The poetical forms in The Nights are as follows:--The Misra'ah or hemistich is half the "Bayt" which, for want of a better word, I have rendered couplet: this, however, though formally separated in MSS., is looked upon as one line, one verse; hence a word can be divided, the former part pertaining to the first and the latter to the second moiety of the distich. As the Arabs ignore blank verse, when we come upon a rhymeless couplet we know that it is an extract from a longer composition in monorhyme. The Kit'ah is a fragment, either an occasional piece or more frequently a portion of a Ghazal (ode) or Kasidah (elegy), other than the Matla, the initial Bayt with rhyming distichs. The Ghazal and Kasidah differ mainly in length: the former is popularly limited to eighteen couplets: the latter begins at fifteen and is of indefinite number. Both are built upon monorhyme, which appears twice in the first couplet and ends all the others, e g., aa + ba + ca, etc.; nor may the same a.s.sonance be repeated, unless at least seven couplets intervene. In the best poets, as in the old cla.s.sic verse of France, the sense must be completed in one couplet and not run on to a second; and, as the parts cohere very loosely, separate quotation can generally be made without injuring their proper effect. A favourite form is the Ruba'i or quatrain, made familiar to English ears by Mr.

Fitzgerald's masterly adaptation of Omar-i-Khayyam: the movement is generally aa + ba, but it also appears as ab + cb, in which case it is a Kit'ah or fragment. The Murabba, tetrastichs or four fold-song, occurs once only in The Nights (vol.i. 98); it is a succession of double Bayts or of four lined stanzas rhyming aa + bc + dc + ec: in strict form the first three hemistichs rhyme with one another only, independently of the rest of the poem, and the fourth with that of every other stanza, e.g., aa + ab + cb + db. The Mukhammas, cinquains or pentastichs (Night cmlxiv.), represents a stanza of two distichs and a hemistich in monorhyme, the fifth line being the "bob" or burden: each succeeding stanza affects a new rhyme, except in the fifth line, e.g., aaaab + ccccb + ddddb and so forth. The Muwwal is a simple popular song in four to six lines; specimens of it are given in the Egyptian grammar of my friend the late Dr. Wilhelm Spitta.[FN#444] The Muwashshah, or ornamented verse, has two main divisions: one applies to our acrostics in which the initials form a word or words; the other is a kind of Musaddas, or s.e.xtines, which occurs once only in The Nights (cmlx.x.xvii.). It consists of three couplets or six-line strophes: all the hemistichs of the first are in monorhyme; in the second and following stanzas the three first hemistichs take a new rhyme, but the fourth resumes the a.s.sonance of the first set and is followed by the third couplet of No. 1, serving as bob or refrain, e.g., aaaaaa + bbbaaa + cccaaa and so forth. It is the most complicated of all the measures and is held to be of Morisco or Hispano-Moorish origin.

Mr. Lane (Lex.) lays down, on the lines of Ibn Khallikan (i. 476, etc.) and other representative literati, as our sole authortties for pure Arabic, the precedence in following order. First of all ranks the Jahili (Ignoramus) of The Ignorance, the : these pagans left hemistichs, couplets, pieces and elegies which once composed a large corpus and which is now mostly forgotten. Hammad al-Rawiyah, the Reciter, a man of Persian descent (ob. A.H. 160=777) who first collected the Mu'allakat, once recited by rote in a seance before Caliph Al-Walid two thousand poems of prae-Mohammedan bards.[FN#445] After the Jahili stands the Mukhadram or Muhadrim, the "Spurious," because half Pagan half Moslem, who flourished either immediately before or soon after the preaching of Mohammed. The Islami or full-blooded Moslem at the end of the first century A.H ( = 720) began the process of corruption in language; and, lastly he was followed by the Muwallad of the second century who fused Arabic with non- Arabic and in whom purity of diction disappeared.

I have noticed (I -- A.) that the versical portion of The Nights may be distributed into three categories. First are the olden poems which are held cla.s.sical by all modern Arabs; then comes the mediaeval poetry, the effusions of that brilliant throng which adorned the splendid Court of Harun al-Rashid and which ended with Al-Hariri (ob. A.H. 516); and, lastly, are the various pieces de circonstance suggested to editors or scribes by the occasion. It is not my object to enter upon the historical part of the subject: a mere sketch would have neither value not interest whilst a finished picture would lead too far: I must be contented to notice a few of the most famous names.

Of the prae-Islamites we have adi bin Zayd al-Ibadi the "celebrated poet" of Ibn Khallikan (i. 188); Nabighat (the full- grown) al-Zubyani who flourished at the Court of Al-Nu'man in AD.

