as Titles of the Suras in which they occur, he writes, "Now the Titles of the Suras for the most part consist of more or less striking catch words which are taken from the Suras concerned, so we would consider the above cases as abbreviations of ingenious catch-words also," He then goes on to give some suggestions as to possible solutions. Thus,—

YS of Sura xxxvi he interprets as an abbreviation of YaS'a (he who runs),²⁸ of verse 19.

S of Sura xxxviii is similarly the Safinatu of yerse 30 "the chargers" which so interested the Moslem commentators (e.g. Baidawi v. 18).

Q of Sura 1, he refers to Qarinuhu (he who is at his side)²⁹ of verse 22 and 26.

TH of Sura xx he takes as two names, not one, the T is for Tuwa of verse 82, the holy valley in which God appeared to Moses; and H is for Harun, the biblical Aaron who is mentioned several times in this Sura.

N of Sura lxviii is for majNin (the demented) which in a way is the subject of the Sura.

Now this is a very big step on the way, but it is not quite satisfactory, in that it still leaves some things un-explained, and has no consistent theory as to the reason of the signs being these.

The most recent investigation is that of Eduard Goossens in an article "Ursprung und Bedeutung der Koranischen Siglen" published in the latest fascicule of vol. xiii of Der Islam. - (pp. 191-226)³⁰ Goossens agrees that the signs must be considered as abbreviations, and the work of the collector, or may be some later redactor of the Koran, but he claims that he has a consistent principle for explaining them, a principle that will do away with the arbitrariness of previous suggestions as to their interpretation.

Seeing that they all stand at the beginning of Suras, and never within the Text, he makes the first point, that they may be conceived of as some sort of introduction to the Suras. They would thus be abbreviations of some such technical introduction as the Bismillah. Now what sort of a technical introduction besides the Bismillah are we likely to find at the head of a Sura? Goossens points out that each Sura already has such in its title, e. g. Sura ii is called "Al-Baqara," Sura vi "Al-Anfal," Sura xii "Yusuf" and so on. Here he finds the key to the problem. We know that the present names for the Suras were not always or universally accepted, for there is still good tradition for the currency of other names for some of them; e. g., Sura ix in our editions is called "At-Tauba," but there is another name "Al-Bara'a" current in Tradition; Sura xvii "Al-Isra'" is also known as "Bani Israil," and Sura xxxii, "As-Sajda," as "Al-Madaji'," etc." Also at the present day the names differ somewhat in different parts of the Islamic world: e. g., Sura xl, which in the east is called "Al-Mu'min," is widely known in the west as "Al-Ghafir," and similarly Sura xlvii, "Mohammed," as "Al-Qital," etc. This variety of names, Goossens argues, must have been much greater in the past than it is now, when everything Islamic has become so stereo-typed, and so he proposes to regard the letters in question as remains of old names which once were current.³²

Supposing then that they are abbreviations of old Titles for the Suras, what principle are we going to use to solve the riddle of their meaning? Goossens says we shall probably find it by examining the titles at present in use to discover the principle on which they were given. Examining these, we notice, that with the exception of Sura i and cxii³³ the Titles are invariably chosen from words occurring within the Suras themselves, in seventy two cases it is the first word,³⁴ or one of the first words that is chosen, and in the remaining forty it is some striking word occurring further on in the Sura. This is a characteristically Semitic procedure,³⁵ and suggests that the same principle would apply to earlier selections of Titles. Now when we look a little more closely at this, we find the remarkable fact that of the Suras with mystic letters, only seven are found among the seventy-two named from the beginning of the Sura, and of these seven, four, viz. - xx (TH), xxxvi (YS), xxxviii (S) and 1 (Q), are themselves the Titles of their Suras, leaving us only three, viz. xxx (ALM), xli (HM) and lxviii (N). Of these, xxx and xli belong to groups, and lxviii has a special reason, as we shall see later; so the conclusion is forced on us that the seventy-two are so well and clearly distinguished by their naming that any other name for them hardly ever comes into question, whereas there is and has been much more uncertainty about those named from within the Sura.

An obvious objection here crops up. If these letters are abbreviations of old Titles, how does it come about that in one case (ALR) there are five Suras with the same letters, and in two cases, (ALM) and (TH) there are six of them? Is it likely that so many different Suras would have had the same name? Goossens provisionally answers this by pointing out that it does not follow that the abbreviation in every case is for the same name. Thus to take ALM, it might be used as the abbreviation of no less than fourteen of the present names of Suras,³⁶ all of which begin with ALM.

