1^ 
 
 p J 
 
 3251 
 H56 
 <866 
 MAIN 
 
 UC-NRLF 
 
GIFT OF 
 
ROYAL ASIATIC BOCIETY 
 
 OP 
 
 GREAT BRITAIN AND IRELAND. 
 
 SPECIMEN CHAPTERS 
 
 AN ASSYRIIN GRAMMAR 
 
 The Eev. E. HINCKS, D.D., 
 
 HON. M. R, A. S. 
 
 Decbmbbk, 1866.] 
 
 J 
 
UM^ 
 
 tll^feiifJlSTIC PUBLICATIONS 
 
 TRUBNER & CO., 
 
 60, PATEENOSTEE EOW, LONDON, E.G. 
 
 Messes. TRtJBNER & Co. respectfully solicit orders for all classes of 
 Publications connected with the History, Antiquities, Geography, and 
 Languages of the East, published abroad. Messrs. Trubnee & Co. have 
 established agencies in all parts of the East, of Europe, and America, and 
 are thus enabled to furnish such publications with as little delay as possible, 
 and at prices below the hitherto customary charges. 
 
 Authors desirous of printing tbeir own books may be furnished with 
 Specimens, and Estimates for Paper, Printing, and Binding, and every 
 facility will be afforded them, if works are intended for publication, of 
 bringing the same advantageously before the public. 
 
 Alcock- — A. Peactical Geammae of the Japanese Language. By Sir 
 EuTHEiiFORD Alcock, Resident British Minister at Jeddo. 4to., pp. 61, sewed. 18«. 
 
 Alcock- — Eamiliae Dialogues in Japanese, with English and French 
 Translations, for the Use of Students. By Sir Rutheki'ord Alcock. 8vo., pp. viii. 
 and 40, sewed. Paris and London, 1 863. 5s. 
 
 Benfey- — A. Peactical Geammae op the Sanskeit Language, for the Use 
 of early Students. By Theodore Benfey, Professor Pull. Ord. of Sanskrit in the 
 University of Goettingcn, etc., etc. 8vo., pp. xviii. and 228, cloth. 7s. 6d. 
 
 Goldstiicker- — A. Dictionaey, Sanskeit and English, extended and im- 
 proved from the Second Edition of the Dictionary of Professor H. H. Wilson, with 
 his sanction and concurrence. Together with a Supplement, Grammatical Appendices, 
 and an Index, serving as a Sanskrit-English Vocabulary. By THEODOK-GoLDSTiicKER. 
 Parts I. to VI. 4to. 1866—1864. Each Part 6fi. 
 
SPECIMEN CHAPTERS LLr-/ 
 
 A^ ASSYRIAN GRAMMAR, j^^^ 
 
 By the Eev. E. HINCKS, D.D., M/^\ A 
 
 HON. M. K. A. S. 
 
 A great number of years have now elapsed since I began 
 to collect materials for an Assyrian grammar ; an object, of 
 wliich I have never since lost sight. Of late, I have been 
 preparing my materials for publication; but a preliminary 
 question has suggested itself : — " If I were to pubKsh a gram- 
 mar, who would read it ?" The persons for whose instruction 
 I should naturally write would be either persons who were 
 acquainted with other Semitic languages, and who were de- 
 sirous of comparing the grammars of these with that of the 
 oldest and the best developed language of the family ; or else 
 persons, who desired to study this language for its own sake, 
 and who sought the aid of a grammar, in the absence of an 
 oral instructor, to teach them the first principles of this 
 " Sanskrit of the Semitic tongues," which they might after- 
 wards improve upon by their own studies. 
 
 I believe that persons of both these classes would derive 
 benefit from such a grammar as I should publish, if they 
 would only make use of it. I fear, however, that no person 
 of either of the classes has so much faith in me, and in my 
 knowledge of Assyrian grammar, as to make use of it. My 
 only readers would probably be my critics ; — those, who have 
 attained some knowledge of the Assyrian language — con- 
 siderable knowledge, I may say, so far as respects the mean- 
 ing of ivords, — but whose published translations show that, as 
 respects grammar, their views are extremely different from 
 mine ; and, of course, if my views be correct, extremely err 
 roneous. My only readers would thus be, with perhaps one or 
 two exceptions, hostile critics, who must condemn what I 
 have written, because, by laying down rules which they have 
 disregarded and violated in their translations, it indirectly 
 condemns those translations. 
 
 In the case of one eminent Assyrian writer on the con- 
 
 siei&i 
 
tinent, the certainty that I should meet with hostile criticism 
 is more obvious than it is in the case of others. Dr. Oppert 
 has published an Assyrian grammar, of which I have already 
 stated that, besides minor errors, it was pervaded by three er- 
 roneous general principles, "so as scarcely to leave a page 
 free from what I consider pernicious error." One of these 
 three principles. Dr. Oppert, in a late article in the Journal 
 Asiatique, intimates that he has abandoned, or is disposed to 
 abandon. To the other two he clings pertinaciously. Now, 
 as I cannot retract the unfavourable opinion above expressed, 
 and as, according to my views of Assyrian grammar, I cannot 
 think the translations from Sargon, which MM. Oppert and 
 Menant have published, to be even approximately correct, I 
 could not possibly expect any criticism from those gentlemen 
 but of the most adverse description. 
 
 Under these circumstances, I have thought it advisable 
 to deviate materially from my original plan. Instead of 
 publishing an entire grammar, I will, in the first instance, 
 publish specimen- chapters only ; treating of the declensions 
 of substantives and adjectives, and of the permansive forms 
 of verbs ; — a subject on which Dr. Oppert, in the Journal 
 Asiatique for last year (Tome vi. p. 297), has mis-stated both 
 my views and the facts to which he appeals. Iti stead, too, 
 of stating what I believe to be the grammatical rules of the 
 Assyrian language, dogmatically, as would be the natura 
 course of the writer of a grammar, I will deduce them from 
 those leading positions on which all are agreed, by inductive 
 proofs, in the order which appears to me the best for this pur- 
 pose ; different as this order is from what I should adopt if I 
 were exempt from the necessity of writing otherwise than dog- 
 matically. My examples must be numerous, because they are 
 not merely illustrative, but justificatory. It will readily be 
 understood that where I bring forward a number of proofs in 
 support of any assertion, it is one which is not generally ac- 
 cepted, and is likely to be controverted ; but except in the 
 one case of the permansive forms of the verb, which I have 
 already mentioned, I do not mean to point out directly what 
 I believe to be the errors of others. It will answer my end, 
 
and will be pleasanter to myself, if I merely seek to establish 
 tbe correctness of my own views. This being done, the in- 
 correctness of any views that are inconsistent with them will 
 follow as a matter of course. 
 
 A few words should be said here on the principles on which 
 I have conducted my investigations. I have sought the rules 
 of Assyrian grammar by induction from passages, the mean- 
 ing of which I considered to be perfectly known, occurring in 
 writings of a good age; the induction being guided by a 
 general knowledge of Semitic grammar. I have taken no 
 account of passages, however well their meaning may be 
 known, which are found in texts of a late age, probably written 
 by persons who had a very imperfect knowledge of Assyrian, 
 and of whom it was not the native tongue. Neither have I 
 assumed the identity of Assyrian and Hebrew rules of gram- 
 matical construction in cases where there is no inductive evi- 
 dence of it. 
 
 As to the transcriptions of Assyrian words and texts which 
 I give, I think that it is unnecessary to say much. Were I 
 to publish a grammar, it should contain a list of all the cha- 
 racters representing syllables consisting of a single vowel, or a 
 vowelpreceded or followed by a consonant, and of all those whose 
 values were syllables consisting of a vowel both preceded and 
 followed by a consonant, and which occurred in any text quoted 
 in the grammar. To each character its transcription should 
 be attached. I represent, as every one else does, D, X *1* T, 
 D, b, t2, X ^, 'n. and n, by b, g, d, z, ky I, m, n, p^ r, and t 
 I also represent H by A, tD by t, D by 5 (which was anciently 
 sounded either st, or sk ; the former being, like ^, repre- 
 sented by the Arabic (jw, the latter by ^ji, and by the Greek 
 I, wJiich- occupied the same position in the alphabet, and had 
 the same figure as the Phoenician D), ^ by :2, p by $', and ^ 
 by s. The remaining five letters, K, H, \ ^, y, are indiscri- 
 minately transcribed as ' with the following exceptions : 
 ^^YI is transcribed by ya ; ^llf^ by 'u or yu; y|[ after an- 
 other yi by ^a or yd ; "gY, when a copulative, by va ; and 
 ^>->-Y by h or ih. I use the vowel e alone to represent ^Yy, 
 
€ standing for ^YTYY- The grave accent over a vowel is not 
 intended to mark a distinction of sound, but to show that a 
 homophone of the character usually representing the syllable 
 without the accent is used in place of that character. I use 
 four vowels in transcriptions, a, i, u, and e, which I believe the 
 Assyrians pronounced as long i, or i. I think that no Assyrian 
 scholar will find any difficulty in restoring the text of my tran- 
 scriptions ; but if he do, he can consult the original text, to 
 which I refer him. To explain my references, I will observe 
 that B. means Botta's plates; L., the first series of British 
 Museum plates, edited by Mr. Layard ; I. and II., the two 
 volumes of the second series of British Museum plates, edited 
 by Sir H. Rawlinson. The number which precedes this is the 
 number of the plate in each series, and the number at the end 
 is the line in the plate ; with a distinctive mark before it, if 
 necessary, the meaning of which will be seen when the plate 
 is referred to. In some of the plates of the second series, I 
 use r, c, and / for right, centre, and left. In these references, 
 and in the transcriptions, I have followed as closely as I could 
 Mr. Norris. 
 
 I believe that the main point in which I differ from Mr. 
 Norris in my transcriptions is that I do not attempt to dis- 
 tinguish in most cases between the several breathings and 
 semivowels. I do not think that we can distinguish them by 
 inspection of the Assyrian characters. We can only do so by 
 knowledge derived from some other source. I find, for in- 
 stance, 1 1 occurring in four words ; between ma and du, be- 
 tween na and du, between ta and bu, and between da and 
 nu. I transcribe it in the four instances by 'a. I happen to 
 know from the Hebrew that in the first word the 'a represents 
 X, in the second H, in the third 1, and in the fourth *• ; the 
 
 T ' T ' t' T ' 
 
 four words signifying muck or ^reat from HX^, glorious from 
 *ini, (/ood from y\t2, and a judge from JH. In another word 
 which had no Hebrew equivalent, I should not know how to 
 render the characters so as to make a distinction ; and I 
 therefore do not make it in the cases where I am able to do ■ 
 60. Some persons would in all these instances represent the TI 
 
by accenting the preceding vowel. In most cases, however, 
 I think that this would be an error. Where it is plain to me 
 that two characters represent together but one syllable, I cut 
 off one of the vowels by an apostrophe ; but I do not consider 
 it a matter of course to do so, nor do I use an accent in all 
 cases where I do so. 
 
 When two syllables are joined together without a point, it 
 will be understood that they are, in the original, not repre- 
 sented phonetically, but by monograms, ideograms, or Accadian 
 roots, as they have been variously called. In such cases I 
 generally give the cuneiform characters in brackets, with i.e. 
 after them, before I give the words that I read. When, 
 however, the signification of the monogram is well known and 
 universally admitted, I omit the cuneiform characters. 
 
 [Chapter V.] On Verbs, and first on their permansive parts. 
 
 1. Generally speaking, a verb admits of seven principal 
 conjugations, of which the first six correspond to the first six 
 of the recognized Hebrew conjugations; the seventh being 
 the causative of the third. I denote them by the seven first 
 Homan numerals. Other conjugations are occasionally met 
 with, but are less frequent. It will, of course, be understood 
 that no one verb is used in all the conjugations. 
 
 2. Each of these seven admits of a secondary or augmented 
 conjugation, which I denote by the proper Eoman numeral 
 followed by t. The augmented conjugation regularly inserts t 
 or te after the first radical letter, or in the conjugations 
 Y.-YII. after the preformative s. In some cases the insertion 
 of this letter or syllable is the only difference between the 
 augmented conjugation and its principal one ; but in other 
 cases, other changes are introduced which will be mentioned 
 in the sequel. 
 
 3. The inserted t is sometimes changed into d or t, through 
 the influence of the letter with which it comes in contact ; 
 and sometimes through its influence that letter undergoes a 
 change. Another irregularity, which is apt to be more puz- 
 zling, is that in most verbs defective in the second radical 
 
6 
 
 (concave verbs) the dental precedes the first radical in place 
 of following it. Thus, we have from N^^, in the aorist of 1. 1, 
 it.bu.ni, instead ofib.tzc.ni, "they came on" (90 L. 63). 
 