580-602, and whose poem is compared with the "Suspendeds,''[FN#446] and Al-Mutalammis the "pertinacious"

satirist, friend and intimate with Tarafah of the "Prize Poem."

About Mohammed's day we find Imr al-Kays "with whom poetry began," to end with Zu al-Rummah; Amru bin Madi Karab al-Zubaydi, Labid; Ka'b ibn Zuhayr, the father one of the Mu'al-lakah-poets, and the son author of the Burdah or Mantle-poem (see vol. iv.

115), and Abbas bin Mirdas who lampooned the Prophet and had "his tongue cut out" i.e. received a double share of booty from Ali.

In the days of Caliph Omar we have Alkamah bin Olatha followed by Jamil bin Ma'mar of the Banu Ozrah (ob. A.H. 82), who loved Azza.

Then came Al-Kuthayyir (the dwarf, ironice), the lover of Buthaynah, "who was so lean that birds might be cut to bits with her bones :" the latter was also a poetess (Ibn Khall. i. 87), like Hind bint al-Nu'man who made herself so disagreeable to Al-Hajjaj (ob. A.H. 95) Jarir al-Khatafah, the n.o.blest of the Islami poets in the first century, is noticed at full length by Ibn Khallikan (i. 294) together with his rival in poetry and debauchery, Abu Firas Hammam or Homaym bin Ghalib al-Farazdak, the Tamimi, the Ommiade poet "without whose verse half Arabic would be lost:"[FN#447] he exchanged satires with Jarir and died forty days before him (A.H. 110). Another contemporary, forming the poetical triumvirate of the period, was the debauched Christian poet Al-Akhtal al-Taghlibi. They were followed by Al- Ahwas al-Ansari whose witty lampoons banished him to Dahlak Island in the Red Sea (ob. A.H. 179 = 795); by Bashshar ibn Burd and by Yunus ibn Habib (ob. A.H. 182).

The well known names of the Harun-cycle are Al-Asma'i, rhetorician and poet, whose epic with Antar for hero is not forgotten (ob. A.H. 2I6); Isaac of Mosul (Ishak bin Ibrahim of Persian origin); Al-'Utbi "the Poet" (ob. A.H. 228); Abu al-Abbas al-Rakashi; Abu al-Atahiyah, the lover of Otbah; Muslim bin al- Walid al-Ansari; Abu Tammam of Tay, compiler of the Hamasah (ob.

A.H. 230), "a Muwallad of the first cla.s.s" (says Ibn Khallikan i.

392); the famous or infamous Abu Nowas, Abu Mus'ab (Ahmad ibn Ali) who died in A.H. 242; the satirist Dibil al-Khuzai (ob. A.H.

246) and a host of others quos nunc perscribere longum est. They were followed by Al-Bohtori "the Poet" (ob. A.H. 286); the royal author Abdullah ibn al-Mu'tazz (ob. A.H. 315); Ibn Abbad the Sahib (ob. A.H. 334); Mansur al-Hallaj the martyred Sufi; the Sahib ibn Abbad, Abu Faras al-Hamdani (ob. A.H. 357); Al-Nami (ob. A.H. 399) who had many encounters with that model Chauvinist Al-Mutanabbi, nicknamed Al-Mutanabbih (the "wide awake"), killed A.H. 354; Al-Man.a.z.i of Manazjird (ob. 427); Al-Tughrai author of the Lamiyat al-'Ajam (ob. A.H. 375); Al-Hariri the model rhetorician (ob. A.H. 516); Al-Hajiri al-Irbili, of Arbela (ob.

A.H. 632); Baha al-Din al-Sinjari (ob. A.H. 622); Al-Katib or the Scribe (ob. A.H. 656); Abdun al-Andalusi the Spaniard (our xiith century) and about the same time Al-Nawaji, author of the Halbat al-k.u.mayt or"Race course of the Bay horse"--poetical slang for wine.[FN#448]

Of the third category, the pieces d'occasion, little need be said: I may refer readers to my notes on the doggrels in vol. ii.

34, 35, 56, 179, 182, 186 and 261; in vol. v. 55 and in vol.

viii. 50.

Having a mortal aversion to the details of Arabic prosody, I have persuaded my friend Dr. Steinga.s.s to undertake in the following pages the subject as far as concerns the poetry of The Nights. He has been kind enough to collaborate with me from the beginning, and to his minute lexicographical knowledge I am deeply indebted for discovering not a few blemishes which would have been "nuts to the critic." The learned Arabist's notes will be highly interesting to students: mine ( --V.) are intended to give a superficial and popular idea of the Arab's verse mechanism.