Thus we are ready now to take up the question of the individual signs. We are to seek the explanation of each one in its own Sura, looking for a personal name or some other striking word, key-word, in the Sura, and looking for it sooner in the body of the text than at the beginning. Goossens' interpretations are as follows:

N. Sura lxviii. Schwally had already pointed out³⁷ that there is an Arabic tradition in the Mufaddaliyat that this N is for Al-Hut — "the fish" from the North Semitic nun — fish.³⁸ Now in lxviii 48, Jonah is called Sahib al nun, and Sura xxi. 87 knows him as dhu 'n-Nun. Thus we are on very safe ground in interpreting N as an abbreviation for "Al-Nun," and this forms a key for the further solutions.

Q. Sura 1. of which it is the name. In Morocco at the present day this Sura is known as "Al-Majid," a name taken from the second word in the Sura. When we look at the preceding word we find that it is Qur'an, and here is our Q for us.

YS. Sura xxxvi. This Sura is used at Moslem funerals, and yet there is in the contents of the Sura no apparent ground for this usage. But when we look at Sura xxxvii we find it full of eschatological matter, just such as we should expect to be used at obsequies. Sura xxxvii is in fact the dies irae of Islam. How is it then that Sura xxxvi is used instead of xxxvii? Goossens suggests that they were once joined together. They both have the same rhyme, save for the introductory verses of xxxvii (1-11), and if we omit these introductory verses we find that xxxvii 12 joins right on with the end of xxxvi. Now in xxxvii 130 we come upon the name of one of the servants of God, Elias (Al Yas) v 123 or Al Yasin, v 130, a name unknown to the Arabs who made out of it Al-Yasin, (parallel with Al-'Amran) as is still done in Morocco. Here then is the Ys,³⁹ an abbreviation of a personal name and fitting in with our principle.

S. Sura xxxviii. The uniting of the two previous Suras left us with verses 1-11 of Sura xxxvii on our hands; Goossens thinks they originally belonged at the head of Sura xxxviii where they fit on e xcellently. The S would then come from the first word of the Sura, "As-Saffat" and its present use as the name of the Sura fits in exceedingly well with the theory; the S at the head of Sura xxxviii, which has no other name, being a tradition of the original place of the eleven verses there.

ALR. Suras x; xi; xii; xiii (ALMR); xiv; xv. These all (except xiii) deal with the same subject, the history of the Apostles whom God had throughout the ages sent to mankind to warn them, e. g. Sura x deals with — Noah, Moses, Aaron and Jonah.

Sura xi deals with — Noah, Hud, Salih, Abraham, Lot, Shu'eib, and Moses.

Sura xii deals with — Joseph and the sons of Jacob.

Sura xiv deals with — Moses, Noah, Hud, Salih, and Abraham.

Sura xv deals with — Abraham, Lot, Shu'eib (to Midian) and Salih (in the valley Al-Hijr)

Moreover they are known by the names of the Apostles — Sura x — Jonah; xi — Hud; xii — Joseph; xiv — Abraham; xv — Al-Hijr (i.e. Salih); and they all have much the same introduction. Thus Goossens thinks they once formed a Sura-group known as the Apostle-group, and called by the general name "Al-Rusul," i. e., the Apostles, and arranged in Order of their length. Later each of them became known by its own prominent Apostle's name, but the ALR attached to them all preserved the old tradition.

Sura xii stands a little out of the group, its introduction being the only apparent connection it has with them. It mentions no individual Apostle, but it does more than once refer to the previous Apostles in general. So maybe ALMR is but "Al-Mursal", a participle from the root RSL, of Mohammed is called "Al Mursal" in verse 43.

ALM Suras ii iii, xxix, xxx, xxxi, xxxii. Like the previous set of Suras these are arranged according to their length, and also probably once formed a group, which the redactor divided into two because of the unusual length of ii and iii, which according to his principle of arrangement needed to take a place near the beginning. For such a group the word Mathal forms the connecting link, and this would suggest for ALM the meaning "Al-Mathal" — the parable. We need not follow Goossens in his working this out in detail for each Sura.⁴⁰

ALMS. Sura vii. In verse 10 we come across the word Sawwarnakum (we fashioned you) which links on with the well-known name of God "Al Musawwir" (the Fashioner) found e.g. in Sura lix. 24. Now as Islamic tradition explains this S as "Al-Musawwir," and as we find in v. 179 of this Sura vii the statement, "Allah's are the most excellent titles," we are fairly safe in this identification.