 4. Generally speaking, again, the verbal forms belonging 
 to each conjugation may be divided into two great classes, 
 which I call permansive and mutative. The former denotes 
 continuance in the state which the verb signifies in that con- 
 jugation ; the latter denotes change into that state. Each 
 portion of the forms belonging to the conjugation has tenses 
 and verbal nouns. The permansive tenses are analogous to 
 the so-called Hebrew preterite, having no preformatives, but 
 having terminations added, (except in the third person singu- 
 lar masculine, which has none, as in Hebrew), by which the 
 number, person, and gender are distinguished. The mutative 
 tenses, on the contrary, are analogous to the so-called Hebrew 
 future, having one or other of the preformatives K^i^ always 
 prefixed to the root, whether or no there be any distinctive 
 addition at the end. 
 
 5. The distinction here made between permansive and 
 mutative forms is in my judgment a fundamental one, on the 
 proper development of which all accurate knowledge of the 
 Assyrian language must depend. Now, Dr. Oppert, has in a 
 recent number of the Journal Asiatique (Tome vi. p. 297), 
 ridiculed the existence of any such forms as what I call per- 
 mansive, characterising them as being a mere fancy of mine. 
 The first thing then that I have to do is to establish the 
 existence of such forms by a number of clear examples. 
 
 6. I will begin with comparing two sentences from adjoin- 
 ing columns of Taylor's Cylinder, in which the same root 
 occurs, — a perfect root having three letters incapable of 
 change, — namely, DflD . It signifies in Arabic celavit (Frey. 
 iv. 10), and this may be the meaning in the single passage in 
 the Hebrew Bible where it occurs (see Ges. 723). At any rate 
 this seems clearly the meaning in the two Assyrian passages 
 which I will cite. In the former of these, 40 I. 68, Senna- 
 cherib, after mentioning the destruction of certain towns con- 
 cluding with \ik.mu, " I burned,'' proceeds thus. I omit a 
 simile which is evidently parenthetical. Qu.tur na.ak.mu,ti. 
 
su,nu pa.an [_*->-\ i.e.'] same!e rap.su.ti 'u.sak.tim, 
 
 "the smoke of their being burned (like a ponderous cloud) 
 the face of the wide heavens I made to conceal." The verb is 
 here in the 1. s. of the aorist of (the causative) Conjugation Y., 
 having two objects expressed, that which is made to produce, 
 and that which is made to undergo, the change. The preceding 
 verb 'akmu is the 1. s. of the aorist of I. of the verb HM. Let 
 us now proceed to 41 I. 45. I omit another comparison to a 
 cloud, which is parenthetic. [t^TT -^Yy ^^TT ^•^•] ^P^'"' 
 sepi.su.nu .... p'a.an same!e rap.su.ti ka.fi.im, "the dust 
 of their feet (like, etc.) the face of the wide heavens is conceal- 
 ing.'* In my judgment, nothing can be clearer than that 
 katim is the 3. m. s. of the principal permansive tense. Dr. 
 Oppert says that I " substitute participles for the 3. s. m." 
 By a participle I presume he means what I call the nomen 
 mutantis. No doubt the latter would be katim, or, as I should 
 write it for distinction, kdtim ; but this would stand before 
 what it governs; it should then be translated " the concealer of 
 (the face of the wide heavens)," and it would require a verb to 
 complete a sentence. It might as well be said that the second 
 word in Genesis was a participle, or any other 3. m. s. of the 
 preterite in the Bible. In this portion of Taylor's Cylinder 
 we have no less than six long lines and a half in which there 
 is nothing that he acknowledges to be a verb. In 1. 43, 
 Sennacherib begins to describe an attack upon him by an 
 immense multitude of his enemies, whom he compares to 
 locusts. All are doing the same things in constant succes- 
 sion, and accordingly, in speaking of what they do, he uses 
 only permansive tenses. In 1. 44 we have te.hu.'u.ni, " they 
 are coming on," the 3. m. p. of the permansive of I.t. of the 
 verb ^^m. It has exactly the same relation to it.hu.m, "they 
 came on," cited in § 3, as ka.fi.im in 1. 47 has to iktum, " it 
 concealed." Passing over a permansive verb in 1. 48, the 
 discussion of which would lead to a digression, we have in 
 1. 49, after the description of the position taken by his ene- 
 mies, zao.tu.va, "they are occupying, and," followed by a 
 mutative verb. Surely it cannot be maintained with any 
 
X 
 
 8 
 
 show of reason that zah.tUy followed as it is by the copulative 
 enclitic which connects verbs, and verbs only, and this by au 
 acknowledged verb, is anything else than a verb, and yet it 
 has no preformative. The verb H!!^ is a very common one 
 in the mutative tenses of both I. and Y. 
 
 7. Before going further, I will give paradigms of the de- 
 clension of the three principal tenses of the imaginary verb 
 7215 ; — the permansive present^ the aorist, and the mutative pre- 
 sent, or present simply. The latter is written indifferently 
 with g or gg; the duplication of the second radical being 
 euphonic, and not characteristic of the third conjugation. 
 What really characterizes this conjugation is the use of u as 
 the vowel of the preformative. It is invariahly so used in 
 Conjugations III., Y., and YII., and their augmented con- 
 jugations ; and never so in any other conjugation, unless the 
 first radical of the verb be 1 . Each of these three principal 
 tenses admits a secondary tense formed by the addition of a 
 to forms which end in the third radical, and of ni to those 
 which end in an added vowel. The most important of these 
 secondary tenses is that derived from the present, which is 
 clearly o. future. 
 
 8. Paradigms of the three principal tenses of Conjuga- 
 tion I.: — 
 
 1. 8. 
 
 2. m.s. 
 
 2. /.s. 
 
 3. m.s. 
 3. f.s. 
 hp. 
 
 2. m.p 
 2. /.p. 
 
 3. m.p 
 S.f.p. 
 
 The forms of the second person plural may require correction ; 
 but I believe that all the others may be depended upon. I use 
 i as the vowel after the second radical in the permansive and 
 ordinary present, and u in the aorist. These are the most 
 
 Permansive Present. 
 
 Aorist. 
 
 Present, 
 
 1. 5. paglaku, or 
 
 paglak 
 
 'apgul 
 
 'apaggil, or 'apagil 
 
 2. m.s. pagilta 
 
 
 tapgul 
 
 tapaggil 
 
 2. f.s. pagilti 
 
 
 tapguli 
 
 tapaggili 
 
 3. m.s. pagil 
 
 
 ipgul 
 
 ipaggil 
 
 3. f.s. paglat 
 
 
 tapgul 
 
 tapaggil 
 
 1. p. pagilnu 
 
 
 napgul 
 
 napaggil 
 
 2. m.p. pagiltunu 
 
 
 tapgulu 
 
 tapaggilu 
 
 2. fp. pagiltina 
 
 
 tapgula 
 
 tapaggila 
 
 3. m.p. paglu 
 
 
 ipgulu 
 
 ipaggilu 
 
 3. fp. pagla 
 
 
 ipgula 
 
 ipaggila 
 
9 
 
 common vowels, but many verbs have different ones. Per- 
 haps the third person feminine singular sometimes ends in a, 
 that is 'a, like the Hebrew H . An apparent example is in 
 18 I. 62, where the nominative is singular, sa, referring to 
 \i.ha.nu 'a.zi.tu, "a projecting summit;" and we have 
 su.qaJu.Ia in the printed text, and lat as a variant. The 
 latter is, however, the reading of far the most copies, if not 
 of all. In 33 I. 2*48, we have su.qaJu.Ia ; but there the nomi- 
 native is plural, ^u.ba.na.at sad.e, "the summits of the moun- 
 tain." In 24 I. 51, we have also su.qa.lu.la, but here the no- 
 minative is masculine singular; and the final a is, as very 
 often happens, a representative of va, the copulative enclitic, 
 and not an inflexion of the verb. 
 
 9. What are called the personal pronouns of the first and 
 second persons are really, as I have stated long since, per- 
 mansive presents of a verb signifying "to be here," aclsum ; 
 and 'anay " ad," " to," or " for," is to be referred to the same 
 root, p^5 . The true personal pronouns are the afformatives of 
 the permansive present, 'aku, nu, and ta, with its derivatives. 
 It is these which the comparative grammarian should compare 
 with the Indo-European pronouns. By the waj^, I never said, 
 as Dr. Oppert represents me to have said, that the second 
 person of the permansive present ended in ka and ki, as in 
 Ethiopic. What I said was, that the fact of the first per- 
 sons ending alike in Assyrian and Ethiopic ought not to be 
 relied on as proof of a special connection between those two 
 languages, because the Ethiopians had k in the second person, 
 as well as in the first, while the Assyrians had not. 
 
 10. I will now give some other examples of permansive 
 presents: In the descriptions of permanent features of a 
 country, which are so frequently found in the historical in- 
 scriptions, the permansive tense is constantly used. Thus e.g. 
 in 11 I. 43, etc., we read of " high mountains, which, like the 
 edge of a sword, sam.tu, are piercing ; which, for the pass- 
 ing of chariots, la.'a na.tu!u, are not fitted." The exact force 
 of the roots tODC^ and ^tOi or XtOi may be questioned, but 
 the general sense is clear. The king tells us that he left his 
 chariots behind, and crossed the mountains without them; 
 
10 
 
 and here he uses tlie aorist. In 22 I. 105, we have Mr.hu.su 
 kima ^u.Va.an sad.e sa.kin, " Its top (or head) like the summit 
 of a mountain was lying," or simply "was." The verb p^ 
 in its mutative parts signifies, actively, "to place," or "make;" 
 in the permansive parts " to lie," or " be." This is something 
 like what happens in several Greek verbs, as lctttj/jll, and in 
 Latin, where facio, as well as Jio, is etymologically connected 
 Vfichfiio. 
 
 11. The permansive present is also habitually used in the 
 passive conjugations ; for the state denoted by these forms is 
 generally a permanent one. So in Greek we have rervfi^iivoi 
 eicriy and in Latin, to a greater extent, pulsatus sum, etc., as 
 well as pulsati sunt Thus, we have 42 I. 11, ra.ki.hu.si.in 
 di.kUy " their riders were killed," and soon after, si.na 
 7nus.su.ray " they (/) were abandoned." In 38 L. 8, we 
 have z'u.uh.hu.rat su.batsa, " its site (was made, i.e. as often 
 in Hebrew) was judged to be small." Examples might be 
 quoted almost without end. 
 
 12. I will now pass to the first person. Dr. Oppert objects 
 to an example which I gave formerly. In 19 I. 101, we read, 
 as it is printed, ina li.me u.may inaNinua uz.ba.ku, which I have 
 translated " In the eponymy of a certain person, in Nineveh 
 I am stopping." I ought, however, to have translated it, " In 
 the same eponymy." The character in two copies that I have 
 collated is not that for u, the copulative conjunction, but that 
 which occurs so commonly on^the tablets, signifying "the 
 same." There can, I think, be no doubt that this is the cor- 
 rect reading in all the inscriptions ; and that the copyists 
 have substituted a common character for one which was not 
 familiar to them. In 18 I. 69, we have Ina li.me an.ni.may 
 " in the eponymy of this person," that is, " of myself." This 
 is one of the many substitutes for the pronoun " my," which 
 we meet with in the Assyrian inscriptions. This is sometimes 
 contracted into an. ma, as in 40 L. 50. The *^y seems to be 
 used to indicate that a peculiar mode of expression is used ; 
 or perhaps it may signify " and so forth." Speaking of these 
 substitutes for the possessive pronouns, I consider it due to 
 Dr. Oppert to acknowledge the great merit of his late dis- 
 
11 
 
 covery of the signification of raman^ which had been so long 
 a puzzle to Assyrian students. I adopt his explanation as 
 perfectly correct, and I have met with five or six passages, 
 besides those cited by himself, which it explains in the most 
 satisfactory manner. si sic omnia ! 
 
 13. The explanation of m.ba.ku given above, " I am stop- 
 ping," clashes with Dr. Oppert's view in two respects. It is 
 a permansive present, a tense which he does not acknowledge ; 
 and it is that of the verb ^^1, a defective in "*|S, of which he 
 has afiirmed in his grammar that the Assyrians had none. It 
 appears then absolutely necessary for the maintenance of his 
 system that he should do away with this passage ; and I will 
 quote from p. 297, what he says on the subject : " D'abord, on 
 ne lit jamais usbaku, mais toujours iishakuni dans les inscrip- 
 tions de Sardanapale III., dans la phrase; * Pendant qu'ils me 
 retinrent k Mnive, etc.* Ushakuni est la 3™® pers. du pluriel de 
 "l^D a riphteal avec le suffixe. Sans suffixe, ce serait yussahaku 
 ^^5D\ ou meme yusabaku ^^ID.'' (G.A. § 128), et contracte avec 
 le sufiixe (ibid. § 197) ^J^lDSD^ . Yoila a quoi se reduit la V^ 
 personne en ku. Et ce seul example, si memo il etait avere, ne 
 pourrait avoir une grande portee, quand on le compare aux 
 milliers de formes des aoristes fournies par les textes." 
 