"The principle of Arabic Prosody (called 'Aruz, pattern standard, or 'Ilm al-'Aruz, science of the 'Aruz), in so far resembles that of cla.s.sical poetry, as it chiefly rests on metrical weight, not on accent, or in other words a verse is measured by short and long quant.i.ties, while the accent only regulates its rhythm. In Greek and Latin, however, the quant.i.ty of the syllables depends on their vowels, which may be either naturally short or long, or become long by position, i.e. if followed by two or more consonants. We all remember from our school-days what a fine string of rules had to be committed to and kept in memory, before we were able to scan a Latin or Greek verse without breaking its neck by tripping over false quant.i.ties. In Arabic, on the other hand, the answer to the question, what is metrically long or short, is exceedingly simple, and flows with stringent cogency from the nature of the Arabic Alphabet. This, strictly speaking, knows only consonants (Harf, pl. Huruf). The vowels which are required, in order to articulate the consonants, were at first not represented in writing at all. They had to be supplied by the reader, and are not improperly called "motions" (Harakat), because they move or lead on, as it were, one letter to another.

They are three in number, a (Fathah), i (Kasrah), u (Zammah), originally sounded as the corresponding English vowels in bat, bit and b.u.t.t respectively, but in certain cases modifying their p.r.o.nunciation under the influence of a neighbouring consonant.

When the necessity made itself felt to represent them in writing, especially for the sake of fixing the correct reading of the Koran, they were rendered by additional signs, placed above or beneath the consonant, after which they are p.r.o.nounced, in a similar way as it is done in some systems of English shorthand. A consonant followed by a short vowel is called a "moved letter"

(Muharrakah); a consonant without such vowel is called "resting"

or "quiescent" (Sakinah), and can stand only at the end of a syllable or word.

And now we are able to formulate the one simple rule, which determines the prosodical quant.i.ty in Arabic: any moved letter, as ta, li, mu, is counted short; any moved letter followed by a quiescent one, as taf, fun, mus, i.e. any closed syllable beginning and terminating with a consonant and having a short vowel between, forms a long quant.i.ty. This is certainly a relief in comparison with the numerous rules of cla.s.sical Prosody, proved by not a few exceptions, which for instance in Dr. Smith's elementary Latin Grammar fill eight closely printed pages.

Before I proceed to show how from the prosodical unities, the moved and the quiescent letter, first the metrical elements, then the feet and lastly the metres are built up, it will be necessary to obviate a few misunderstandings, to which our mode of transliterating Arabic into the Roman character might give rise.

The line::

"Love in my heart they lit and went their ways," (vol. i. 232)

runs in Arabic:

"Akamu al-wajda fi kalbi wa saru" (Mac. Ed. i. 179).

Here, according to our ideas, the word akamu would begin with a short vowel a, and contain two long vowels a and u; according to Arabic views neither is the case. The word begins with "Alif,"

and its second syllable ka closes in Alif after Fathah (a), in the same way, as the third syllable mu closes in the letter Waw (w) after Zammah (u).

The question, therefore, arises, what is "Alif." It is the first of the twenty-eight Arabic letters, and has through the medium of the Greek Alpha nominally entered into our alphabet, where it now plays rather a misleading part. Curiously enough, however, Greek itself has preserved for us the key to the real nature of the letter. In ? the initial a is preceded by the so called spiritus lends ('), a sign which must be placed in front or at the top of any vowel beginning a Greek word, and which represents that slight aspiration or soft breathing almost involuntarily uttered, when we try to p.r.o.nounce a vowel by itself. We need not go far to find how deeply rooted this tendency is and to what exaggerations it will sometimes lead. Witness the gentleman who, after mentioning that he had been visiting his "favourite haunts"

on the scenes of his early life, was sympathetically asked, how the dear old ladies were. This spiritus lends is the silent h of the French "homme" and the English "honour," corresponding exactly to the Arabic Hamzah, whose mere prop the Alif is, when it stands at the beginning of a word: a native Arabic Dictionary does not begin with Bab al-Alif (Gate or Chapter of the Alif), but with Bab al-Hamzah. What the Greeks call Alpha and have transmitted to us as a name for the vowel a, is in fact nothing else but the Arabic Hamzah-Alif,(~)moved by Fathah, i.e. bearing the sign(~) for a at the top (~), just as it might have the sign Zammah (~) superscribed to express u (~), or the sign Kasrah (~) subjoined to represent i(~). In each case the Hamzah-Alif, although scarcely audible to our ear, is the real letter and might fitly be rendered in transliteration by the above mentioned silent h, wherever we make an Arabic word begin with a vowel not preceded by any other sign. This latter restriction refers to the sign ', which in Sir Richard Burton's translation of The Nights, as frequently in books published in this country, is used to represent the Arabic letter ~ in whose very name 'Ayn it occurs.