KHY'S Sura xix. This is a composite Sura. Verses 1-34; 42-75 are connected by similarity of rhyme, and deal with the history of the Prophets, John, Jesus, Abraham, Moses, Israel and Idris. Verses 35-41 are an interpolation about Jesus in a different rhyme. Verses 76 to the end are also in different rhyme, and look like a later embellishment. This suggests that this long list of signs is also composite, and Goossens unravels them as follows: Y — Yahya (John), ' — 'Isa (Jesus). That clears up two of the sections. The third section v. 76 ff. is a fragment of a sermon of Mohammed to the Meccans, and was once known as "Ad'-Dalal" (the Error), so this gives us our S.⁴¹ But what of the KH? There is clearly nothing with which they can be related in Sura xi; but if we look for a moment at Sura xviii, we find its name is "Al-Kahf" and that is also a composite Sura containing stories much after the style of Sura xix. So the conclusion is that they formed one Sura, and the KH at the beginning of our group of signs is explained.⁴²

TH. Sura xx. If we take the two letters as an abbreviation of one word, there is no likely word in the Sura, but if we separate them, the two obvious words are those Bauer had already suggested, viz. T — Tuwa, and H — Harun. Prof. Grimme makes a suggestion here which may be an improvement on this, He thinks that the T of Tuwa, having been made the Title of the Sura, and not being understood, was given in pronunciation the feminine — ah ending.

TSM. Suras xxvi, xxvii (TS) and xxviii. This would look like another group, and Bauer notices that they all begin with the story of Moses, so that the M of at least two of them might be from Musa (Moses).

In xxvi Goossens finds the S in "Ash-Shu'ara'" "the poets," of v. 224⁴³ which has given the Sura its name, and the T in the South-Arabian Tud, the mountain of v. 63.

Sura xxvii deals mostly with Solomon, so it is natural find the S in Sulaiman, and the T would be for the Tair (the Birds) who play such a prominent part in the Solomon story.

There is much more difficulty with Sura xxviii and Goosens almost gives it up in despair. For the T he suggest Adh-Dhill" (the Shadow) of v. 24⁴⁴ or "Attur" (the mountain) of v. 29, 46, or "At-Tin" (the clay). S there is only "Ash-Shatil' (the Brink) of v. 30; and the M either Musa, or Midyan (v. 21, 22, 45).

HM. Sura xl; xli; xlii (HM'SQ)'; xliii; xliv; xlv; all having the same characteristic introduction. They form an eschatological group and doubtless the HM has the same meaning in each case.⁴⁵ (Bauer noticed that words for Hell, Jahannam, Jahlm, Nar occur particularly frequently in these Suras, and Goossens notes further that they remark on the dreadful Hamim (the boiling water) which is to be so essential a part of the punishments in the life beyond.⁴⁶ So he reads HM from that as perhaps a typical word for future punishment.

Sura xlii is peculiar in having in addition to the HM the group 'SQ. It is to be noticed that in writing the two groups are-always kept separate. HM of course would obviously link up with the rest of the group, and Goossens agrees with Bauer following Schwally⁴⁷ and Loth⁴⁸ in taking the 'SQ as an abbreviation of "la'alla 's-sa'ata qarib' (maybe the hour — of the hamim' — is near) of v. 16.⁴⁹

That is the complete list, and students of the Koran will greet it with mixed feelings. Some of the identifications certainly appear very dubious, and at times one wonders whether the author has escaped the arbitrariness of which he charges others. But there can be no doubt whatever that this is the biggest advance yet made toward the solution of the problem, and we cannot but feel that here at least we have the right line to follow if ever an entire solution of the mystery can be obtained. Discoveries like Mingana's of fragments of seemingly pre-Othmanic Korans⁵⁰ give us hope that there may yet come to hand manuscript evidence that will give us yet older tradition and more certain evidence. The present writer has tried this present theory on some Azhari-trained Sheikhs in Cairo, and while they are forced to admit the cogency of some of the arguments, they all fall back on the position that these letters are part of the very word of God written from all eternity on the Preserved Table, whereas this theory would make them out to be the work of man.