 14. I have given the whole of this passage from M. Op- 
 pert's text ; and I must say, that I think I have never read 
 a passage of the same length, in which there are so many 
 mis-statements as to matters of fact. The word uzbaku does 
 occur in the passage cited ; and when he denied that it did 
 so. Dr. Oppert must have trusted to a treacherous memory. 
 There is no ni after it in the printed text ; there is no various 
 reading noted in 19 I. 101 ; and, morover, I collated, some 
 years ago, very carefully, all the copies of the inscriptions 
 brought over by Mr. Layard, which substantially coincide 
 with that which is printed, and I am thus in possession of some 
 important various readings which are not printed ; and I can 
 say most positively that there is not a single copy which con- 
 tains any other reading than uzbaku, which Dr. Oppert says "on 
 ne lit JAMAIS ! '' Elsewhere, e.g. in 1. 94 of the same plate, ni 
 is added; but there the sentence begins with ki, "when;** 
 
12 
 
 and I translate " when in Sur I was stopping.'* This is the 
 secondary permansive tense noted at the end of ^ 7. Where 
 there is no ki at the beginning of the sentence, there is no ni 
 at the end of tizbaku. In the second place, the existence of 
 the verb *liD is, so far as I am aware, a pure fiction. I never 
 recollect to have met it, and I believe no one else has met it. 
 It has been imagined purely for the purpose of neutralizing a 
 text, the existence of which is inconsistent with two gram- 
 matical dreams. But in the third place, even if the existence 
 of this verb be granted, it could not produce by any legi- 
 timate process such a grammatical monstrosity as yushakuni. 
 Dr. Oppert cannot produce any similar form. Isbakuni would 
 be legitimate ; and so would isahkuni, or with st or ss in the 
 place of s ; yusahkuni might pass also for Conj. III. ; but the 
 substitution of yu for i before sh — such a form as yupgaluni — 
 is unparalleled. Yu and i are not interchangeable, as he says 
 in his grammar, the former is used in Conj. III., the latter 
 in Conj. I.; but these two conjugations, the Pihel and Qal of 
 the Hebrews, he has blended together in hopeless confusion. 
 
 15. But, perhaps, the most extraordinary sentence in the 
 passage I have quoted is the last. The first question at issue 
 is, whether the Assyrians had, like the Hebrews, at least one 
 tense in each conjugation, in which there were no prefer ma- 
 tives. It is in regard to this tense that we are at issue ; that 
 they had also a tense or tenses in each conjugation, which 
 had preformatives, we are agreed. Now if I can establish 
 even a single instance, in which a verb has no preformative, 
 my case is proved. The fact is, however, that there are 
 hundreds of instances of permansive verbs in inscriptions of 
 every age, from Tiglath Pileser to Darius. Dr. Oppert says 
 that sarrakUy even if it should not be read sarratus, — a sup- 
 position which the variant form in ak (of which presently) 
 proves to be untenable, — may mean " I am a king," and yet 
 not be a verb. Here I differ from him. I say, in common I 
 believe with all grammarians, that a word which includes in 
 itself a pronominal subject, a copula and a predicate, is essen- 
 tially a verb. It might as well be denied that maJakta in 2 Sam. 
 iii. 21, was a verb, because malkcy " kings of,'' is a noun. In 17 
 
13 
 
 I. 32, etc., there are, after the completion of a sentence, no 
 less than eleven words ending in aku. I take them to be per- 
 mansive presents in the 1. s., ten of them belonging to tri- 
 literal verbs in the first, third or fourth conjugations, and the 
 eleventh to a quadriliteral. It is not the slightest objection 
 to this view that substantives or adjectives are in use, con- 
 taining the former part of the alleged verb. So it is in all 
 the Semitic languages. There is a nomen permanentis, and a 
 permansive present, the beginnings of which generally agree, 
 though their endings are generally as different as the cases of 
 reoe and the persons of rego. In the proper names Nahu- 
 naUdy " Nebo is glorious," and NaUd-Mar.duk, " Marduk is 
 glorious," as well as in na.h.da.ku in 1. 32, we have the per- 
 mansive present ; in E.I.H. I. 3, on the contrary, ru.ha!a 
 na!a.dmn, " the glorious prince," gives us the nomen per- 
 manentis, here an adjective. 
 
 16. In the Babylonian inscriptions the u at the end of the 
 1 s. is dropped. Thus in E.I.H. I. 39, we haye p'i.it.lu.fi'a.ak 
 he.Vu.ut.fuMn, " I am habitually worshipping their lordships." 
 The u, it will be observed, is also dropped in this affix, as it 
 sometimes is in the affix su, when added to a word ending 
 with a vowel. This is the 1. s. of the permansive present in 
 the first augmented conjugation of Tw^. Before this, in 
 1. 10, the king had used the permansive past (badly spelled), 
 h'i.it.l'u.hu b'i.e.fuAit.su.un, "he has been habitually worship- 
 ping their lordships." In the Nakshi-Rustam inscription, 1. 9, 
 we have sa.al.fa.ak, "I am ruling," as the translation of 
 patiyakhshiya, or rather of the latter part of this Persian word ; 
 for the preposition at the beginning is expressed in the 
 Babylonian text by a separate preposition. I was curious to 
 see what Dr. Oppert would make of this, and I turned to his 
 transcription of it into Hebrew characters in the Z. D. M. G. 
 XI. 136. I was no little surprised to find that he makes. two 
 words out of these four characters, namely, ^J^NI tO^t?^*- 
 The latter of these two words he has substituted for the ter- 
 mination ak, which he did not understand ! This led me to 
 look further into his transcriptions. In 1. 24 there is a pas- 
 sage, the true reading and signification of the first word of 
 
14 
 
 which, was suggested to me by Mr. Talbot. ^^ is zub as well 
 as lib. The reading is ziib.bufu sa 'ana.ku zi.ba!a.ka. I take 
 the ka to represent ak, which would look very awkward after 
 ha!a\ while if the jy were omitted it would be natural to 
 read zibak ; the word is really zibdak. I translate, *' they are 
 always made to will what I am willing." Here, again, Dr. 
 Oppert transcribes the four characters by ^^) ^55V instead of 
 *nX!lV > introducing a second imaginary word in place of a ter- 
 mination which he does not understand. 
 
 17. These are not the only instances of permansive verbs 
 in the IN^akshi-Eustam inscription. In 1. 26, we have kid.lu, 
 " has been holding," a deponent verb like dominatus erat ; 
 and in the following line we may restore the damaged word 
 of Conj. TV. na.su. 'u, " are carrying," from X^^ . This word 
 occurs again 1. 18 (" The lonians who maginat on their heads) 
 are carrying ;" as does the preceding word in 1. 11 (" my laws) 
 they are holding ;" -<^>^>-y, which must be here read ha^ being 
 added to the singular, to form the feminine plural. An eighth 
 permansive form occurs in 1. 21, where in speaking of the 
 depraved state of the people before Darius became king, the 
 obscure word mmmuhu is used. To these eight instances of 
 permansive words I think myself entitled to add two others 
 in which 'anaku is used as a verb, there being no other in the 
 sentence. J^ow, I can only count thirty-one mutative verbs 
 in this inscription ; so that the permansive verbs are in the 
 inscription about a fourth of all the verbs. In the inscriptions 
 generally they are perhaps a fifth or sixth of the whole ; but 
 this is a very different proportion from " one to thousands." 
 
 18. I will now give some examples of permansive forms of 
 two verbs, the meanings of which have been strangely mis- 
 conceived, ^\^^, "to have," and K^l, "to be." Both are 
 used in the mutative, as well as in the permansive tenses ; 
 and in the former verb, there is a great liability to confusion, 
 the aorist being in both the first and third persons singular 
 i.siy as in E.I.H. 10, 15, where we have the derivative form 
 'a.ya 'i.si na.ki.ri, " let me not have enemies." The 3 s. of the 
 permansive form would, I presume, be written precisely alike ; 
 and i.su!u and i.sa!a would be the 3 p. m. and f. in both the 
 
15 
 
 aorist and the permansive tense. We have, however, in 9 I. 58, 
 a form which is clearly permansive, ma.JiLra (in another copy 
 sa.ni.na) 'as tahazi la i.sa.'a.kuy " a confronter (or rival) in the 
 close combat I have not." In 1. 44 of the same plate we have 
 ma.hi.ru la i.su.'u, " a confronter has not had ; " and in 42 L. 
 40, the name given by Sennacherib to his palace is sa sa.ni.na 
 la i.su!u, " which has not had a rival." 
 
 19. The verb " to be " occurs frequently in a variety of 
 mutative forms; as in 3 m. s. Conj. I. sa.nin ul ib.si, "rival 
 (or adversary) there was not," 18 I. 43 ; mal.ku gah.ra.a!a ul 
 ih.siy "king prevailing against me there was not," 145 B. 1 ; 
 for which we have in 171 B. 7, gah.ra.a.su ul ih.su! u, " one 
 prevailing against him (whom) there was not." Here the u 
 at the end of the verb is not the formative of the preterper- 
 fect, but a relative enclitic ; and it may be so in some of the 
 instances of the permansive past which I have given. It is 
 to be observed that the u and ni of all the secondary tenses 
 are annexed immediately to the primary tense ; whereas, if an 
 objective affix follow, the enclitic is placed after it. Thus, in 
 89 L. 40, we have, after m, i.qa.hu.su.ni, " which (the people 
 of Hatti) call." Here i.qa.hu is the 3 p. m. of the present 
 of T\'2p\ su the affix "it," which, in combination with the 
 preceding sa, must be translated "which," as in Hebrew; 
 and ni the relative enclitic is after the affix. Were the 
 perfect to be expressed, we should have i.qa.hu.ni. 
 
 20. Other forms of this verb are i.b'a.as.su.'u, 3 m. p. present 
 "do not exist," E.I.H. 2, 20; I have also met i.b^a.as.si in the 
 singular; and on Bellino's cylinder 1. 31 we have the infini- 
 tive *a.di la ba.s'i.e 'u.sa.lik.su.nu.ti, " till there was none (left) 
 I made them depart." In 38 I. 18, this is repeated with two 
 various readings, i for e, and su for su.nu.ti. The singular su 
 is very often used, when there is no definite antecedent, in the 
 sense of "people," as in 47 B. 88, etc. In Qo I. 2, 13, we 
 have 'u s'a.al.mi.is (for 'usalmisu) to express the nia.da 'u.s^a.al. 
 mi of E.I.H. 6, 43, " the people I caused to see." Once more 
 we have the 1 s. of the aorist of i^^^ in Conj. Y. in 42 L. 44, 
 *a.gam.mu 'u.sab.si " a lake I made " (lit. " I caused to be.") 
 
 21. The permansive form of this verb is most commonly 
 
16 
 
 found in the parenthetic formula ma.la (or mal) ha.su^u, "as 
 many as there are." I must here observe that while I have 
 always, in common with Sir H. Rawlinson, assigned to mala 
 an affirmative signification, instead of a negative one, as Dr. 
 Oppert assigned to it in his grammar, I never attributed 
 to basuu the signification of number, as I am represented to 
 have done in Dr. Oppert's late paper. He now admits that 
 mala basuu means "all," though he does not yet see his mistake 
 in supposing ba.su.'u to mean " bad" or " contemptible." That 
 this word and its Accadian equivalent *^y<T5 *^ET nal.la, as I 
 read it with some small doubt, means simply " to be," I feel 
 perfectly sure. But, as this is rather a matter for a lexico- 
 grapher than a grammarian, I will not discuss it here. 
 
 22. It remains for me to give the forms of the permansive 
 tense in the difierent conjugations; of pagil Conj. I. enough 
 has been said ; I have also given examples of pitgul Conj. 1. 1. 
 Another occurs in 37 I. 66, where Sennacherib speaks of 
 people, " who to the kings my fathers," la kit.nu.su, " were 
 not submissive," from ^^D. The conjugation generally in- 
 dicates repeated action, and the tense that this repeated action 
 was continued for a length of time. 
 