The 'Ayn is "described as produced by a smart compression of the upper part of the windpipe and forcible emission of breath,"

imparting a guttural tinge to a following or preceding vowel- sound; but it is by no means a mere guttural vowel, as Professor Palmer styles it. For Europeans, who do not belong to the Israelitic dispensation, as well as for Turks and Persians, its exact p.r.o.nunciation is most difficult, if not impossible to acquire.

In reading Arabic from transliteration for the purpose of scanning poetry, we have therefore in the first instance to keep in mind that no Arabic word or syllable can begin with a vowel.

Where our mode of rendering Arabic in the Roman character would make this appear to be the case, either Hamzah (silent h), or 'Ayn (represented by the sign') is the real initial, and the only element to be taken in account as a letter. It follows as a self- evident corollary that wherever a single consonant stands between two vowels, it never closes the previous syllable, but always opens the next one. Our word "Akamu," for instance, can only be divided into the syllables: A (properly Ha)-ka-mu, never into Ak-a-mu or Ak-am-u.

It has been stated above that the syllable ka is closed by the letter Alif after Fathah, in the same way as the syllable mu is closed by the letter Waw, and I may add now, as the word fi is closed by the letter Ya (y). To make this perfectly clear, I must repeat that the Arabic Alphabet, as it was originally written, deals only with consonants. The signs for the short vowel-sounds were added later for a special purpose, and are generally not represented even in printed books, e.g. in the various editions of The Nights, where only quotations from the Koran or poetical pa.s.sages are provided with the vowel-points. But among those consonants there are three, called weak letters (Huruf al-?illah), which have a particular organic affinity to these vowel sounds: the guttural Hamzah, which is akin to a, the palatal Ya, which is related to i, and the l.a.b.i.al Waw, which is h.o.m.ogeneous with u. Where any of the weak letters follows a vowel of its own cla.s.s, either at the end of a word or being itself followed by another consonant, it draws out or lengthens the preceding vowel and is in this sense called a letter of prolongation (Harf al-Madd). Thus, bearing in mind that the Hamzah is in reality a silent h, the syllable ka might be written kah, similarly to the German word "sah," where the h is not p.r.o.nounced either, but imparts a lengthened sound to the a. In like manner mu and fi are written in Arabic muw and fiy respectively, and form long quant.i.ties not because they contain a vowel long by nature, but because their initial "Muharrakah" is followed by a "Sakinah," exactly as in the previously mentioned syllables taf, fun, mus.[FN#449] In the Roman transliteration, Akamu forms a word of five letters, two of which are consonants, and three vowels; in Arabic it represents the combination H(a)k(a)hm(u)w, consisting also of five letters but all consonants, the intervening vowels being expressed in writing either merely by superadded external signs, or more frequently not at all. Metrically it represents one short and two long quant.i.ties (U - -), forming in Latin a trisyllable foot, called Bacchius, and in Arabic a quinqueliteral "Rukn" (pillar) or "Juz"

(part, portion), the technical designation for which we shall introduce presently.

There is one important remark more to be made with regard to the Hamzah: at the beginning of a word it is either conjunctive, Hamzat al-Wasl, or disjunctive, Hamzat al-Kat'. The difference is best ill.u.s.trated by reference to the French so-called aspirated h, as compared with the above-mentioned silent h. If the latter, as initial of a noun, is preceded by the article, the article loses its vowel, and, ignoring the silent h altogether, is read with the following noun almost as one word: le homme becomes l'homme (p.r.o.nounced lomme) as le ami becomes l'ami. This resembles very closely the Arabic Hamzah Wasl. If, on the other hand, a French word begins with an aspirated h, as for instance heros, the article does not drop its vowel before the noun, nor is the h sounded as in the English word "hero," but the effect of the aspirate is simply to keep the two vowel sounds apart, so as to p.r.o.nounce le eros with a slight hiatus between, and this is exactly what happens in the case of the Arabic Hamzah Kat'.