Goossens goes on to discuss the probable origin and date of the signs, and their relation to the ordering of the Suras, linking on again here with Bauer's paper in ZDMG. But this equally fascinating and important discussion, must be postponed for a later issue.

Cairo, Egypt.

ARTHUR JEFFERY.

¹ The arrangement followed here is that of Nöldeke-Schwally Geschichte des Qorans, II 68 69.

² iii 5, but see Baidawi thereon.

³ Arrisala Annairuziyya. Const. 1298. apud Hirschfeld New Researches, p. 141.

⁴ Itqan. passim. Nöldeke-Schwally iia, 70 ff. depends mostly on the Itqan for the account of Moslem interpretations.

⁵ Des Leben and die Lehre des Mohammad. 2 Bd. II 182 ff. See also Journal of Royal Asiatic Society of Bengal. xx p. 280.

⁶ "Monogrammerne i Qoranen," in pp. 32-34 of Festskrift i Anledning of Prof. D. Simonsens 70 aarige Födselsdag, Copenhagen. 1923.

⁷ Les Prolégomčnes, tr. Baron de Slane. iii 68, n. i.

⁸ Op. cit. iii 68. Baron de Slane notes that whereas Zamakhahari deals at length with these mysterious letters in his commentary on the second Sura, this particular remark quoted by Ibn Khaldun is not found there.

⁹ Op cit.

¹⁰ Sale mentions this method of interpretation in cap. iii of his Preliminary Discourse. (p. 64 to the latest edition, whose pagination is quoted throughout this article.)

¹¹Sale op. cit. p. 64.

¹² I use the Cairo edition in two volumes. A.H. 1319? It is true that Siyuti mentions various other explanations which belong more properly to the first type, but his own position I judge to be that of the second type.

¹³ Others, however, think they are titles of the Prophet. Nöldeke-Schwally II 71, from Itqan. Nöldeke's note here is interesting.

¹⁴ iii, 145 of my edition.

¹⁵ This seems a most unlikely solution, and I have been unable to trace its origin. My Arabic Sheikh in Cairo knows it well, and can enter into long explanations of it, but he does not remember on whose authority it is given. Rodwell also knew of it.

¹⁶ is in his Appendix to Erpenius' Arabic Grammar, quoted by Sale, Preliminary Dis. course. p. 64.

¹⁷ The Chandos Edition of Sale writes this as Cob yaas: which is repeated in the most recent edition, that of Sir E. Denison Rosa. cf. p. 65.

¹⁸ p. 215 ff.

¹⁹ Reprinted as the second essay in his Orientalische, Skizzen, 1892. (There he writes "At one time I suggested that these letters did not belong to Mohammed's text, but might be the monograms of possessors of codices, which through negligence on the part of the editors, were incorporated in the final form of the Koran; but I now deem it more probable that they are to be traced to the Prophet himself as Sprenger and Loth suppose... Mohammed seems to have meant these letters for, a mystic reference to the archetypal text in heaven. To a man who regarded the art of writing, of which at the best he had but a slight knowledge, as something supernatural, and who lived amongst illiterate people, an ABC may well have seemed more significant then to us who have been initiated into the mysteries of this art from our childhood. The Prophet himself can hardly have attached any particular meaning to these symbols; they served their purpose if they conveyed an impression of soemnity and enigmatical obscurity.")

²⁰ ZDMG vol. xxxv 588 ff. The section "Die Monogramme" pp. 603-610.

²¹ Hirschfeld points out, however, (New Researchs p. 242), "that Jewish mysticism of this kind does not go back as far as the period in which these initials were written."

²² Leben. ii. 182 n. i.

²³ It is curious to note that it is not necessary in a Semitic language to choose the initial letters for making an abbreviation. A Hebrew example is in the memoria technical OaNDaG, used to denote the books supposed by Jewish tradition to have been written by the "Men of the Great Synagogue." viz. Ezekiel, the Minor Prophets, Daniel, Esther. To form QaNDaG, they have taken the fourth Hebrew letter in Ezekiel, the second in "The Twelve", the first in Daniel, and the second in "roll of Esther."

²⁴ London 1902, pp. 14r ff.

²⁵ The most interesting piece of evidence in favor of this is the fact that in a British Museum MS of Al Dani's Commentary, the letters TH at the bead of Sura xx are followed by the common Muslim phrase "on Whom be peace," which of course could only be used in reference to persons. (This is evidently a reference to Ta-Ha used as a name for Mohammed, cf. Lane's "Modern Egyptians," chap. xxii, p. 426 of ed. in Everyman's Library. D. B. Macdonald.)