 23. The permansive of Conj. II. is of the form napgul ; 
 and of II. t. I presume it would be naptegul. I have, how- 
 ever, met with no instance of the latter, and only one of the 
 former. It occurs in 40 L. 3, sa.qis n a.an.zu.zu, " they were 
 fixed on high." The root is TT^. The n of the root is here 
 preserved, but it is often assimilated to the z when it comes 
 in contact with it ; and we have thus apparently, though not 
 really, a root of which all the letters are T. In the present 
 text, which is to me in great measure obscure, the verb quoted 
 lies between two other permansive ones, pit.hc.lu Conj. I. t. 
 and n'u.us.su.ru Conj. III., which, again, is followed by 
 nu.um.mu.ni in the same conjugation. 
 
 24. The permansive form of Conj. III. is but rare : the 
 form is paggal. Examples are qar.ra.da.ku, " I am very 
 strong," 17 I. 32 ; and 'al.la.ka bir.lia.a.'a, " my knees are 
 moving briskly," 16 II. c. 30. This is the 3 f. p. from "^Sn- 
 In the following line we have another permansive verb, la.^a 
 
17 
 
 ni.ha se.p'a.a.'ay "my feet have not any rest." This is from Hli 
 in a conjugation peculiar to concave verbs, which is in the mu- 
 tative parts analogous to Conj. III. of perfect verbs, but in the 
 perraansive parts has generally a passive or neuter signification. 
 Another example will be found in § 11. Examples of Conj. 
 lY., of which the form is puggul, have been already given in 
 §§ 11 and 17. The verbs in the last section are Tw'2 and H^D- 
 
 25. I have not met with the permansive form of Conj. Ill.t. 
 or lY. t., nor of Y. or Y. t. The former of these I should 
 from analogy expect to be sapgal ; and the latter, perhaps, 
 satpegal. Conj. YI. makes supgul; so, at least, I infer from 
 the defective and otherwise irregular verbs in this tense which 
 alone I have hitherto found. From \)T\ we have certainly 
 8u.lu.hu y " they were made to go," 49 I. 4, 30 ; and I suspect 
 su.qu.ru f in an obscure passage, 39 L. 44, to be another 
 3 m. p. from *np^. On the other hand we have su.qa.lu.la 
 and su.qu.lu.la from 7/p in 18 I. 62, as if the form had an 
 additional syllable. The fact, however, of the additional vowel 
 being in difierent copies a and u, suggests that it is irregularly 
 introduced ; perhaps in order to distinguish to the ejQ a deri- 
 vative of this root from one of 77H or 77^, which were both 
 in use ; or perhaps, because the liquid / admitted a vowel be- 
 fore it (see § 24 of Chap. II.) I would therefore read suqlula. 
 
 26. There are two verbs in a permansive tense beginning 
 with ^TTT, of which the conjugation has been doubted. In 
 the one this character is followed by qui, in the other by kun. 
 There can be no doubt that the verbs are 7p^ and p^ ; but 
 the initial character admits two readings. It may be sit, and 
 the verbs would then be in Conj. I. t. ; or it may be mus, 
 which, as a variety of examples prove, may be interchanged 
 with mus before p or ^. Thus, mus.ki and mu.us.ki are both 
 used for the name of the country, Musuk, Heb. ^K^^* Think- 
 ing that the significations of the verbs in question were not 
 such as suited the Conj. I. t., it occurred to me that m was 
 here substituted for s as a pre formative, on account of the 
 first radical being s, the conjugation being YI. As, however, 
 I have found no such tense beginning with mu, or mus, and 
 as 'u.s*a as.kin is met with in Conj. Y., with the preformative 
 
 2 
 
18 
 
 s before a first radical s, I am now inclined to think tliat the 
 sixth conjugation has never m for its preformative, and that 
 we should read sit.ku.nat su.bat.sun in 41 B. 39, "their 
 dwelling is lying ;" and in the Equinox tablet, K. 15, sit.qu.lu, 
 (the day and night) " are balancing one another/' 
 
 27. Of the permansive form of Conj. YI. t. I have as yet 
 met only one example. It is in some respects a very valuable 
 one, as it clearly establishes the existence of a permansive 
 past, terminating in u. The nominative is singular, and there 
 is no relative particle on which it could depend. It occurs in 
 51 I. 1, 32, where we have la su.te.su.ru mu.ze.'e mi.'e.sa, 
 "the exit of its waters had not been made straight." The 
 verb is ^^% and the form is probably sutpegul; but from 
 the weakness of the first radical, this is not certain. I have 
 not met with the permansive form of either YII. or YII. t. 
 
 [Chapter II.] — Nouns. 
 
 1. In the present chapter I propose to consider the declen- 
 sion of Assyrian substantives and adjectives. In the follow- 
 ing chapter I will treat of their syntax, endeavouring to show 
 under what circumstances each of the forms, the connexion of 
 which is explained in the present chapter, is found to be used. 
 In a future chapter I will treat of the connexion between the 
 primitive forms of nouns and their significations ; nouns which 
 have the same primitive form bearing, in a very great de- 
 gree, the same relation in their meanings to their verbal roots. 
 
 2. I think it best to reserve the full consideration of the 
 forms of nouns, as connected with their significations, till I 
 have treated of the verbal forms of roots ; but as I shall have 
 to speak of these forms occasionally in connexion with declen- 
 sional difierences, I feel it necessary to define the terms of 
 which I shall have to make use. 
 
 3. A perfect Assyrian root consists of three consonants, 
 which are called its radical letters ; and every Assyrian word 
 derived from such a root consists of three consonants at least, 
 and two vowels at least. Words which difier from one an- 
 other only in their radical letters are said to be of the same 
 
19 
 
 form. Thus marzi and namri are of the same form ; as are 
 puluhti and tukulti ; musaskin and musalhir ; gahsaati and 
 rapsaatif etc. 
 
 4. If we assume a root of three letters, we can always ex- 
 hibit a derivative of this root of the same form as any deriva- 
 tive of any other root that we may meet with. It has been 
 the custom with Hebrew and Arabic grammarians to assume 
 7j?^ as a standard root. It was chosen on account of its 
 signifying "to do" or "to make," which seemed the most 
 natural type of a verbal root ; but it is objectionable on 
 account of the weakness and uncertainty of sound of the 
 second radical. I substitute for this a H? which gives an 
 imaginary root, but one which is very convenient. The words 
 in the preceding section would become, by substituting the 
 three letters of this root for the three in each of them, pagliy 
 puguUi, musapgil, pagla'ati; and I take these as the names of 
 the forms to which these words respectively belong. 
 
 5. All the forms given in the preceding section are forms of 
 nouns ; but there is a difference between them. One of the four, 
 musapgil, is Si primary form, or theme; the three others are de- 
 rived forms, or cases. Observe that, besides the principal theme, 
 which is singular, there are plural and dual themes, and some- 
 times feminine themes, singular and plural. In the present 
 chapter I will show how the cases are derived from the theme, 
 and to a certain extent how the theme may be recovered from 
 one of its cases. The same derived form may, however, be 
 derived from different primary forms. For example, pagli is a 
 case, the theme of which may be pagal, pagil, or pagul. Had we 
 only the words marzi and namri before us, we should have no 
 means of judging to which of three possible themes we should 
 trace each of them. Other derivatives of the themes in ques- 
 tion are required before we can say that the former theme is 
 maruz, the latter namir. 
 
 6. The themes last given differ from musapgil, in that they 
 contain no consonant but the three letters of the root. I call 
 such themes as consist of the three radical letters, with two 
 short vowels intervening, simple themes. Those which contain 
 any addition, reckoning as such a quiescent letter, X, ^> or V 
 
20 
 
 which may be considered as lengthening a vowel, I call 
 augmented themes. 
 
 7. The following are the different kinds of augmentation 
 which an Assyrian root may have, and it may have two or 
 more of them at the same time. First, it may have a prefixed 
 addition consisting of one or more servile letters with a vowel 
 or vowels, as in miisapgil; 2nd, it may have a medial addition 
 of a servile letter, as in pitguly or in pdgil, where the the first 
 syllable would be in Arabic li , in Hebrew iS or b ; 3rd, it 
 may have the middle radical doubled, as in paggal; 4th, it 
 may have the final radical doubled, as in paglal ; or 5th, it 
 may have one or more servile letters added at the end as in 
 pugldn or paglut. I do not consider the at or it at the end of 
 feminine nouns as constituting an augmented theme, but as a 
 declensional modification of a theme. The form pugulti in 
 § 4 I consider to be a feminine case of puguly or a case of 
 puglat, the feminine theme oi pugnl. 
 
 8. There are many feminine themes of this description. 
 Some of them are substantives denoting females, or what are 
 considered to be such, where the true themes denote the cor- 
 responding males. Thus sa.pH.in is " a sweeper away," applied 
 to a king, 40 B. 25 ; sa.pi.na.at is the same, applied to a 
 chariot, which is conventionally feminine, 41 1. 82 ; mu.rap.pis, 
 33 L. 9, and mu.rap.pi.sat, 38 L. 5, mean "enlarger," and 
 are applied to the king and to the sceptre respectively. 
 Adjectives, which always agree with their substantives in 
 gender, form feminine themes of this description : they are, 
 however, rarely used in the theme, either masculine or femi- 
 nine. A few nouns are used in the masculine and feminine 
 forms without distinction, as puluh and pidhat, " fear ; " and 
 there are several which are only found in the feminine form, 
 as hirat, "a wife;" ^rzit, "a land," or "the earth;" 'imt, 
 "fire;" ''amat or tamat, "a sea." 
 
 9. Besides these, there are many feminine themes, which have 
 a collective signification, and may very often be translated as 
 plurals. In 42 B. 70, we have'^^.A•m lib.na.as.su, "I made 
 fast its bricks ;" as.su stands for at.su by a euphonic change 
 which will be explained in § 56 of this chapter. The word 
 
21 
 
 is lihnaty the feminine theme of lihiny which would signify " a 
 single brick." It is not a plural, as might perhaps be thought. 
 The plural would be, according to analogy, assuming lihin to 
 be a feminine, lih.na' a.ti.sii. Besides, it is decKned as a 
 feminine singular; and similar nouns are accompanied by 
 adjectives in the feminine singular. It is, as I have stated, 
 a collective singular, used for a plural. There are above a 
 score of such collectives, from perfect and imperfect roots, in 
 frequent use in the texts. 
 
 10. All nouns terminating in at or it servile are feminine ; 
 other nouns 7nay he so. Examples of such are haluz, " a 
 castle" (I put the second vowel of the theme in roman, be- 
 cause I am not sure what it is ; see § 5). In 52 I. N°- 3, 2.16, 
 we have ha.al.zi raMMm, " of the great castle ; " and the 
 plural occurs with feminine adjectives (>->- being used 
 for the first syllable) 146 B. 6, 7. In 17 II. 32 I, we have 
 li.sa.an li.mut.tUf " a sore tongue," the adjective being femi- 
 nine. Other examples of feminine nouns, not so by syllabic 
 addition (or, as Hebrew grammarians express it, by motion), 
 are' urn, "a mother;" 'istar, "a goddess;" (I believe, a loan 
 word, originally signifying *'a star;") 'umman, "an army;" 
 'uzun, "an ear; qat or qaat, " a hand " (we have qa.as.su, " his 
 hand," 49 B. 32 and 8 II. 45 r ; the plural in the principal 
 case is qa.ta.tu 8 II. 40 r, which proves that the t is radical), 
 and all augmented forms ending in ut, as sarrut, " a kingdom" 
 or "reign." 
 
 11. Nouns have three numbers, the singular, the dual, and 
 the plural. The dual is not often used, and only, I beheve, 
 in the theme (§ 5) ; the cases of the dual do not seem to be 
 distinguished from those of the plural. The dual is of course 
 most frequently used for nouns which express objects that are 
 in their nature double; but I think that I have met with 
 duals of other nouns. 
 
 12. Nouns in the singular number have three cases in 
 addition to the theme. I call these the first or principal 
 case, the second case, and the third case. They are used so 
 differently from the nominative, genitive, and accusative cases 
 of European languages, that I think, on mature considera- 
 
22 
 
 tion, the use of tliese terms decidedly objectionable, as likely 
 to mislead. Their use will be explained in the following 
 chapter. Here it will suffice to say that the first case ends 
 in u or um : it appears from the grammatical tablets that the 
 Assyrians considered this to be the leading form of the noun. 
 The second case ends in i, e, or im ; and the third in a or am. 
 The third case is only used in the singular number. 
 
 13. Before I speak of the formation of these cases from the 
 theme, I must go back to the classes of roots. I stated in § 3 
 that a perfect Assyrian root consisted of three radical con- 
 sonants. It does not follow, however, that every root which 
 consists of three consonants is perfect. Not to speak of roots 
 having for the first radical a X which is often assimilated to 
 the following consonant, and which sometimes disappears 
 altogether ; nor of surd roots, as they are called, of which the 
 second and third radicals are the same, and which have some 
 peculiarity consequent upon this ; there are certain weak 
 letters, the existence of which in a root renders it imperfect. 
 
