Some say that this is a singular form.—In the Seventh Assembly Ḥarîri uses the singular form ; the contrary opinion appears to rest on a Tradition of the Prophet, who said to his followers, (p. 78 De sacy’s edition). Ibn Mâlik treats as a singular having the form of a plural, and therefore imperfectly declinable. Others think that either declension is lawful. Alfîyeh, v. 660. Similar words are , a name for the hyena, and , a man’s name.

As for the which when it attaches itself.—This requires little explanation. As the formal secondariness which is neces­sary to the imperfect declension arises from divergence from a singular form, the addition of the , which gives the plural a singular form, takes away this condition, and makes the declen­sion perfect. Alfîyeh, v. 658.

As for the which deposes the regent.—The particle is of two kinds. Firstly, , that is, “forming a proposition equivalent to the masḍar;” in this case it combines with the muḍâri‘, which it causes to be manṣûb; thus: , I desire that thou stand, is equivalent to , I desire thy standing. So in Koran ii. 180, , And that ye fast is better for you, is equivalent to “your fasting.” In this case is an , regent, and naṣbs the verb. But may also be considered a contraction of , in which case it is called , because the original form is lightened by taking away the teshdîd. is one of the sisters of , and not distinguished from it by Sîbawayh, who is of opinion that the former is originally . The property of and its sisters is that they annul () the mubtada, their action being the reverse of and its sisters; for whereas the latter make their ism marfû‘ and their khabar manṣûb, the former make their ism manṣûb and their khabar marfû‘, as . These par­ticles are considered to partake of the nature of the verb, and are therefore called particles assimilated to verbs This resemblance consists, firstly, in their adhering to the noun, since it is their peculiar property to be prefixed to the mubtada; secondly, in a logical analogy, since they have the signification of verbs in expressing emphasis, like , or comparison, like ; thirdly, in a formal analogy, since they are composed of three or more letters, the last of which is meftûḥ. If this resemblance to the verb be clearly perceived, the peculiar action of these particles on the mubtada and the khabar will be readily understood.

Now when it is said that is the lightened from the heavy, that is, a contraction of , it is meant that is used in the signification of ; but this can only happen under certain con­ditions. When is thus lightened it preserves its regency, i.e. its grammatical effect, but its ism can only be the elided (the ḍamîr as sha’n is the affixed pronoun in , for which see the grammars), and the khabar can only be a pro­position, . Thus in the phrase , the is the lightened from the heavy, and its ism is an elided ; the virtual expression being . This ḍamîr as sha’n is manṣûb as the ism of , while the khabar of is the pro­position , which is in the position of raf‘. In this case the proposition is nominal, , and does not require a dividing particle between itself and its khabar, except in the case of a negative. But if the proposition be verbal, , distinctions are made. In the case of undeclined verbs, as or , there is no fâṣil or dividing particle; so also where the verb expresses prayer or imprecation. But in other cases a dividing particle is lawful, and to be preferred. In the case in the text the dividing particle is what is called , that is or . These particles are described as of or , because they extend the time from the present to the future. We are now able to perceive the meaning of Ḥarîri’s riddle. If be prefixed to the verbal proposition then is the lightened from the heavy; and the sentence , , He (God) knows that some of you will be sick, Koran lxxiii. 20, is equivalent to ; the elided ḍamîr as sha’n is the ism of and is manṣûb, while , etc. is the khabar of and in the state of raf‘. But if there were no the must be considered maṣdarî, and a nâṣib of the verb. It is hardly correct to say that here has no regency, since it naṣbs its elided ism the ḍamîr as sha’n, and places the proposition in the state of raf‘. But though exercises a regency on the proposition it does not exercise it on the verb itself, which within that proposition is grammatically unaffected, and is therefore marfû‘. Or it may be that Ḥarîri takes the view that the entirely destroys the regency of that which precedes it on that which follows it; if so we must consider that governs an elided khabar as well as an elided ism; in that case the taḳdîr would be . Compare Bayḍâwi on Koran v. 75.

That which is naṣbed as a term of circumstance.— On this ques­tion it will be sufficient to quote Ḥarîri himself, who says in the Durrah that when one pronounces , he commits a fault, because is mejrûr after only. Such is the constant use in the Koran; see ii. 73, 83, 95, etc. The reason, according to Ḥarîri, that has this peculiarity is that it is the mother (the origin or principle) of the particles which make the noun mejrûr; for in each class of words that which is the original is distinguished by properties which the others do not possess. Thus , with the kesr, is distinguished from its sisters by the prefixion of to its khabar; thus alone admits a verb in the preterite as its khabar; thus, in swearing, the particle alone can be used when the verb, “I swear” is expressed, and alone can be joined to the pronoun, as . Whatever may be thought of this reasoning, the usage in the Arabic language is certainly as the author states.

As to the annexed noun.—The meaning of the author is that , though its property is to be followed by a noun mejrûr, yet in one case is followed by a manṣûb, and therefore cannot be looked upon as in annexion to the noun following; it may, consequently, be said to have lost one of its handles or links of annexion. The word or belongs to the class of nouns which are constantly annexed. These are divided into two classes; firstly, such as require annexion both logically and formally, and are never used separately; among them are , and . Secondly, such as require an­nexion logically though not formally, like , and , and . Alfîyeh, v. 396: Al Ashmûni, Part II. p. 152. is made indeclinable by the generality of the Arabs, because it has only one use, and that an adverbial one; and it only quits its quality of ẓarf, when it is mejrûr by . Indeed it never appears in the Koran except after , as at xi. 1; xviii. 2, 64. The tribe of Ḳays decline it; so that at xviii. 2, it is also read . The noun which follows it is mejrûr, except in the case of , which is manṣûb by , or specification; it is also said that in this case the noun is the khabar of an elided , the virtual sense being . Yet the jerr is lawful, and is according to rule, the naṣb being the more rare. The Kufians teach that may be marfû‘ after , through elided, the taḳdîr being : Alfîyeh, v. 409. To understand the construction of these sentences it must be noticed that is one of the nouns of annexion which may be annexed to the proposition, being, with the exception of , the only ẓarf of place which has this property. Al Ashmûni, Part II. p. 153. Another way of accounting for the naṣb of is that it is assimilated to a mef‘ûl by reason of being assimilated to the ism al fâ‘il, and the latter is so assimilated, since it sometimes has the final , and sometimes not; the so far having the nature of tanwîn. You would then say , as you say . There are other explana­tions, but the above are the most satisfactory.