²⁶ "Über die Anordnung der Suren und über die geheimnisvollen Buchstaben im Qoran". His theory of the order of the Suras is as interesting as his theory of the letters, but does not concern us here.

²⁷ ZDMG xxxv p. 609. See the criticism on his suggestions in Nöldeke-Schwally.

²⁸ This was Habib the carpenter, according to Baidawi (iv. 186), who reproved their idolatry, and thus yes'a here could in some sense be looked on as a personal name.

²⁹ Baidawi (v. 93) interprets this in v. 22 as the angel who has charge over him and in v. 26 55 Satan.

³⁰ Goossens says that his theory was worked out in the summer of 1920 and talked over with his teacher Prof. Grimme, who encouraged him to elaborate it during the autumn of the same year as an Academic dissertation. Thus, though it has appeared later, it is chronologically earlier than Hans Bauer's discussion of the question.

³¹ Flügel gives them in the Table of Contents to his Corani Textus arabicus, Leipzig.

³² In support of this proposal he points out that a statistical examination of the Suras reveals that just those Suras whose names have varied the most are those which have an abbreviation at the beginning.

³³ Goossens' explanation of these two exceptions is that the first one was given its special name as the "Opening" Sura of the Koran, and Sura cxii, as a sort of Credo or succinct profession of faith, was placed last. Suras cxiii and cxiv, which are only incantations, were added later as a kind of Appendix. This explanation follows Ibn Mas'ud, as preserved in the Fihrist.

³⁴ As in the Hebrew names for the books of the Pentateuch, e. g. Genesis "Bereshith" "in the beginning."

³⁵ Prof Grimme suggests this, as a way out of the difficulty of ii Sam. i 18. If we take "the Bow" to be the name of the song taken from v. 22, we can read the Hebrew text without any emendation, "he bade them teach the children of Judah 'the Bow'," i.e. the song which then follows.

³⁶ i.e. v; xxiii; xxxv; xl; lviii; lxi lxiii; lxvii; lxx; lxxiii; lxxiv; lxxvii; lxxxii, cvii.

³⁷ Nöldeke-Schwally, II.70, Goossens confesses that it was this note of Schawally's that first put him on the track of his idea for the interpretation of these signs.

³⁸ Baidawi calls this Sura, Nun, and interprets it as "Al-Hut," v. 143.

³⁹ Goossens, however, has a tendency to favor Bauer's reading as not needing the combination of the two Suras.

⁴⁰ xxix is called "the Spider" and in v. 40 we find 'parable of the Spider." xxx is all about the signs and mathal of Allah, xxxi contains the amthal (plu. of mathal) of Luqman to his son. ii is also full of proverbs and apothegms, and iii may be named from v. 113 or v. 52, or the ALM might be from "Al-'amran"". xxxii gives the greatest difficulty, and if it be too great a stretch of the imagination to class it in the mathal-group, Goossens suggests its ancient title was "Al-Madaji'" — the beds, which tradition gives us as so alternative title for the Sura.

⁴¹ In Arabic script the difference between S and D is only the matter of a dot as also between H, Kh, J. and T, Z etc., and such diacritical marks did not exist in the old Cube alphabet, so the above identification presents no difficulty on that ground.

⁴² There are two other points in favor of this view — (1) that Sura xix shows the characteristic beginning of a new section of a Sura, but not of a new Sura, whereas Sura xviii has this introduction; (2) the length of Sura xix deranges the system of ordering the Suras according to length, whereas the uniting of them both brings it to the requisite length.

⁴³ Again the difference between S and Sh is a matter of dots only.

⁴⁴ So also between T and Dh.

⁴⁵ Note also they follow one another in order of length.

⁴⁶ Vide xl 73; xliv 46, 48; xii 34 (in connection with "friend," another meaning of hamim. cf. xi 19).

⁴⁷ Nöldeke-Schwally, II 73.

⁴⁸ ZDMG xxxv p. 609.

⁴⁹ Curiously enough tradition tells us that Ibn Mas'ud's copy read only SQ. which is got by dropping the La'alla which is only an introductory particle.

⁵⁰ Leaves from three Ancient Qu'rans. London, 1914.