 14. I consider ^, T\^ V ^> and J? to be weak letters; I do 
 not include T\ in the list ; preferring to regard those roots in 
 which a weak letter is found where cognate languages have 
 n, as substituting an H or ^ for it. This is the case in a few 
 roots, of which the most common are pITl, hHS? and Hp/- 
 I consider the corresponding Assyrian roots to be pj?^? yn3> 
 and yp7 ; as I consider the Assyrian root corresponding to 
 the Hebrew ^J?^^ to be ^H^. In this root the second radical 
 never exhibits any symptom of weakness. 
 
 15. All these letters are capable of causing the assimilation 
 of the preceding and following vowels, which sometimes causes 
 the contraction of two syllables into one ; for when a weak 
 letter has the same vowel before and after it, the two may 
 coalesce, though they do not necessarily do so. An example 
 may be given in the third person singular of the present of a 
 verb with a weak letter for its first radical, suppose \)T\. 
 Instead of lallak^ which would be the regular form, the 
 Assyrians first substituted i for a, assimilating the vowels, and 
 then contracted Vil into simple il. Thus they wrote illahy a 
 dissyllable, where a regular verb would have three syllables. 
 
23 
 
 16. So, when the second radical was weak, they wrote 
 ru^iiq for ra\iqy like maruz, and perhaps pronounced 
 
 >^TTT ^ ^'^^^ >^y>--< ru.uqJi as a dissyllable, when the 
 corresponding form of a perfect root would have three 
 syllables. It must be observed, however, that when the two 
 vowels of the form were characteristically different, an as- 
 similation could not take place; for example, in the form 
 pagil, the nomen mutantis, no assimilation is permitted. 
 We have Xf ^flf qais, 17 I. 9, and XJ ^^ V qd'Lsat, 
 66 II. 9, " an ensnarer (of the living)," applied to a god and a 
 goddess. See Gesenius under SJ^Ip, 2. Examples of assimila- 
 tion when the third radical is weak will be given presently. 
 
 17. Besides these irregularities, which are common to the 
 ^Ye weak letters, three of them, K> V and *^, are liable to be 
 dropped altogether. In the case of one of these being the 
 middle radical, this may always be considered a case of con- 
 traction, and it may be so sometimes when the first or third 
 radical disappears ; but the dropping of a letter of which I 
 speak here, and of which I will shortly give examples, is not 
 the result of contraction. Where a perfect root would have a 
 complete syllable, though a short one, a root beginning or 
 ending with one of the letters in question will sometimes drop 
 ijt,<^pparently on the principle that a short unaccented vowel, 
 paving no substantial consonant to support it, is a nullity. 
 Such Hebrew forms as 75?5' HX^Jj and K^ll will show what 
 I mean. In the first and second, weak letters, called by 
 Hebrew grammarians H and *• , are altogether omitted ; of 
 these letters, the latter is in Assyrian certainly 1, and the 
 former is possibly ^ . In the second and third examples an &< 
 is written, but regarded as a nullity. 
 
 18. I now come to consider the different modes of declining 
 themes. The most natural mode is simply to add to the 
 theme the three terminations, whjch I will here caU u, i, and 
 a, reserving tiU the next chapter an explanation of the modi- 
 fications of these terminations indicated in § 12. This natural 
 declension is always used when the theme terminates in a 
 strong consonant preceded by a long vowel, such as that of 
 
24 
 
 the form pag^l, tlie infinitive, or nomen mutationis, or tliat of 
 the augmented forms in an or ut, whatever may precede the 
 last syllable. It is also used when the last consonant of a 
 monosyllabic theme is strong, provided that the theme be not 
 derived from a surd root. Examples are mut, *' a husband," 
 whence muti; ah, "a father," whence abu, ahi. Lastly, it is 
 used when, the last consonant being as before, the last vowel, 
 though short, is so completely separated from the preceding 
 vowel, that a contraction is impossible. This happens in 
 many forms where a double letter intervenes, as tapgal, pitgul, 
 musapgil, etc. 
 
 19. There are three cases indicated in the preceding section 
 in which the noun is not declined in the natural manner there 
 indicated. 1st, The last consonant may be a weak one ; 
 2nd, the theme may be a monosyllable derived from a surd 
 root; and 3rd, the theme may terminate in a strong con- 
 sonant, preceded by a short vowel ; and this may be separated 
 from the preceding vowel by so thin a partition that a con- 
 traction becomes possible, and generally takes place ; the last 
 vowel being dropped. I will treat of these three cases in 
 their order. 
 
 20. If the theme terminates in a weak letter, the preceding 
 vowel is assimilated to that of the case. Thus from 'azi, 
 "going out" (root ^^"l), we have ]] tr"^ ^!!T^ 'a.zw*w, 
 67 I. 2. 37. So we have from the theme which signifies 
 "a crown'' — it is imcertain what it is — y][ ^^ ^*a,gu-Uy 
 
 Tr ^yiA ^Tt '^^-Oe'e, and ^^ ^]]]^ ]] 'a.ga.'a in the 
 three cases. In 30 II. 19 r. we hoNQ^a.gu.^u e.lu.'u, "the 
 high crown," where both adjective and substantive end in 
 weak consonants. Examples of the other cases will be found 
 in 9 I. 5 and 21. So again from the theme for " mouth," we 
 have V^ Jryyyjz pn:u in 39 II. 1 /, ^y^ Jr£ pi:i 39 II. 5 /, 
 and ^ y^ pa' a 1 L. 14. ^ Here also the actual theme is 
 uncertain; nor is the root itself less so. Perhaps it is 
 ^KS. If so, two weak letters would come together, and 
 a contraction would take place, as it certainly does in 
 ^Tr "^T^ KT^ It -^^ ^'^^-^^ *^''^^> "another work," 16 I. 20. 
 
25 . 
 
 I cannot think, however, that a contraction takes place as a 
 matter of course, and that we should read 'agu, elUy pu, etc., 
 when a second vowel is expressed. 
 
 21. In monosyllabic themes derived from surd roots, or 
 from which surd roots are derived (for I suspect that 
 many of these themes are adopted from foreign languages, 
 and that the verbal form, if in use, is derived from the sub- 
 stantive or adjective), the final consonant is repeated before 
 the case-ending. Thus from sar^ "a king," we have sar.ru; 
 from lib, " a heart," lib.bu ; fvom' um, " a womb" or " mother," 
 'um.mu. A few dissyllables also double their last letter, as 
 *agam, " a lake," whence^ a.^am.mu, 41 L. 44. Perhaps this 
 should be considered as derived from a quadriliteral root. 
 The word is singular, for it is referred to by ^ sa, 
 
 22. The omission of the last vowel before the case-ending is 
 very common. It always takes place in the forms pagal (but 
 not pagal, the infinitive), pigil, and pugul, and in augmented 
 forms like muptagil, in which the second radical stands alone 
 between the last two vowels. It takes place for the most part 
 in the forms pagil and pagul, as in n^a.am.ri, from namir, 
 mar.zi from maruz, and many others ; but there are excep- 
 tions. In 7 I. F. 24 we have bit.su la.bi.ru, "his old house ;" 
 older inscriptions have la.be.ru , 20 1. 3 ; and it is only in the 
 inscriptions of Nebuchadnezzer that I find ^^ ^Z^^ hT'^T 
 ra.ab.ri, Q6 I. 2. 39. We must, however, distinguish the 
 form pagil, the nom^en mutantis, and pagul, nomen mutati, 
 from pagil and pagul, which like pagal belong to the nomen 
 permanentis. Both forms may belong to the same root, and 
 to the eye may be the same. Thus ^y Jjy may be 
 "ruler," md.lik, which would make md.li.ku, and in the 
 feminine md.li.kat, Q>Q II. a. 4:\ or "king," ma.lik, which 
 would make maLku, 145 B. 1. So dal.hu, 74 B. 13 (whence 
 da.li.'h.tu), means " troubled, being in disturbance ; " but 
 ddlih, dd.Ii.hat, means "the putter in disturbance, the dis- 
 turber," 66 II. a. 4. 
 
 23. The three forms piggal, piggul, and puggal require 
 special notice. The two former are derived from the third 
 
26 
 
 conjugation, and the last from the fourth. It will appear 
 very strange to an Hebraist, but it is certainly the case, 
 that the duplication of the second radical is very apt to dis- 
 appear, and the initial vowel to be the sole characteristic of 
 the conjugation; the forms thus become pigal, pigul, and 
 pugal, and the first two of these generally, though not always, 
 drop the second vowel in their declension. Thus for gissar, 
 the intensive form of gasar, we have gisar ; and in the second 
 
 case ^y >^TT<T gis.riy 17 I. 1, a little more forcible than 
 g^a.as.ri, which we have in the parallel text, 32 I. 1. The 
 meaning is "bold, daring;" and we have the feminine gi.sar.tUy 
 66 II, a. 1. The word which is constantly used for "enemies" 
 in the Behistun inscriptions is similarly contracted. It is 
 nikrutu ; in the singular it would be nikrUy from nikar. So 
 ^y>- >i^ lim.nu in 48 B. 31 is from liman, for limman, 
 "annoying" or "injurious." But this word gives occasion 
 for another remark. The word limnu may come from a theme 
 limuriy as well from liman ; from the nomen mutati as well as 
 from the nomen per manentis. We have, in fact, in 17 II. 32. 
 "i*^*- ^ ^y>- ^ pu'u lim.nu, "an injured mouth;" fol- 
 lowed by li.s'a.an li.mutM (for limimtu), " an injured 
 tongue;" the text evidently speaks of wounds. The two 
 adjectives are distinct in the feminine singular, which would 
 be with active signification li.mat.tu, hut are confounded in all 
 other parts. 
 
 24. In some adjectives, however, of .these forms the final 
 vowel of the theme is not dropped. I have met i.sa.ru from 
 isar for issar. The reason of this I take to be that the initial 
 vowel is here virtually formed by contraction. The root is 
 ^tJ^^ ; and the i of the root and the i of the form combine 
 together, so as to form an obstacle to further contraction. 
 The retention of the vowel in sufuru may be explained on the 
 same principle, the root being ^Hl and the form supgulu, so 
 that two u'*s combine in the first syllable. It is less easy to 
 account for pumalu ; perhaps the a is long by nature, or per- 
 haps the liquid third radical may have rendered it desirable 
 that a vowel should precede it. I am the more inclined to 
 
27 
 
 think this, because I find ha.ta.lum in 38 II. 74 r. ; but in 
 50 I. 4, 28, I find Va.at.lu.ti. The theme is hatil, as the 
 feminine form ha.fi.il.tii in 172 B. 5 proves conclusively ; the 
 a before the I must therefore be euphonic. 
 
 25. The feminine theme may be said to be formed from the 
 principal case of the masculine by changing u into at or it. 
 I am not able to give any rule by which it can be absolutely 
 determined which vowel should be used ; and in most adjec- 
 tives it cannot be known, the theme not being in use, and its 
 vowel being dropped in the cases. On the whole, however, I 
 find at to be more frequently used than it ; and I believe it 
 is exclusively used when the theme is a nomen mutantis of 
 any of the conjugations, or a monosyllable derived from a 
 surd root. 
 
 26. The feminine cases are generally formed by adding tu, 
 tiy and ta to the theme, properly so called ; sometimes, how- 
 ever, by adding u, i, and a to the feminine theme. The 
 latter mode is the only one possible, if the theme be a mono- 
 syllable derived from a surd root. Thus from sar, " a king," 
 we have sarrat, "a queen," in the theme, and sarratu in the 
 principal case. From hirat, " a wife," on the contrary, we 
 may have hirttc, though hiratu is also in use. So also from 
 napsat, " a life," theme napis, we have sometimes napsatu, but 
 more frequently napistu. Adjectives almost always form their 
 feminine cases by adding tu, etc. to the proper theme, at least 
 if the root be a perfect one, though they sometimes admit the 
 other form also. Thus we have saplitu, "low," as well as 
 sapiltu ; mahritu, " old," as well as mahirtu. From dan, 
 however, we could only have dannatu ; and, on the contrary, 
 from eli only elitu. 
 
 27. The addition of the feminine case-endings' to the 
 theme sometimes causes a change in its last letter. For 
 instance, n is changed into ty as in the instances already 
 given in § 23 ; the masculine case ^>- ^ lim.nu cor- 
 responding to the feminine *"£B[ *"|I-^ I ^*-^^^^«^^> for 
 limuntu. So we have >^£J<J ^11^1 *"££! li.hit.tu^ for UhintUy 
 49 B. 52. The feminine theme is lihnat. It is probable that 
 
28 
 
 d and t when final were also assimilated to t when they pre- 
 ceded it in the feminine cases ; but the Assyrians did not in 
 writing distinguish these letters from t when terminating a 
 syllable. 
 