With regard to the Wasl, however, Arabic goes a step further than French. In the French example, quoted above, we have seen it is the silent h and the preceding vowel which are eliminated; in Arabic both the Hamzah and its own Harakah, i.e. the short vowel following it, are supplanted by their antecedent. Another example will make this clear. The most common instance of the Hamzah Wasl is the article al (for h(a)l=the Hebrew hal), where it is moved by Fathah. But it has this sound only at the beginning of a sentence or speech, as in "Al-Hamdu" at the head of the Fatihah, or in "Allahu" at the beginning of the third Surah. If the two words stand in grammatical connection, as in the sentence "Praise be to G.o.d," we cannot say "Al-Hamdu li-Allahi," but the junction (Wasl) between the dative particle li and the noun which it governs must take place. According to the French principle, this junction would be effected at the cost of the preceding element and li Allahi would become l'Allahi; in Arabic, on the contrary, the kasrated l of the particle takes the place of the following fathated Hamzah and we read li 'llahi instead. Proceeding in the Fatihah we meet with the verse "Iyyaka na'budu wa iyyaka nasta'inu," Thee do we worship and of Thee do we ask aid. Here the Hamzah of iyyaka (properly hiyyaka with silent h) is disjunctive, and therefore its p.r.o.nunciation remains the same at the beginning and in the middle of the sentence, or, to put it differently, instead of coalescing with the preceding wa into wa'yyaka, the two words are kept separate by the Hamzah, reading wa iyyaka, just as it was the case with the French Le heros.

If the conjunctive Hamzah is preceded by a quiescent letter, this takes generally Kasrah: "Talat al-Laylah," the night was longsome, would become Talati 'l-Laylah. If, however, the quiescent letter is one of prolongation, it mostly drops out altogether, and the Harakah of the next preceding letter becomes {he connecting vowel between the two words, which in our parlance would mean that the end vowel of the first word is shortened before the elided initial of the second. Thus "fi al-bayti," in the house, which in Arabic is written f(i)y h(a)l-b(a)yt(i) and which we transliterate fi 'l-bayti, is in poetry read fil-bayti, where we must remember that the syllable fil, in spite of its short vowel, represents a long quant.i.ty, because it consists of a moved letter followed by a quiescent one. Fil would be overlong and could, according to Arabic prosody, stand only in certain cases at the end of a verse, i.e. in pause, where a natural tendency prevails to prolong a sound.

The attentive reader will now be able to fix the prosodical value of the line quoted above with unerring security. For metrical purposes it syllabifies into: A-ka-mul-vaj-da fi kal-bi wa sa-ru, containing three short and eight long quant.i.ties. The initial unaccented a is short, for the same reason why the syllables da and wa are so, that is, because it corresponds to an Arabic letter, the Hamzah or silent h, moved by Fathah. The syllables ka, fi, bi, sa, ru are long for the same reason why the syllables mul, waj, kal are so, that is, because the accent in the transliteration corresponds to a quiescent Arabic letter, following a moved one. The same simple criterion applies to the whole list, in which I give in alphabetical order the first lines and the metre of all the poetical pieces contained in the Mac.

edition, and which will be found at the end of this volume. {This appendix is not included in the electronic text}

The prosodical unities, then, in Arabic are the moved and the quiescent letter, and we are now going to show how they combine into metrical elements, feet, and metres.

i. The metrical elements (Usul) are:

1. The Sabab,[FN#450] which consists of two letters and is either khafif (light) or sakil (heavy). A moved letter followed by a quiescent, i.e. a closed syllable, like the afore-mentioned taf, fun, mus, to which we may now add fa=fah, 'i='iy, 'u='uw, form a Sabab khafif, corresponding to the cla.s.sical long quant.i.ty (-). Two moved letters in succession, like mute, 'ala, const.i.tute a Sabab sakil, for which the cla.s.sical name would be Pyrrhic (U U). As in Latin and Greek, they are equal in weight and can frequently interchange, that is to say, the Sabab khafif can be evolved into a sakil by moving its second Harf, or the latter contracted into the former, by making its second letter quiescent.