 28. The letters s, z, s and z, were sometimes, but not neces- 
 sarily, changed to / before the t of the feminine cases. I be- 
 lieve, however, that none of the three last-mentioned letters 
 is ever found preceding the t. It is changed either into s or 
 into /. Thus in 17 II. 27 I. we have ma.ru.us.tu, evidently 
 the feminine of mar.zu. In 16 I. 76 we have mayu.us.ta; 
 and in a parallel text, 27 I. 92, we have ma.ru.ul.ta. In the 
 upper part of 27 I. 10 we have ris.ii for rizti; and, again, in 
 12 I. 90 we have ri.hH.il.ti, evidently for rihizti. As to s itself 
 we have Iu.Vums.U, " clothing," 38 II. 48 m., but much oftener 
 fcTf *"*"*^Tyy^ >^y»^ lu.huLU. a comparison of the text 
 last quoted with 67 II. 62 must, I think, satisfy every one 
 that these are the same word. I at one time read the initial 
 character tib^ and translated it, "what was dyed with;" and 
 I perceive that Dr. Oppert and M. Menant have done the 
 same ; but here we have luhulti matisunu, " clothing of their 
 people," just as in 38 II. 48, 49 m. we had luhusti ilutisunUy 
 " clothing of their godships." It follows that hirmiy which 
 usually follows luhulti^ does not signify a dye-stufi" but a 
 material ; no doubt ** wool," FepFtov ; compare vellus for velves 
 (as mollis for moMs). The other word, which is commonly 
 joined with birmi, T^T >+- , ku.kum, or kiim ? for I suspect 
 that the first character is a nonphonetic determinative of 
 names of plants, I take to mean "flax" or "cotton." 
 
 29. The Assyrians formed their plurals in several ways, of 
 which I will treat in succession. Some plurals are formed by 
 the addition of n for the theme, and ^ nu and J^ ne for 
 the case ending, to the third case of the singular. These are 
 sometimes written in full phonetic characters ; but very often 
 the theme, or the principal case of the singular, is accom- 
 panied by the character \«<. This combination must be 
 understood to denote the proper plural form, which the reader 
 is presumed to know;— and this constitutes one of the chief 
 
29 
 
 difficulties in reading Assyrian. Sometimes, however, as a 
 guide to him Jf^ is added, which I read ne • because, while 
 ne and ni were expressed alike, te and ti were distinguished ; 
 and it is ^J te which is used similarly to £yy" when the 
 plural theme terminates in t 
 
 30. Examples are ^]]]] ^ ]] £f^ dup.p'a.a.ne, "clay 
 tablets," K. 116; where K. 131 has ^]]] fn^ 5f^ and 
 K. 136 ^yyyy \^ only. All these tablets contain the 
 same text. So we have >->- ][][ ^^ hal.za.ne, "castles." 
 32 I. 50 and >->- t*"^ T"^^ ^«^^.^^^ with plural sign (to be 
 read halzan or halzane) 146 B. 6, 7, 8. Observe that the 
 >^^yy prefixed to this word is a nonphonetic determinative. 
 In 28 1. 1, 12, we have har.s^ a.a.nu, " woods ; '' and in 145 B. 2, 
 har.sa.ne, in the second case. In 39 II. 11 Lypa'a.nu. from 
 pa.^a (% 20) is "mouths," as the Accadian equivalent proves; 
 but in 17 II. 31, pa^an is used for "a face;" it takes, how- 
 ever, a plural adjective. In 33 L. 6, e.mu.q^a.an^ "powers," 
 and in 30 II. 14 r., riJaMn^ "heads," are written in full. 
 The singulars emuq and tib are in common use. 
 
 31. The plural in an only appears in a limited number of 
 substantives, some of which admit also different plurals. It 
 is, I believe, never used in the case of an adjective or of a 
 substantive which is feminine by termination. Several plurals 
 in an are, however, feminine, as well as the singulars from 
 which they are formed, as appears from the adjectives which 
 accompany them. Such are emuqan^ risan, and halzan, cited 
 in § 30. It would, therefore, be incorrect to say that an was 
 the termination of the masculine plural. I believe that origi- 
 nally, in the language from which all of the Semitic family 
 have diverged, the addition of am or av to the singular theme 
 rendered it plural ; and that at was added as a feminine ter- 
 mination to both singular and plural. Thus the theme alone was 
 of either gender, as was the plural in am or av ; but at in the 
 singular and ami in the plural were distinctively feminine. 
 From the original am, which was retained in Phenician, came 
 the Hebrew im, and the Assyrian and Himyaritic an ; from 
 
30 
 
 the av came the Egyptian u. The Aramaic and Arabic 
 forms, in which both vowel and consonant are different from 
 what they were originally, are probably of later origin than 
 the others. 
 
 32. Another form of the plural ends in t in the theme, and 
 in tu and ti in the cases. Originally, I conceive, as I stated 
 in the last section, this termination was avat or awat^ and was 
 peculiar to feminines by termination. This was in time 
 changed to daty sometimes contracted to at; but in the cases 
 
 there is generally, though not always, an YI between what 
 
 belongs to the theme properly so called and the preceding ; 
 and this should, I conceive, be sounded as a distinct syllable. 
 In Arabic, as well as in Masoretic Hebrew, the termination 
 CL>1 or ni is pronounced as one syllable, but the quiescent 
 letter which is always inserted indicates contraction ; and the 
 Hebrew holem manifestly stands for awa. 
 
 33. This plural in at is almost always used for feminine 
 adjectives, and for substantives feminine by termination. It 
 is also used for many substantives which are feminine other- 
 wise than by termination. Thus we have ^um.ma.n^a.atf 
 "armies," 146 B. 4, and ' iim.ma.na.tey 151 B. 12, from 
 
 *umman. In 37 II. 11, the plural of " mouths" is stated to be 
 
 ji|^ TI iZ^Y^ -^ '^ pa^a.tum or pa.'a.nu; pu.^u in the 
 
 singular and pa' an have masculine adjectives, 17 II. 32 and 
 31 1. I presume, however, that pa.'a.tu would take a feminine 
 one. Irregularities of this kind are found in all languages. 
 
 34. Examples of this plural when the singular is feminine 
 by termination are zi.ra.'a.te "high, supreme,'' 33 L. 6; 
 eJaJum, " high,'* 30 II. 14 r. ; dan.na.' a.ti, " strong," 
 146 B. 6; mar.za.'a.tiy "difficult of access," 146 B. 7; all 
 which adjectives are in concord with the plurals in an men- 
 tioned in § 31 ; hi.ra.ti ra.ha^ adi^ "great wives," 153 B. 12, 
 from hirat. This is a theme feminine by termination ; but 
 it must not be supposed that hir would mean " a husband." 
 
 The latter I take to be ha^ar, whence ][][< Y|[ Y! ^YYT 
 
 Ka.ala.rUy 36 11. 40 /. ; hura.tum and hH.ir.tum in line 43 /. 
 
31 
 
 are equivalent forms (see § 26), or perhaps dialectic variations, 
 of tlie principal case of Mr at, " a wife,'* which is in the same 
 relation to the former theme as din or dinat, " a law" or 
 "judgment," to da" an, " a lawgiver" or "judge." 
 
 35. Besides the two plurals of which I have hitherto 
 spoken, which I take to have been the original forms common 
 to all the Semitic languages, there are other forms which the 
 Assyrians appear to have developed after their separation 
 from those who used the other languages of the family. One 
 of these is the masculine plural in ut in the theme, and utu 
 and uti in the cases. It differs from the feminine plural last 
 mentioned in having ^* in place of a at the end of the theme, 
 and ►^TTT^ *«* for YI "a. The'w is, however, oftener omitted 
 than expressed, contrary to what takes place in the case of ' a, 
 
 36. This plural is used by all adjectives and by the nomina 
 mutantis of all the conjugations. All these admit feminines 
 by termination ; and it may be laid down as a general rule 
 that a masculine plural in ut can be changed into a feminine 
 plural in at ; or if not into that, into et (see § 37) ; but many 
 feminine plurals in at do not admit masculine plurals in ut. 
 For example, sarrob ate is good Assyrian for " queens," but for 
 "kings" they would say sflrra?2e ; sarruti would come from 
 sarruty " a kingdom" or " reign." Examples of these plurals 
 need not be given here. They will appear in the following 
 chapter, and are everywhere to be met with. 
 
 37. Some feminine substantives and adjectives form their 
 plurals in etu, ete, in place of atu, ate ; or, perhaps, I should 
 say, as well as in atu, ate. In 33 1. 10 we have 'a.na ru.q^e.e.tiy 
 "to distant places (he fled) ;" in 153 B. 12 we have is.re.ti 
 nam.ra' a.ti ', the latter word is certainly an adjective in the 
 feminine plural, " shining," or the like. Whatever, then, 
 may be the meaning of isreti, it is clearly used as a substantive 
 feminine and plural. Nebuchadnezzar speaks E. I. H. 3. 13 
 of having made and purified (?) the isreet of Babylon, and 1. 
 65 those of Borsippa. The spelling is not exactly the same 
 in any two of the three passages, but there can be no reason- 
 able doubt that the word is the same. These examples esta- 
 blish the existence of the feminine plural in et ; and I confess 
 
32 
 
 that I cannot affirm with confidence anything more than that 
 it exists. 
 
 38. I may, however, mention a conjecture which has oc- 
 curred to me. Can et be a feminine plural of adjectives used 
 in place of at when they are not accompanied by substantives, 
 but are used as substantives, with " persons, places, or things," 
 understood after them ? This explains the ruqeti of 33 I. 10, 
 and the muqalleti of E. I. H. 10, 16, "may I not sinfully in- 
 cline (ar.si, from y^^) to the blasphemous persons!" the 
 feminine plural being used to express contempt. In a similar 
 context in 68 I. 2. 30, 7ii.te.tl, *' sinful persons," is substituted. 
 It explains also the >->-Y Jpp >^T*^ '^ '^ y-<^ of N.R. 8, 
 where the adjective an.ne.fi cannot be attached to the sub- 
 stantive which follows it ; but we must translate, " these (are) 
 the countries ; " and so in similar contexts. But how, it may 
 be asked, can isreti be explained on this supposition ? May 
 it not be the feminine plural of *1ti^^ isar, meaning "the 
 straight places, the avenues ? " " lightsome avenues" well suits 
 the context in 153 B. 12; and Nebuchadnezzar may have 
 said, "avenues of Babylon (and Borsippa) I caused to be 
 made, and I cleansed, or kept clean." In a future chapter 
 I will give reasons for assigning to pi the primary sense of 
 " cleansing." It does not appear to me that the spelling of 
 isret is inconsistent with its derivation from isar. So far as 
 I have observed ^^^ h is preferentially used when a word 
 begins with a radical letter, and ^TT when it begins with a 
 servile i. In the Assyrian text, and in one of the Babylonian 
 ones, the word begins with ^^^ ; in the other Babylonian 
 one the first syllable is expressed by ^YI ^^^ e.hy which, 
 however, may have only meant that the initial i was very 
 long, and which, on the authority of the other two passages, 
 we may safely pronounce to be bad spelling. I do not re- 
 collect any place in which et occurs as a formative of the 
 feminine plural, which is inconsistent with the theory which 
 I have here proposed. JSTevertheless, I am far from having 
 the same confidence in it as I have in my other statements. 
 I can only say that if the difierence between et and at be not 
 
33 
 
 what I have suggested, I am quite unable to explain what 
 that difference is. 
 
 39. A very common form of the Assyrian plural remains 
 to be spoken of, namely, that in which no syllable is added 
 to form it, but the plural cases are either the same as the 
 singular, or modifications of the singular forms. The dis- 
 tinction between the two numbers is indeed sometimes made 
 by a change of case, as will be shown in the following chapter ; 
 but even this is often not to be perceived, and the context 
 becomes our only guide. I will first give examples in which 
 the singular and plural are identical. 
 
 40. As this identity of the singular and plural is what most ' 
 persons would consider very improbable, it is necessary to 
 give some very clear instances of it. I do not rely on such a 
 passage as ' i.luM.su.nu, "of their godships," already quoted 
 in § 28. In iact,' iluti is here singular ; derived nouns in id 
 not admitting a plural. In 15 I. 106, 113, it occurs joined 
 to an adjective in the feminine singular. Neither do I rely 
 on lih.hl.khi.un, "your hearts," 9 I. 19, though I have myself 
 no doubt that lihhi is here plural, because words which are 
 certainly singular occur elsewhere, accompanied by plural 
 afiixes. I will, however, bring forward clear instances of 
 plurals, the same as the singular, accompanied by plural 
 adjectives. 
 
 41. In 152 B. 2,1, we have Tnf <I^ -::]^ HT -^f-C 
 suh.ki nakJu.tiy which I would translate "well-built houses;" 
 but whatever be the exact translation, it is clear that suhki 
 is here used as a plural, and equally clear that it is the 
 second case of the singular. Again, we have in 41 B. 57, 
 mal.ki la.hLru.ti, " ancient kings," another example that 
 cannot be contested. And so in 41 I. 74, s'u.u.ri ma.ruM, 
 "fatted oxen;" in 144 B. 10, 11, yu.me ma.'h.duM, "many 
 days." 
 
 42. In other cases forms identical with the singular are 
 substituted for the theme with a plural sign. Thus in 38 B. 67, 
 we have >:-^fZ^ ]mm for " gates ; " in another copy of the 
 same inscription, 50 B. 74, ba.bi is substituted. This noun 
 
 3 
 
34 
 
 admits also a plural in at. We have in 39 L. 22, ba.ba'a.ti. 
 Variations of this sort occur in almost all languages; and 
 there does not appear to me to be any difference in the use of 
 these plurals, so that one could be called definite and the 
 other indefinite. 
 
 43. A similar example occurs as to Sy^ y ^Y ^ y£*T ^^^ , 
 ba.tulj with the determinative of males prefixed, and the 
 plural sign added, 21 I. 43 ; in 22 I. 109, ba.tu.li, with the 
 same determinative, but without the plural sign, is substi- 
 tuted. To both of these words is added ba.tu.la.ti with the 
 determinative of females ^ prefixed. The meaning is ob- 
 viously " pure boys" and " pure girls." Other instances occur 
 in which two noims are coupled together, one of which is 
 evidently plural from its form, while the two are evidently 
 in the same number. Even if the passage 21 I. 43 did not 
 exist, we might infer that ba.tu.li in 22 I. 109 was plural, 
 from its being coupled with ba.tu.la.ti^ which is manifestly so. 
 
 44. The same principle applied to the formula "oxen" and 
 " sheep " gives us two more plurals of this sort. In ^1 II. 
 41 we have >^Ti6, the monogram for an ox, followed by 
 y-<^, the plural sign, and ^f'^TpTy . ''/, the copulative con- 
 junction; then comes z'e.e.ne, "sheep," which the context 
 proves to be a plural. In 41 I. 82 we have, in place of this, 
 rag.ge H ze.ne, and in E.I.H. 2. 28 r^a.ag.ga u z'e.e.nim. 
 The construction, it will be observed, in the last two passages, 
 is different. Surely, it is a fair inference that raggi is plural 
 as well as z^e.e.ni. I am rather disposed to draw a further 
 inference, namely, that 'a/ap, which corresponds to ►=:Yi^, 
 was only used in the singular number, and that raggu was 
 used as its plural. On this question, however, I do not think 
 that I have evidence which warrants a positive opinion. 
 
 45. The above examples, to which might be added many 
 others, must, I think, satisfy every one that the Assyrians 
 had a plural of the same form with the singular ; though 
 there was some slight difference in the use of the cases. I 
 now proceed to consider plurals which are modifications of 
 the singular, not identical with it. The principle of these 
 
35 
 
 modifications is that in dissyllables, the last of which is not 
 lengthened by a quiescent letter, the singular inclines to have 
 the accent on the former syllable, and the plural on the latter. 
 Monosyllabic themes admit no such modification ; nor those 
 in which, the middle radical being weak and the two vowels 
 the same, they coalesce in the declension into one, as JKIC, 
 ffm, whence zenu. Nor again is this distinction possible 
 where the last vowel of the theme is necessarily long, as in 
 hatul, § 43. 
 
 46. As a general rule, however, such a noun as pigil would 
 add its case-ending, in the singular to pigil with the accent 
 on the first syllable, and would thus form, piglu, etc., with the 
 second vowel suppressed ; while the plural would retain the 
 second vowel ; and would perhaps admit also another change 
 consequent on this. For example nakru is used for " an 
 enemy," but nakiri for " of the enemies," the theme being 
 nakir, 
 
 47. I have spoken of this distinction being rendered more 
 marked by another change consequent on this. The Assyrians 
 were accustomed to double the consonant of an accented syl- 
 lable. This is the reason why the second radical is commonly 
 doubled in the present of the first conjugation, as in i.qah.bi, 
 " he says," and numerous other instances, which must not be 
 supposed to be Pihel forms. In conformity with this usage 
 the last consonant of the noun is often doubled in the plural. 
 Thus, in 43 I. 43, we have sal.gu na.haUim, " the snows of 
 the valleys ; " while in the parallel text, 40 I. 77, we have 
 naJh.li, " of the valley." 
 
 48. Where the first radical was a very weak one, that ad- 
 mitted of being altogether dropped, it was, I think, dropped 
 in plurals of this kind. Instances of this are neceesarily rare ; 
 and I am not sure that there are sufiicient to establish the usage. 
 I remark, however, that in the same nouns that could drop the 
 initial syllable in the plural, that syllable is dropped in the 
 singular when the noun is in what Hebrew grammarians call 
 the state of construction. It may have appeared strange that 
 I have spoken of the noun, when without a case- ending, as 
 the theme and not as the construct form. I have done this_ 
 
36 
 
 advisedly. It will be seen in the following cliapter that the 
 second case is repeatedly used, — almost, if not quite, as often 
 as the theme, — where the Hebrews would use the construct 
 form ; and it will be seen also that the theme is repeatedly 
 used where the Hebrews would use the absolute form. If I 
 were, therefore, to give the name of construct form to the 
 theme, I should be using a term that would certainly mislead. 
 I think, however, that the theme was in some instances pro- 
 nounced differently when it did and when it did not indicate 
 a state of construction ; that the theme pigil, for instance, was 
 pronounced pigil when in construction, and pigil when not so ; 
 and I think that, consequent upon this difference, when the 
 first radical was so weak that the first syllable might be 
 dropped, it was dropped when the noun was in construction. 
 
 49. Taking, then, 'agal or 'agil as a type of such a noun 
 as I have described, I think we should have for the theme 
 when absolute ''dgal or 'dgil; for the theme in construction, 
 gal or gil; for the first case of the singular, 'a^/i^ ; and for 
 the first case of the plural, galu or gilu. Such is the con- 
 clusion at which I have arrived by induction ; but I give it 
 doubtingly ; the examples being few, and what others would 
 probably explain differently from what I do. I think there 
 are three nouns following this type which occur with and 
 without the initial syllable, namely, those which signify " a 
 son," " a servant," and " a bull." From the first we have 
 'a.b'i.il.su, "his son," E. I. H. 1. 33; 'ab.lam, "son," 51 I. 
 "No. 1, 2. 16, in one of the copies, the other having the usual 
 monogram for " son." Without the initial syllable we have, 
 I think, baJu.sa, "her sons," in 6Q II. 5. It must be owned, 
 however, that this passage is obscure ; and others would in- 
 terpret it otherwise. On Hebrew and Greek transcriptions 
 but little reliance can be placed, so far as the vowels are con- 
 cerned ; but I am disposed to take "Baladan" as authority 
 for the theme being 'abal as well as 'abil. 
 
 50. As respects the second noun, we have in 95 B. 6, 'ar.du 
 for " servant," before kan.su, " obedient," where 145 B. 24 
 has the monogram >Ii^T ; also in 10 II. 15 /. we have 'ar.dtiy 
 " a servant," in the third case. On the other hand, in 1 L. 1, 
 
37 
 
 we have, I think (but I admit that there is no positive proof, 
 and others take it differently), rid for " servant of (Assur).'* 
 As for the third word, ri.i.mu is given in Porter's transcript 
 of portions of the E. I. H. inscription as the equivalent of 
 Jl^^ \<M ^ 3. 59. This group certainly signifies "bulls." 
 On the other hand, 'ar.rn^" is a singular noun in the second 
 case, denoting some male animal ; and it appears to me pro- 
 bable that it is the singular of the other word. A fourth 
 example may perhaps be the plural ni.si from anis^ see § 57. 
 This is a case in which the few examples bearing on the ques- 
 tion appear to throw light on one another. "Without, there- 
 fore, maintaining that the positions advanced in the last three 
 sections are established, I think that they are highly probable ; 
 and I state them here in order that they may be tested by 
 further examples, should such occur. 
 
 51. It remains for me to speak of the dual number of nouns ; 
 but here again I confess that I can only speak doubtingly. It 
 appears to me that a dual theme in <2, formed precisely as the 
 third case of the singular, must be admitted to exist. I have 
 only met with it, however, in a few nouns ; qa.ta, " hands," 
 (where the t is radical, for we have the plurals qa.ta.tu, 
 qa.ta.te in 6 II. 40, 41), se.pa, "feet," hir.kay "knees," and 
 perhaps ma.ta or ma.da^ " peoples." I once thought that a 
 distinction might be established between the second case of 
 the dual and that of the plural ; but I rather think now that 
 such is not the case. 
 
 Appendix on the possessive pronominal affixes. 
 
 52. It has been necessary in several places to assume a 
 knowledge on the part of the reader of the possessive pro- 
 nominal affixes attached to nouns. A complete knowledge of 
 these affixes is essential to a grammatical knowledge of the 
 inflexions of nouns ; and I believe that no complete and cor- 
 rect list of them has yet been published. I will therefore 
 endeavour to supply the deficiency in the following sections. 
 
 53. The affixes attached to the case endings are always 
 
38 
 
 expressed by the addition of one or two characters to those 
 which express the noun. They are the following : 
 
 1. s. Tr '^ ; instead of which we find more commonly after 
 
 the second case ^Efy ya, " my." 
 
 2. s.m. ^j:|td[ ha, and f, <]^ kl, "thy." 
 
 3. s.m. ^y or \ su, " his," f. V or *^ sa, " her." 
 
 1. p. Sf^ ni, "our." 
 
 2. p.m. jfcj *^ ku.nu; f. ^fcj >--^y ki.na, "your." 
 
 3. p.m. ^1 ^ su.nu or ^1 ^y^fy s'u.un; f. ^y>^ >"-^y 
 5e.w«x or ^y>- ^£J^ 5'i.i72, ** their." 
 
 '^lyvy is used for the affix of the 3. p. of both genders ; 
 it is interchanged with su.un, cf. 164 B. 23 and 165 B. 17; 
 and with s'i.in, cf. 38 B. 69 and 46 B. 76. 
 
 54. The affix of the 3 s.m. after an unaccented u is often 
 shortened to s ; this happens most commonly after the femi- 
 nine first case tu ; and the tus is then expressed by J^ . 
 
 On the other hand, after an accented u, an additional 
 character is generally introduced after the first case, as it 
 would be written without the affix. This character is >=yyy^ 
 or ^ before '6?, the affix of the first person singular (which, 
 however, is not to be read as an additional syllable) and be- 
 fore those of the second and third persons ^ j6Y>->- uh and 
 J^y us respectively. Before the affix of the first person 
 plural I should from analogy expect >^X\ un, but I have as 
 yet met with no example of it. 
 
 55. Although this reduplication, which may be compared 
 with that of § 48, may be used after the u of the first case 
 whenever it is not preceded by a ty formative of the feminine ; 
 there are some words, after which it occurs, which should be 
 particularly noticed. Such are the prepositions, kirhu, " with- 
 in," ziru, " upon," and panu, " before," which are used before 
 pronominal affixes in place of the kirih, zir, and ^:>«/^, which 
 are used before nouns. Examples are zi.ru.u'a, " upon me," 
 41 I. 45 ; pan.ukM, " before thee (f.)*," 66 II. 2. 7, 8 ; in 
 a similar inscription addressed to Nebo, 85 L. 15, pan.iik.ka ; 
 
39 
 
 k'Lir.b*u.us.sa, " witHn it," E. I. H. 10, 12 ; ziyu.us.su.un, 
 " upon them," 39 I. 78. Such also is the noun of non- 
 existence panUf whence yaM^uMla^ "I am not," 42 II. 14 r, ; 
 and in the two preceding lines, ya.n^uk.ha, " thou art not," 
 ya.n'u.us.sUi "he is not." I may also notice ki.hl.t^u.iik.ka, 
 " it is thy wiU " or " in pursuance of thy wiU," E. I. H. 10, 1 ; 
 ki.bi.tus.sUf "in pursuance of his will," 16 B. 141. The t of 
 this word is radical. I think it is invariably used of the 
 divine will, which was not to be resisted. 
 