2. The Watad, consisting of three letters, one of which is quiescent. If the quiescent follows the two moved ones, the Watad is called majmu' (collected or joined), as fa'u (=fa'uw), mafa (=mafah), 'ilun, and it corresponds to the cla.s.sical Iambus (U - ). If, on the contrary, the quiescent intervenes or separates between the two moved letters, as in fa'i ( = fah'i), latu (=lahtu), taf'i, the Watad is called mafruk (separated), and has its cla.s.sical equivalent in the Trochee (- U)

3. The Fasilah,[FN#451] containing four letters, i.e.

three moved ones followed by a quiescent, and which, in fact, is only a shorter name for a Sabab sakil followed by a Sabab khafif, as mute + fa, or 'ala + tun, both of the measure of the cla.s.sical Anapaest (U U -)

ii. These three elements, the Sabab, Watad and Fasilah, combine further into feet Arkaan, pl. of Rukn, or Ajzaa, pl. of Juz, two words explained supra p. 236. The technical terms by which the feet are named are derivatives of the root fa'l, to do, which, as the student will remember, serves in Arabic Grammar to form the Auzan or weights, in accordance with which words are derived from roots. It consists of the three letters Fa (f), 'Ayn ('), Lam (l), and, like any other Arabic root, cannot strictly speaking be p.r.o.nounced, for the introduction of any vowel-sound would make it cease to be a root and change it into an individual word. The above fa'l, for instance, where the initial Fa is moved by Fathah (a), is the Infinitive or verbal noun, "to do," "doing." If the 'Ayn also is moved by Fathah, we obtain fa'al, meaning in colloquial Arabic "he did" (the cla.s.sical or literary form would be fa'ala). p.r.o.nouncing the first letter with Zammah (u), the second with Kasrah (i), i.e., fu'il, we say "it was done"

(cla.s.sically fu'ila). Many more forms are derived by prefixing, inserting or subjoining certain additional letters called Huruf al-Ziyadah (letters of increase) to the original radicals: fa'il, for instance, with an Alif of prolongation in the first syllable, means "doer"; maf'ul (=maf'uwl), where the quiescent Fa is preceded by a fathated Mim (m), and the zammated 'Ayn followed by a lengthening Waw, means "done"; Mufa'alah, where, in addition to a prefixed and inserted letter, the feminine termination ah is subjoined after the Lam, means "to do a thing reciprocally."

Since these and similar changes are with unvarying regularity applicable to all roots, the grammarians use the derivatives of Fa'l as model-forms for the corresponding derivations of any other root, whose letters are in this case called its Fa, 'Ayn and Lam. From a root, e.g., which has Kaf (k) for its first letter or Fa, Ta (t) for its second letter or 'Aye, and Ba (b) for its third letter or Lam

fa'l would be katb =to write, writing; fa'al would be katab =he wrote; fu'il would be kutib =it was written; fa'il would be katib =writer, scribe; maf'ul would be maktub=written, letter; mufa'alah would be mukatabah = to write reciprocally, correspondence.

The advantage of this system is evident. It enables the student, who has once grasped the original meaning of a root, to form scores of words himself, and in his readings, to understand hundreds, nay thousands, of words, without recourse to the Dictionary, as soon as he has learned to distinguish their radical letters from the letters of increase, and recognises in them a familiar root. We cannot wonder, therefore, that the inventor of Arabic Prosody readily availed himself of the same plan for his own ends. The Taf'il, as it is here called, that is, the representation of the metrical feet by current derivatives of fa'l, has in this case, of course, nothing to do with the etymological meaning of those typical forms. But it proves none the less useful in another direction: in simply naming a particular foot it shows at the same time its prosodical measure and character, as will now be explained in detail.

We have seen supra p. 236 that the word Akamu consists of a short syllable followed by two long ones (U - -), and consequently forms a foot, which the cla.s.sics would call Bacchius. In Latin there is no connection between this name and the metrical value of the foot: we must learn both by heart. But if we are told that its Taf'il in Arabic is Fa'ulun, we understand at once that it is composed of the Watad majmu' fa'u (U -) and the Sabab khafif lun (-), and as the Watad contains three, the Sabab two letters, it forms a quinqueliteral foot or Juz khamasi.

In combining into feet, the Watad has the precedence over the Sabab and the Fasilah, and again the Watad majmu' over the Watad mafruk. Hence the Prosodists distinguish between Ajza asliyah or primary feet (from Asl, root), in which this precedence is observed, and ajza far'iyah or secondary feet (from Far'= branch), in which it is reversed. The former are four in number:- -

1. Fa'u.lun, consisting,as we have just seen, of a Watad majmu'

followed by a Sabab khafif = the Latin Bacchius (U - -).

2. Mafa.'i.lun, i.e. Watad majmu' followed by two Sabab khafif = the Latin Epitritus primus (U - - -).

3. Mufa.'alatun, i.e. Watad majmu' followed by Fasilah = the Latin Iambus followed by Anapaest (U - UU -).