 56. The affixes annexed to the theme are the same as those 
 annexed to the case endings, except for the first person 
 singular ; and for the third person in all its forms, when the 
 theme ends in a dental or sibilant. In the last case, s is sub- 
 stituted for s in the affix: ; the preceding consonant being 
 sometimes retained, sometimes omitted, but most frequently 
 changed into another s. That is to say, the affixes of the 
 third person given above can never follow ^^J , ^^^ , '^ J , 
 
 5^2EEI' -I' ^"-y<T^ ^^ ^n> or J^f; oranycha- 
 racter, the value of which terminates with any of these ; but 
 after these letters ^]] is substituted for ^, ^^ for ^, 
 and *^YY for ^Y»- . Examples are hiyi.it. su, "its ditch," 
 E.I.H. 6. 60, or hi.ri.m, 65, 1. 2. 7, or Uyi.is.m. E.I.H. 6. 30. 
 All these are used in parallel texts, and are evidently equi- 
 valent. The last of the three forms is, however, the com- 
 monest. Other examples are ru.pu.us.su, "its breadth," 
 from rupus, 7 I.E. 23 ; e.pi.su.un, " their work," from epis, 
 132 B. 18. It is needless to multiply them, as they are every- 
 where to be met with. To the rule here laid down there are 
 no exceptions ; and the student, if he thinks that he sees it 
 violated, may be quite sure that he is mistaken as to his read- 
 ing of the text. 
 
 57. The affix of the first person attached to a theme 
 which ends in a consonant is generally i ; sometimes a is 
 substituted, but only, I believe, in Babylonian inscriptions. 
 Examples are 'a.ha, " my father," E. I. H. 7, 48 ; b'i.e.lay 
 " my lord," m I. 18. This affix, whether i or a, is not repre- 
 sented by a separate character, but by a change of th& last 
 
40 
 
 character of the theme, which, with this affix, is the same as 
 the second or third case. Examples are very numerous ; but 
 they appear to have been overlooked, or set down as mistakes, 
 by others. I will give a very few out of a long list. In 
 151 B. 16, Sargon says, 'ak.zu.ra 'us.ma.ni, "I prepared my 
 camp." Five lines after, speaking of his enemy, he says, 
 ik.zu.ra ' us. m' a. an.su, "he prepared his camp." These trans- 
 lations may be only approximate ; but there can be no doubt 
 as to the "I," "my," "he/' and "his." Again, b'i.in.ti 
 'ad.din.sUy 145 B. 18, is, " my daughter I gave to him;" 
 'u.s'a.az.bil ra.ma.ni, 49 1.4, 11, is, " I caused myself to carry." 
 Both these texts are correct as they stand ; and the emenda- 
 tions that have been proposed would render them the con- 
 trary. We have also qa.ti for " my hand," 10 I. 98 ; mu.ti, 
 "my husband, 10 II. 4 l/as.sa.ti, "my wife," do. 10 /. This 
 last is for an.sa.ti, the feminine theme of anis, "a man," 
 from which I derive the plural ni.si mentioned in § 50. All 
 these nouns occur with other affixes, qa.as.su, "his hand," 
 49 B. 32, mu.us.su, " her husband, 10 II. 2 /.; ojidi' as.sa.ti.su, 
 " to his wife," do. 9 /. 
 
 58. The use of su for " her " in this ancient text must not 
 be passed over. We have oho'^at.ta for "thou (woman);" 
 do. 10 /. This fragment of the ancient laws of Assyria is 
 probably the oldest text in the language that we possess. The 
 tablet, indeed, is not older than the seventh century B.C. ; but 
 it is a transcript of one of the highest antiquity. Now, it is 
 very remarkable that, in the Hebrew Pentateuch, the mascu- 
 line pronoun of the third person singular is often used for the 
 feminine ; that is to say, if we go by the written letters and 
 neglect the Qeri and vowel points. The distinction between 
 the pronouns of the second person, masculine, and feminine is 
 also ill several places only made in the vowel points. I think 
 it is a fair inference from this, that in the earliest stage of the 
 language " thou " and " thy," masculine and feminine, had 
 but one representative ; and that " his '* and " her " were ex- 
 pressed alike, as they were in Latin, and as they still are in 
 the languages derived from it. 
 
i" 
 
 Linguistic Publications of IVilbner ^ Co, 
 
 Goldstucker. — Panini : His Place in Sanskrit Literature. An Investigation 
 of some Literary and Chronological Questions which may be settled by a study of his 
 "Work. A separate impression of the Preface to the Facsimile of MS. No. 17 in tlie 
 Library of Her Majesty's Home Government for India, which contains a portion of the 
 Manava-Kalpa-Sutka, with the Commentary of Kumakila-Swamin. By Theodok 
 GoLDSTxicKER. Imperial 8vo., pp. 268, cloth. 12«. 
 
 Goldstiicker. — ^ Compendious Sanskkit-English Dictionary, for the Use 
 
 of those who intend to read the easier Works of Classical Sanskrit Literature. By 
 Theodor GoLDSTiJCKER. Small 4to., pp. 900, cloth. {In preparation. 
 
 G-rammatography. A Manual of Refeeence to the Alphabets of Ancient 
 and Modern Languages. Based on the German Compilation of F. Ballhorn. In 
 one volume, royal Svo., pp. 80, cloth. Is. 6d, 
 
 The " Grammatography" is offered to the public as a compendious introduction to the reading? 
 of the most important ancient and modern languages. Simple in its design, it will be consulted 
 with advantage by the philological student, the amateur linguist, the bookseller, the corrector of 
 the press, and the diligent compositor. 
 
 Haug. — Outline of a Grammar of the Zend Language. By Mabiin Haug, 
 Dr. Phil. 8vo., pp. 82, sewed. 145. 
 
 Haug. — Essays on the Sacred Language, Writings, and Religion of the 
 Parsees. By Martin Ha^g, Dr. Phil., Superintendent of Sanskrit Studies in the 
 Poona College. 8vo., pp. 278, cloth. 21«. 
 
 Hang. — The Aitareya Brahmanam of the Rig Yeda : containing the Earliest 
 Speculations of the Brahraans on the meaning of the Sacrificial Prayers, and on tl'" 
 Origin, Performance, and Sense of the Rites of the Vedic Edition. Edited, Trar ' 
 and Explained, by Martix Haug, Ph.D., Superintendent of Sanskrit Studk . 
 Poona College, etc., etc. In 2 Vols., cr. Svo. Vol. I. Contents, Sans^-^ UOOKS 
 Preface, Introductory Essay, and a Map of the Sacrificial CoraB5r 
 Sacrifice, pp. 312. Vol. II., Translation, with Notes, pp. 54' . ^ Hav*; 
 
 Hernisz. — ^ Guide to Conversation in the English ai 
 guages, for the Use of Americans and Chinese in California a* 
 Stanislas Hernisz. Square 8vo., pp. 274, sewed. 
 
 The Chinese characters contained in this work are from the collections of 
 
 engraved on steel, and cast into moveable types, by Mr. Marcellin Legrand, Ei., 
 
 Imperial Printing Office at Paris. They are used by most of the Missions to China. 
 
 Hoffman. — Shopping Dialogues, in Japanese, Dutch, and English. By 
 Professor J. Hoffman. Oblong 8vo., sewed. 3s. 
 
 Justi. — Handbuch der Zendsprache van Ferdinand Justi. Altbactrisches 
 w'aerterbuch. Grammar Chrestomathie. Four parts 4to. sewed, pp. xxii. and 424. 
 Leipzig, 1864. 24,v. 
 
 Kidd.-^CATALOGUE of the Chinese Library of the Royal Asiatic Society. 
 By the Eev. S. Kidd. 8vo., pp. 58, sewed. U. 
 
 Legge. — The Chinese Classics. With a Translation, Critical and Exe- 
 
 fetical, Notes, Prolegomena, and Copious Indexes. By James Legge, D.D., of the 
 .ondon Missionary Society. In Seven Vols. Vol. I., containing Confucian Analects, 
 the Great Learning, and the Doctrine of the Mean. 8vo., pp., 526, cloth, price 21. 2s. 
 Vol. II., containing the Works of Mencius. 8vo., pp. 634, cloth, price 21. 2s. 
 
 Manava-Kalpa-Sutra ; being a portion of .this ancient Work on Yaidik 
 Rites, together with the Commentary of Kumarila-Swamin. A Facsimile of the MS. 
 No. 17 in the Library of Her Majesty's Home Government for India. "With a Preface 
 by Theodor GoldstUckek. Oblong folio, pp. 268 of lettei-press and 121 leaves of 
 facsimiles. Cloth. 4/. 4*. 
 
 Medhurst.— Chinese Dialogues, Questions, and Familiar Sentences, 
 literally "rendered into English, with a view to promote commercial intercourse and 
 assist beginners in the Language. By the late W. H. Medhurst, D.D. A new and 
 enlarged edition, pp. 226. 8vo. 15*. 
 
 Morley. — ^ Descriptive Catalogue of the Historical Manuscripts in the 
 Arabic and Persian Languages preserved in the Library of the Royal Asiatic Society 
 of Great Britain and Ireland. By William H. Morley, M.R.A.S, 8vo., pp. viii. 
 and 160, sewed. London, 1854. 2«. 6rf. 
 
Muir.— Ori 
 
 of India, 
 illustrated 
 Vedic witl 
 440, cloth 
 
 Bam Eaz.- 
 
 Native Ju^ 
 Great Brit= 
 1834. Ori 
 
 •^^ 
 
 and AssYB 
 brief Notic. 
 Asiatic Soc 
 1850. 
 
 Eawlinson.- 
 
 Nineveh. ]- 
 Layard, Es 
 
 Eenan.—An - 
 
 Agricultuh 
 Shemitic Na- 
 I'lnstitut. J 
 
 "^.^eda Saf 
 
 ^ifthto] 
 last --la and" 
 
 from whiL. 
 these noi 
 49B.32,- 
 
 "*°%;^- 
 
 58esof J 
 
 Vislihu-Puranr 
 
 lated from the 
 Puranas. By " 
 Notes, by Dr. 
 
 Wilson.— WoRK_ 
 
 of the Royal i 
 Germany, etc., 
 and II.— Also r 
 HiNDus, by the 
 Dr. Eei.vhgld - 
 
 Wilson.— Essays 
 
 connected with ~ 
 Edited by Dr. R 
 
 Wilson.— The Pr" 
 A Lecture delive 
 fessor H. H. Wi] 
 
 THIS BOOK IS DUE ON THE LAST DATE 
 STAMPED BELOW 
 
 AN INITIAL FINE OF 25 CENTS 
 
 WILL BE ASSESSED FOR FAILURE TO RETURN 
 THIS BOOK ON THE DATE DUE. THE PENALTY 
 WILL INCREASE TO 50 CENTS ON THE FOURTH 
 DAY AND TO $1.00 ON THE SEVENTH DAY 
 OVERDUE.* 
 
 fiawlinson. JU^ ^J^ 1944 
 
 JOl IC 1944 
 
 3l3ajv 
 
 t^lish, 
 
 tions cfi 
 by A. Bji 
 1*. 
 
 « of t.]) 
 
 Bmbre d 
 
 itutin, 
 ' for th' 
 Sanskri 
 imes E 
 Benares 
 Press. 
 
 iment 
 ecedin; 
 tes, anl 
 I. 2s. 
 
 trans 
 [■Q othe 
 .^ witl- 
 , Qd. i 
 h'ess. 
 
 lety o| 
 ols. If' 
 of thq 
 ted bjl 
 
 jectsiijj 
 1 audi] 
 i6*. 
 
 LD21-10w-5,'43 (6061s) JcrHEj 
 
 i-ondon, 1852. " ' ^-^r, Pro- 
 
 Wise.— Commentary on tbp TTtxt^^ a « , ^ ^^*' 
 
 M.I.., Bengal Medical Semite! ^^Zt^^^f.^^^-- % T. A. W.k„ 
 
 LONDOK 
 
 : TRirBNEE & CO.. 60. PATERNOSTER ROW. 
 
U. C. BERKELEY LIBRARIES 
 
 RETURN TO the circulation desk of any 
 
 University of California Library 
 
 or to the 
 
 NORTHERN REGIONAL LIBRARY FACILITY 
 BIdg. 400, Richmond Field Station 
 University of California 
 Richmond, CA 94804-4698 
 
 ALL BOOKS MAY BE RECALLED AFTER 7 DAYS 
 2-month loans may be renewed by calling 
 
 (510)642-6753 
 1-year loans may be recharged by bringing books 
 
 to NRLF 
 Renewals and recharges may be made 4 days 
 
 prior to due date 
 
 DUE AS STAMPED BELOW 
 
 OCT 16 1995 
 
 MAY 2 9 1996